Full text of "The collected works of Edward Sapir"

■■-;kk<.;;:

■?■:■■-■■• ■■■ ,r',:'rl „

,, /.'■•'J'!' I- '," ■-' ■ ."iwi''..)..'....'."';''''*-'

^!^;'-^^f^'ti'^^:^.

:'*r''5^-'. . .^ .„

Mouton
de Gruyter

Vi •■;..;>-(;

Ihc

(\)lIcLicd Works

F.dwarcl Saj^ir

The Collected Works of Edward Sapir

Editorial Board

Philip Sapir
Editor-in-Chief

William Bright

Regna Darnell

Victor GoUa

Eric P. Hamp

Richard Handler

Judith T. Irvine

Pierre Swiggers

The

Collected Works

Edward Sapir

General Linguistics

Volume Editor
Pierre Swiggers

With contributions by

Philip Sapir (t)
Zellig S. Harris (t)

John Lyons
Stanley Newman (t)

MoLiton dc CiruNlcr
Berlin • New York

Mouton dc (iru\ur (li)rmcrly Mduidii. Hic Hague)

is d DiviMon of Walter dc Ciruylcr CimhH cV: Co. K(i. Berlin.

@ Printed on acid-free paper which falls within the guidelines
of the ANSI to ensure permanence and durability.

Library of Congress Cataloging-in- Publication Data

Sapir. Edward. 1 884- 1939.
The collected works of Edward Sapir. 1, General linguistics / edited by Pierre Swiggers.

p. cm. - (The collected works of Edward Sapir)
Includes bibliographical references and index.
ISBN 978-3-1 1-019519-4 (hardcover : alk. paper) 1. Linguistics. 2.
Language and languages. I. Swiggers, Pierre. II. Title.
P27.S325 2008
410-dc22

2007047474

ISBN 978-3-11-019519-4

Bibliographic information published by the Deutsche Nationalbibliothek

The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie;
detailed bibliographic data are available in the Internet at hltp://dnb.d-nb.de.

All rights reserved, including those of translation into foreign languages. No part of this
book may be reproduced or transmitted in any form or by any means, electronic or
mechanical, including photocopy, recording or any information storage and retrieval
system, without permission in writing from the publisher.
Printed in Germany.

Edward SAPIR, around 1938
(Courtesy of Sapir family)

Edward Sapir ( ISS4 I W.^'M has been rclcrrcd to as "one of ihc nu)sl bril-
liant scholars in linuuislics and anlhropolouy in our counlrv" (Franz Boas)
and as "one ot the greatest figures in American humanistic scholarship"
(Frankhn l^dgerton). Mis classic book. J.iini^uuf^c ( 1W21 ). is still m use. and
main of his papers in general linguistics, such as 'Sound Patterns in
Language" and "Hie Psychological Reality of Phonemes. ' stand also as clas-
sics. The development of the American descriptive school ol structural lin-
guistics, including the adoption of phonemic principles in the study of non-
literary languages, was primarily due to him.

The large body (.)f work he carried out on Native Ameneaii languages lias
been called "ground-breaking" and "monumenlal" and includes descripli\e.
historical, and comparative studies. Iliey are of continuing importance and
relevance to today's scholars.

Not to be ignored are his studies in Indo-Luropean. Semitic, and ,-\trican
languages, which have been characterized as "masterpieces of brilliant asso-
ciation" (Zellig Harris). Further, he is recognized as a forefather of ethno-
linguistic and sociolinguistic studies.

in anthropology Sapir contributed the classic statement on the llKH)ry and
methodology of the American school of Franz Boas in his monograph. "Time
Perspective in Aboriginal American Culture" (1916). His major contribu-
tion, however, was as a pioneer and proponent for studies on the interrelation
of culture and personality, of society and the individual, providinii the the-
oretical basis for what is known today as symbolic anthropology

He was, in addition, a poet, and contributed papers on aesthetics. Iiteia
ture, music, and social criticism.

Note to the Reader

'Ihroughout The Collected Works of Edward Sapir, those publications whose
ivpographic cDmplcxity would have made new typesetting and proofreading
difticult have been photographically reproduced. All other material has been
newly typeset. When possible, the editors have worked from Sapir's personal
copies of his published work, incorporating his corrections and additions into
the reset text. Such emendations are acknowledged in the endnotes. Where the
editors themselves have corrected an obvious typographical error, this is noted
by brackets around the corrected form.

The page numbers of the original publication are retained in the pho-
tographically reproduced material; in reset material, the original publication's
pagination appears as bracketed numbers within the text at the point where the
original page break occurred. To avoid confusion and to conform to the existing
literature, the page numbers cited in introductions and editorial notes are those
of the original publications.

Footnotes which appeared in the original publications appear here as foot-
notes. Editorial notes appear as endnotes. Endnote numbers are placed in the
margins of photographically reproduced material; in reset material they are
inserted in the text as superscript numbers in brackets. The first, unnumbered
endnote for each work contains the citation of the original publication and,
where appropriate , an acknowledgment of permission to reprint the work here.

All citations of Sapir's works in the editorial matter throughout these vol-
umes conform to the master bibliography that appears in Volume XVI; since
not all works will be cited in any given volume, the letters following the dates
are discontinuous within a single volume's references. In volumes where
unpublished materials by Sapir have been cited, a list of the items cited and the
archives holding them is appended to the References.

Conlcnls

Frontispiece: I\l\var(.i S;ipir. nrouiul 1M3S " 6

Note to the Reader 7

1 orewoid. In V\iilip Saplr II

ChriMiologieal list of Sapii "s \Miliiigs eoiilaiiieJ in \t>UmiL I IS

Prelace

INIKODI c 1 K)N: Sapii's Lite and Work: Iwo appraisals

Introductory Note, by Philip Sapir and /'icrre Swif^f^ers 21

Introduction to Zellig Harris's text, by Pierre Swi^f^ers 23

<Sapir"s approach to hmguago, by Zcllii^ S. Harris 26

Introduction to Stanley Newman's text, by Pierre Swii^iiers 47

<Ed\vard Sapir's work and stylo, by Stanley Newfntin 49

Si (1 ION ()\i: I 111 PkoHI 1 \1 oi 1 III ()RI(.I\ OI LANC.lAdi- ( 1^)07)

Introduction to Sapir's "Herder's "Ursprung der Sprache"" ( 1^0'').

by Pierre Svviggers 57

I lerder's "Ursprung der Sprache" ( I'-K)?) 65

Section Two: History, Variety and Setting of Language

(U)ll. U)12)

Introduction: History, Variety and Setting o\' Language, by Pierre Swifif^ers . 103

The History and Varieties of Human Speech ( UH I) HI

Language and Environment (1912) 1 ^^

SfCIION ThRI:E:ThEORETI( AI , Di S( KIIM I\ I A\n HiSIORK AI LlNtil ISIKS

(1923-1929)

Introduction: Theoretical, Descriptive and Historical

Linguistics, 1923-1929. by Pierre S\\ii^i;er\ 153

An Approach to Symbolism (review oi ( "K. Ogden and I A. Kichardv,

The Meaning, of Meanini^) (1923) I<^3

ITie Grammarian and his Language ( 1924) 167

Review of A. Meillet and M. Cohen (eds.). I.e\ lam^ue.s dit numdc (1925). . 177

Sound Patterns in Language (1925) 179

Philology (1926) 195

Review of (). Jespersen. Vfankind, Nation and Individmil from

(I l.ini^iiisiic Point ol \'ie\\ { 1^^2(i) .... -'^-^

Language as a lorm oi Human Heh.iMor { T^J/) 2()4

Review of R.( i. Kent. Pani^uai^e and Philoloi^y ( 1928) 217

The Status of Linguistics as a Science ( 1929) 219

A Studv in Phonetic Svnibolism ( |92W). . 227

IQ General Linguistics I

Section Four: The Problem of an International Auxiliary Language

(1925-1933)

Introduction: The Problem of an International Auxiliary Language,

hv Pierre Swif^gers 245

Memorandum on the Problem of an International Auxiliary

Language (1925) 251

ITie Function of an International Auxiliary Language (1931) 264

Wanted: a World Language (1931) 276

The Case for a Constructed International Language (1933) 284

Annex: The Statement of the International Auxiliary Language Association made

at the second International Conference of Linguists (Geneva, 1931) 287

Section Five: Stlidies in Universal Conceptual Grammar
(1930.1932,1944)

Introductory Note: Sapir's Studies in Universal Conceptual Grammar,

by Pierre Swiggers 291

Introduction to Sapir's texts 'Totality", "Grading", and

"The Expression of the Ending-Point Relation", by Sir John Lyons 294

TotaUty (1930) 300

The Expression of the Ending-Point Relation in English, French, and

German (1932) 326

Grading, a Study in Semantics (1944) 447

Section Six: Patterns of Language in Relation to History and
Society (1931-1933-[1947])

Introduction: Sapir's General Linguistics in the 1930s, by Pierre Swiggers . . . 473
The Concept of Phonetic Law as Tested in Primitive Languages

by Leonard Bloomfield (1931) 484

Communication (1931) 494

Conceptual Categories in Primitive Languages (1931) 498

Dialect (1931) 499

Language (1933) 503

La realite psychologique des phonemes (1933) 518

Corrections to the French version published in 1933 538

The Psychological Reality of Phonemes (1933/1949) 539

Editorial notes on the English version published in 1949 554

The Relation of American Indian Linguistics to General Linguistics (1947) . 556

Acknowledgements 561

Indices

Index of Personal Names 563

Index of Concepts 568

Index of Languages 578

Foreword

In January 1982, David Mandelbaum wrote to David Sapir. suggesting that "a
new volume of your father's writings might be prepared for puhheation in I9S4,
the centenary of his birth." He suggested that "it might contain some of the
papers that were not included in the Selected Writings of Edward Sapir: perhaps
some of his letters; and possibly some of the papers that have discussed his
work." In April he further suggested, as possibilities, "a biographical memoir,
recollections by some who knew him, an essay on his influence and coniinuiiiu
stimulus, selections from his poetry, selections from his letters to Lowie, Kroebcr
and others, and a bibliography of writings about him."

Later, in 1982, a complete list of Edward Sapir's major scholarly writings was
circulated to some four dozen anthropologists and linguists, with a request that
they rate each paper on a four-point scale. Some 25 replies were received, llierc
was hardly a single paper that two or more had not rated as a "must" or "highl\
desirable." Also, a number of individuals expressed their preference for a
"Complete Works" rather than a "Selection."

At the 1982 meeting of the American Anthropological Association, an ad hoc
meeting of some 10 people was convened to discuss how to go about making a
proper selection of additional papers not included in the Selected Writings, and
where a publisher might be found. Among those present were Dell Hymes.
Regna Darnell, Victor GoUa, Keith Basso, Harold Conklin, Lita Osmundsen o{
the Wenner-Gren Foundation, and myself. After a relatively brief discussion it
was unanimously recommended that, rather than a second "SWES," a "Complete
Works" should be published, if at all possible. Two university presses were men-
tioned as possible publishers, but neither proved to be interested in a "Complete
Works," and both required a significant financial contribution to support a much
more limited publication.

At the same time, a joint ad hoc committee of the American Anthropological
Association and the Linguistic Society of America had been established to de-
velop plans for celebrating the centennial of Edward Sapir's birth. Its membership
included the following: Dell Hymes (Chairman), William Bright. William (\)wan,
Regna Darnell, Paul Friedrich, Margaret Langdon. Vicli^ria I ronikin. Joel
Sherzer, and Judith Irvine.

In the meantime it had become known that plans were well advanced lor an
Edward Sapir Centennial Conference, to be held in Ottawa. Canada, where
Edward Sapir had served as the first Chief of the Anthropological Division
within the Canadian Geological Survey, Department of Mines, from 1910 to 1925.
This conference, planned and organized by William Cowan. Michael K. Foster,

12 General Linguistics I

and Konrad Koerner, was held on October 1-3, 1984, and was very well attended
by participants from Canada, the United States, and Japan. The Conference
Proceedings were published in 1986.'

Not wishing to attempt to duplicate this event, Dell Hymes, then President of
ihc Linguistic Society of America and Past President of the American
Anthropological Association, and Chairman of the abovementioned ad hoc com-
mittee, agreed to have the committee assist in selecting a Board of Editors for
The Collected Works, as well as in finding a publisher. At the 1983 meeting of the
IJneuistic Society of America, the ad hoc committee duly met and appointed the
following to serve as a Board of Editors: William Bright, Regna Darnell, Judith
Irvine, \'iikov Malkiel, and myself, as Editor-in-Chief. Dr. Malkiel later found it
necessary to resign, and was succeeded by Eric Hamp. Victor Golla, Richard
Handler and Pierre Swiggers were added later to the Board.

Present at the annual meeting of the Linguistic Society of America that year
was Dr. Marie-Louise Liebe-Harkort, the then newly appointed Editor-in-Chief
of Mouton de Gruyter, a recently added division of the Berhn pubhshing house
of Walter de Gruyter & Co., and an Americanist in her own right. Upon learning
of the plans of the ad hoc committee, she met with me and indicated that she was
very interested in the possibility of Mouton de Gruyter serving as the publisher
of The Collected Works, and would be happy to propose this to the Board of
Directors of Walter de Gruyter & Co. This she did and the Board gave its ap-
proval.

The Editorial Board for The Collected Works held its first meeting at the
Berkeley Campus of the University of California, courtesy of Dr Malkiel, in July
1984. a few months before the Ottawa Centennial Conference, where, later, the
Board was able to meet with Dr Liebe-Harkort, and get the project off to a suc-
cessful start. During the twenty years that have passed since, the members have
taken the responsibility of editing the several volumes, with the assistance of
other anthropologists and linguists on those volumes devoted to a single lan-
guage or group of related languages. At present, eight of the volumes have been
published, with an additional five in various stages of preparation.

Funding has been received from the National Science Foundation for support
of the scholarly work on six of the nine volumes devoted to specific languages or

New Perspectives in Language, Culture, and Personality. Proceedings of the Edward Sapir Centenary
Conference (Ottawa, 1-3 October 1984). Edited by William Cowan, Michael K. Foster, and Konrad Koerner
(SiHoLS 41 ), Amsterdam/Philadelphia: John Benjamins, 1986, xiv-627 p. [Reviews of the volume were publish-
ed m: Lingua e Stile 22 (1987). 623-624 (L. Rosiello); American Anthropologist 90 (1988), 219 (RK. Bock);
Historiographia Linguistica 15 (1988), 405^09 (M.B. Emeneau); Anthropos 84 (1989). 269 (J.W. Burton);
i'nTnLv ^^^^^^' ^^^383 (R.H. Robins); Tijdschrift voor Filosofie 51 (1989), 160-162 (R Swiggers); Semiotica
79 (1990). 273-300 (review article by A.S. Kaye and H. Waltz).]

Foreword 13

groups of languages; and from the Phillips lund ol ihc American i'hilosophieal
Society for the project expenses of the Editor-in-Chief. We are pleased to express
our gratitude to these two institutions.

7Tie members of the Board have received assistance Ironi a luinihcr ol ct)llea-
gues in anthropology and linguistics, who have worked with the noIuihc editors
and helped to edit a number of volumes, or sections within solumes. Of particu-
lar value has been the editing of previously unpublished linguistic and ethnolo-
gic materials which Edward Sapir left unfinished at the time of his death. Ilie
Editorial Board wishes to thank all these scholars who have contributed to or are
contributing in this way to provide a "Complete Works."

I wish personally to thank Ms Jane McGary, Associate Editor, for her valuable
help in the editing of these volumes during the past decade, as well as the mem-
bers of the Board of Editors themselves, who have given much of their time and
effort to seeing this project through to a successful conclusion. I also wish to
thank Dr Liebe-Harkort, and her successor Dr Anke Beck, for the interest and
support over the years.

Philip Sai'ik
Editor-in-Chief, The Collected Works of Edward Sapir

Chronological list of Sapir's writings contained in Volume 1

1907 "Herder's "Ursprung der Sprache"". Modern /'hiloloi^v >. 10^^ 142.

1911 "The History and Varieties of Human Speech". I'opnhu Science Monthly
79. 45-67.

1912 "Language and Environment". American Aniliropoloi^isi 14. 226-242.

1923 "An Approach to Symbolism". The Freeman 7. 572-573.

1924 "The Grammarian and his Language". American Mercury 1. 149-155.

1925 "Memorandum on the Problem of an International Auxiliary Language".
The Romanic Review 16. 244-256.

1925 Review of Antoine Meillet - Marcel Cohen (eds.), Les langues du niomle.
Modern Language Notes 40. 373-375.

1925 "Sound Patterns in Language". Language 1. 37-51.

1926 "Philology". Encyclopaedia Britannica, Supplementary volumes ( 1 3lh cd. )
vol.3, 112-115.

1926 Review of Otto Jespersen, Mankind, Nation, and Individual from a
Linguistic Point of View. American Journal of Sociology 32. 498^99.

1927 "Language as a Form of Human Behavior". The English Journal U\
421-433.

1928 Review of Roland G. Kent, Language and Philology. The Classical Weekly
21.85-86.

1929 "The Status of Linguistics as a Science". Language 5. 207-214.

1929 "A Study in Phonetic Symbolism". Journal of Experimental Psychology 1 2.
225-239.

1930 Totality. (Linguistic Society of America, Language Monographs 6).

1931 "Communication". Encyclopaedia of the Social Sciences 4. 78-81. New
York: Macmillan.

1931 "The Concept of Phonetic Law as Tested in Primitive Languages by
Leonard Bloomfield". Methods in Social Science: A Case Book (ed. Stuart
A. RICE), 297-306. Chicago: University of Chicago Press.

1931 "Conceptual Categories in Primitive Languages". Science 74. 578.

1931 "Dialect". Encyclopaedia of the Social Sciences 5. 123-126. New York:
Macmillan.

1931 "The Function of an International Auxiliary Language". Psyche 11.4-15.

1931 "Wanted: a World Language". The American Mercury 22. 202-209.

1932 (with Morris SWADESH) The Expression of the Ending- Point Relation in
English, French, and German. (Linguistic Society of America. Language
Monographs 10).

1933 "The Case for a Constructed International Language". Actes du deuxieme
Congres international de Linguistes, Geneve, aout h)M. S6-S8. Paris:
Librairie d'Amerique et d'Orient Adrien Maisonneu\e.

1933 "Language". Encyclopaedia of the Social Sciences 9. 155-169. New ^ork:
Macmillan.

15 Cicncral Linguistics I

1933 "La realiie psychologiquc dcs phonemes". Journal de Psychologie norma-

If et palhologique 30. 247-265.

Sec also \^)4').
1944 -(irading, a Study in Semantics". Philosophy of Science 11. 93-116.
1947 •• I"he Relation of American Indian Linguistics to General Linguistics".

Southwestern Journal of Anthropology 3. 1^.
1 949 "Ilie Psychological Reality of Phonemes". Edward Sapir: Selected Writings

in Language, Culture and Personality (ed. D.G. MANDELBAUM), 46-60.

Berkeley/Los Angeles: University of California Press.

(F.nglish original of 1933. "La realite psychologique des phonemes"]

Preface

Volumes I and II of The Collected Works of Edward Sapir conlaiii Sapirs writ-
ings in the field of general linguisties, general-deseriptive linguislies and histori-
cal linguistics. Volume I includes Sapir's papers in general linguistics (the papers
deal with themes in the history of linguisties and the philosophy of language, with
general issues in the study of language, and with the relationship between linguis-
tics, anthropology, psychology and sociology, or they bear on the foundations of
general and theoretical linguistics).'

Sapir's introduction to linguistics. Language: An Introduction to the Study of
Speech (1921), is included in volume II, together with Sapir's historical-compara-
tive papers (in the field of Indo-European and Semitic), his publications on
African languages, and his progress reports on the project of a new, semantical-
ly-based grammar of English. The wide historical and comparative scope of
Sapir's Language, with its chapters on genetic and areal relationships, on the
historical forces behind the evolution of languages, and on the supposed links
between language, culture and race, justifies its inclusion in volume II, although
Language also, naturally, deals with basic concepts of general linguistics; as a mat-
ter of fact, several of Sapir's writings included in volume I either foreshadow or
elaborate upon issues discussed by Sapir in his 1921 classic.

The papers contained in volume I have been arranged into six sections, rough-
ly corresponding to six chronological stages or sequences. In section I Sapir's
master thesis on Herder's views on the origin of language is reprinted. Section II
contains two early papers (1911, 1912) by Sapir on the historical, cultural and
social setting of languages; readers familiar with Sapir's Language (1921) will
note the continuity between these papers and the later book. Section III, corres-
ponding to a crucial phase in Sapir's intellectual development, includes papers
and reviews of general linguistic interest. This section contains Sapir's short, but
incisive review of Ogden and Richards's The Meaning of Meaning, Sapir's classic

1 For general surveys of Sapir's career and his contribution to general, theoretical, descriptive and historical lin-
guistics and to anthropological linguistics, see: Edwin Ardener, "Edward Sapir 1884^1939", Journal of the
Anthropologiccil Society of Oxford 1 S/1 (1 987), 1-12; Ann E. BertholT, "Sapir and the Two Tasks of Language".
Seniiolica 71 (1988), 1^7; Regna Darnell, Edward Sapir: Lint^iiist. Anthropolo)iisi, Humanist (Berkeley. 1990)
[see Michael Silverstein, "Problems of Sapir's Historiography". Hislorioiiraphia t.iniiuistica 18 ( 1991 ). 181-204];
Regna Darnell - Judith T. Irvine, "Edward Sapir, January 26, 1884 - February 4, 1939". .\ational Academy of
Sciences. Biographical Memoirs (1997), 281-3()(); Maria Xose Fernande/. C'asas, Edward Sapir en la linnuisiica
actual, l.ineas de continuidud en la hisloria de la lingiiistica (Verba. Anexo >A) (Santiago de Conipostela. 2004);
Mikio Hirabayashi. "Studies on the Concepts of Language. Culture, and Personality Expressed in Sapir's
Papers", Bulletin of Daito Bunka University: The Humanities 21 (1983). 4.V52; David J. Sapir. "IntrixJucing
Edward Sapir", Language in Society 14 (198.5), 289-297; Michael Silverstein. "llie Diachrony of Sapir's
Synchronic Linguistic Description", in New Perspectives in Language. Culture, and Personality. Proceedings of
the Edward Sapir Centenary Conference (Ottawa. 1-3 October IW4I. edited by William Cowan. Michael K.
Foster, and Konrad Koerner (Amsterdam/Philadelphia, 198(i),67-l 10; Pierre .Swiggers, "Note sur la linguistiquc
generale en 1921-22. Avec I'edition de deux Icttres de Joseph Vendryes a Edward Sapir". Heitrdge zur
Geschichte der Sprachwissenschaft 1 (1991), I8.'S-192, and "■Svnchri>n\' and -Diachronv' m Sapir's Language
( 192 1 )", Neuphilologische Miiteilungen 94 ( 1 993), 3 1 3-322.

jv; General Linguistics I

p-,pcr on "Sound Patterns in Language," as well as a more traditionally slanted
encyclopedia article on "Philologv."The papers in this section reflect major chan-
ges' and developmenls in Sapir's personal and social life and his mtellectual
career, owinu in part to the interest he took in psychology and psychiatry, the
study of symbolism and social structures; these developments are reflected in the
increasing number o{ publications by Sapir in these domains (see volumes III and
IV of The Colleaed Works of Edward Sapir). Section IV and section V reflect
Sapir's interest, in the second half of the 1920s and first half of the 1930s, in the
problem ol an international auxiliary language, and in the theoretical grounding
ot a Uumuaiic fi>r international communication; Sapir's theoretical involvement
in the International Auxiliary Language Association (lALA) is amply testified to
b> three major papers in the field of "conceptual grammar," which constitute a
solid contribution to general semantics. Section VI, covering the last decade of
Sapir's life, contains another of Sapir's classic papers, viz. his article on the psy-
chological reality of phonemes, three substantial entries ("Communication,"
"Dialect" and "Language") from the Encyclopaedia of the Social Sciences, and a
tew papers of methodological and theoretical interest, in which very often use is
made of American Indian materials (see also the various volumes of The
Collected Works of Edward Sapir devoted to his work on American Indian lan-
guages, and especially some of the texts contained in volumes V and VI).

Almost all of the papers included in volume I were published during Sapir's
litelime; the only exceptions are "Grading, a Study in Semantics" (1944; see sec-
lion V). the English version of "The Psychological Reality of Phonemes" (1949;
see section VI). of which a French translation had been published earlier in 1933,
and "The Relation of American Indian Linguistics to General Linguistics" (1947;
see section VI). Both the French and the English version (first published in the
Selected Writings) of the paper on the psychological reality of phonemes are
reprinted here.

ITie texts have been either reset or anastatically reprinted; whenever necessary
or relevant, editorial corrections or notes have been integrated or added. For a
few papers, an offprint with Sapir's handwritten corrections was available, and
was graciously put at my disposal by Philip Sapir; in such cases, Edward Sapir's
corrections are explicitly mentioned. In all cases, except one (viz. "The Function
of an International Auxiliary Language" [1931]; see section IV), we had access to
the original publication. The page numbering of the original publication or -in
the case of the just mentioned paper- of the version taken as the basis for the
reprint, has been maintained. For all papers, information has been provided con-
cerning the original publication and possible later reprints. The page references
in the introductory texts are always to the page numbering of the original pub-
lication, and not to the new, continuous pagination in this volume.

The editor gratefully acknowledges the help and encouragements of Philip and
Midge Sapir. and of J. David Sapir, during various stages in the preparation of this
volume.

Pierre Swiggers

Introduction

Sapir's Life and Work: Two Appraisals
(Zellig S. Harris; Stanley Newman)

Introductory Note

Wc found it appropriate to have this et)llection of Sapir's wrilinus m eeneral
hnguisties preeeded by extracts from two important book reviews of the Stlccit'd
Writings, by two major linguists from the first generation alter Sapir. Ihe latter
collection. pubMshed by David G. Mandelbaum in 1949 (and several times reprint-
ed since then), remains — in spite of some regrettable omissions — an extremely
useful anthology of Sapir's major articles in linguistics, in anthropology, and in
cultural and personal psychology. The Selected Writings received reviews in the
leading linguistic and anthropological journals. Among these reviews two stand
out for their penetration and insight into Sapir's approach to language, culture,
society, and personality. One was written by Sapir's student Stanley Newman, the
other by Zellig Harris, who was neither a student of Sapir nor of Bloomfield. but
whose work was inspired by both these men. Whereas Harris's review article
emphasizes the continuity and the uniformity between Sapir's study of language
and his approach to culture, to society, and to the individual, Newman notes spe-
cific contributions, and pays much attention to Sapir's style and temperament. In
both reviews, however, the continuity and homogeneity of Sapirs wide-ranging
approach are highlighted, and the two reviewers concur in idcnUfying form-pat-
terning as the clue towards a just understanding of Sapir's perspective and inten-
tion, whatever the object of study — language, culture, society, behaviour.

Philip Sapir - Pierre Swiggers

Introduclion lo Zclliu llairis's text

Zcllig Harris (19(W-1W2)' published his review of the SrhrifJ Wniint^s of
Edward Sapir al a lime when he was already developing his iransfornuilionai
approach to language, having left behind him the ideas presented in his j Methods
in] Structural Lini^uistics.' Although in his review Harris does not use the term
"transformation." there is at least one passage' containing an allusion lo uhat
could be called transformational operations.

The entire text of Harris's review article merits rereading, bui it was decided
to publish only part of it, for two reasons:

(a) .some passages refer directly to the specific occasion t)f republication of
Sapir's texts in the Selected Wriiini^s, and would have been out of place here;

(b) given that this volume, and volume II. deal with Sapir's general linguistics,
and that his anthropological and culture-psychological writings have for the lar-
ger part been published in other volumes of Uw Collected VVV;rA:.v of i.dwurd
Sapir* it was felt that only the first (also the largest) section of Harris's review
article, which deals with the linguistic articles in the Selected Writings, should be
reprinted here.

Tlie bottom-line of Harris's review article is that Sapirs approach and
methods were uniform, whether he dealt with language, culture or jursonality —
although Sapir himself stressed the difference in time-span, evolutionary rhythm,
and intrinsic content of these three objects.' As to the methods and working pro-
cedures of "the whole Sapir" Harris identifies three characteristic features:
(i) Sapir's capacity of extracting results from elusive data. i.e. his capacity o\. and
intuition of. structural depth;

' For information on Zellig S. Harris's career, sec the obituaries in ( iilijornui l.in\;iiiMu \t-i\sU'iitr 2.^:2 (1992).
bO-bA (by liriicc E. Nevin). S'ewslelier ofilw Socieiv for the Study of the liuli^^fnous luinfiuaiifs of the Amencas
1 1:2 ( 1992). .^-4 (Victor Ciolla. with help from Dell Hymes and Bruce Nevin). (}rh>is ?>> (1992). .Mh-.VS3 ( Pierre
Swiggers).an(J lAini;iiiii;c 75 ( 1999). 112-119 (Peter H. Matthews). For a comprehensive bibliography of his writ-
ings, compiled by Konrad Koerner. see llislorioi>rnphui l.ini;iiistic(i 20 ( 1993). 509 522

- Zcllig .S. Harris, Mcihocis m Sirmtiinil /,/;(t;«;,\7/V.v (Chicago, 1951 ): the book was later reprinted under the title
Siritciiiral l.ini;iiistics.

' llie passage in question is the following: "Hut the |iossibility of including the results (output) of one relatio-
nal statement into the terms of another, by means of successive definitions, makes it possible for mathcmalic.il
statements to carry a far greater communication load than linguistic statements on the same subjects"

•* See The Collected Works of Edward Sapir. volumes 1 1 1 .uul I \'. On Sapir's psychology of culture(s). see Judith
T. Irvine (ed). Edward Sapir: The Psycholofiv of i ultiire. A Loiirse of Lectures. Rectmstrucled and eJilcJ hy
J. I. Irvine (Fierlin/New York. 1994): l.jiljana Hibovic. 'F^dward Sapir's Concept «>f Culture and its Presenl-day
Implications". International Review of Slavic l.ini'iiistics 2 (1977), 125-I.V5: Michael Silverstem. "Sapir's
Psychological and Psychiatric Perspectives on Culture ". California l.infiuislic Soies 21 ( 1992). .^Sl 4()h.

^ Sec Edward Sapir, "ITie History and Varieties of Human Speech", and "l.iinguage and linvironmcnt" (both
reprinted in section II of this volume), and especially his biMik l.ani^iiai^e (New York, 1921 ) | reprinted in vol-
ume II of The Collected Works of Edward Sapir\. pp. 229-235.

24 General Linguistics I

(ii) the ••dramatic wa\" in which Sapir's conclusions followed from the data:
Harris points here to Sapir's sense of holistic perspective, combined with remar-
kable argumentative skill;

(iii) Sapor's "sensitivity and critical independence," which appears perhaps most
clearly in his treatment of modern society and the modern individual. Here lies
Sapir's capacity of unraveling presuppositions, tacit convictions, and unfounded
beliefs, and of making his reader conscious of the need for authentic reflection
and responsible behaviour.

In discussing Sapir's methods of work in Hnguistics, Harris highlights the follow-
ing points:

( 1) Sapir's overarching interest was in discovering the structure of language;

(2) structure in language is, from the viewpoint of language itself, the result of
processes (in fact, linguistic entities are the result of processes of change); this
explains the -process-like" nature of Sapir's statements;'

(3) the structure of (a) language is, from the viewpoint of the linguist, the result
of the structuring of "structure in language" by the linguist, who characterizes
relationships between elements and processes in specific ways;

(4) apart from being characterized by a process-oriented approach, Sapir's work
is characterized by the recognition of patterning in language; the greatness of
Sapir's work lies in the establishing of "total" patterns, and in showing the inter-
play between organized structures at various language levels;

(5) the combination of process and pattern allowed Sapir to move constantly
from form to function, and from structure to history: much of his work is both
syn- and diachronic, and his linguistic analysis is never confined to pure forms,
but always starts out from forms and their use(s).

Harris rightly notes that Sapir's concept of patterning made possible the
distinction between grammar and grammaticalness, and grounded his interest in
language as (formal) completeness, or unlimited "constructivity."

Sapir's "functional" conception of form followed from his approach to lan-
guage as a form of behaviour, defined by its use as a symbolic system of reference.
This system of reference is constituted by content-units and by form-units, as well
as by syntactic relationships and contextual insertion. Both units and relation-
ships are dynamic concepts for Sapir: in his analysis of word meanings, he showed
the capacities of meaning, and their exploitation in use.^ The structuring (or, if

^ Harris's reader is supposed to be familiar with the distinction between "item-and-arrangement" and "item-
and-process" models (the item-and-arrangement model is associated with a strictly Bloomfieldian approach);
the classic statement on this methodological issue is the article by Charles F. Hockett, "Two Models of
Grammatical Description", Word 10 (1954), 210-234.

Harris discusses some of the factors of meaning: absolute vs. relative comparison, graduality, directionality,
(ir)reversibilitv.

Sapir's Life and Work: Two Appraisals 25

one may venture the term. struetura(li/;i)ti()n ) ol languaue and slruetiiie in lan-
guage are complementary here: "1 he lormal analysis ot language is an empirical
discovery of the same kinds ol relations and eomhinalions which are devised in
logic and mathematics; and their empirical discovery in language is of value
because languages contain (or suggest) more complicated types of combinatit)n
than people have invented for logic" (Z.S. Harris, p. 301 ). Moreover, the linguist
does not operate in isolation from the speakers of the language.' but makes use
of the speakers' behaviour as a heuristic tool.

In his review article Harris points to the perfect continuity in Sapir's linguistic
interests, ranging from language description to reflections on semantic structure
(which is never approached in an a priori way), and to the interest in (the prin-
ciples and conditions of) the construction of an artificial language (one that is to
be effectively used, and thus correlates with a "world view").

In the last paragraph (of the section on language reprinted here), Harris deals
with Sapir's diachronic work — which became increasingly important in the
1930s. Here also, patterning provides the key towards a deep and true under-
standing of how a particular structure came about, and underlying the patterning
of (sets of) forms, there are unconscious macro-processes — for w hich Sapir aptly
used a term from psychology, viz. drift" — , processes which eventually ha\ e to be
explained by larger configurations in and from a distant past.

Pierre Swiggers

** As Harris puts it; "The decision ol what to inckidc in the linguistic structure rests with the linguist, who has to
work out that structure, and is simply a matter of what can be fitted into a structure of the linguistic type. The
question of what activities constitute what kind of communication is largely an independent one. and is answer-
ed by observing the kind of use people make of the various communicational and expressive activities" (p. 30.^).

'' See his book I .uni^imi^c ( 1^21 ). pp. 160-163, I65-16S. 174-IK2. IS.Vl V. On Sapirs eimcepl of "drift." see Dell
Hymes - John Vou^hX. Amcricun Siruciunilism (The Hague/Paris, IWI; originally published in Current Trends
in Linf-ui.siiis vol. 13. part 2 |1^>75|). pp. 232-233; Yakov Malkiel. "Drift. Slope, and Slant: background of. and
Variations upon, a Sapirian llieme". Lim^ua^e 67 ( 1 Wl ). 53.>-.s7(); Ciiovanna Marotl.i. 'Sulla no/ione di deri-
va" in Sapir". Qiuulcrm (IcH'Isiiiiilo di Lin,i;iiistiai delVVnivcrsitd di Vrhmo 4 (IMSh). .sM-^l; Michael Shapiro.
"Sapir's Concept of Drift in Semiotic Perspective". Scmioiiai 67 (1^87). 159-171; Henning Ander>en. " llie
Structure of Drill", in H. Andersen - E.F.K. Koerner (eds.). Hi.sioricid l.inf;ui.slics /VcS'7 (Amsterdam/
Philadelphia, IWO). 1-20; Michael Silverstein."'I'he Diachrony of Sapir's Synchronic 1 inguislic Description", in
New Perspectives in I .cuiiiiuit'e. Cidlure. and Persomditw I'roeeedinf^s of the luiiuird Sapir Centenary Conference
(Ottawa, /-? October IW4). edited by William Cowan, Michael K. foster, and Konrad KiK-rncr
(Amsterdam/Philadelphia, 1986), 67-1 10; Pierre Swiggers, "Svnchronv' ami Diachrony' in Sapir's I ans^uafie
(1921 )". Neiiphiloloiiische Mitteiliin^en 94 ( I9i)3). 313-322.

Zellig S. HARRIS
<Sapir's approach to languago*

[...] The writings of Edward Sapir are invaluable for their complete grasp
of linguistics, for their approach to language and culture and personality.
for the wonderful working of data which they exhibit. We all know what a
never-ending source of learning and delight this was to Sapir 's students and
friends. [...]

<1 > Descripttve linguistics: process; analysis in depth. Sapir puts the es-
sential statements of modem linguistics in postulational or definitional form:
'Not only are all languages phonetic in character; they are also phonemic'; and
morphemes are 'conventional groupings of such phonemes' (8-9).^ But by the
side of this, we find his characteristic approach in depth. Phonemes are presented
not as a classification of phonetic events or tj-pes, but as the result of a process
of selection: 'Between the articulation of the voice into the phonetic sequence
... and the compHcated patterning of phonetic sequences into ... words, phrases,
and sentences there is a verj-- interesting process of phonetic selection and gener-
alization.' And concerning the phonemic constituency of morphemes we find:
'the limiting conditions [of morphemes] may be said to constitute the phonemic
mechanics, or phonologj', of a particular language.' The term 'limiting condi-
tions' aptly relates the range of morpheme construction to the range of phoneme
combination.

Sapir thus sees the elements of linguistics and the relations among them as
being the results of processes in language. The descriptive structure of a lan-
guage can, of course, be regarded as the result of many processes of change, as
de Saussure pointed out in his example of the cross section of a tree-trunk in
relation to the growth and vertical axis of the tree.^ This kind of interest appears
in Sapir's Glottalized Continuants, and -^ill be discussed below.

Process or Distribution. Sapir, however, also used this model of an 'entity as

' Page numbers refer to <the Selected Writings. 1949>. without specifying the particular article involved.
- Ferdinand de Saussure. Cours de linguistique generate 125.

Sapir's Life and Work: Two Appraisals 21

290

a result of process* within descriptive linguistics proper. Consider, for example,
those environmental ranges by virtue of which two sound types never contract:
say the fact that in a certain language no morpheme contains two vowcLs in
succession; and that in any word which contains one morpheme ending in a
vowel, followed by a second morpheme beginning with a vowel, a glottal stop
is pronounced between these two vowels. When we speak in terms of distribu-
tion and classification, we would say that no morpheme contains the VV se-
quence, and that all morphemes which end in V before consonant or juncture
have alternants ending in V? before vowel (before any following morpheme which
begins with a vowel). Hence the VV sequence never occurs across morpheme
junction, just as it doesn't occur within a morpheme. In contrast with this,
Sapir would say that no two vowels could come together (within a morpheme),
and that when a particular morpheme conjunction would have the effect of
bringing two vowels together a glottal stop comes in as a protective mechanism
to keep them apart. This kind of model appears in much of Sapir's grammatical
work and in the work of some of his students, as for example in Newman's
handsome analysis of Yokuts.'

We can consider this simply as a method of description, an alternative to our
present formulations, which we make in terms of the classifying of occurrences.
The process model has the advantage of being more dramatic, and often of re-
flecting the actual historical changes (the inter-morphemic glottal stop may well
have been a later development).* It has the greater advantage of opening the
way to a more subtle descriptive analysis — something always dear to Sapir's
heart — by giving a special secondary status to some parts of the descriptive
structure. For example, we may be missing something when we say innocently
that VV does not occur across morpheme boundary (while V?V and VCV do):
the V?V which we find there may not be fully equivalent to the VCV which
result from morphemes ending in -VC plus morphemes beginning in V- (or
from -V plus CV-) ; for one thing, these VCV alternate with -VC and V- when
their morphemes occur separately, whereas the V?V alternate with -V and V-;
for another, the frequency of V?V (differently from VCV) may be much greater
in those positions where morpheme boundaries can occur than in other posi-
tions.^ On the other hand, the process model has the disadvantage of bringing
into descriptive analysis a new dimension — the relations of one distribution to
another distribution — which does not fit well into the algebraic character of
the present bald statements of distribution. There is need for further elaboration

'Stanley S. Newman, Yokuts language of California (New York, 1944).

*Cf. Sapir's article on glottalized continuants (225-50), and Henry M. Hoenigsw&ld,
Sound change and linguistic structure, Lg. 22.138-43 (1946).

'To make this more explicit: Suppose all word-initial morphemes h.avc two or more
syllables (vowels). Then the probability of finding ? rather than some other consonant after
the FIRST vowel of a word is related simply to the frequency of the medial glottal stop. The
probability of finding ? after the second vowel is related to the frequency of the glottal
Btop (medial and at the end of morphemes) plus the frequency of morpheme* which end
with a vowel (and of morphemes which begin with a vowel). However, the probability of
finding other consonants (not ?) after the second vowel is related merely to the frequency
of those consonants medially and at morpheme-end.

28 General Linguistics I

391

of descriptive techniques, in order to make room for such refinements among
our direct distributional statements.

The Process and its Result. We can also consider the use of the process model
as an activity of the linguists who use it; and we can then say that aside from
such personality reasons as may have dictated Sapir's use of it, it also occupies
a determinate position from the point of view of the history of science. It seems
to constitute a stage in the separation of descriptive method both from historical
analysis and from the older psychologizing of grammatical forms. The older
grammars did not distinguish descriptive from historical statements, so that
the history of the glottal stop at word boundary would have been combined
with the statement of the absence of vowel sequences there. The older grammars
assigned reasons for speech forms: people said V?V (with 'intrusive glottal stop')
in order to avoid W which they did not otherwise pronounce.^ Finally, the
older grammars frequently failed to distinguish morphological from phonological
considerations, so that the morphophonemic fact about V?V appearing for
-V + V- would be given together with the phonemic fact about the absence
of W. The formulations in terms of process give expression to all this while
at the same time separating descriptive linguistics from the rest. This is achieved
by the dual character of these formulations: the 'process' of protecting the cross-
boundary -V + V- yields the 'result' that V?V occurs.

The process section of this formulation takes cognizance of such factors as
were brought out by the older linguistics (or by Sapir's interest in descriptive
detail); the result section gives the distributional statement as an item in a
separate science of distributions.'^

Process in Language Structure. The process model led to a characterization of
linguistic structures in terms of the types of process involved in them. A grammar
was viewed as consisting of so much prefixation and sufl&xation, so much internal
change or reduplication, used at such and such points.* Much of what was
called process concerned the changes in or near a given form as its environment
varied. For example, there is an internal change in knife (to knive-) when -s
'plural' appears in its environment. There is another internal change in sing
(to sang) which can occur without any change in environment: You sing well
~ You sang well. (But if we vary the environment to / like to ( ), we exclude
sang and find only / like to sing.) There is a process of suflBxation that adds -ed

• How different Sapir's psychologism is from this will be discussed in Part 3 below. For
the moment, it is worth noting that Sapir's grammatical formulations stayed within lin-
guistic categories. In descriptive linguistics he would not say that people inserted a glottal
stop so as to avoid the sequence W, but that the glottal stop constituted, in respect of
medial VV, a 'protection' (in cross-boundary position) of that non-occurrence of VV. The
primacy of medial W over the cross-boundary case is maintained, but in terms of the
structure rather than in terms of people's intervention in their own speech behavior.

^ We can say that the use of base forms in morphophonemics — as in Leonard Bloom-
field's Menomini morphophonemics, TCLP 8.105-15 (1939)— is a further step from history
or process toward purely distributional statements.

' It is interesting that Bloomfield's work, which (as suggested above) represents a later
stage m this particular development, presents phonemes no longer as the result of process
but as direct classification, whereas the morphology is still largely described in terms of
process. Cf . the chapters on phonology and on morphology in his book Language.

Sapir's Life ami \\'i)rk: fwo Appraisals 29

292

to many English words without any accompanying chanp;e in environment, or
when the environment is changed to include yesterday, but never directly after
will or to: I walk, I walked, I walked yesterday, I will walk, I want to walk. Today
we would say that knife and knive are alternants of one morpheme, and tliat
the internal change there is a morphophonemic alternant of zero (other mor-
phemes, like spoon, have no change before -s). We would say that sang con8iHt«
of sing plus some other perfectly respectable morpheme, and that this other
morpheme (change of /i/ to /ae/) is an alternant of the morpheme -ed.

To speak only of the presence of internal change, suffixation, reduplication
in a language is to tell merely what is the phonemic history of a morpheme and
its neighborhood, as the morpheme is tracked through its various environments.
To speak only of the fact that some nouns have alternant forms before -« (or
that some nouns before -s are complementary to other nouns not before -«),
and that -ed has various alternant forms, is to give bare distributional state-
ments with the merest nod to the phonemic composition of the morphemes.

To speak of internal change and suffixation and the like a.s occurring under
particular environmental conditions is to give a detailed distributional statement
of morphemes as phonemic groupings. This last can be described as a combining
of today's distributional interests with the interest in process of Sapir (and,
in morphology, Bloomfield) and various European linguists; it is a direction of
development which would be fruitful in the present stage of linguistics. It would
be fruitful because linguistics has at present one technique for stating the rela-
tion of phoneme to morpheme (morphemes are arbitrary combinations of pho-
nemes) and another for stating the general relation of morpheme to utterance
(utterances are composed of stated distributions of morphemes). To take greater
cognizance of the phonemic composition of morphemes is to come nearer to the
direct relation of phoneme to utterance (utterances are composed of stated dis-
tributions of phonemes). This goal will presumably never be reached, because
there will always be arbitrary elements in the phonemic composition of mor-
phemes. But if we can make general statements about part of this field, as by
noting when the morphemes or alternants consist of added new phonemes or
of repeated phonemes or of exchanged phonemes, we leave lei>s that is arbitrary
and outside our generalized statements.

<2.> Linguistic structure: Pattern. Sapir's greatest contribution to lin-
guistics, and the feature most characteristic of his linguistic work, was not the
process model but the patterning of data. Both of these analytic approaches
were of course used by many linguists beside Sapir, but Sapir made major con-
tributions to both lines of development. For patterning we have, first of all,
his famous Sound patterns in language (1925). (.1

Here he pointed out that what is linguistically significant i.s not what
sounds are observed in a given language but under what linguistic circum.stanccs
(i.e. in what distribution) those sounds occur. The phrivseology of course is
pre-phonemic, but (or since) the article is one of the cornerstones of phonemic
analysis.

Sapir's search for patterns pervaded not only his phonemic but also hLs mor-
phological work, as anyone would know who saw him working over his largp

30 General Linf^nistics /

293

charts of Navaho verb forms. His morphological patterning may be seen in
his analysis of paradigms in his book Language (Ch. 5), and in his Navaho work,
and in his published and unpublished American Indian material. His phonemic
patterning is amply evident in the articles reprinted in this volume.

Since the original appearance of his articles, patterning has become an every-
day matter for linguistics. Phonemic analysis seems quite obvious today. Mor-
phological analysis is more procedural now than in Sapir's book Language (1921).
Some of the earliest organized work in morphophonemic patterning was carried
out by Sapir' or under his influence.'"

Today the distinction between phonemic and morphophonemic patterns is
quite prominent. In La r^alit^ psychologique des phonemes (1933; English ver-
sion printed here on pp. 46-60), Sapir includes both kinds without explicit
distinction. Phonemic examples (from native responses) are: writing /hi/ in
Nootka for phonetic he, e being the allophone of i after h (54) ; reconstructing
the Southern Paiute allophone p when post-vocalic -/3a' 'at' was experimentally
pronounced after pause (49; initial p and post-vocalic /3 are positional variants
of each other) ; writing [p'] with prior release of oral closure and ['m] with prior
release of glottal closure equivalently as /f)/ and /m/, because the distributional
features of [p'] and ['m] are equivalent (56-7; both occur at syllable beginning
where clusters do not occur, neither occurs at syllable end where other types
of consonants occur, plus a morphophonemic equivalence). Morphophonemic
examples (from native responses) are: recognition of the difference between the
phonemically identical Sarcee /dini/ 'this one' and /dinl/ 'it makes a sound'
based on the form of the stem before suffixes, e.g. /-!/ 'the one who', where we
find/dind-V, /dinit'i/, morphophonemic stem nW (52-3); writing Nootka mor-
phemic s-s (with morpheme boundary between them) as morphophonemic ss,
and phonetic [Vs-V] as containing phonemic /s/ — [s-] being the allophone of
/s/ after short vowel and before vowel — even though this ss is phonemically
/s/: in the morphophonemic writing tsi' qHClassatlni 'we went there only to
speak' (containing 'as 'to go in order to' and sa 'only') the ss is phonemically
identical (and phonetically equivalent) with the /s/ of /tlasatl/ 'the stick that
takes an upright position on the beach' — phonetically [tlas-atl] and with mor-
phemic boundary tlasatl (54-5).

Language Classification. The variegated kinds of patterning, once recognized,
invited attempts at some kind of organization. To organize the patterns of each
language into a total structure of that language, and to investigate and compare
the kinds of structuralization, was not possible until much more work had been
done around these patterns. What was done instead by Sapir and others was to
classify patterns (case system etc.) and to classify language types on this basis.
To a large extent this was what Sapir did in his famous classification of (North)
American Indian languages into six major groups (169-78). It is clear from the
considerations explicitly presented by Sapir in this article (and also from the
difficulty of conceiving any discoverable genetic relation among some of the

• In Sapir and Swadesh, Nootka texts 236-9 (Philadelphia, 1939).

"» As in Morris Swadesh and C. F. Voegelin, A problem in phonological alternation, Lg.
15.7 (1939; written some years earlier).

Scipir's Life and Work: Iwo Apprmsals 31

294

families, for example in the 'Hokan-Siouan' group) that this classification \r
structural rather than genetic, though in many cases it suggests possible genetic
connections that can be supported by further research.

Sapir also proposed a general method of classifying languages on the basis of
types of grammatical patterning (in his book Language), but neither he nor others
followed it up. For since there was no organizing principle for all patterninga,
such as would arise out of an analysis of the full possibilities of linguistic pat-
terning and of their structural interrelations, the classification work was a uweful
but temporary way of noting what formal features occur in languages, and
which of them occur together. The classification results could not in them.selvcs
be used for any further work, except to suggest distant genetic relationships as
in the American Indian classification. (In contrast, if a fully organized — though
not necessarily one-dimensional — classification of complete language structures
is ever achieved, the results would be useful for understanding the development
of linguistic systems, for discovering the limitations and further possibilities
of language-like systems, etc.) The piling up of research in distribution and its
patternings has made it possible by now to talk about the place of one pattern
relative to others, and about the way these fit into a whole structure. With more
work of this type we may be able to say wherein and to what extent two lan-
guages differ from each other, and thus approach a structural classificatory
principle.

Descriptive Function. This structural Umitation did not affect the general lin-
guistic approach that was made possible by recognition of patterning. Sapir's
patterning is an observable (distributional) fact which he can discover in his
data and from which he can draw those methodological and psychological con-
siderations which he cannot observe directly, such as function and relevance,
or perception and individual participation. He can the more readily do this
because his patterning is established not directly on distributional classification
but on an analysis in depth of the way in which the various elements are used
in the language. The 'way the elements are used' is equivalent to their di.stri-
bution; but talking about such use gives a depth which is lacking in direct
classification of environments.

Thus Sapir uses the patterning of elements in order to express their function
(their functional position within the language): 'to say that a given phoneme
is not sufficiently defined in articulatory acoustic terms but needs to be fitted
into the total system of sound relations peculiar to the language is, at bottom,
no more mysterious than to say that a club is not defined for us when it is said
to be made of wood and to have such and such a shape and such and such di-
mensions. We must understand why a roughly similar object, not so different
to the eye, is no club at all ... To the naive speaker and hearer, sounds (i.e.
phonemes)'^ do not differ as five-inch or six-inch entities differ, but as clubs
and poles differ. If the phonetician discovers in the flow of actual speech some-
thing that is neither "club" nor "pole", he, as phonetician, has the right to
set up a "halfway between club and pole" entity. Functionally, however, such

"Sapir means: sounds as phonemically heard (perceived, atructured) by the n»ivr
speaker and hearer.

32 General Linguistics I

295

an entity is a fiction, and the naive speaker or hearer is not only driven by its
relational behavior to classify it as a "club" or a "pole," but actually hears and
feels it as such' (46-7)."

Perception. In a related way, patterning is used as a basis for the structuring
of perception. Sapir reports that English-speaking students often mistakenly
hear p, t, or k instead of a final glottal stop; and after learning to recognize a
glottal stop, they often mistakenly hear a glottal stop at the end of words ending
in an accented short vowel (they write sme' for srru). He then points out (59-60)
that the second type of error is simply a more sophisticated form of the first.
Since words ending in accented short vowel do not occur in English, the students
who fail to recognize the glottal stop in sme' cannot perceive the words as sme
(since such words are out of their pattern) and therefore (selecting a consonant
nearest ') hear it as smtk or the hke. Later, when they know about glottal stops
and hear sme, they can still perceive only a word ending in a consonant and
(selecting a consonant nearest zero) hear it as sme'.

This effect upon perception is claimed not only for such phonemic hearing,
but also for the structuring of experience in terms of the morphological and
vocabulary patterns of the language: 'Even comparatively simple acts of per-
ception are very much more at the mercy of the social [more exactly : linguistic]
patterns called words than we might suppose. If one draws some dozen lines,
for instance, of different shapes, one perceives them as divisible into such cate-
gories as "straight," "crooked," "curved," "zigzag" because of the classificatory
suggest iveness of the linguistic terms themselves' (162).

System. Sapir goes on to recognize patterning as one of the basic characteristics
of language: 'Of all forms of culture, it seems that language is that one which
develops fundamental patterns with relatively the most complete detachment
from other types of cultural patterning' (164). Had he used the descriptive
word 'consists of instead of the process word 'develops', he might have gone
beyond this to add that we can even use this linguistic patterning to determine
what is to be included in 'language'. There are scattered bits of speech-Hke
noises — coughing, crying, shrieking, laughing, clucking — which may or may not
be considered part of 'language' on one basis or another, but which we count out
of language because they do not fit into its detached patterning.

Out of all this Sapir was able to make important generalizations about lan-
guage as a system. Recognition of the detachment of Unguistic patterning leads
to the statement that 'the patterning of language is to a very appreciable extent
self-contained and not significantly at the mercy of intercrossing patterns of
a non-linguistic type' (165). This explicit talk about the fact of patterning
makes possible the distinction between the grammar (specific pattern) and gram-
maticalness (degree of patterning) of language: 'In spite of endless differences
of detail, it may justly be said that all grammars have the same degree of fixity
One language may be more complex or difficult grammatically than another,

" Note 'relational behavior' for our 'distribution'. The hearer might also classify it as a
'bad pole', so that even if the difference between the halfway sound and the regular sounds
is noticed and not lost, it is nevertheless referred to (i.e. structured in terms of) the fxmc-
tionally (distributionally) determined points of the pattern.

Sapir's Life and Work: I wo Appruisuls 33

296

but there is no meaning whatever in the statement which Ls sometimes
made that one language is more grammatical, or form bound, than another'
(9-10).

From this, Sapir could go on to an interesting formulation of the adequacy
of language. We all know the statement that any language can be used as the
vehicle for expressing anything. Sapir removes the air of triviality from this by
saying, 'New cultural experiences frequently make it necessary to enlarge the
resources of a language, but such enlargement is never an arbitrary addition
to the materials and forms already present; it is merely a further application
of principles already in use and in many cases little more than a melaijhorical
extension of old terms and meanings' (10). In other words, the adequacy of
language is not simply definitional, but derives from the possibilities of exten.sion
and transference within the language structure, without either disregarding or
destroying the structure. 'The outstanding fact about any language is its formal
completeness ... No matter what any speaker of it may desire to communicate,
the language is prepared to do his work ... Formal completeness has nothing to
do with the richness or the poverty of the vocabulary ... The unsophisticated
natives, having no occasion to speculate on the nature of causation, have proba-
bly no word that adequately translates our philosophic term "causation," but
this shortcoming is purely and simply a matter of vocabulary and of no interest
whatever from the standpoint of linguistic form ... As a matter of fact, the
causative relation ... is expressed only fragmentarily in our modern European
languages ... [but] in Nootka ... there is no verb or verb form which has not
its precise causative counterpart' (153-5). Sapir might have continued here to
point out that the work of language in communication and expression can be
carried out both by grammatical form and by vocabulary (though with dif-
ferent effect), since one can insert to cause to before any English verb somewhat
as one can add a causative element to every Nootka verb." Hence what is im-
portant is not so much the distinction between grammatical form and vocabulary,
as the fact that the distribution of grammatical elements, and so the grammatical
structure, can change in a continuous deformation (the structure at any one
moment being virtually identical Avith the immediately preceding structure),
and that vocabulary can be added without limit (and changed in meaning).
What we have, therefore, as the basic adequacy of language is not so much the
static completeness of its formal structure, but rather its completability, or
more exactly its constructivity without limit.

<3 > Language as social activity. The fact of patterning. \ person who is
interested in the various kinds and relations of patternings, for their own sake,
can establish pattern and structure as bland distributional arrangement. m, and
thence move toward the mathematical investigation of the combinatorial jKia-
sibihties. Sapir, however, was interested in the fact of patterning, and what

" We omit here the important difference that an English verb by itaolf contraata mo«l

immediately with the small class of affix combinations (e.g. verb plus rd), and only aec-
ondarily with a vast class of phrasal sequences in which that verb could U act (of which to
cause to do so-and-so is one), while a Nootka verb by itself contraats with a few upocjfic
combinations of verb plus affix (of which the causative affix is one), and only 8econd*nly
with the large class of phrasal sequences.

34 General Linguistics I

297

could be derived from the discovery that language was so patterned a bit of
human behavior. This was not only because Sapir was above all an anthropolo-
gist, but also because of the particular development in linguistic science at the
time.

From de Saussure to the Prague Circle and Sapir and Bloomfield, the fact
of patterning was the overshadowing interest. In the later work of this period
in linguistics we find attempts to analyze and classify these patterns, but the
big result was still the very existence of structure. This was the big advance in
several sciences at the time. In the late depression years, when neither admiration
of Russia nor war preparations in America had as yet obscured the scientific
and social results of Karl Marx, Leonard Bloomfield remarked to me that in
studying Das Kapital he was impressed above all with the similarity between
Marx's treatment of social behavior and that of linguistics. In both cases, he
said, the activities which people were carrying out in terms of their own life
situations (but in those ways which were socially available) turned out to con-
stitute tight patterns that could be described independently of what people
were about. In language, they communicate, or pronounce words they have
heard, but with the descriptive result of maintaining a patterned contrast be-
tween various subclasses of verbs or the like. In economic behavior, they may
do various things just in order to make profit, but with the descriptive result
that the producing population becomes increasingly removed from control over
its production. Sapir saw this fact of patterning even more clearly — in language,
in culture, and later in personality. Throughout his writings one sees how im-
pressed he was with this fact, one which was also being stressed at the time
(but with less happy success) in other social sciences. In his comments about
language as patterned behavior he reached the heights of his subtlety, and
pioneered a form of research which few have as yet taken up.

Talking as pari of behavior. About the very act of talking he says: 'While
it may be looked upon as a symbolic system which reports or refers to or other-
wise substitutes for direct experience, it does not as a matter of actual behavior
stand apart from or run parallel to direct experience but completely interpene-
trates with it ... It is this constant interplay between language and experience
which removes language from the cold status of such purely and simply symbolic
systems as mathematical symbolism or flag signaling ... It is because it is learned
early and piecemeal, in constant association with the color and the requirements
of actual contexts, that language, in spite of its quasi-mathematical form, is
rarely a purely referential organization' (11-2). This understanding of the rela-
tion of language to other experience is involved also in the view that psycho-
logical suggestion (and, in extreme form, hypnotism) is in essence the same as
talking. In The psychology of human conflict (174), E. <R.> Guthrie says: 'Sug-
gestibility is the result of learning a language. When we acquire any language,
such acquisition lies in associating the sounds of the language with action. The
use of suggestion is merely the use of these acquired cues ... There is no essential
difference between causing a man to perform some act by suggestion and causing
him to perform that act by request.' Arthur Jenness amplifies:'^ 'In the past,

" Hypnotism 496 (where the Guthrie quotation is given in full) = Chap. 15 of J. McV.
Hunt (ed.), Personality and the behavior disorders, Vol. 1.

Sapir's life and Work: I'wo Appraisuls 35

298

the subject has been drowsy when the word "drowsy" has been spoken, and the
state of drowsiness has thereby become conditioned to the word "drowHy". The
word "drowsy" repeated later under the proper circumstances tends to elicit
drowsiness.'

Sapir's point has the merit that instead of referring language back to an un-
defined and dangerously over-used 'symbolism', he presents it a-s a direct item
of behavior, associated with other behavior: 'If language is in its analyjiod form
a symbolic system of reference, it is far from being merely that if we con-sider
the psychologic part that it plays in continuous behavior' (12).'* In order to
treat of the 'symbolic' character of language, he says that symbols 'begin with
situations in which a sign^^ is disassociated from its context' (566); and he adds,
'Even comparatively simple forms of behavior are far less directly functional
than they seem to be, but include in their motivation unconscious and even
unacknowledged impulses, for which the behavior must be looked upon as a
symbol' (566-7). Language, then, is just an extreme type (and a physiologically
and structurally separable portion) of the associations and dissociations that
occur in all behavior.

Sapir goes on to distinguish two characteristics (and origins, and types) of
symbols: the 'substitute for some more closely intermediating type of behavior',
and the 'condensation of energy' (565-6). His first or 'referential' symbolism,
like telegraphic ticking, is the one we all know in science and technology;" hia
second, like the washing ritual of an obsessive, is that which occurs in psycho-
analysis. In ordinary behavior, and even in language, both are blended.'*

Forms and meanings. Sapir's interest in language as patterned behavior, in
some respects continuous (associated) with other behavior and in some respects

" This was published in 1933. The novelty of this view may be seen from the fact that in
1929 Sapir had given it a more traditional formulation: 'If I shove open a door in order to
enter a house, the significance of the act lies precisely in its allowing me to make an easy
entry. But if I "knock at the door," a little reflection shows that the knock itself does not
open the door for me. It serves merely as a sign that somebody is to come to open it for me'
(163-4). His later understanding would suggest that the knock can be viewed instead aa a
tool, an indirect step in the course of getting the door opened (like the stick with which
Kohler's ape knocks down the banana, or the lever with which we pry up a rock). It is p.irt
of the continuous behavior which makes the person inside unlock the door for us, or which
makes him ready for our intrusion. It is not a 'substitute for shoving' but rather the equiva-
lent for shoving in a society where people are customarily apprised of a visitor's arrival.
In social situations where this is not customary (as among intimates), one indeed opens the
door without knocking.

'• For 'sign' we should say: any associated behavior, such as a noi.se.

" Note Martin Joos's statement of it in the last paragraph of his pajK-r Description of
language design, Jour. Acoustic Soc. America 22.707 (1950).

'* It is conceivable that there might have been yet another clement of eymboliani in
language, if the noise behavior that became dissociated hud had such a relation to the Bitu
ation with which it was associated as would be independently arrived at by ovpry •jx'jikcr
(or by every speaker in the given culture). Such associations occur in onomalopoolic do-
ments (14), and they would have made words more a matter of individual oxprwwion tliao
of arbitrary social learning. Sapir found some traces of such phonetic symboliam by a ncal
use of the methods of experimental psychology; part of this work appears in the prewsnl
volume (61-72), part is as yet unpublished.

36 General Linguistics I

299

dissociated from it (symbolic), enabled him to use readily the morphological
approach current at the time. Grammars were usually organized not only on the
basis of the formal (distributional) relations of elements,'^ but also on the basis
of the major relations between form and meaning — such as whether there are
gender or tense paradigms. Sapir accepted this as a basis for grammatical de-
scription, and used it in distinguishing language types.

This kind of consideration is quite different from the purely formal one. The
formal typology would note to what extent linguistic elements have positional
variants (i.e. environmentally determined alternants), what kinds of combina-
tions of classes there are to be found, at what points in the structure we find
domains of varying lengths (as against unit length of operand), and the like.
The form-meaning typology notes the importance of noun classification on the
basis of gender, or the Uke; to this Sapir added the criterion of 'the expression
of fundamental syntactic relations as such versus their expression in necessary
combination with notions of a concrete order. In Latin, for example, the notion
of the subject of a predicate is never purely expressed in a formal sense, because
there is no distinctive symbol for this relation. It is impossible to render it with-
out at the same time defining the number and gender of the subject of the sen-
tence' (21).

The correlation of form and meaning is, however, only one side of linguistic
typology. It can tell us whether certain meanings are always either explicitly
included or explicitly excluded (like the plural in hook -^ hooks), or are unde-
fined when absent (as in Kwakiutl, where nothing is indicated about number if
no explicit plural morpheme is given). It can tell whether some meanings are
very frequently indicated, as any paradigmatic morpheme hke the English plural
would be. It can tell what meanings are expressed together, as in the Latin
example cited above. But the differences are largely in degree. As Sapir recog-
nized, even a meaning which is not paradigmatically expressed can be expressed
in any given language, even though absence of the morpheme would not then
mean presence of its paradigmatically contrasted meaning (as absence of -s indi-
cates singular, or absence of -ed and the vrill-dass indicates present). The fact
that a particular meaning is expressed as a grammatical category (rather than,
say, in a separate noun) is of interest to cultural history (443), but is not essen-
tially different from having the meaning expressed by any morpheme, of any
class (100).

Which meanings or kinds of meaning are expressed by which kinds of struc-
tural elements (paradigmatic sets, large open classes hke nouns, etc.) is never-
theless of considerable interest in discussing a language as social behavior. It
may affect perception, and may in part determine what -can be efficiently said
in that language. Sapir pointed out, for example, that the Nootka translation
for The stone falls would be grammatically equivalent to It stones dovm (some-
thing like the difference between Rain is jailing and IV s raining), and com-
mented that such differences show a 'relativity of the form of thought' (159).

"E.g. what large open classes there were (such as stems, or distinct verb and noun clas-
ses) which occurred with small closed classes (such as affixes, or distinct verb and noun
affixes in various environmental subclasses).

Sapir's Life uiul Work: Iwo Appnusals yj

300

Meanings. This line of interest led to research of a purely Hcmantic character.
Around 1930, Sapir wrote three long semantic papers as preliminar>' r««Koarchc«
toward an international auxiliary language: Totality (I,angua«e Monograph
No. 6); The expression of the ending-point relation in English, French, and
German (in collaboration with Morris Swadesh; Language M(jnograph No. 10)-
and Grading (reprinted here on pp. 122-49). We can distinguLsh .several prohlcma
in these investigations. First, there was some analysis of the purely .sfmanlic
relations among the meanings themselves. For example, Sapir says: 'Gnuiing aa
a psychological process precedes measurement and counting ... The term four
means something only when it is known to refer to a number which Is "less than"
certain others' (122). And farther on: 'Judgments of "more than" and "leas
than" may be said to be based on perceptions of "envelopment" ' (i.e. of suc-
cessively inclusive bounds). Such analysis could be aided by the abstract study
of relations in mathematics and logic (as in the relation between order and quan-
tity which is involved on p. 124), and perhaps also by investigations along the
lines of experimental psychology into basic (not culturally determined) percep-
tion and behavior.

Second, we find analysis of the precise meanings of the relevant word.s of a
given language. Sapir was always an artist at bringing out the complexities of
meanings hidden in a particular word, or in someone's use of the word in a given
situation. Here he does this in a more formal way. He shows, for example, that
there are two different uses of good, near, and other grading terms (12t>-8):
referred to an absolute norm (e.g. brilliant, or better in Thanks. This one is better) ;
and referred to comparison (e.g. better in My pen is better than yours, bid I confess
that both are bad) ; note that one wouldn't say A is more brilliant than B, but both
are stupid. In this second category we have good in the sense of of ivhat quality
{How good is it? Oh, very bad), and near in the sense of at what distance {How
near was he? Still quite far). Similarly, he points out that many grading terms
'color the judgment with their latent affect of approval or disapproval (e.g. "as
much as" smuggles in a note of satisfaction; "only" and "hardly" tend to voice
disappointment)' (139) .2°

Third, from his analysis of the total meanings which are expressed in each
word, Sapir isolates various factors of meaning, chiefly the following: the dis-

" It is always possible, of course, to overlook varioua environmental factors in analyi-
ing the meanings of words. Sapir says (140): 'if a quantitative goal is to be rpach<*d by
increase, say "ten pages of reading," more than necessarily has an approving ring (e.g., "I
have already read more than three pages," though it may actually be less than four), Ust
than a disapproving ring (e.g., "I have only read less than eight pages," though it may »c-
tually be more than seven). On the other hand, if the quantitative goal is to be rrached
by decrease, say "no more reading to do," more than has a disapproving ring (e.g., "I h»ve
still more than three pages to do," though actually less than four remain to \ye t\onc). Ui*
than an approving ring (e.g., "I have less than eight pages to do," though more ih.nn nevta
pages remain to be done out of a total of ten).'— If the form of the verb were taken into
consideration here, it might be possible to show that the approving ring comca from tbe
conjunction of more with the past tense and less with to do, the disapproving ring from mort
plus to do and less plus the past tense. To isolate the 'afTect in grading', which Sapir ar«)a
here, we extract an element 'approval' out of more plus paat and Uss plus future, »nd an
element 'disapproval' out of the opposite combinations.

38 General Linguistics I

301

tinction between grading with reference to a norm and grading with reference
to terras of comparison (125-6), noted above; open and closed gamuts of grading
with one central or two end norms (127-30) ; reversible and irreversible sets
(132-3) ; direction of increase or decrease (and also goal) implied in the grading
word, as in good : better versus good : less good (134-5, exemplified in fn. 20 above) ;
the intrusion of affect in regard to the grade (and the goal) (139-44, and cf. fn.
20 above). Such isolating of 'elements of meaning' is not subject to the usual
criticisms directed against semantic work, because it is an empirical linguistic
investigation. It does not derive elements of meaning from some deductive sys-
tem of presumed basic meanings, but discovers what elements can be separated
out from the total meaning of each word ; and it discovers this by comparing the
various words of a semantic set, by seeing the hnguistic environment in which
these occur, and the social situation or meaning of each use.

All these investigations involving meaning, when carried out with the kind of
approach that Sapir used, have validity and utility. The formal analysis of
language is an empirical discovery of the same kinds of relations and combina-
tions which are devised in logic and mathematics ; and their empirical discovery
in language is of value because languages contain (or suggest) more complicated
types of combination than people have invented for logic. In much the same
way, we have here an empirical discovery of elements of meaning in natural
languages, instead of the seemingly hopeless task of inventing basic elements of
meaning in speculative abstract semantics.^^ True, the particular elements we
obtain depend on the languages considered and upon the degree and type of
analysis. But it serves as a beginning, to suggest what kind of elements can be
isolated and arranged in varied patterns, which ones can be combined within a
single morpheme (with what effect), what would result from expressing some of
them in grammatical forms and others in ordinary words, and so on. We thus
obtain both a picture of how meanings are expressed in languages, and a sug-
gestion of how other ways can be constructed.

Communication and expression. Having surveyed the relation of talking to
other behavior, and the meaning of talk, we turn now to the place that talking
occupies in the life of a person — what might be called the function of speech.

Sapir points out that talking fills various functions beside communication.
There is first the direct expressive effect to oneself of talking and of the way one
talks. To this Sapir adds the symbol of social solidarity that is expressed by
having speech forms in common— in the nicknames of a family, in professional
cant, in all sorts of small and large common-interest groups: 'No one is entitled
to say "trig" or "math" who has not gone through such familiar and painful
experiences as a high school or undergraduate student ... A self-made mathema-
tician has hardly the right to use the word "math" in referring to his own in-
terests because the student overtones of the word do not properly apply to him'
(16). Finally, because of the dissociated character of language, there is 'the im-

" As is well known, logic and especially semantics are also based in part upon the lan-
guage of their practitioners, and are limited by their linguistic experience. However, this
linguistic basis is not explicit because usually unacknowledged; narrow because usually
limited to European languages; and arbitrary because not subject to explicit empirical and
analytic techniques or to controls.

Supir's Life and Work: f'no Appraisals },^

302

portant role which language plays as a substitute means of expression for thim
individuals who have a greater than normal difficulty in adjusting to th(? en-
vironment in terms of primary action patterns' (18). Such functions of language
though episodically mentioned by linguists, merit further study, even though
these functions are often filled more adequately by other behavior— gewture
symbol, art, and the like. As a method of communicating, however, no other
behavior compares with language. Writing originated as an independent method
of communicating, but Sapir points out that 'true progress in the art of writing
lay in the virtual abandonment of the principle with which it originally started'
(13) : the pictorial and direct symbolization of experience was replaced by sym-
bolization of words; and we may add that in most systems the direct symboliza-
tion of words was replaced by signs for the sounds of speech.

Of non-verbal communication, such as railroad lights or wig^vagging, he adds
that 'while they are late in developing in the history of society, they are very
much less complex in structure than language itself (107). This statement holds
only in certain senses. It is true that each field of mathematics, and all of them
together, can deal with but a small range of subjects. And the symbols and
statements (equations) and sequences of statements of mathematics may each,
taken individually, be less complex than those of language. But the possibility
of including the results (output) of one relational statement into the terms of
another, by means of successive definitions, makes it possible for mathematical
statements to carry a far greater communication load than linguistic statements
on the same subjects: compare any mathematical formula but the most trivial
with its translation into English. Furthermore, developments in electrical circuit
systems, in electronic control instruments, and in electronic computers open the
possibility of highly complicated activities equivalent to communication. The
ultimate communicational operation in these instruments is simpler than in
mathematics (and much simpler than the countless experiential associations of
language), since it is generally reducible to yes-^o (closing or opening a circuit)
or to a distribution of a given current as among several branches in the circuit
(depending on the resistance of each branch). Nevertheless, the innumerable
possible lay-outs of paths, and the rapid and numerous occurrences of the basic
operation, may enable these instruments to carry more complex communication
than language can, within a limited range of subject-matter.

Sapir notes, indeed, that non-verbal communication may be more useful even
when it is not more complex (or because it can be more simple) ; namely 'where
it is desired to encourage the automatic nature of the response. Because language
is extraordinarily rich in meaning, it sometimes becomes a little annoying or
even dangerous to rely upon it where only a simple this or that, or yes or no, is
expected to be the response' (107),

Behind the discussion of language as a method of communication IIc.h the less
important but still relevant question of just how much of language-like com-
munication is language proper. This is largely the question of the intonations
and gestures which occur with speech. Sapir says: 'The consistent mes.sago de-
livered by language symbolism in the narrow sense may flatly contradict the
message communicated by the synchronous system of gestures, consi.sting oi

40 General Linguistics I

303

movements of the hands and head, intonations of the voice, and breathing sym-
bolisms. The former system may be entirely conscious, the latter entirely uncon-
scious. Linguistic, as opposed to gesture, communication tends to be the ofl&cial
and socially accredited one' (105).

While all this is quite true, a few cautions may be in place. Some of the intona-
tions may be reducible to patterned sequences of a few contrasting tones (tone
phonemes), and may thus be considered morphemes no less than the ordinary
morphemes with which they occur: in English this may be true of the assertion
or command intonations, but not of the ones for excitement or for irony.

This means that the question of which intonations are part of language and
which are gestural sounds is simply the question of which of them can be de-
scribed like the other elements of language — as combinations and sequences of
phonemic elements (in this case phonemic tones). In turn, this means that at
least some of the distinction between gesture and language is a matter of the
linguist's methods of analysis. This is not to say that the distinction is not im-
portant. The fact that ordinary morphemes and some intonations can be de-
scribed as fixed combinations of fixed phonemic elements, while other intonations
and all gestures cannot be so described, reflects a difference in the explicitness
and type of use of these two groups of communicational (and expressive) ac-
tivities.

For the linguist, one group is language, the other is not. For the hearer and the
speaker the difference may be one of degree, with decreasing awareness and
explicitness as we go from morpheme to morpheme-like intonations to other
intonations and gestures. But there is still considerable awareness of gesture and
intonation, which most people can understand with nicety. And there is often
great unawareness of the 'accredited' hnguistic communication and expression,
as when a person reveals his attitudes or wishes by what we call his 'natural
choice of words' (with or without the hearer's understanding of what lies behind
this choice).

The decision of what to include in the linguistic structure rests with the
hnguist, who has to work out that structure, and is simply a matter of what can
be fitted into a structure of the hnguistic type. The question of what activities
constitute what kind of communication is largely an independent one, and is
answered by observing the kind of use people make of the various communica-
tional and expressive activities.

Constructed language. So far the description and analysis. It is fine to do this
for its own sake. It is fine to obtain from this work generalizations and predic-
tions about language, or interconnections with more general problems about the
patterning of behavior. However, the linguist who has all these results in his
hands is also able to construct something with it, to synthesize something by
means of his knowledge. He can carry out critiques of people's language and
communication activities, showing what is being effected by them, or how they
fall short by one standard or another. He can use his particular analytic experi-
ence in devising combinatorial techniques, not only of linguistic material. He
can try to construct a communication system (and perhaps a representation
system) more eflBcient and free than existing languages.

Sapir's Life and Work: f'wo Appraisals 41

304

This last is always an attractive task to any linguist who is interested in the
productive potentialities of his work. It is little wonder that Je.si>er8en and Sapir
two linguists who were avidly interested in life and in their work, were each con-
cerned with the construction of a superior language.

The most obvious source of interest lay in the need for international communi-
cation. Because Sapir's anthropological horizons were naturally wider than Jcs-
persen's, the problem was more complicated for him because 'intemalionai*
meant for him more than just the western world: 'As the Oriental peoples \ic-
come of more and more importance in the modern world, the air of .sanctity that
attaches to English or German or French is hkely to seem less and less a thing
to be taken for granted, and it is not at all unlikely that the triumph of the
international language movement will owe much to the Chinaman's and the
Indian's indifference to the vested interests of Europe' (119) Furthermore, an
international language meant more than a pidgin auxiliary: 'It is perfectly true
that for untold generations to come an international language must be au.xiliary,
must not attempt to set itself up against the many languages of the folk, but it
must for all that be a free powerful expression of its own, capable of all work
that may reasonably be expected of language' (113). Special audiences for it
already exist, as in the 'social unity' of the scattered scientific world (108); but
Sapir recognized the social blocks: 'Any consciously constructed international
language has to deal with the great difficulty of not being felt to represent a
distinctive people or culture. Hence the learning of it is of very little symbolic
significance for the average person' (31). Under possible future political circum-
stances, however, such a language might conversely be 'protected by the power-
ful negative fact that it cannot be interpreted as the symbol of any locali.sm or
nationality' (113). And Sapir's comment quoted above about the possible effect
of the Asiatics on the establishment of an international language is an example
of the kind of social need which alone would bring such a languii^e into currency.

The need for a language of international communication arises not only from
the fact that communication without it may be impossible (where people do not
know each other's language), but also from the fact that it may be inefficient
(where one depends on translation, interpreters, or one's limited knowledge of
a foreign tongue). We are here dealing with the question of information lass in
translation. On this subject Sapir says: 'To pass from one language to another
is psychologically parallel to passing from one geometrical system of reference
to another. The environing world which is referred to is the same for either
language; the world of points is the same in either frame of reference. But the
formal method of approach to the expressed item of experience, as to the given
point of space, is so different that the resulting feeling of orientation can be the
same neither in the two languages nor in the two frames of reference' (153).

There is however a difference between the two cases. One might claim that
what is said in one geometric frame (or language) is different from what i.s said
in another, or that the relation of the given information to its univense (or to
other bits of information) is different in one from its translation in the other.
Still, any identification of a point or relation in, say, Cartesian coordinates CAn
be given completely in, say, polar coordinates, and conversely (though the 'trans-

42 General Lini^iiistics I

305

iation' may be more complicated than the original statement). This does not in
general hold for language translation. Except for relatively simple parts of the
physical world (like the smaller numbers), or very exphcitly described parts of
it (like the set-up of a scientific experiment), we carmot get a description of the
physical world except as variously perceived by the speakers of one language
or another." It is therefore not in general possible to see how two language sys-
tems depart from their common physical world, but only how they depart from
each other. The question of translation is the question of correcting for the dif-
ference between the two systems. But neither system can be referred to an abso-
lute physical system (as is possible in the case of scientific terminology), nor is
there at present any general method for establishing equivalence relations among
them (as can be done among geometric frames of reference). Therefore it does
not seem possible to establish a general method for determining the information
loss in translating from one language to another, as Wiener would do on the basis
of his measure of 'amount of information'. ^^

These two types of difficulty in international communication may have been
the major stimulus to the many attempts at forming auxiliary languages. To
Sapir, however, as to some linguists and logicians, there was also the incentive
of fashioning a superior language system. He was well aware of the limitations
of our language, which both narrows our perception and prevents us from ex-
pressing adequately some of the things we have perceived: 'As our scientific
experience grows we must learn to fight the implications of language ... No mat-
ter how sophisticated our modes of interpretation become, we never really get
beyond the projection and continuous transfer of relations suggested by the
forms of our speech. After all, to say "Friction causes such and such a result"
is not very different from saying "The grass waves in the wind" ' (10-1). He
was also able to show that linguistic systems are much less satisfactory than
might appear: 'The fact that a beginner in English has not many paradigms to
learn gives him a feeling of absence of diflBculty ... [but] behind a superficial
appearance of simplicity there is concealed a perfect hornet's nest of bizarre and
arbitrary usages ... We can "give a person a shove" or "a push", but we cannot
"give him a move" ... We can "give one help", but we "give obedience", not
"obey" ... "To put out of danger" is formally analogous to "to put out of school",

" See E. Sapir and M. Swadesh, American Indian grammatical categories, Word 2.103-12
(1946) — an item not included in the bibliography. On p. Ill Swadesh quotes a perfectly
valid note of Sapir's: 'Naivete of imagining that any analysis of experience is dependent on
pattern expressed in language. Lack of case or other category no indication of lack func-
tionally ... In any given context involving use of language, lang. response is not to be split
up into its elements grammatically nor sensorimotorly but kept as unit in contextual pat-
tern.' Elsewhere, however, Sapir says: 'The "real world" is to a large extent unconsciously
built up on the language habits of the group ... The worlds in which different societies live
are distinct worlds, not merely the same world with different labels attached' (162). There
is no contradiction here, since the 'environing world' is the physical world, whereas the
'real world', in quotes, is also called 'social reality' (162) and constitutes the physical world
as socially perceived: 'Even the simplest environmental influence is either supported or
transformed by social forces' (89) ; 'The physical environment is reflected in language only
insofar as it has been influenced by social forces' (90).

» Norbert Wiener, Cybernetics, Chap. 3, esp. 75-9.

Sapir's Life and Work: l\\<> Appiuisuh 43

306

but here too the analogy is utterly misleading, unless, indeed, one defincH school

as a form of danger' (114-5).

Because of his sensitivity to these limitations, Sapir had in mind 'an engine of
expression which is logically defensible at every point and which tend.s to cor-
respond to the rigorous spirit of modern science' (112). He pointed out that the
inadequacies of language systems have led to the development of Bcparate hj-b-
tems of symbolism in mathematics and symbolic logic (118). The problem wa«
therefore one of constructing a language system which by its structure would
avoid ambiguities and inefficiencies, would be a conformable vehicle for our
present scientific understandings, and would be able to change with growth of
our understanding. However, there may well be a distinction between the con-
struction of an international language for flexible use in ordinary life, and that
of a scientific language which would not only express in its structure the various
types of relations, of operations and operands, known to science, but would alao
have the truth- value retention of a logical system.-*

The program called for a language that would be easy to learn for people
coming with the background of the existing languages, and that would be as
simple as possible in its structure, while selecting the kind of structure that would
fit the scientific understanding of the world. Because these were his interests,
Sapir did not try to construct a language, like Jespersen's Novial, but tried
rather to find out what should go into the construction of such a language. Even
his investigation of phonetic symbolism is relevant here, as showing what mean-
ings might be less arbitrarily expressed by particular sounds. The investigations
which he made specifically for the International Auxiliary Language .Vssocialion
were the semantic papers mentioned above, which would show how aseful or
harmful it was to have certain meanings expressed together within a morpheme,
and what component factors of meaning could be extracted from given words
by seeing how they are used. The questions of what meanings could be conveni-
ently expressed by what kinds of structural elements, and of what pattemings
and formal structures were possible, were not touched by Sapir.

<4.> Change in language. Sapir's tendency toward analysis in depth, which
he could express within descriptive linguistics by means of the process tj-pe of
formulation, led also to the historical investigation of patterned features. In the
process formulation, time was not involved, and depth was a matter of various
analytic layers of the system. We now consider investigations in which depth
was a matter of historical time, of various successive forms of the system through
time.

A descriptive pattern can of course be viewed as being just nn interesting
arrangement of the data. However, since Sapir saw it n-s the result of various
distributional processes (such as protective mechanisms) among the rlcments,

** For an example of how particular logical relations can bo built into a constructed
language, consider the 'newspeak' of George Orwell's novel Nineteen oiKhtyfour. One of
the distributional features which is only lightly suggested in his nyKtem in the technique
(not unknown in our real languages) of letting opposites equal or replace each other in cer-
tain environments, with the result that no distinction between oppoeilea (say between war
and peace) can be made in the language.

44 General Linguistics I

307

he could readily see it also as the result of various historical processes affecting
the elements. An instance is the historical addition of a glottal stop between
morpheme-final vowel and morpheme-initial vowel in the example cited earlier:
in terms of descriptive process, the ? in -V -f V- was based on a descriptively-
prior absence of -W-; in terms of history the ? in -V + V- may actually have
been a late development, due analogically to the absence of -W-.

A detailed example of this is the discussion of glottalized continuants in cer-
tain west-coast languages. After making it clear that all or most of the types
y, w, m, and n are distinct phonemes in the languages under consideration,
Sapir points out that they are 'so singular that it is tempting to seek evidence
accounting for their origin' (226-7). Their singularity is partly distributional
(in Navaho, these alone of all consonants do not occur as word-initial), partly
morphophonemic (in Navaho, these occur in morphemic environments which
can be otherwise shown to have once contained a d morpheme, 228-9). For
Wakashan (Nootka and Kwakiutl), he shows that these consonants go back to
coalescences of ^ or h with neighboring continuants (244) ; the argument is far
too involved and detailed to be summarized here (230-44) . In the course of his
analysis, Sapir shows that additional glottaUzed continuants probably existed
once in Wakashan (231), and that Boas' 'hardening' process is not the opposite
of his 'softening' but is simply a glottalized softening (233). The whole recon-
struction, based on comparative evidence, is then used to suggest that when
phonetically 'weak' consonants drop they may leave influences in neighboring
phonemes, i.e. that they are absorbed rather than dropped (244). With this
background, Sapir than reconstructs Indo-European laryngeal bases out of vari-
ous sets of irregular cognates (245-50), by explaining the various consonantal
irregularities as regular reflexes of the effect of lost larjnigeals (i.e. of their ab-
sorption) ,

The same methods of investigation are apparent in the famous series of articles
on word cognates and word borrowings in Indo-European, Semitic, and other
Mediterranean languages, which began to appear in 1934. Two of these are re-
printed here (285-8, 294-302) ; all are of course listed in the bibliography. Studies
of loanwords were prominent in this series, because they made it possible to
consider the effect of each language system on the form of the word, and to ex-
plain otherwise unexplained forms. These papers, together with that on glot-
talized continuants, are masterpieces of brilliant association, bringing together
all sorts of apparently unrelated data, and of meticulous responsibility to every
possibly relevant consideration or counter-argument. To discuss what Sapir does
in them would take as much space as the original articles ; only a careful reading
can reveal their remarkable craftsmanship. Some aspects of the method of work
used in them, however, will be discussed in Part 4 below.

Much of this brilliance and craftsmanship went into Sapir's painstaking work
on Tocharian, which was one of his main projects during those years, and most
of which is as yet unpublished.

In addition to all this work, which was of a unique character and bore the
stamp of his personality, Sapir also carried out standard work in comparative lin-
guistics, as for example in the Encyclopaedia Britannica article on Philology,

Sapir'.s Life and Work: Two Apjniiisals 45

308

or in The concept of phonetic law as tested in primitive languages by Leonard
Bloomfield (73-82), in which he presented Bloomfield's Algonkian n»conHtructiona
and his own Athabascan ones.

Sapir being what he was, he not only carried out historical linguiatic invewti-
gations but also made historical linguistic interpretations. In his hook Language
(Chapter 7), he suggested that similarities among genetically related lunguagcH
which were too late for their common ancestry, but which could not eawily be
explained as diffusion, might be explained by a 'drift' which rx-curs in each of
these languages independently of the other but along parallel lines of develop-
ment. This view has been generally questioned and disregarded by linguista,
although data that may support it are not lacking." Sapir granted that such drift
could be explained only on the basis of what he sometimes called 'configurational
pressure' in the structure with which each of the sister languages started. That
is to say, the parent structure may have contained certain imbalances or irregu-
larities, or may otherwise have favored the occurrence of certain changes rather
than others; and as this structure developed in various separate places (in what
became the various daughter languages) it underwent some of these structurally
favored changes in several places independently of each other. Elsewhere, Sapir
uses the concept of drift, i.e. of structural favoring as a source of change, to
explain the bulk of changes — differentiating ones as well as parallel ones (23).
Little, however, can be done with this concept until we can say what kind of
structure favors what kind of change in it, i.e. until we can specify 'configura-
tional pressure' and then test to see if it operates.

In addition to this tentative suggestion about the direction of Unguistic change,
Sapir commented on the even more general problem of the rate of change. There
have been various conditional suggestions, as for example that languages with
tightly knit structures (e.g. Semitic) change more slowly than those with looser
structures (e.g., in comparison, Indo-European). To this Sapir added the general
statement that all languages change much more slowly than culture (26-7) and
at a more even rate (433),^' although he thought that changes in both rates might
be interconnected: 'The rapid development of culture in western Europe during
the last 2000 years has been synchronous with what seems to be unusually rapid
changes in language' (102). He then used this statement for a possible explana-
tion of why there is no structural correlation between the patterning of language
and the patterning of culture: even if there was once a 'more definite association
between cultural and linguistic form, 'the different character and rate of cliange
in linguistic cultural phenomena ... would in the long run very materially disturb
and ultimately entirely eliminate such an association' (101, also 26 and 102).| ]

» Cf . Zellig S. Harris, Development of the Canaanite dialects 99-100 (Npw Haven. 1939)

" An echo of this appears in the work of Sapir'autudent Morris Swadeah on rale of vocab

ulary change. Cf. in particular hia Saliah investigations, carried out under the auspice* o(

the Boas Collection in the American Philosophical Society Library, and published in S»luh

internal relationships, UAL 16.157-67 (1950).

46 General Linguistics I

Editorial Note

First published, as a review of Edward Sapir: Selected Writings in Language,
Culture and Personality (ed. D.G. Mandeibaum, 1949) in Language 27 (1951),
288-333. Of the three sections, "Language", "Culture", and "Personality" (follow-
ed by a "Conclusion"), only the first is reprinted here.

Editorial interventions: passages deleted within a sentence are indicated with [...];
passages deleted between sentences with [...]; all editorial additions or changes
are put between < >.

Introduction to Stanley Newmans ic\i

Stanley Newman (1905-1984)' was one ol Sapirs most gificd and brilliant stu-
dents, whose interests covered the fields of linguistics, anthropology, and psycho-
logy, domains to which he made significanl coiilrilnitions. as Sai^ir did also.

Whereas Zellig Harris's review of the Selected Writings of Edward Sapir of
which extracts are reproduced here — offers, so to speak, an "analytical arrange-
ment" of Sapir's working methods and procedures. Newman's review - reprodu-
ced here with deletions of those passages that directly relate to the occasion of
making Sapir's writings available in ihc 1949 selection — provides us with a
"process" view of Sapir's interests and scholarly career, and also of his academic
prose. Newman pays particular attention to the scientific (and literary) genres
which Edward Sapir practiced, and to his expert handling of various styles.

Newman's appreciation of Sapir's book Lnni^uai^c ( 1921 ) appears lo he raiher
unenthusiastic — and one could easily question his statement thai Sapir '"later
abandoned [...] many of the problems discussed in its pages" — but there is much
praise for Sapir's opening up of linguistics to the study of modes ai behaviour,
in language as well as in adjoining fields, and of his extension o\ the study o\ lin-
guistic patterns to the total range of social patterns of form.

Newman's review was written against the background of the e\oluliO!iar\ ten-
sions in American linguistics during the late 4()s and early ri()s. marked b\ the
refinement of linguistic techniques and a restrictive practice of linguistics as a
descriptive (or descriptivist) doctrine. It is not so much with the unfortunate
overemphasizing of methods opposite to those of Sapir and his students that
Newman has a problem, but rather with the trend towards a narroued perspec-
tive and towards the reductionist practice of linguistics as "microlinguistics." a

I For an obituary of Stanley Newman, see l.(iniiuiii;f 6.^ (1487). .^46-36() (obituary and scicclnc biblH>ftr.iphy
by Michael Silverstein). For a full bibliography, the edition of a number of unpublished biocraphual lc\t»
of Stanley Newman, an inventory of his linguistic materials by Mary Ritchie Key. an appr.iivil •-% Mi.h.icl
Silverstein. an obituary by Philip Fiock and Harry Baseharl. a histi>riographical Mud\ oi ' c

within the "Sapir school of linguistics" by Regna Darnell, and various articles in honi>ur .>t
see the volume Cicncral ancl Ameriiulian i.{hiu>lin\iuisius: In Htnif/nhrumt' of SltinU-\ S < ^

Ritchie Key and Henry M. Hoenigswald (Berlin'New York. IWW) (see my review in D'

- Ilie full text of the reviews bv Harris and Newman has been reprinted in Konrad KiK*mcr (cd.). Edward Sapir
Apprcisdls of his Life ami Work ( Amsterdam/Philadelphia. 1^84). pp. .'>9- ft.S ( Newman's review) and pfK W-1 M

(Harris's review).

^ See Newmans posthumous article "I he Development of Sapir s INvchology of Human Bch.i>io„r- m V.^
Perspectives in Lani^iui\ie. Ciiliiire. and I'ersonulilv. edited by Willi.im ('t>w.in. Muh.ul K I
Koerner (Amsterdam/Philadelphia, I W6), 40.'^ 427 (and discussion, pp 427 4 M > In this .utu
a more positive appraisal of Sapir's l.anf^iiage (see pp. 411 4 1 2 1

48 General Linguistics I

term which Newman borrows from George L. Trager/ To this (neo-)Bloom-
fieldian hnguistics — of which he recognizes the methodological rigour, as well as
the practical necessity — , Newman opposes Sapir's "cosmopolitan" linguistics, of
which he perceives signs of revival (Newman was probably alluding here to work
in anthropological linguistics and to anticipatory efforts in what came to be call-
ed ''sociolinguistics," especially in the study of bilingualism and multilingualism).
It is in the light of Newman's intuition of evolutionary trends in American lin-
guistics that we should read the conclusion of his text, where he draws an admit-
tedly oversimplified contrast between the "centripetal" Bloomfield and the "cen-
trifugal" Sapir.

Pierre Swiggers

4 See George L.Trager, The Field of Linguistics (= Studies in Linguistics: Occasional Papers, 1; Buffalo, 1949).

Stanley NEWMAN
<Edward Sapir's work and sivlo*

|1X0| |...] <l>n spile 1)1 Sapirs shorl lilc. his monographs, arliclcs anJ rcvic\^^
flowed in a voluminous and steady stream over a productive period of nearly
thirty-five years. His writings encompassed a wide range of topics it! -I

distinct disciplines. And the quality of his writing was mamlained al a ie\ . i

ginality and richness that was as steady as its volume. Sapir did not seem lo ex-
perience the ups and downs of inventiveness that normally pleague a untcr. Tven
in a brief review, where he would ostensibly be discussing a specilic lH>ok. his
fresh insights illuminated a circle of new problems with unsuspected significance.
[...] <In> [...] Sapir's papers in the field of American Indian languages |...| • i -he
historical evidence is emphasized [...] <, as> in internal Linguistic l^Nidcnce
Suggestive of the Northern Origin of the Navaho," |...l illustrat<ing> how com-
parative linguistic [181] data can be utilized lo reconstruct the history of group
migrations. <His two> papers "Abnormal Types o\ Speech in Nootka" and ".Male
and Female Forms of Speech in Yana," deal with linguistic devices characterizing
certain socially defined groups in these two cultures. Tlie Noolka article lakes up
the problem of the historical development of these abnormal lypes of speech,
which resemble speech defects but function as mocking forms or as styles of
speech identifying certain folktale characters. Tlie possible similarilies in the pho-
nological development of glottalized continuants in several unrelated languages
are examined in "Glottalized Continuants in Navaho, Noolka. and Kwakiull
(with a note on Indo-European)." The "note" of some half-dozen pages is a suc-
cinct presentation of Sapir's views on the Indo-European larvnueal hvpolhesis

The first ten years <of Sapir's scholarly career>. from U^O^ through U>15. were
primarily devoted to descriptive studies in American Indian languages. Dunne
this time he published texts, vocabularies, descriptive sketches or fragments on
Kwakiutl.Chinook.Yana.Wishram. Wasco. Takelma. Ule. Paiule. Noi>lka. Iiik-lt*.
Chasta Costa, Comox. Toward the end of this period another aspect ol .American
Indian linguistics was brought into focus. Sapir's background of training in
Semitic and in Indo-European comparative linguistics <was> now applied lo
American Indian languages. In 1913 he published the first of his papers on
"Southern Paiute and Nahuatl. a Study in Uto-Aztecan." |...| Hiis suhstanli.il
study of nearly a hundred pages represents, as far as I know, the first apphcaiion
to American Indian languages of the comparative method based upon the aua
lysis of systematic phonetic correspondences and directed toward the reconstruc-
tion of the sound system in a parent language. |...| It is a revealing comment i!\
on Sapir's character that when he wrote an article. nearK twenlv \ears l.itci.
demonstrating the application of the comparative appri>ach lo American Indian
languages, he entitled it "Hie Concept ol I'hoiictic I aw as Tested in Primilive
Languages bv Leonard Bloomlield."

ITie period of 1916 through 1925. which c*>\ered the last ten years of his fif-
teen-year Slav in Otlawa, brought significant new currenls inU> (he broadening

50 General Linguistics I

stream of his interests. He continued, though less intensively, to publish descrip-
tive studies in American Indian languages. The full-length grammar of Takelma
appeared in this period, based upon data collected some ten years earlier. His
detailed and meticulous description of Southern Paiute, not published until 1930,
was completed in 1917. He also wrote descriptive articles on Nootka, Yana,
Kutcnai, Chimariko, Haida, Sarcee. Comparative linguistics drew more of his
attention than it had previously, but his interest turned increasingly toward struc-
tural comparisons rather than phonological analyses. He kept on publishing com-
parative studies in Athabascan and Algonkin, and it was during these years that
he wrote all of his articles on the Hokan problem and his one paper on Penutian.

Sapir's contributions to American Indian linguistics should correct the impres-
sion that he was a writer who produced only one book. Language, with the re-
mainder of his work appearing in the form of brief articles. It is true that he had
a special flair for condensing a problem or a point of view in the ten-to-twenty-
page article which is the favored literary form of scholarly journals. But he did
not by any means confine himself to this form. He also wrote many longer
articles, and he produced about a dozen monograph-length or book-length
grammars, text collections, and [182] comparative studies in American Indian
languages.

But, in addition to his linguistic work in the American Indian field, Sapir's writ-
ings during this period reveal the new trend that was to become the absorbing
interest of his life. He began to venture beyond the strict confines of linguistics
and to seek new perspectives for the phenomena of language that would relate it
to other forms of human behavior. About half of his monograph of 1916, Time
Perspective in Aboriginal American Culture: A Study in Method, discusses the
types of linguistic evidence which can be utilized for reconstructing culture
history. His book Language, published in 1921, contains sections and chapters
which show the same tendency to explore wider problems. But the book should
be regarded as merely an evidence of his early attempts in this direction, for he
later abandoned or completely restated many of the problems discussed in its
pages, such as the relation between language and thought, or the characteristics
of language as a form of art — an idea in which he was apparently stimulated at
the time through his reading of Croce.

In 1925 he wrote his first article on an international auxiliary language. The
same year saw the publication of "Sound Patterns in Language," the first article,
I beheve, in which he used the term "phoneme." To Sapir the phoneme concept
was significant, not so much as a methodological tool for the linguist, but rather
as a powerful and clear demonstration of the unconscious patterning of human
behavior. Essentially, he attempted to show in this article that speech sounds can-
not be fruitfully understood as a mere set of articulatory motor habits: two lan-
guages "may have identical sounds but utterly distinct phonetic patterns; or they
may have mutually incompatible phonetic systems, from the articulatory and
acoustic standpoint, but identical or similar patterns."

One can gain some notion of the new sources of stimulation and vitaHty that
entered Sapir's work during the 1916-1925 period by examining his writings out-
side of linguistics. Ethnological papers continued to appear as before. But in 1917
he pubhshed reviews of Freud's Delusion and Dream and of Oskar Pfister's The

Sapir's Life and Work: Two Apprmsals 51

Psycliounalvtic Method, llicsc \scrc llic lirsl indicaluuis in his wnlmgs of an
interest that was to continue throughout his hte. Articles and reviews on music
and literature also began to appear during these years. And wi I«^17 he In-yan to
publish poetry, whose volume and whose signilicanee to his thinking sht>uld not
be overlooked. He published one book of verse and <hundreds of> poems ( .| m
many literary journals of (\mada and the United Slates. Iliese were no amateur-
ish effusions which he tossed o{[ now and then m his lighter momenlv In fact, it
might be said of Sapir that he could not approach any task in the true spirit of an
amateur or a dilettante. He worked at poetry with the same unrelaxing energy
and incisiveness of mind that characterized his efforts in linguistics or elhnt»logv.

His experience with poetry had a distinctive influence upon his prose style.
Sapir was always a competent writer of expository prose. Even his earliest papers
show that he never lacked the ability to write the clear, precise, well-organ*'
though somewhat colorless prose characteristic of the better academic wrii
But his prose from about 1920 began to take on new dimensions. One can notice
a growth in the apparently effortless and graceful fluency of his expression.
Certain verbal habits peculiar to poetry invaded his prose. liven passages pulled
out of context from his later writings are eminently quotable, for he became skill-
ful in the use of the packed phrase, the vibrant word, the familiar image reset in
an unfamiliar context to evoke fresh and unsuspected im[Tliealions of a theme.

His writing continued to be clear and |183| ordered in its conceptual exposi-
tion, but he emphasized more and more the control (^{ e\ oealive overtones in an\
topic he discussed. He set out to capture, not only the intellects of his readers, but
their feelings and attitudes as well, and anyone who knew Sapir can have little
doubt that he did this with utter frankness and a full consciousness of what he
was doing. Instead of continuing to master the one style of conventional aca-
demic writing, he became adept at handling many styles. He preferred to play a
variety of stylistic tunes in one and the same article, shifting imperceptibly from
a sober argument, to an imaginative play with words and concepts, to an inter-
lude of wit and humor — and Sapir became increasingb fond of indulging in pas-
sages of academic leg-pulling — back to the sober line of argument again. It is this
breadth and variety in his control over language which gives his writing its coN>r
and refreshing vitality.

The implications of Sapir's holistic use of language |...l uere realized more
fully in his publications after he returned to the United Stales m l^^^."^. He prac-
tically stopped writing descriptive and comparati\e studies in .-Xmerican Indian
languages. Most of the few American Indian papers which appeared were appar-
ently based upon previously collected materials and mereb etiited ft>r publica-
tion during this period. He became more inlererested m utilizing this data to ilhis
trate socially and psychologically significant modes of behavior in language. |...|
Throughout many articles he drew upon his American Indian linguistic data {ox
examples to pinpoint a broader theme. Iliis technit|ue the presentation of con-
crete examples, followed by an explanation oi then meaning and significance in
a more inclusive frame o{ reference - became a fa\onie mode of ex|>oMtion with
Sapir.

This period saw a revival of his earlier mleresi m historical and comparative
studies of the Indo-European and Seniilie languages Hu lliltite problem stimu-

52 General Linguistics I

lated him to examine Hittite-Indo-European relationships and to publish sev-
eral papers on his results. In one article he traced certain influences of Tibetan
on Tocharian. which he believed to be a "Tibetanized Indo-European idiom." He
had additional data on the Tocharian-Tibetan problem, and early in his career he
had collected Sinitic materials in exploring Sinitic-Nadene relationships. [...]

He also continued publishing articles on the problems of an international con-
structed language. And it was during this period that he wrote his three papers in
the field of semantics — Totality, The Expression of the Ending-Point Relation in
English, French, and German, and "Grading, A Study in Semantics."

The bulk of his articles after 1925, however, reflected his primary interest in
pushing language study beyond the conventional boundaries of linguistics. Some
of his general articles during these years — "Philology" (1926), "Communication"
(1931), "Dialect" (1931), "Language" (1933), "Symbolism" (1934)— outlined the
multiple facets of linguistic phenomena as they impinge on problems of indivi-
dual and group behavior. This point of view was presented in a programmatic
manner in "The Status of Linguistics as a Science" (1929), whose purpose, in spite
of its title, was "not to insist on what hnguistics has already accomplished, but
rather to point out some of the connections between linguistics and [184] other
scientific disciplines." In this paper he stressed the strategic importance of lin-
guistics for the methodology of social science. [...]

The content of language was, to Sapir, significant as "a symbolic guide to cul-
ture." "We see and hear and otherwise experience very largely as we do because
the language habits of our community predispose certain choices of interpreta-
tion." The individual's behavior in language was also important as symptomatic of
his personality, and this theme he discussed in detail in "Speech as a Personality
Trait" (1926). But it was the evidence of form in language which impressed Sapir
as having the deepest implications for an understanding of human behavior.
Linguistic form was a patterned phenomenon; in the individual or the group
these formal configurations were adhered to or recreated unconsciously and
intuitively. Sapir unceasingly hammered at this theme in his articles, whether
written for linguists, psychologists, or social scientists. In "The Unconscious
Patterning of Behavior in Society" (1927) [...] Sapir used linguistic data as his
prize exhibit, but he also attempted to show that unconscious patterning was char-
acteristic of non-linguistic forms of behavior as well. He translated this concept
into psychoanalytic terms, [...] when he wrote in an earlier book review of the
need for discovering a social psychology of "form-libido." In short, language pro-
vided the clearest and most easily described evidence of the fundamental human
tendency to mold behavior into unconscious patterns of form. [...]

<I>t is somewhat arbitrary to divide Sapir's writings into the categories of
[...] Language, [...] Culture, and [...] The Interplay of Culture and Personality.
[...] <T>o Sapir these were not separate fields, and his writings, particularly
during the last 15 or 20 years of his life, explain and reiterate his reasons for
considering them as an indissolubly fused whole. [...] <His> Time Perspective
paper <for instance> [...] gives a concentrated presentation of the methods for
using linguistic evidence to work out time perspectives. This paper should be
required reading for students in linguistics, as it is for most students of ethnol-
ogy. Similarly, [...] such papers as "Speech as a Personality Trait," "Symbolism,"

Sapir's Life and Work: I wo Appraisals 33

and "The Unconscious Paltcrning ol IkhaMor m |IS5| Societv" '-nrc imporlanl
for> the hnguisl.

At the time that Sapir was seeking to expand the hori/tms ..t i.r '^

beyond the hnguist's traditional universe ot discourse, hisiors pliJNcii k

on him by directing hnguistics into contrary channels. Under ihc influence of
Bloomfield. American Hnguists in the h>3()s turned to rm intensive cultivation of
their own field, sharpening their metht)dological tools and rigorously defining ihc
proper limits of their science in terms of what Trager has identified as "microlin-
guistics."71iey became increasingly efficient microlinguisis. Certainly no one can
deny that this involutionary trend has gi\en linguistics a disciplined claril\ and
power of analysis that it never had before. But it is equally true thai ihis trend
carries with it the seeds of an ever-narrowing parochialism. And it was Sapir's
main purpose to make linguistics a more cosmopolitan member of the commu-
nity of sciences.

[...] Sapir was as thoroughly committed as Bloomfield to the view that a valid
linguistic science must be a coherent and self-consistent body of concepts. It must
not look for extra-linguistic formulations to support or, still worse, to validate its
findings. [...]

Sapir's policy in seeking interdisciplinary linkages between linguistics and psy-
chology was simply to present linguistic formulations and to allou psychologists,
of whatever brand, to make their own reinterpretations. Many of his articles were
addressed to psychoanalysts and psychiatrists, for he saw that the operations o{
the unconscious as manifested in language could provide data t>f particular inter-
est to these specialties. In "The Status of Linguistics as a Science" he pointed
out that the configurated character of language, which "develops its fundamental
patterns with relatively the most complete detachment from other types of
cultural patterning" should have a special value for Gestalt psychoU\iiists. He
spoke to experimental psychologists in their own lingo in "A Study in Phonetic
Symbolism," where he reported the results of his use of experimental technujues
in studies of sound symbolisms.

In making this manifold approach to psychologists Sapir realized that, if lin-
guistics is the body of formulations made by professional linguists, then psycho-
logy is the body of formulations made by psychologists. He did not try to select
or construct a linguist's psychology, which, like a psychologist's version of lin-
guistics, would be neither fish nor ftnvl, but a spurious bod\ i>f diKtrine irrelevant
to both disciplines. [...J

Sapir's approach to this delicate interdisciplinary problem is especially impor-
tant [...] because there are signs that a renewed effort in this ilirection is wow
being made.<'> Fruitful results can be achie\ed if interested linguists and special-
ists in the other sciences of human behavior are willing to respect and to lr>' to
understand |186| one another. This type of endeavtir is. o{ course, fraught uilh
misunderstandings and disillusionmcnts. But it is the only way in which linguists

' <Newman"s statement, applying to the situation in the early U>5(K is (n«>Nv
ment between models of grammar and models of cognition. In addition, ihi
guistics has expanded significantly since the mid-century. >

54 General Linguistics I

and other specialists can cooperate to find concrete problems in which both can
contribute and to formulate concepts relevant to both fields. Because Sapir
understood the necessity of this approach, his linguistic writings are particularly
meaningful to non-linguists. [...] <I>t may turn out that Sapir's major contribu-
tion in the long run will be as the linguists' spokesman to psychologists and social
scientists. Although Sapir used linguistic methods and procedures with consum-
mate skill, he was an artist rather than a scientist in this regard. It was Bloomfield
who formulated the methods of linguistic science into a clearly defined and tight-
ly coherent body of doctrine.

Linguistics has been fortunate indeed in claiming two men of this stature of
genius, who could provide such utterly different and complementary impulses to
their field. The one might be considered the centripetal force in hnguistics; the
other's impulse was decidedly centrifugal. One pointed the way to a more inten-
sive and logical analysis of linguistic phenomena; the other indicated the broader
perspectives within which linguistic science could contribute to a richer under-
standing of human behavior.

Editorial Note

First published, as a review of the Selected Writings of Edward Sapir (ed. D.G.

Mandelbaum, 1949), in International Journal of American Linguistics 17 (1951),

180-186.

Editorial interventions: passages deleted within a sentence are indicated with

[...]; passages deleted between sentences with [...]; all editorial additions or

changes are put between < >.

Section One

The Problem ofthe Origin of LANCriAdi ( \'-n)l)

Introduction to Sapir's ^^Hcrdcr's "Urspruno dcv

Sprachc''^' (1907)

This essay is Sapir's first scholarly wrilinu. ihouuh ii u.is |nililishcd ai a
later date than his note on the "Rival ( hiets" ( l^^O.S) and onK in the same
year as his first articles on the Takelma Indians. Ilie text is the somewhat
revised version of Sapir's master's thesis in (iernianic Philology at
Columbia University (1905);' it was published in the journal Modern
Philology.^

Sapir's article on Herder is significant in at least two respects:

(a) As the elaboration of a rather marginal theme within (iermanic
philology, it reflects Sapir's interest in general linguistics . and in the
philosophy of language;

(b) The topic chosen allowed Sapir to bring in part of his background in
Biblical Hebrew studies (given the fact that Herder takes Hebrew to be
a primitive language); see, e.g., the references to Hebrew ^N "nostril,
anger," dual D"'QK "nostrils, nose, face" (p. 127). and the allLisii)n to
Hebrew m~l "breath, life, wind, soul" (pp. 127. 129).'

In addition the topic provided Sapir with the opportunits to profit from the
widening of his linguistic horizon which he owed to Franz Boas. Although
Sapir does not offer specific examples from American Indian languages."
he refers, in very general terms, to the "elaborate formal machinery, parti-
cularly in regard to the verb, of the Semitic and of many North American

' "The Rival Chiefs, a Kwakiutl Story Recorded by George Hunt", in Boas Anniversary Voltune (New
York, 1906), pp. 108-136 [Reprinted in The Collected Works of Edward Sapir. vol. \'I. pp. 323-351).

2 See: "Rehgious Ideas of the Takelma Indians of Southwestern Oregon", yf>»rmj/ of American Folk-
lore 20 (1907), 33^9 [Reprinted in The Collected Works of Edward Sapir. vol. IV. pp. 297-313). and
"Notes on the Takelma Indians of Southwestern Oregon", American Anthropolof^tsl n.s. 9 (1907).
251-27.^ [Reprinted in The Collected Works of Edward Sapir. vol. IV, pp. 267-291 1.

^ Sapir majored in (lermanics; he received his master's degree in the spring of l^X).s. Sec Stephen O.
Murray - Wayne Dynes, "Edward Sapirs Coursework in Linguistics and Anthroptilogy".

ffisiorioi^raphia Lini^iiisfica 13 (19<Sh), 125-129.

■* Modern Thilology .S ( 1907), 109-142.

5 Note also the conclusion of the article, where Sapir speaks of the •fundamental projvrius .<( l.m
guage" (p. 142).

''The link established between the meanings "holy" (Hebrew root B•^p) and "set apart" (p. 12V) should

be attributed to Herder.

■^ Contrary to what is stated by Regna Darnell. Edward .Sapir l.ini;ui\i. Anihmpol.-csi Humantst

(Berkeley! 1990), p. 1 1 ("the inclusion of Hskimo and Indian examples, which muM Iv .i' !«»•»

years of study with Boas") and p. 12 ("Americanist examples supplement H.r.l.r s s cn to
demonstrate the grammatical complexity of all languages").

58 General Linguistics I

Indian languages" (pp. 129-130), to the "complexity" of "the Eskimo verb"
(p. 130), and to "startling cases of linguistic conservatism [...] found among
certain primitive peoples, such as the Eskimos" (p. 134). However, no con-
crete examples are adduced, and we should also keep in mind that Herder
himself had referred to American Indian languages. It may therefore be
historicallv incorrect to claim a high amount of Boasian influence in Sapir's
master's thesis. Neither does the fact that Sapir shows familiarity with the
Humboldtian trend (Humboldt, Steinthal, Haym)*" constitute conclusive
evidence for strong Boasian influence: in any linguistic-philosophical ana-
lysis of Herder's 1772 texf and its reception, mention had, and still has to
be made of the relationship of Herder's text to Humboldt's writings on the
nature of language and the diversity of language structures, as well as to
Heymann Steinthal's classic Der Ursprung der Sprache (first edition,
1851),"* while Rudolf Haym's two-volume work (1880-84)" remains an
indispensable source-book on Herder's life, his intellectual background, his
hterary and philosophical contacts, and on the writing-history and publica-
tion of Herder's texts.

Sapir's analysis of Herder's text'- is basically a linguistic one, in that Sapir
reflects, as a hnguist, on the theses and (pseudo-)arguments of Herder.
Sapir does not approach the text in its philosophical dimension, as this was
done by Carl Siegel in a book which appeared in the same year as Sapir's

^ On the affinities between Sapir's linguistics and Humboldt's philosophy of language see Emanuel J.
Drechsel. "Wilhelm von Humboldt and Edward Sapir: Analogies and Homologies in their Linguistic
Thoughts", in William Shipley (ed.). In Honor of Mary Haas: from the Haas Festival Conference on
Native American Linguistics (Berlin/New York, 1988), 225-264; Jon Erickson - Marion Gymnich -
Ansgar Nunning, "Wilhelm von Humboldt, Edward Sapir, and the Constructivist Framework",
Historiographia Lingiiistica 24 (1997), 285-306.

■^ Herder's Abhandliing iiber den Ursprung der Sprache, which won the 1770 contest of the Berlin
Academy of Sciences, first appeared in print in 1772 (Berlin, published by C.F. Voss). It was reprinted
in the two editions of Herder's collected works: Sdmmtliche Werke (ed. by Johann von Miiller;
Karlsruhe, 1820-29) and Herders sdmmtliche Werke (ed. by Bernhard Suphan: Berlin. 1877-1913). Sapir
used both the original 1772 edition and the one in volume 5 of Herders sdmmtliche Werke (see p. Ill);
his page references are always to the 1772 edition.

1** Heymann Steinthal, Der Ursprung der Sprache, in Zusammenhang mit den letzten Fragen alles
Wissens. Fine Darstellung der Ansicht Wilhelm von Humboldts, verglichen mit denen Herders und
Hamanns (Berlin, 1851, second ed. 1858, third ed. 1877, fourth ed. 1888). Steinthal's work was written
as a response to Friedrich Schelling's call (in 1850) to reopen the question addressed by Herder in his
prize essay.

11 Rudolf Haym, Herder nach seinem Leben und seinen Werken dargestellt (Berlin. 1880-84. 2 vols.).

12 The literature on Herder and on his Abhandlung iiber den Ursprung der Sprache is extensive. The
two main bibliographical instruments are: Gottfried Gunther - Albina Volgina - Siegfried Seifert,
Herder-Bibliographie (Berlin, 1978) [Sapir's master's thesis is hsted there, p. 513, as nr. 3713], and Tino
Markworth, yo/jaAj/2 Gottfried Herder A Bibliographical Surx'ey 1977-1987 (Hurth-Efferen, 1990). On
the occasion of the 175th anniversary of Herder's death an international colloquium was held in Berlin,
the proceedings of which constitute an important reference work: see H. Scheel {ed.}, Johann Gottfried
Herder Zum 175. Todestag am 18. Dezember 1978 (Berlin, 1978).

One: The f'rohirm of ihc ()ii\^in oi /,<//; v/«/i,'f 59

article. In lad. ihc philDsophical and acsihciic tlmK-iiMnii ot Herder's Icxl
has altraclcd iiuich inoic attcnlioii than ihc properly linguistic conicni. as
can been seen Iroiii ihc lonu list oi "cxciiclicar" articles on Henler's
Ah/idndliing, starling with W.ll. Jacohi (\71}) ■ up to Albrecht tind
Matuszewski (1978),'- Franck ( l^)S2-3)" and (iaier (19KX). Sapir's analysis
offers an interesting approach to Herders Ah/uindluni^ from a linguistic
point of \ic\v.

Sapir's analysis testifies to his philological background: he offers a
"close reading" of Herder's text, of which the RiuUiktionsf^csihuhif is first
recounted, followed by a brief sketch of the intellectual context. Ilic
choice of Herder (and not Grimm. Humboldt, or Steinthal) as a landmark
in the history of reflections and debates on the "origin of language" —an ever
fascinating theme""— is made clear from the beginning: Herder, reacting
against the views of Sussmilch," wrote a "pathlinding work" (p. 112), pre-
cisely by introducing a turn in perspective. As a matter ol lact. Herder
paved the way for the modern view on the problem of the origin of lan-
guage, in that he linked language with the specificity and the evolution of
humankind. More specifically, the insistence on language as a (biological)
faculty which has undergone a gradual evolution is taken In Sapir to be
Herder's main contribution: Herder was responsible lor shifting the issue
from a conservative (be it orthodox, rationalist, or even materialist) settmg
to a modern (biologically informed) one. With hindsight, we could say that

•-^ See Carl Siegel. Herder als rhilosoph (Sluttgarl/Bcrlin. 1^07). pp. .^7-4.^.

'•' Willielm Heinricti Jacohi. ' Bctractitung ulier die von Herrn Herder in seiner .-Xbhandlung M>m
Ursprung der Spractie vorgelegte genetische Erklarung der Iliierischen Kunslferligkeilcn und

Kunsttriehe". Der Teuische Sh-rkitr i7>3. vol. 1. lasc. 2. pp. W-121.

''' Erhard Albrechl - Jt)zel Matuszewski, "Herder Qber den I'rsprung und das Wi^. n >!' i Spr.ulu-"
Deutsche Zeitschrift fiir Philosophie 26 (1978). 1297-1300.

"' Luanne Frank. "Herder's 'Essay on the Origin of Language". Forerunner of conlcmporar> view* in

history, aesthetics, literary theory, philosophy". Forum lAnf^uisticum 1 ( 19K2 3). 1.^ 26

'^ Ulrich Ciaier. Herders Spriuhplulosoplue und irkenntiuskritik (Slutlgarl/Bad < .mnsi.m l^>S.Si. cm>
pp. 75-156.

'** The topic of the origin(s) of language has given rise to a considerable literature. !• •••

survey (with several thousands of titles), see Ciordon W. Hewes, /.<//iv"<'>.'»" tfriam^ \ '>c

Hague. 1975. 2 vols.). ¥ox a survey of theories on the origin of language, sec i! '♦

in Joachim ( Jessinger and Wolfert von Rahden (eds.). Thcoru-n v<»m Lrspruny ,; \ *

York. 1989, 2 vols.), ["he best monographic treatmenl is still James H. Slam. Inquines mio ihf tfngm of
Luiii^ucii^e. ihc fate of u question (New Y«)rk, 197h).

''' Johann Peter Siissmilch. Versueh ernes Heweises. dass du- crsic Spracht ihrrn Inrmng nnhi «•#»!
Menschen. sondern allein vom Schdpfer crhallen fuihe (Berlin. 1766).

60 General Linguistics I

Herder paved the way for the interdisciplinary research^" — profiting from
insights provided by linguistics, biology,-' anatomy, psychology, chemistry
and semiotics — which characterizes present-day investigations on the ori-
gin of language. --

Having singled out Herder as a turning-point, Sapir sketches the intellec-
tual background with respect to the problem of the origin of language. He
deals with the three prevailing 18th-century doctrines (pp. 112-115), viz.
(i) The "divine origin" view, called the "orthodox" view, which was spe-
cifically held by Siissmilch, who was the target of Herder's criticisms
against this "God's gift" view;"
(ii) The contract-theory, defended by Rousseau {Discours sur Vohgine et
les fondements de Vinegalite parmi les hommes, written in 1753; Essai
sur Vorigine des (ungues, oil il est parle de la melodie et de limitation
musicale, written between 1749 and 1760, and posthumously pubhshed
in 1781), where language is viewed as resulting from a mutual agree-
ment within society. Sapir refers to this theory as the "rationalist" view
(a rather unfortunate designation for Rousseau's general stand);
(iii) The sensualist theory of the origin of language, associated with the
work of I'abbe Condillac {Essai sur Vorigine des connoissances humai-
nes, 1746), which views language as originating in instinctive expres-
sions, which are analyzed and systematized by the (developing) human
mind.-^

The shortcomings — circularity, insufficiency of empirical evidence, lack of
historical perspective — of these doctrines were noticed by Herder; Sapir

20 The importance of interdisciplinary orientation is apparent from the various contributions contained
in JiJrgen Trabant (ed.). Origins of Language (Budapest, 1996), in which scientists from different fields
approach the problem of language origins, and from the state-of-the-art discussion in Guy Jucquois.
Pourqiioi les hommes parlent-ils ? L'origine dii langage hiimain (Brussels, 2000).

21 On the importance of "allometric" studies in evolutionary biology see, e.g., Emmanuel Gilissen,
"devolution du concept d'encephalisation chez les vertebres" and "L'encephalisation chez les prima-
tes", in Guy Jucquois and Pierre Swiggers (eds.), Le comparatisme devant le miroir (Louvain-la-Neuve,
1991), pp. 85-100, 100-117.

22 See, e.g., Glynn L. Isaac and Alexander Marshack, "Origins and Evolution of Language and Speech",
Annals of the New York Academy of Sciences 280 (1976). 275-31 1.

23 On Herder's criticisms of Sussmilch's views, see: James H. Stam, ox. [note 18], pp. 115-116, 127-128,
and Bruce Kieffer, "Herder's Treatment of Sussmilch's Theory of the Origin of Language in the
'Abhandlung uber den Ursprung der Sprache'. A re-evaluation", Germanic Review 53 (1978), 96-105.

2'' Sapir also refers (p. 118) to the French scientist and philosopher Pierre-Louis-Moreau de Maupertuis
[1698-1759], whose lectures at the Berlin Academy of Science constituted the starting-point of the dis-
cussions on the origin of language in the 1750s and later. For a study of Maupertuis's views on the ori-
gin of language (set out in two pamphlets, published in 1748 and 1756), see Pierre Swiggers,
"Maupertuis sur l'origine du langage". Studies on Voltaire and the Eighteenth Century 215 (1982),
163-169.

Onr The rrohlcni of ilw Orti^in nj Language 61

pills ihcni III proper (linmiislic) pcrspcclivc. aiul Jdcs iinl hcsilalc ti) qual-
ity thcni as ■■ridiciiU)iis'" or ■inraiitilc."

Herder's apprt)aeli \shieh ilsdl is enliealK evaluated In Sapu
then analyzed in greater detail (pp. I JS 136). Sapir starts from Herder's
basic thesis: language did not originate in emotional erics (in fad. man
being inferior in instinctive power to other animals, one could hardly
explain the specificity of human language while advoeating .m insiinctiNc
origin of language), but it i)riginated wiiinn the "larger sphere" in which
humans live and operate. While being at birth the '"most helpless" animal.
man is characterized by his capacity of attention, his propensitv to\\ards
diversity (and, at the same time, universality), and his higher svmbolic
power.-" The menial disposition underlying this (capacity of) complex
behaviour, is called Bcsonnenhcit by Herder (Sapir translates the term as
"reflection:"" one could also propose "pondering"). Although this
Besonncn/wii is a divine gift, it develops within, and with, the human spe-
cies as a historical phenomenon.

This Bcsonnenhcit proceeds, in Herder's view, by singling out fragments
of experience, i.e. (experienced) properties, primarily auditory impressions;
these "sounding actions," emanating from an animated nature, are first
echoed, and later systematized in the formation o'i grammatical categories.
Sapir shows himself extremely skeptical with regard to the thesis of an "ori-
ginal singing-speech" and of the chronology of the various parts of speech
(in fact, such hypothesized chronology was rather typical of ISth-century
philosophy of language and grammatical theory): in Sapir's opinion, there
was no differentiation in word-functions at an early stage. Sapir also raises
an interesting "relativistic" issue: on what griiunds can we claim that (lin-
guistic) symbolization relates to a fragment of an experience (and on what
basis can such a part be delimited), and not to the total experience itself .'
This is a major philosophical as well as linguistic problem. In part. Herder
avoids this problem, by positing the cenlralit\ o{ hearing. ' as being inter-

ns Cf. Sapir's terms: •erroneously" (p. 124). "with a gram of salt" (p. 124). "mere spcculahon" (p IN'
"wildest and most improbable fancy" (p. 124). "antiquated and suhjcctivcly contused p»ychol.'

127). For Sapir's appreciation of Herders style, see p. 1.^7 of his article.

^^ On this topic, see Paul Salmon. "Herder's Essay on the Origin of l^ngu.if. .uul ihi- PLicc of Man m
the Animal Kingdom ". (icrman Life ami Letters n.s. 22 (l«^6H-l>9). 5y-7()

2^ See. e.g.. Wallace Chafe, "l-anguage as Symboli/ation". Luniiiiaiie 4.^ ( l«M7)
sic statement of the philosophical problem, see Willard \ <> ( )imu- \\,>r,l .;■
[Mass.], 1960).

-'^ Herder gives six reasons for this cenlralitv (cf. .Sapir's article, pp 126-127).

62 General Linguistics I

mediary between the other senses, and by postulating synaesthetic proces-
ses, which characterize sensation, apperception, and subsequent symboHza-
tion.-"

ITie interaction of the various senses, characterizing perception and sym-
boHc expression, pervades — in Herder's view — the primitive and original
languages;"' in these, imbrications and mutual transitions between the
various senses abound, and metaphor plays an important role. Again. Sapir
— as a linguist (and "modern semasiologist", p. 128)^' — expresses strong
reservations: metaphor is not specifically typical of older language stages,
and instead of explaining the history of the lexicon by metaphorical pro-
cesses, it is better to admit "an indefinite number of gradual semantic trans-
mutations" (p. 128; italics ours).

As noted by Sapir, Herder's account leaves little room for grammar at
the original stages of a language, but — what is more problematical — it
hardly can provide insight into the "growth" of grammar (or of a grammat-
ical system), although the assumption of such a growth is a central hypo-
thesis in Herder's account. To this, Sapir opposes the view of grammar as
developing "from within."

Sapir then proceeds to a discussion of the second part of Herder's text.
Whereas in the first part of his text. Herder had answered the question of
the possibility of the human origin of language, in the second part he tack-
les the question of the path along which language has (or would have)
developed. Here Sapir limits himself to mentioning the four natural laws
which Herder had formulated in order to account for the development of
language:

(a) language undergoes growth in the individual;

(b) language undergoes growth in the family (or: in the cultural stock);

(d) the growth of language is continuous throughout the human race and
throughout human culture.

29 Sapir does not go into a discussion of Herder's epistemology; for a penetrating study, see Marian
Heinz, Sensiialistischer Idealismiis: Untersuchimgen zur Erkenntnistheorie des jiingen Herder
(1763-1778) (Hamburg, 1994); cf. also U Gaier. ox. [note 17], pp. 61-63, 81-82, 167-168, 191-194.

^0 Herder presents five criteria allowing to characterize and identify "original languages;" see Sapir's
discussion, pp. 127-130.

^' The term semasiology was frequently used in the late 19th and early 20th century, especially by
Germanic and Romance philologists, with reference to research in (historical) semantics.

32 According to Sapir. this is the most interesting of the four natural laws. On the issue of dialectaliza-
tion, see also Sapir's Language (New York, 1921), pp. 159-164, 184-193.

One: The Prohlcni of flu- Orit^in <>t I .alienage 63

With ihc lourlh law. 1 krdcr in spiic i)l personal hcsilalion-- adheres to
ihc thesis ol linguistic monogenesis (based on the general similarily of
grammatical structures): Sapir does not conceal his disagreement \s\\h
Herders lightly lormulated conclusit)n and his patent neglect ol historical
perspective.

llie last pages ol Sapir's article deal with the icn.piM.ii lespeciallN tn
J.G. Hamann) ' ol Herder's text and its impact on later scholars, and ihey
contain a brief history ol the language origins theme. I iere \se have lo note
Sapir's reliance on secondary literature, ' especialK the works of
Nevinson/' Hayni.'" and Steinthal.

Sapir's master thesis, while testifying to his broad linguistic interests and
to his fascination with the basic problems in the study of language, is a
theoretically modest — and moderate — contribution. His critical rem.irks
on Herder, and on Herder's predecessors, are well-taken, but not \er\
innovative. Also, Sapir does not go into the intricacies of the relationship
between Herder, Humboldt and (neo-)Humboldtian linguistics, and his
treatment of the late 19th-century "naturalistic"" views is confined to a
brief mention of Friedrich Max Muller (and the criticisms lormulated b\
William Dwight Whitney ').

Sapir does not offer a methodological contribution to the debate ^m the
origin of language, as was to do Otto Jespersen^" by distinguishing between

•^■' On Hamann's reaction, and Herder's "conversinn" see J. Slam. o.c. (nute 1S|. pp. \y\ -170.

^•* A cursory reference is made to Friedrich I.aucherl. "Die Anschauuneen Herders utx.r den t'rsprung
der Sprache, ihre Vorausset/ungen in der Philosophic seiner Zeit und ihr lorlwirkcn". l.iiplu<n,>n
Zeii.schrift fiir Literatur^esvhichie 1 (1894). 1A1-11\. with respect t() Herders p.isiii..ii m ilu tii«>i..r\ of
linguistics. In Lauchert's article the influence of Herder is slightly overrated.

- Henry W. Nevinson. A Sketch of Herder and his limes (London. 1884).

'''R. Haym.ru-. [note ll|.

^'' H. Steinthal. (u. [note 1()|.

^'^ For a comprehensive study of "naturalistic linnuistics" in the 19th century, sec Picl !)<.
guistiijue naturidiste en frtinee USh7~l^2J). Muntrr. orifiine el evohilutn tin /.j/h-.u-. iI v

1996): in this book the views of August Schleicher |lS21-18(vS|. Honore ( 1 >•

Friedrich Max Muller | I82.^-19(K)| on the t)rigin of language are discussed m .1 , , *'

119-125), as a preliminary to the extensive analysis of the ideas of their French lollowcrv.

''* Whitney's criticisms of Max Miiller's views on language and relign>n w> ''

American Review 1 (K) ( 1 865 ). 565-58 1 . 11 .M 1 «7 1 ). AMi^} 1 . 1 1 9 ( 1 874 ). M > ^

(1870). 242-244; thev are brought together in his Max Muller ami the Scienct of Ltttfiua^ A tniHum

(New York. 1892).

■"' Otto Jespersen. I'ronress in l.any;iuii;e. With special reference u> I nnltOi (I ondon/New Yofk. I8»4|.
chapter 9 ("Origin of Language").

64 General Linguistics I

the origin of language and the origin of speech. Present-day research should in
fact take into account the more subtle distinction between the origin of language
and the origin of speech," although one cannot but note that very often dis-
cussions on the origin of language blurr the distinction between concepts such as
"communication," 'language" and "speech". Nevertheless we should credit Sapir
with having grasped the necessity of appealing much more strongly to (psycho)-
biology and to (comparative and typological) linguistics, for further reflection
on the origin of language, not in order to "solve" the problem, but with the pur-
pose of putting it in proper perspective and context. "Despite Max Miiller, how-
ever, it seems to me that the path for future work on the prime problems, more
especially the origin, of language lies in the direction pointed out by evolution. A
new element, the careful and scientific study of sound-reflexes in higher animals,
must now enter into the discussion. Perhaps this, with a very extended study of
all the various existing stocks of languages, in order to determine the most fun-
damental properties of language, may assist materially in ultimately rendering
our problem more tractable" (p. 142).

Finally, while it may seem that this article — maybe because of its "compul-
sory" academic raison d'etre — stands apart in Sapir's scholarly production, one
should not forget that

(1) in his Language, "^^^ Sapir also discusses, though very briefly, the problem of the
origin of language;

(2) already in his master's thesis Sapir proposes the definition of language"^ that
he would later use in his general-linguistic discussions of "language" (in 1921 and
in 1933);

(3) in dealing with the problem of the origin of language, Sapir hit upon his dee-
pest and most pervasive linguistic interest: the history and variation of language
[see the writings reprinted here in section II].

Pierre Swiggers

•*! See the arguments advanced by Thomas A. Sebeok, "Signs, Bridges, Origins", in Jiirgen Trabant (ed.). Origins
of Language, o.c. [note 20], pp. 89-115.

^'^ Language (New York, 1921), pp. 4-6.

^^ See p. 109: "the communication of ideas by means of audible, secondarily by means of visible, symbols."
Compare Language (New York, 1921), p. 7: "Language is a purely human and non-instinctive method of com-
municating ideas, emotions, and desires by means of a system of voluntarily produced symbols;" and
"Language" {Encyclopaedia of the Social Sciences, 1933, vol. 9. p. 155) [reprinted here in section VI]: "In the
first place, language is primarily a system of phonetic symbols for the expression of communicable thought
and feeling."

HERDER'S ''URSPRUNG DER SPKACHK"

In 1769 the Berlin Academy gave out as the nubject of a com-
petitive prize essay the following, couched, as was customary, in
French: *'En supposant les hommes abandonn<^^s k leurs facult^^ij
naturelles, sont-ils en 6tat d^nventer le langage? et par quels
moyens parviendront-ils d'eux-m6mes k cette invention?" There
are two points in the wording of this theme which are interesting
and characteristic of the time. In the first place, it will be
noticed that the pivot of the discussion was to be this: whether
language was of natural — i. e., human — or supernatural origin.
It could by no means be taken for granted then, as normal in-
vestigators would nowadays, that the gift of speech was acquired
by man in a purely normal way; the burden of proof, in fact,
lay upon those who disputed the divine origin of language. In
the second place, the query of the academy speaks of language
as an "invention." Just as a machine is a tool, a means for
bringing about certain desired mechanical effects, so language
was looked upon as a tool, a means for bringing about certain
other desired mechanical effects — namely: the communication of
ideas by means of audible, secondarily by means of visible, sym-
bols. And since history and experience showed, or seemed to
show, that machines were "invented" by the application of cer-
tain powers of intelligence, the logic of parallelism seemed to
require that also that most admirable tool called language should
have been the "invention" of some intelligence. The only ques-
tion, then, was this: Was the human mind intelligent and
resourceful enough to invent so fine a machine, or did the latter
require the master-hand of the Deity? Voila tout.

The attitude of a modern linguist toward the projK^sed subject
is certainly very different from that of the eighteenth-century
philosopher. To the first half of the question he would unhesi-
tatingly answer "Yes;" to the second he would reply: ''Language
was not invented in any true sense at all;" or, as Topsy would

jOgi [Modern PniuJixxiT. July, 1907

^(^ General Linguistics I

put it: "It wasn't born, it growed." It is in these two points,
after all, that the chief progress from the older to the modern
view of the question lies; and for both of them, the doing away
with the conception of divine interference, and the introduction
of the idea of slow, but gradual and necessary, development from
rude beginnings, we are very largely indebted to Herder. The
very answer that Herder gave to the question posed made the
question itself meaningless; henceforth there could be no serious
and profitable discussion of the divine origin of language, while
the crude conception of the "invention" of a language had to
give way more and more to that of the unconscious, or, as we
should perhaps say now, largely subconscious, development of
speech by virtue of man's psychic powers. The question resolved
itself into another: Just what factor or factors were most promi-
nent in that exceedingly slow process of mental evolution that
transformed a being giving vent to his emotions in inarticulate
cries to one giving expression to a rich mental life by an elaborate
system of auditory symbols? Despite the vast accumulation
of linguistic material that has been collected since Herder's time,
and the immense clarification that has been attained in linguistic
conceptions, processes, and classifications, we cannot today make
bold to assert that this problem is satisfactorily answered, or
apparently in a way to be satisfactorily answered in the immediate
future. Bearing this in mind, we shall be able more justly to
value the great service Herder accomplished in merely shifting
the point of view. That alone was an inestimable service.

It was to be expected that the proposed subject should appeal
strongly to a mind of Herder's stamp, occupied, as it was, with
problems touching the most important phases of human culture.
We thus find him, while still in Nantes, writing to his publisher-
friend Hartknoch that he was intending to work up the theme the
following year. He speaks of it as "eine vortreffliche, grosse und
wahrhaftig philosophische Frage, die recht fur mich gegeben zu
sein scheint."' The latest time at which the competing essays
could be handed in was January 1, 1771; yet Herder did not set
to work at the actual composition of his treatise until well on in

1 R, Haym, Herder nach aeinem Leben und seinen Werken dargestellt. Vol. I, pp. 400-403.

110

One: The Problem of (he Orifjin of I.anf'uaf^e 67

December, 1770, when his Strassbiir^ jxTicKi was druwin^' Ui a close.
To excuse the peculiar defects of it, lio wrote to Nicolai early in
1772 that it was written "flQchtig, in Kile, in dm l.-t/t.-n Ta^on
des Decembers.'" So rapidly, indeed, was this Uj()-pago' cKnay
composed that it was finished even before Christmas and waa Bent
away anonymously, with accompanying billet, to Tourney, the
secretary of the academy. This almost incredible rapidity of
composition can be explained only by assuming, as we have every
reason to do, that Herder had thought out the whole problem in
considerable detail long before, or, as Suphan suggests, oven
made the first rough draft' at Nantes, and had now simply to
mold these ideas into literary form. Although few would ven-
ture to call the Preisschrift a model of literary form, yet the dis-
tinctness of the theme and the short time in which it had to be
got ready give the treatise "a directness and ease that is too often
wanting in his other works."* In few other subjects had Herder
been so deeply interested up to 1771 as in this one concerning
the origin and development of language. As early as 1764,
according to Suphan, he had drawn up a plan for the somewhat
elaborate investigation of the origin of language, writing, and
grammar. One of his contributions to the Riga Gelehrte Anzei-
gen dealt with the problem, "Wie weit alte und neue, fremde
und die Muttersprache unsem Flciss verdienen" — an essay that
anticipates some of the striking phrases found in the Preisschrift.
In his first important work, the famous Fragmeiiie, Herder had
already given expression to some of his later thoughts, among
other things maintaining the human origin of language. When
Stlssmilch's Beweis, doss die Sprache gdttlich sei, which had
been read at the Berlin Academy ten years before publication,
appeared in 1766, Herder was deeply interested, and wrote to
SchefiPner (October 31, 1767) : "Da SQssmilch sich in die Sprach-
hypothese neulich gemischt und es mit Rousseau gogen Moses
[i. e., Moses Mendelssohn] aufgenommen, so hatto ich wohl Lust,

' Haym, op. cit.

'As contained in Herder's Collected Workt, edited by Johnnn von Mollor (Carlsmb*.
1820).

3Distineruished by Suphan as "a." Soo p. xii. Vol. V. of his rnlition.

* Novinson, Herder and His Times, [> 162.

Ill

58 General Linguistics I

auch ein mal ein Paar Worte offentlich zu sagen." ' With what
had been said on the subject by Rousseau, Condillac, Abbt, Lam-
bert, and others, Herder was well acquainted, so that his own
Preisschrift^ while in every sense a pathfinding work, takes a
definite historically conditioned place in the linguistic-philosophic
speculations of the eighteenth century.

Before proceeding with the detailed analysis of Herder's
epoch-making work, we must briefly consider the theories on the
origin of language which prevailed at the time he wrote it. By
far the most commonly held theory, at least in Germany, was that
supported by Sfissmilch, the orthodox view. According to this,
language was given or revealed to man by God. The power to
create the subtle mechanism of speech was considered by the
supporters of this theory beyond the earliest human beings ; they
had to receive the first rude concept of language, the first fruitful
suggestions, at least, from without. In earlier stages of linguis-
tic speculation, particularly at the time of the Reformation, it
was believed, on inferred biblical evidence^ that this earliest God-
made language was the Hebrew tongue, from which all other
idioms, the Greek and Latin as well as the Chinese, were derived
by processes of corruption, transposition of letters, or what not.
In Herder's day, however, it was no longer considered necessary
by all supporters of the orthodox view to maintain the absolute
primitiveness of Hebrew, although Hebrew was regarded, among
others by Herder himself, as a peculiarly primitive or "original"
language. Many deemed it sufficient to assert the revelation to
man of some form, however imperfect, of speech, and were willing
to concede the possibly somewhat late advent of the Holy Tongue.
We can easily understand some of the reasons that led to the sup-
port of this, it is needless to say, now wholly antiquated view.
In the first place, scriptural evidence, in general, seemed to imply
the divine origin of language ; although we are told that the Lord
brought the various denizens of the field and forest before Adam,
that he might give them names, still this appears to have been
done under careful paternal supervision. In the second place,
there was good, in fact irrefutable, evidence, from an orthodox

•See Haym, op. ctf., Vol. I, p, 402.

112

One: flic I'rohlcni of (lie Origin of I (iniiiuifie 69

point of view, for the divine authorship of the indu.strios; and it
seemed illogical to sup^wse that a much higlior factor of civiliza-
tion than the industries, namely language, liad Ihm'ii l«-ft to the
ingenuity of primitive man. In the third place, the less than bix
thousand years which had elapsed since the creation of the world
were looked upon, and with good reason, as quite inadequate for
the development from the crudest |x)ssible beginnings of our
modern, richly organized languages. Even long after Herder
had demonstrated the untenableness of the orthodox theory, many
scholars still clung to a view which made God, as Goethe put it,
"a kind of omnipotent schoolmaster." I note, by way of illus-
tration, that our Noah Webster still considered it the most prob-
able explanation. From a psychological point of view, granting
the possibility of revelation, the theory is, of course, absolutely
useless. Its advocates do not seem to have perceived that the
imparting of a language to the first speechless human beings,
accompanied, as it presumably was, by grammatical instruction,
must have been a fairly impossible task, implying, in fact, lin-
guistic training in the recipient; moreover, the theory begged the
question, by assuming the existence of what it set out to explain.
The modern critical standpoint has been well, if somewhat cyni-
cally, formulated by Fritz Mauthner in his entertaining Kritik
der Sprache:

Wir wissen kaum, was der abstrakte Begriff Sprache bedeutpt, wir
wissen noch weniger, wie wir den Begriff Ursprung zcitlich Ix'grenzen
sollen, wir wissen gar nicht mehr den Gottesbegriff zu definieren; da
kOnnen wir mit dem "gOttlichen Ursprung der Sprache" wirklich nicht
mehr viel anfangen.'

A second theory, supported notably by Rousseau and the
German Rationalists, was very similar in character to the contract-
theory of the origin of government, also held by Rousseau. They
conceived the matter approximately thus: Primitive men, after
having long been compelled to get along without B{>oech, at last
awoke to a consciousness of the manifold inconveniences of their
then condition; were in particular troubled by the important
problem of communicating ideas. To remedy, if |M)s.sil)le, this

I Fritz Mauthnor, Kritik der Sprache, Vol. II, " Zur Sprachwissonschafl." p. 3M.

113

70 General Linguistics I

deplorable state of affairs, our primitive ancestors, or perhaps
only the wisest of them, put their heads together to devise ways
and means for the more practicable interchange of thoughts.
After much cogitation — not deliberation, for language was not as
yet — they hit upon the excellent device of representing things
and actions to each other by means of arbitrarily chosen symbols,
presumably auditory. Henceforth they had no difficulty in under-
standing each other, civilization progressed more rapidly than
heretofore, and all was well. One is amazed to find that men in
the eighteenth century were willing to maintain so ridiculous a
theory, even if not presented in quite so absurd a light as above.
It is not difficult to point out the vicious circle implied therein.
Man could not conceivably have advanced so far as to perceive
the advantages of speech as a means of communication without
already being possessed of it ; on the other hand, if primitive man
could already successfully communicate such abstract ideas as
those of symbols, one fails to see the necessity of a change in
method.

A far more valuable theory than these two was that held by
the English and French "naturalists," though a crude, mechan-
istic psychology makes their speculations often seem rather infan-
tile today. The "naturalists," generally speaking, were inclined
to look upon language as a reflex, expressed in cries, of the sen-
sations and perceptions imprinted upon the human mind by man's
environment. They considered the growth of a vocabulary abso-
lutely co-ordinated with the growth of experience, and were pretty
sure, most of them, that untaught children, if isolated from the
companionship of their fellow-beings, would develop a language
of their own. Condillac, probably the most profound of the
philosophes, attempted in his Essai sur Vorigine des connois-
sances humaines to show how two human beings of opposite sex
might naturally be led to acquire speech. He supposes that at
the outset all their desires and emotions are expressed by purely
instinctive cries, accompanied by violent gestures. By the
psychic processes of memory and association these cries gradually
come to serve as the fixed means of communicating the more ele-
mentary feelings, such as fear, joy, and the like. Says Ck)ndillac :

114

One: The Problem (tf the Or'ni'ui of I un linage 71

Meanwhile, while these human Ixiin^s have acquired th«- lu'ibil of
associating several ideas with arbitrary signs, tho natural cries Kcired
them as model to make a new langujige. They articulat<!<i new wjunds
and, repeating them several times and accompanying them by some
gesture which indicated the objects they wished to call atU^ntion to, they
accustomed themselves to giving names to things. At first the prrjgreiwj
of this language was extremely slow. The organ of spfvch wan so infl»*x-
ible that it could articulate only a few very simple sounds. Tlie ol>stacle«
that presented themselves in pronouncing others even prevented them
from suspecting that the voice was capable of moving Ix-yond the small
number of words they had imagined. This pair had a child, who, pretisod
by wants that he could give expression to only with difBculty, violently
moved all the limbs of his body. His very flexible tongue curled itself
in an extraordinary manner and pronounced an entirely new word. Tho
want, still continuing, again gave rise to the same effects; this child
moved his tongue as before, and again articulated the same sound. The
surprised parents, having finally guessed what he wished, attempted,
while giving it to him, to repeat the same word. The difficulty they had
in pronouncing it made it evident that they would not of themselves have
been able to invent it.'

In snch wise C5ondillac thinks their language would be slowly
and painfully enriched; not until after many generations would a
language in our sense be approximated. A fairly ingenious
theory, emd much to be preferred to either the orthodox or the
rationalist views, yet not truly convincing. The great difficulty
that Herder found with it was the failure to draw a sharp line
between the instincts of the animals and the higher mental {>ower8
of man. One does not clearly see why, according to Cond iliac,
the animals should not have likewise developed a language.
Herder, although he inclined on the whole to the views of the
French "naturalists," attempted to avoid their shallow mechan-
istic psychology, and was chiefly concerned in showing that the
peculiarly human faculty of speech was a necessary correlative
of certain distinctly human psychic conditions.

The analysis of Herder's views here given is based, not on the
second edition of the prize essay (Berlin, 1789), but on its first
published form (Berlin, 1772), as given in Suphan's edition of
Herder. Following the formulation of the academy's theme, he
divided his treatise into two parts, the first answering th«' <]ue«-

I CEuvre* de Condillac (Paris, 1798), Vol. 1, pp. 2G4, '.Jto.

115

72 General Linguistics I

tion: "Haben die Menschen, ihren Naturfahigkeiten uberlassen,
sich selbst Sprache erfinden kOnnen?"' the second dealing with
the problem: "Auf welchem Wege der Mensch sich am fuglich-
sten hat Sprache erfinden kSnnen und mussen?"^

Herder begins his treatise with the postulation of a "natural
law." All the higher animals involuntarily respond to their emo-
tions, particularly the more intense ones, such as pain, by cries.
As Herder formulates the "law": "Here is a sentient being,
unable to inclose within itself any of its intense feelings; which,
in the first moment of surprise, must give expression in sound to
every feeling, even without intention and purpose."' But this
sentient being is not an isolated phenomenon. There are other
beings, besides itself, similarly constituted, that respond to like
stimuli in the same way. Hence the instinctive cries of each sen-
tient being find a responsive echo in other beings of like organ-
ism, very much as a vibrating string will cause other strings to
vibrate that are pitched in harmony to itself. These tones con-
stitute a species of language, "a language of feeling" directly
given by nature (unmittelhares Naturgesetz) ; its genesis it does
not occur to Herder to explain. Such natural cries are not
peculiar to the animals, but are shared also by man. No matter
how highly developed a language may be, it always includes a
number of vocables that are intelligible per se as emotional expres-
sions. These are represented on paper — with miserable inade-
quacy, as Herder strongly emphasizes — by the interjections
(such as ach, O, and so forth) ; their real existence, however, is
in their utterance in the appropriate emotional milieu. It is true
that in our modern, metaphysically refined idioms, these emo-
tional elements play a very subordinate rOle, but in the older
oriental and in the primitive tongues Herder thinks to find more
numerous survivals of the earliest linguistic conditions. It may
be noted that all through the essay Herder, quite uncritically
from our modem point of view, considers the oriental, by which
he means one or two Semitic, dialects and the languages of primi-

I Herder'* Sdmmtliche Werke, herausgegeben von Bemhard Suphan (Berlin, 1891),
Vol. V, p. 3.

"i Ibid., p. 91. 3 Ibid., p. 6.

116

One: The Prohlmi of ihc Orii^in of Luni^uage 73

tive peoples as essentially more "originar' tlian our iii.xI.tu ver-
naculars. We should never forget that Herder's lime-|)or8pective
was necessarily very different from ours. Whih; wt; unconcern-
edly take tens and even hundreds of thousands of y^'ars in which
to allow the products of human civilization to deveUjp, Herder
was still compelled to operate with the less than six thousand
years that orthodoxy stingily doled out. To us the two or three
thousand years that separate our languages from the i)U\ Testa-
ment Hebrew seems a negligible quantity, when s{)eculating on
the origin of language in general ; to Herder, however, the Hebrew
and the Greek of Homer seemed to be appreciably nearer the
oldest conditions than our vernaculars — hence his exaggeration of
their Ursj^runglichkeit. The supposedly "primitive," or rather
"original," character of the languages of savages was due to a
very natural, though, unfortunately, on the whole erroneous, con-
clusion from d priori, considerations.

Herder next proceeds to take up and refute one of Sossmilch's
arguments for believing that language is God's direct work.
Stlssmilch contended that it was evident in the alphabetic uni-
formity of all languages, pointing to an original divine simplicity;
all the sounds found in the multiform idioms of the earth, he
thought, could be adequately represented by about twenty letters.
This queer argument Herder conclusively showed to bo a mere
orthographic quibble. It is a huge fallacy, as Herder clearly
saw, to imagine that even one language could be |X3rfectly repre-
sented by an alphabet at all, let alone one of twenty letters. He
recognizes, apparently as clearly as any modern linguist, that the
real elements of language are spoken sounds of which letters are
but makeshift and imperfect substitutions. He quotes from trav-
elers as to the extreme difficulty of representing many of the
dialects of primitive peoples through the medium of our letters;
but he calls attention also to the very faint idea that one geta of
even spoken English and French from the written forms of those
languages. That a comparatively "original" language (I speak
in Herder's terms) like the Hebrew did not orthographic/iUy rep-
resent the vowels is due, he thinks, to their finely modulated,
natural, almost unarticulated character. "Hire Ausprache war so

117

74 General Linguistics I

lebendig und feiiiorganisiert, ihr Hauch war so geistig und
athemiscb, dass er verduftete und sich nicht in Buchstaben fassen
liess.'" Hence he concludes that the nearer a language comes to
the original conditions, the less possible to mirror it in ortho-
graphical symbols.

After this digression on the Sussmilch argument, Herder
returns to a consideration of the natural emotional sounds of man
and the animals: he emphasizes the great influence that these
still have emotionally, and sees in them the closest bond of union
between the various members of animated creation. ** Their
origin," Herder declares, "I consider very natural. It is not only
not superhuman, but evidently beastlike {thieriscK) , the natural
law of a machine capable of feeling."^ But — and here comes a
critical point in Herder's argument — it is impossible to explain
the origin of human language from these emotional cries.

All animals, down to the dumb fish, give vent to their feelings in
tones; but no animal, not even the most perfect, has on that account the
sUghtest genuine disposition toward a human language. Let one form and
refine and fix this natural cry as one will; if no understanding is present,
so as purposely to use the tone, I fail to see how human, conscious lan-
guage is ever going to arise. Children utter emotional sounds, like ani-
mals: is not the language, however, that they learn from men quite
another idiom ? ^

All writers, notably Condillac and Maupertuis among the
French philosophes, and Diodorus and Vitruvius among the
ancients, that have attempted to derive human speech from
instinctive animal cries, are, then, on the wrong path at the very
outset of their investigations. Since, among all living beings,
man alone has developed a language, in the ordinary sense of the
word, and since it is, after all, this power of speech which chiefly
separates man from the animals, any rational attempt to explain
its origin would have to begin with a consideration of the essen-
tial psychic differences between the two. Herder, consequently,
proceeds to seek for the most fundamental of these psychic dif-
ferences, and finds it in this, "that man is far inferior to the
animals in strength and sureness of instinct j indeed, that he lacks

» Herder, op. cit., p. 13. 2 Ibid., p. 17. 3 Ibid., pp. 17, 18.

118

Otic: The rrohh'in t>l ilic ( )ni:iii i>j I iitif^ua^i'

entirely what in eo many animal species w.- t<rin inUorn art-
capabilities or impulses {Ku7i<i{f(lhiykcil<'n und KuTisttriehe).^^^
This inferiority in instinctive power Herder ascribes to a greater
sphere of attention and activity on the part of man. Harder
finds by observation ''that the sharper the scnsos of th»' animale,
and the more wonderful their works of art, the narrower their
sphere, the more uniform their art-work;"' and inversely, "the
more scattered their attention to various objects, the more unde-
fined their mode of life, in short, the larger and more manifold
their sphere, the more do we see their sense-powers divided and
weakened.'" Now all instincts, even such complicated ones ae
those we see manifested in the construction of beehives and spider-
webs, are to be explained by the intense, specialized activities of
the senses within a narrow sphere. Hence Herder feels justified
in assuming that intensity of the senses and perfection of instinct
are in inverse proportion to the amplitude of the sphere of atten-
tion. Since man has the widest possible sphere, is the least
specialized of all creatures in his activity, it follows that he is
least endowed with inborn mechanical dispositions; in other
words, is at birth apparently the most helpless of all living
beings. It is inconceivable, however, that nature should have
acted in so stepmotherly a fashion as to intend man for the widest
field of activity, and at the same time fail to grant him powers
successfully to maintain himself in his complex environment.
There must be some psychic element which secures him his due
position in the world; if we succeed in discovering this psychic
element, we shall also have obtained the distinctive characteristic
of man; and if, furthermore, we can show the human faculty of
speech to be a resultant of this mental characteristic, our problem
is practically solved. The peculiarly human characteristic sought
is conditioned by the wide range of attention; for this latter
implies that the human senses, unrestricted to any narroNvly
specialized field, are left free for development and the acquisition
of clearer impressions. Now, this clearness of view leads to what
is variously termed understanding, reason, judgment. Herder,
who is alone to be held responsible for the psychology of all f^''»»

' Op. cjf., p. 22, 2/6tJ., p. 22. J //..■. f ,. ^?

119

76 General Linguistics 1

is indifferent about the name applied (he himself suggests Beson-
nenheit, "reflection"); he very strongly emphasizes, however,
that this Besonnenheit is not a faculty superimposed upon the
lower animal mind-elements and transforming them into the human
mind, but rather a certain disposition or aspect of the really
unanalyzable unity called the "mind." It is truly refreshing to
find Herder, in the age of neatly pigeon-holed faculties, boldly
asserting these to be but more or less convenient abstractions; to
Herder the human "mind" is an indivisible entity, in no wise
genetically related to the animal mind. As he expresses it: **Der
Unterschied (zwischen der menschlichen und der tierischen
Seele) ist nicht in Stufen, oder Zugabe von Kraften, sondern in
einer ganz verschiedenartigen Richtung und Auswickelung aller
Krafte."* Furthermore, Besonnenheit must have been present in
the human race from the very start, must have been implanted in
it by the Creator; although this, of course, does not mean that it
is not capable of growth with the increase of experience. Any
attempt, then, to bridge over the gulf separating man and the
animal world is to Herder absurd.

Does not reflection or Besonnenheit^ however, imply the inven-
tion — or, better, genesis — of language? Herder proceeds:

[Der Mensch] beweiset Reflexion, wenn er aus dem ganzen schweben-
den Traum der Bilder, die seine Seele vorbeistreichen, sich in ein
Moment des Wachens sammlen, auf einem Bilde freiwillig verweilen, es
in helle ruhigere Obacht nehmen, und sich Merkmale absondem kann,
dass dies der Gegenstand und kein andrer sei. Er beweiset auch Reflex-
ion, wenn er nicht bloss alle Eigenschaften lebhaft oder klar erkennen;
sondern eine oder mehrere als unterscheidende Eigenschaften bei sich
anerkennen kann; der erste Aktus dieser Anerkenntniss** giebt deutlichen
Begriff; est ist das erste Urtheil der Seele.'

Further, the singling out and apperception of any attribute, the
formation of a clear concept, is in itself, in the true sense of the
word, language, even though it be not uttered; for language can
very well be defined as series of associated attributes or concepts,
symbolically interpreted. For the purpose of illustration. Herder

^ Herder, op. cit. p. 29.

2 By AnerkenntnUs Herder meaas aboat as mach as " apperception."

^ Op. cit., p. 85.

120

One: T/w /'rohlcni of the Orii^in of I dUiiuage 77

supposes a sheep to pass by primitive man.' Tin* latU-r, with
mind unobscured by the wolfish instinct to tear the Hhuep to
pieces, will, by virtue of his power of Bcsonnenheit, quietly jxjr-
ceive the white, wooly phenomenon. Suddenly the shoop bloats;
involuntarily primitive man picks out this remarkable Bound as
the characteristic attribute in the complex of sensations pros<«nt«Hl
to him by the sheep. The sheep crosses his path once a«^ain.
Primitive man, not yet fully practiced in the a[)f)erception of
objects, does not at first recognize his wooly friend. Hut the
sheep again bleats, whereupon he remembers the similar sound
heard before; the characteristic attribute then, in this case the
bleating, serves to establish the identity of the two sensation-
groups. Ever thereafter the remembered audible imago of bleat-
ing will associate itself with the totality of images, visual, tactile,
and others, that go to make up the phenomenon sheep. Does not
this mean that the image of bleating becomes the name of the
object, even though the speech-organs of primitive man never
attempt to reproduce the sound? With the acquisition of a num-
ber of constant images of apperceived attributes language is now
fairly begun, and is shown to be, in Herder's opinion, a necessary
corollary of the postulated Besonnenheit^ peculiar to man.

At this point Herder takes up certain arguments advanced by
some to prove the impossibility of the human origin of language.
Silssmilch contended that without the use of language no act of
reason was possible. Consequently man, in order to reason, must
have been in prior possession of the gift of speech. But it is
impossible that he should have himself invented it; for reason,
not yet in operation, would evidently have been necessary there-
for. The only way out of the difficulty, then, is to assume that
God first taught man the use of language, by the employment of
which the exercise of reason followed. Herder has no dilficulty
in showing the circle in the argument. If man was to grasp
the linguistic instruction of God, and not Bimj)ly repeat hie
words in parrot-like manner, he must already have been iu i>o8-
session of an elaborate complex of concepts and, therefore, also of
language; for, according to his analysis, the genesis of the two is

•The illustratiou wtis borrowed by Herder from Moses Mrndclss<.>hn.

121

78 General Linguistics I

simultaneous. To seek a parallel for divine instruction in the
language-teaching of children by their parents is of no avail.
The child recognizes most attributes and acquires his store of
concepts by his own unaided experience — that is, the real acqui-
sition of language is his own unaided work; all that the parents
do is to force him to label his stock of concepts with those arbi-
trary symbols which they happen to use themselves.

So far Herder has discussed only the singling-out of attributes
in general. The question now arises which attributes are most
likely to be picked out originally as the first elements of language.
The sense of sight, Herder believes, develops at the start with too
great difficulty to allow visual attributes to be seized upon as the
characteristic ones of objects ; moreover, light-phenomena are too
"cold" (this is Herder's term) to appeal to primitive man. The
lower senses, notably that of touch, on the other hand, receive
impressions that are too coarse and undefined for the purposes of
speech. It remains, then, for the sense of hearing to give the
characteristic attributes and, as Herder expresses it, become Lehr-
meister der Sprache. Thus the sheep, as we have seen, becomes
to man the "bleater;" the dog, the "barker;" the wind, the
"rustler;" and so on indefinitely. The abstraction, then, of
sound- attributes, coupled with a mechanical and spontaneous imi-
tation of these, forms the first vocabulary of man. The biblical
sentence, "Gott fahrte die Thiere zu ihm, dass er sfihe, wie er sie
nennte, und wie er sie nennen wtirde, so sollten sie heissen,"
Herder chooses to consider a poetical, peculiarly oriental rendi-
tion of the soberly expressed philosophic truth: "Der Verstand,
durch den der Mensch fiber die Natur herrschet, war der Vater
einer lebendigen Sprache, die er aus TOnen schallender Wesen zu
Merkmalen der Unterscheidung sich abzog."*

If language were the invention of the Creator, we should
expect to see his spirit — that is, pure reason — reflected in his
work. But such is by no means the case. Pure reason or logic
would require us to seek nouns as the most original constituents
of the vocabulary of a language; for evidently, in strictly logical
reasoning, the subject comes before the predicate, the thing

1 Herder, op. cit., p. 50.

122

One: I'hc I'rohlcni <>l the Orti^m <>j I iiiii^uage 79

that acts before the action. Ah a iiiattt-r of fn<--t, liowfvwr, the
radical elements of lantruages are nni substantive in chnractor,
but verbal; this we can explain, if we bear in mind that primitive
man was most impressed by sounding actions (iOnemle Htuuilun-
gen). Since these actions were manifested in certain objeels, it
followed that the latter were named by the same natural soundH as
the former; thus we have nouns developing out of vei-bs, and not
vice versa. "The child names the sheep, not as sheep, but as a
bleating animal, and thus turns the interjection into a verb."'
All old and primitive languages clearly show, Herder is very
sure, the verbal origin of nouns, and a philosophically arranged
dictionary of an oriental language would be "a chart of the
course of the human spirit, a history of its development, and
. . . . the most excellent proof of the creative power of the
human soul."' It is somewhat strange to find as keen a mind as
Herder's occupying itself with so useless and at bottom meaning-
less a problem as the priority of the parts of speech. It goes
without saying that in the earliest period of language-formation
there could not have been the slightest differentiation in word-
functioning. Making use of Herder's favorite example, there is
no reason to suppose that the remembered audible sensation
** bleating" should originally have had more reference to the
action of bleating than to the sheep itself. We shall probably bo
nearer the truth if we assume that the word made in imitation of
bleating was employed to signify all that had reference to the
remembered phenomenon. The word, which we may assume for
the sake of argument to be "baa," might in niod(>rn terms signify
"to bleat," "sheep," "wooly," or what not; only we must beware
of imagining that "baa" had any clearly defined grammatical
function. Herder speaks of the sheep as "ein blcickt-yidvs
GeschOpf ;" noting that blOckend is a verbal form, he concludea
that the verb was the original part of s|>eech.

Nature, Herder proceeds, appears to man as n resoundinf^
(tOnend) whole; hence man infers that nature is aniniat^^d. living,
and personifies all the phenomena j^resented to his conaciousneBft.
By this peculiarly human tendency to vivify the inanimate and

> Op. cit., pp. 52, 53. ^ If'Ki ■ P ''■'

123

80 General Linguistics I

relate to his own experience the vast sea of extra-human phe-
nomena can be easily explained the most primitive religions, the
grammatical category of sex-gehder, which Herder, erroneously
of course, seems to imagine is particularly widespread, and, above
all, the genesis of poetry. For what was this earliest language,
imitating the sounds of living nature and expressing ideas in vivid
imagery, but poetry ? Furthermore, this language was song, not
learned, as Herder well shows, from the birds, but "song, that
was as natural to him, as suited to his organs and natural impulses,
as the nightingale's song to herself." "All nature sang and
resounded, and the song of man was a concert of all these voices ;
in so far as his understanding needed them, his feeling grasped
them, [and] his organs could give expression to them."* That the
oldest song and poetry are derived from this primeval condition
of identity of song and language Herder considers proved; he is
inclined to look upon the Homeric poems as survivals from this
earliest time; and even today the originally musical character of
speech is attested by the accents of many savage idioms.

All of this enthusiastic speculation of Herder's on the singing-
speech of primitive man — ideas which he had already developed
in the earlier Fragmente — must now, of course, be taken with a
grain of salt. That song and poetry are among the most natural
forms of expression, and are found, inseparably linked, practically
all over the world, is now fully recognized. Moreover, we have
no difficulty in supposing that the earliest forms of language were
more expressive emotionally than now; the human voice may,
very possibly, have had a more decided pitch-modulation, have
moved at greater musical intervals, than now, and thus have pro-
duced much of the effect of song. Even this, however, is mere
speculation. But to suppose that the earliest speech was, in any
true sense of the word, melodic song, from which the vocal art of
the Greeks could be more or less directly derived, is to be con-
sidered the wildest and most improbable fancy. As to the finished
art-works of Greece being survivals from Herder's hypothetical
period of spontaneous poesy, that needs no comment here.

It is not difficult to understand how objects that have distinctive

1 Herder, op. cit., p. 58.

124

One: Tlic Prohlr/n of ihc Origin oj I an^uage 81

BOund-attributes, such as a bleating sheep, can b«« Bymbolizi'd
by sound images. But how is it witli phonomona that do not of
themselves suggest suitable audible HynibulK'/ How an* the
impressions of sight and feeling, taste and smell, to bo naturnlly
expressed in terms of auditory impressions? Herder seeks the
solution of this puzzling question in a psychologic trutli, wliich
one is somewhat surprised to see grasped so clearly in the
eighteenth century. His remarks are so illuminating that 1 can-
not do better than quote from them :

What are these properties [i. e., of sight, hearing?, etc.] in objects T
They are merely sensed impressions in us, and as such do they not all
flow into one another? We are one thinking scnsorirun rommunt-, (m\\
affected on various sides — therein lies the expliiuatiou. Fct*lin^^' undfr-
lies all senses, and this gives the most disparate sensations, so intimate,
strong, indefinable a bond of union, that out of this combination the
strangest appearances arise. There is more than one instance known to
me of persons who, naturally, perhaps from some impression of child-
hood, could not do otherwise than directly combine by some rapid mutii-
tion [we should say "association" nowadays] this color with that sound
with this appearance that entirely different, indefinite feelinj^, that, when
viewed through the light of slow reason, has absolutely no connection
therewith: for who can compare sound and color, appearance and feeliuj<f ? '

From this now experimentally well-established psychologic law of
an ever-present, at times indefinite but always real, undercurrent
of feeling, accompanying and coloring the ever-fiowing stn^nm of
sensation. Herder derives a somewhat nebulously stated corollary,
an application of the law to the genesis of language. He declares:

Since all senses, particularly in the condition of man's chiidhcKKi. art-
nothing but forms of feeling of one mind; and since all feelin^ir. according
to an emotional law of animal nature, has directly its own vocal expres-
sion; therefore, if this feeling is heightened to the int(nisity of a character-
istic mark {Merkmal}, the ivord of external language is there.-'

Furthermore :

Since man receives the laufjruas"t' of nature only l>y way of he-arinfj.
and without it cannot invent lan^mage, hearing has. in sonu* manner,
become the central one of his senses, the gate, as it wrrc. to thf mind,
and the bond of union between the other sen.ses.'

» Op. cif., p. 61. a/fctff.. p. G4. > Ibid., P.M.

125

82 General Linguistics I

That the sense of hearing does really occupy so relatively
important a place is, of course, at least questionable; but Herder,
undismayed by disquieting doubts, proceeds to give six reasons
for thinking so. First, hearing occupies a medial position
among the senses in regard to range of impressivenesss {Sphdre
der Empfindbarkeii) ; being richer than the tactile, and less over-
whelming and distracting than the optic, sense. Thus it stands
in the closest connection of any with the other senses, and is well
adapted to serve as a transformer into linguistically usable
images of the impressions conveyed through the medium of sight
and touch. Secondly, hearing is a middle sense in the matter of
clearness ; the sense of touch gives too dull, undefined impressions,
while the impressions of sight are too bright and manifold. Hear-
ing effectively heightens the former, modifies the latter. Thirdly,
hearing occupies a middle position as regards vividness (Lebhaf-
tigkeit) ; feeling is too warm a sense, lodges itself too deeply and
overpoweringly in one's consciousness, while visual impressions
are too cold and leave one indifferent. The auditory sense is,
then, for the whole mind, what the color green, as it were, is for
the visual sense — neither too dull nor too intense. Fourthly,
hearing is the middle sense in respect to the time in which it
operates ; tactile impressions are sudden and momentary ; those of
sight, on the other hand, confuse by their simultaneity. As
opposed to these, auditory impressions generally take place in
progression — not very much is offered at any one moment, but
the flow of auditory sensation is fairly continuous. Fifthly, the
images induced by the sense of hearing need outward expression.
Impressions produced by feeling are too dark and self-centered
to need such expression; the objects revealed by the sense of
sight are generally permanent and can be indicated by gesture;
"but the objects that appeal to hearing" Herder says, "are con-
nected with motion: they slip by; but just on that account they
give rise to sounds. They are capable of expression because they
must be expressed, and because they must be expressed, because
of their motion, they become capable of expression.'" This
quotation is a fair specimen of Herder's method of ratiocina-

>0p. ctt.,p. 67.

126

One: The Prohlc/n <>t ilic Orn^tn ,>f f (infituige H3

tion at times. Sixthly and finally, lH'arin«; is the middle Bcnse
in point of order of development, feeling coming before and
sight after. Thus Herder proves, apparently to Ijis own
entire satisfaction, that all the impressions of sense bocome
capable of adequate linguistic expression by way of their close
connection with the supposedly middle sense -that of hearing.
Let us not bother with an unprofitable critique of Herder's anti-
quated and subjectively confused psychology, and consider it
proved, for the sake of his argument, that all intense outward
stimuli, no matter of what description, find their natural reaponse
in vocal expression.

Herder next proceeds to lay down a series of theses on the
characteristics of "original" i^ursjiruncjlicheii) languages, the
purpose of which is to show how clearly the imprint of the human
mind is visible in them. This to us is very axiomatic, but we
should not forget that it was necessary for Herder to demonstrate
it in order to disprove the orthodox theory of divine origin. The
first of these Sdtze reads: "The older and more original language*;
are, the more is this analogy of the senses noticeable in their
roots."' Where we characterize in terms of sight or feeling, the
oriental often prefers to have recourse to the sense of hearing.
"Anger," for instance, is most commonly in later times thought
of in terms of visual phenomena, such as blazing eyes and glow-
ing cheeks; the oriental, however, finds its characteristic mark in
the fierce snort of the nostrils.' Again, "life" is to most of us
moderns best characterized by the pulse-beat, while the oriental
hears in the living breath the most salient element of animated
existence. And so on indefinitely.

Herder's second thesis reads: "The older and more original
languages are, the more do the various shades of feeling cross
each other in the roots of the words." ^ He proceeds:

Let one open any oriental lexicon at random, and one will jx»rwiv<»
the struggling desire to express ideas! How the inventor tore out ideas
from one feeling and borrowed them for another! How ho borrowed
most of all from the most difficult, coldest, cleanest .sense's! How every-

> Op. cit., p. 70.

2 Herder had in mind probably iieb. "aph " auK«r '". dual "uj.p.lvim • uoslriU.'"

^Op. cit.. p. 71.

127

34 General Linguistics I

thing had to be transformed into feeling and sound, in order to gain
expression! Hence the strong, bold metaphors in the roots of words!
Hence the extensions from feeling to feeling, so that the meanings of
stem-words and, still more so, those of their derivatives, when put next
to each other, present the most motley picture. The genetic cause lies
in the poverty of the human mind and in the flowing together of the
emotions of a primitive human being.'

The whole discussion of the metaphorical character of "original"
languages is one of those wonderfully intuitive bits of insight
that one meets with frequently enough in Herder's writings. He
saw clearly the perfectly natural, and, indeed, psychologically
inevitable, play of metaphor that runs through the history of
language. This was remarkable at a time when figures of speech
were thought to be the artistic flowers of polite literature. The
modern semasiologist can, however, be bolder than Herder. He
recognizes clearly that metaphor operates with equal power at
all periods in the development of a language, not chiefly in the
relatively older phases, as Herder thought, but just as well in
times nearer the present. Moreover, he is inclined to believe
that not merely a large part of the vocabulary of a language is
metaphorically transferred in meaning, but that practically all of
it has undergone an indefinite number of gradual semantic trans-
mutations.

Herder's third thesis is as follows:

The more original a language is, the more frequently do such feeHngs
cross in it; the less easily can these be exactly and logically subordinated
to one another. The language is rich in synonyms; alongside of real
poverty it has the most unnecessary superfluity.'

How can so ill-arranged a mass of material be the work of God?
As coming from the hand of man, however, the presence of
synonyms is entirely explicable. Herder argues:

The less acquainted one was with natural phenomena; the more
aspects one could in inexperience observe therefrom and hardly recognize;
the less one invented d priori^ but rather according to circumstances of
sense: the more synonyms.'

This great wealth of synonyms is seen not only in the proverbial
two hundred words of the Arab for "snake" and about one

lOp. ct<., p. 11. 2/6id., p. 75. 3/6id., p. 76.

128

One: The Problem of ihc Oni^in of lun^uage SS

thousand for "sword," \m[ also in most lanj^uaguH of primitive
tribes. Herder points out that the latter, nlthou^h often lackinj^
terms for ideas which seem to us most necessary of expression,
frequently possess an astonishing wealth of words for ideas but
slightly differentiated in our own minds.

As fourth canon Herder enunciates the foUowinir-

Just as the human mind can recall uo al)stniction out of th«< realm
of spirits which it did not obtain by means of opiK)rtuiiitio-s and awaken-
ings of the senses, so also our language contains no abstract word which
it has not obtained by means of tone and feeling. And th«i more original
the language, the fewer abstractions, the more feelings.'

That all abstract ideas are originally expressed in terms of con-
crete images is almost self-evident, and Herder has, indeed, little
difficulty in proving his point. Again he has recourse for his
illustrations to the language of the Orient and of primitive jxioplea.
"Holy" was, for instance, originally "set apart" in meaning;
"soul" meant really "breath." Missionaries and travelers unani-
mously testify to the great difficulty of rendering abstract terms
in the idioms of comparatively uncultured peoples; the history
of civilization shows that many of the terms used in philosophy
and other regions of abstract thought are simply borrowed from
the vocabularies of nations already farther advanced in specula-
tion. All this. Herder rightly emphasizes, points to the operation
of purely human powers; no terms are found for abstractions not
absolutely necessary to the thought of the people who use them,
while in every case such terms are originally of purely sensational
origin. Surely there can be no talk here of divine intervention,
where only human weakness is manifest.
Herder's fifth and last thesis runs:

Since every grammar is only a philosophy of language and a mothcxi
of its use, it follows that the more original the language, the less grammar
there must be in it, and the oldest [language] is simply the vocabulary of
nature."

Herder devotes several pages to a consideration of this matter, but
his whole treatment seems now confused and antiquated. One
acquainted with the elaborate formal machinery, particularly in

1 Op. c«<.. p. 78. J //-id., pp. 82,83.

129

86 General Linguistics I

regard to the verb, of the Semitic and of many North American
Indian languages, both of which Herder considered particularly
** original" in character, will be inclined to deny point blank the
validity of his thesis. Herder, however, is not blind to this
grammatical complexity. On the contrary, he asserts, paradoxi-
cally enough, that this very complexity is a sign of the lack of a
true grammatical sense. He thinks that those languages that
make use of complicated grammatical schemes show thereby their
inability to arrange their material systematically and logically;
the Germans or French, for instance, he implies, with fewer para-
digms, have more of a grammar in the true sense of the word.
Speaking more particularly of verb-forms, he says:

The more conjugations [are found], the less thoroughly one has
learned to systematize concepts relatively to each other. How many the
orientals have! and yet they are not such in reality, for what transplanta-
tions and transferences of the verbs are there not from conjugation to
conjugation! The matter is quite natural. Since nothing concerns man
so much, at least appears to him so linguistically suitable, as that which
he is to narrate: deeds, actions, events; therefore, such a multitude of
deeds and events must originally have been gathered that a new verb
arose for almost every condition.*

Herder's arguments do not, it is almost needless to observe,
bear inspection. Herder thought of grammar, as was very
natural in the eighteenth century, as something which was,
with increasing civilization, brought to bear on language from
without. With this conception in his mind, it seemed that to
admit the existence of complex grammatical form in "original"
languages was playing into the hands of Stlssmilch. Today,
however, owing to the vast stock of comparative and historic
linguistic material at our disposal, we see clearly that grammar,
so far from needing the loving attention of the grammarians,
takes very good care of itself and develops along definite lines.
We need not, therefore, deal in paradoxes and can admit, with a
clear conscience, that many typically "original" languages, to
adopt Herder's now unserviceable terminology, possess truly
grammatical features of incredible complexity, as in the case of
the Eskimo verb or Bantu noun.

» Op. cit., p. 83.

130

One: riw /'rohlnn of ihr Origin of I.uni^uagt 87

Herder ends the first part of liis prizo rssay w,fl, ,, i.nra^'raph
which contains the following stateni.nt:

Ich bilde mir ein, das KOuneu der Enipfindun^- monnchlicher Spracbe
Rci mit dem, was ich gesa^^t .... so lxnvios<;n, class wer d«m Men»ch«u
nicht Vernunft abspricht, oder, was elxai so viel ist, wer nur woi.s.s, wa«
Vernunft ist: wer sich feriier je urn dio EloraenU? der Sprache philorto
phi.sch bekummert, .... derkanu niciit eiuen Au^'enhlick zweif.,hi. wenn
ich auch weiter kein Wort mehr hinzusetzo.'

And, indeed, Herder might have logically cunc'luded the essay
at this point, though we should not like to miss some of tho
thoughts in the second part. The form of the question as put bj
the academy, seemed, however, to Herder to demand an arrange-
ment of his subject-matter into two distinct parts. Inasmuch as
both of the queries posed are practically answered by Herder in
the first part, we need not analyse in detail the trend of his argu-
ment in the second, which ostensibly discusses "auf welchem
Wege der Mensch sich am fOglichsten hat Sprache erfinden
kOnnen und mflssen," but which, in reality, deals chiefly with the
gradual development of language. It is itself subdivided into
four sections, each of these discussing a natural law operative in
this development.

The first of these Naturgesetze, which Herder takes so seri-
ously as to put in the imperative form, reads: "Man is a freely
thinking, active being, whose powers work on progressively;
therefore he shall be a creature of language {GeschOpf cU-r
Spr(iche)y^ As the wording of this law implies, Herder hero
recapitulates, with amplifications, a good deal of what he had
already presented in the first part. But there is a new thought
introduced here — that of development in the line of prooress.
The gift of speech is, it is true, as characteristic of man as the
ability to construct a hive is native to the bee — with this notable
difference: the bee, acting mechanically by virtue of its inlx)rn
instinctive powers, builds as efficiently the first day as the last,
and will build so. Herder believes, to the end of time; tho lan-
guage of man, however, increases in power and efficiency with
every use that is made of it. The reason for tliis law of linguistic
growth in the individual is evident when we consider the relation

' Op. cit., p. 89. 1 1(>i<i., p. 93.

88 General Linguistics I

between thought and language. "There is no condition of the
human mind," Herder says, "which is not capable of linguis-
tic expression or not really determined by words of the mind
( Worte der Seele)?''^ Hence growth in the power of reflexion or
Besonnenheit, conditioned by the growth of experience, entails
also advance in the employment of language. "The growth of lan-
guage," as Herder puts it, "is as natural to man as his nature
itself.'"^ Siissmilch had objected to the idea of a human develop-
ment of language on the ground that such a proceeding would
have required the thought of a philosophically trained mind,
such as it would be utterly absurd to imagine primitive man to
have been in possession of. Herder points out the shallowness
of his argument in very emphatic terms — an argument that,
lacking absolutely all sense of historical perspective, would pic-
ture primitive man as placed in the same environment and
governed by the same conventions as prevailed in the pseudo-
philosophical atmosphere of the eighteenth century. Herder
gives a most excellent characterization of the spirit of his time in
a few sentences which express his profound dissatisfaction with it :

Es ist ftir mich unbegreiflich, wie man sich so tief in die Schatten, in
die dunklen Werkstatten des Kunstmassigen verlieren kann, ohne auch
nicht ein mal das weite, helle Licht der uneingekerkerten Natur erkennen
zu wollen .... Aus den Meisterstiicken menschlicher Dichtkunst und
Beredsamkeit [sind] Kindereien geworden, an welchen greise Kinder und
junge Kinder Phrasen lernen und Regeln klauben. Wir haschen ihre
Formalitaten und haben ihren Geist verloren; wir lernen ihre Sprache
und ftihlen nicht die lebendige Welt ihrer Gedanken. Derselbe Fall
ist's mit unserm Urtheilen iiber das Meisterstiick des menschlichen
Geistes, die Bildung der Sprache tiberhaupt. Da soil uns das todte
Nachdenken Dinge lehren, die bloss aus dem lebendigen Hauche der
Welt, aus dem Geiste der grossen wirksamen Natur den Menschen
beseelen, ihn aufrufen und fortbilden konnten. Da sollen die stumpfen,
spaten Gesetze der Grammatiker das Gottlichste sein, was wir verehren,
und vergessen die wahre gottliche Sprachnatur, die sich in ihren Herzen
mit dem menschlichen Geiste bildete, so unregelmSssig sie auch scheine.
Die Sprachbildung ist in die Schatten der Schule gewichen, aus denen
sie nichts mehr fur die lebendige Welt wirket : drum soil auch nie eine
helle Welt gegeben sein, in der die ersten Sprachenbilder leben, ftihlen,
schaffen, und dichten mussten.^

1 Op. cit., p. 100. 3 Ibid., p. 101. 3 Ibid., pp. 111. 112.

132

One: The f'rohlc/n of i/ic ()ni;in <>t I ini^iui^v 89

Herder's second natural law carries the devi'lopmenl of lan-
guage one step farther; the first law dealt with th(? growth of
language in the individual, the second is devoted to its develop.
ment in the family. The law reads: "Man is, according to his
nature, a being of the herd, of society: The development of a
language is therefore natural, essential, necessary to him.'" The
great physical weakness of the human female as compannl witli
the male, and the utter helplessness of the newborn child, make
it absolutely necessary that human beings, even more bo than iH
the case among the animals, form into families, sharing a common
abode. The primitive man, more experienced than his mate and
offspring, would naturally proceed to teach them that st<K"k of
linguistic information which he had himself so laboriously gath-
ered. The child, entirely dependent as he is upon the exertions
of his father, would babblingly repeat the sounds uttered in his
neighborhood, and in time become inheritor of his parent's entire
vocabulary. Arrived at maturity, he would go on enriching the
store of linguistic knowledge on the basis of his own experience.
In this way the institution of the family becomes an important
means for the perpetuation from generation to generation and for
the gradual enrichment of language. Moreover, in the verv
process of teaching, the language becomes more definitely organ-
ized, the stock of ideas becomes more and more clearly defined ;
Herder, indeed, sees in this earliest process of language-instruc-
tion the genesis of grammar.

The most interesting portion of the second part of Herder's
essay is the discussion of the third natural law, dealing with the
rise of nationally distinct languages. Herder formulates his law
thus: *'Just as the whole race of man could not jK)ssibly remain
one herd, so also it could not retain one language. There arises
consequently the formation of various national tongues.'" Herder
begins his discussion by clearly pointing out that, in the exact
or, as he terms it, "metaphysical" sense of the word, no two |>er-
sons speak precisely the same language, any more than they are
exact physical counterparts. Setting aside, however, such minute
individual differences, it can easily be shown that more distinctly

1 Op. cit.y p. 112. a Ibtd., pp. IM, 124.

133

90 General Linguistics I

marked linguistic groups or dialects would, in the nature of things,
form. Every family-group puts its own characteristic stamp upon
the inherited linguistic stock; differences of climate work upon
the speech-organs, and consequently, Herder supposes, upon the
language itself;* the preference for different words and turns of
expression in different sections of the linguistic field gives rise to
dialects. "Original" languages, moreover, being less hampered,
according to Herder, by grammatical rules, are more liable to
dialectic disintegration than more cultured idioms; although here,
too, the most careful modern linguistic research does not uncon-
ditionally bear out Herder's presumption. The oft-asserted and
oft-repeated statement of the incredibly rapid change of the
languages of primitive tribes is founded chiefly on the untrust-
worthy reports of linguistically inefficient missionaries; many of
the extreme statements formerly and even yet current are absurdly
untrue. Indeed, the most startling cases of linguistic conserva-
tism are found among certain primitive peoples, such as the
Eskimos.

Man, Herder proceeds, with an almost naive anthropomorphic
teleology, is made for all the earth and all the earth for man.
Hence we find him at home as well in the regions of eternal snow
as under the burning sun of the tropics. As is to be expected,
under widely diverse geographical and climatic conditions, the
originally homogeneous race of man differentiates into diverse
races, while the originally homogeneous speech of man differ-
entiates first into dialects, then, with the lapse of ages, into
mutually quite unintelligible languages. Hence is to be explained
the bewildering Babel of tongues; as Herder expresses it: "Die
Sprache wird ein Proteus auf der runden Oberflache der Erde."*
It is to be borne in mind, of course, that the now familiar con-
ception of independent linguistic stocks was in the main a foreign
one to Herder; it did not arise until after a lucky fate discovered
the relationship of the idiom of ancient India with that of far-
distant Greece and Rome. If Herder's view of the gradual

1 It should be stated here, however, that, contrary to all ezpectatioo, anatomical investi-
gation has never succeeded in demonstrating differences of vocal anatomy to be the basis of
differences in dialect or language.

2 Op. cit., p. 127.

134

One: The Problem of the Orii^in of l.iini^tta^c 91

differentiation of speech is correct, we expect tu fjnd lint'uistic
modifications congruent to geographic difTerences. How comes
it, then, that totally diverse languages an* (jfteii found H|}<jken
side by side? Herder thinks to be able satisfactorily to explain
this puzzling condition by the hatred with which neighlx)riiig
tribes frequently regard each other. Such discord would oj>er-
ate quite as effectively as geographical barriers toward linguistic
isolation. The description of the confusion of tongues in Genesis
Herder interprets as a characteristically oriental method of pre-
senting this truth.

The fourth and last Naturgesetz reads:

Just as, according to all probability, the human race forms a pro-
gressive whole from one origin in one great system; so also all languafjos,
and with them the whole chain of culture,*

After a brief r6sum6 of the links in the chain of linguistic develop-
ment — development in the individual, in the family, and in the
nation — Herder most emphatically supports a monogenistic view
of language as the most rational. His chief reason for the con-
tention is the evident similarity he finds in the grammatical
structure of the various languages — a similarity that he believes
to be BO great as to preclude all possibility of polygenism. The
only serious departure from the common grammatical outlines is
found, according to Herder, in the case of Chinese, which how-
ever, is but an exception. Herder advances as another argument
in favor of linguistic monogenism the almost universal use of the
same or very similar alphabets. This universality was more
apparent in Herder's time than now, for the Egyptian and various
cuneiform records had not as yet been deciphered. Still one is
rather surprised to find a man of Herder's stamp so lose sight of
the perspective of history as to present so lame an argument.
One might have expected him to perceive that in any case the
formation of written alphabets must have taken place long after
independent languages had developed, and that the wide spread
of the so-called Phoenician alphabet was due to several stages of
borrowing, to a great extent, within historic times.

The latter part of the essay rapidly summarizes the orthcxiox

I Op. cit., p. 134.

135

92 General Linguistics I

stand taken by Siissmilch and his followers as against the more
philosophical and psychologically sounder view of Herder con-
cerning the origin of language. Herder makes bold to call the
orthodox view "nonsense" (Unsinii), and accuses its supporters
of petty anthropomorphism in their conception of God's activity.
On the other hand, he claims:

Der menschliche [Ursprung] zeigt Gott im grossesten Lichte; sein
Werk, eine menschliche Seele, durch sich selbst, eine Sprache schaffend
und fortschaffend, weil sie sein Werk, eine menschliche Seele ist.'

At the very end of the essay we find Herder's own statement of
his aim, with which I shall close m}' analysis of his work as a
perfectly good formulation of the spirit to be pursued in investi-
gations of such a character even today:

Er [i. e., der Verfasser] befliss sich feste Data aus der menschlichen
Seele, der menschlichen Organisation, dem Bau aller alten und wilden
Sprachen, und der ganzen Haushaltung des menschlichen Geschlechts
zu sammlen, und seinen Satz so'fcu beweisen, wie die festeste philo-
sophische Wahrheit bewiesen werden kann.'^

It is hardly necesssary to go into any general criticism on
Herder's prize essay, particularly as various points in Herder's
argument have been the subject of critical comment in the course
of our analysis. That much of the work is quite antiquated, both
in subject-matter and general attitude, is, of course, self-evident;
it is rather to be wondered at how much in the essay is still valid,
and with what remarkable intuitive power Herder grasped some
of the most vital points both in psychology and language. One
wishes that we today could be so cocksure of the solution of cer-
tain linguistic problems as Herder seems to have been ; but, then,
that was characteristic in a large measure of the age of rationalism.
The philosophers of the eighteenth century, relying very heavily
on pure reason unfettered by hard facts, proceeded, with admir-
able courage, to attack and solve the most obscure and intricate
problems in the history of human culture — problems to the solu-
tion of which we have now learned to proceed quite timidly. Some
of this blind trust in pure reason is apparent in our prize essay;
yet Herder attempted, as much as possible, to make use of what
linguistic material was at hand in the verification of his theories.

» Op. cit., p. 146. 2 Ibid., p. 147.

136

One: The Prohlcni of the Orii^in of I.iini^iiaf^e 93

It is not necessary, either, to p^o into any analysis of the
literary style of Herder's essay, as tlui im{)«)rtaMt (hin^ in our
subject is not Herder's language, but rather his thought. The
characteristic qualities of Herder's style are here in evidence afi
elsewhere; wealth of figurative expression; a lavish use of rhot<3ri-
cal periods, as outwardly indicated by a generous s{)rinkling of
exclamation points, interrogation points, and dashes (it may be
noted that the treatise ends with a dash) ; and a warm, enthusias-
tic diction, which often carries the author away from tlic imme-
diate object of discussion. In general, however, the essay is
remarkable, at least when considered as a Herder document, for
the systematic development of the theme and for clearness of
exposition.

It is certainly very strange, and almost incredible, that one
who succeeded so well in demonstrating the human origin of
language should himself have later been in doubt as to the
validity of his conclusions; yet such was the melancholy case
with Herder. When Herder wrote the prize essay, during the
latter part of his stay in Strassburg (1770), he was still, in the
main, in sympathy with the rationalistic advocates of reason, even
though the bloodless Aufkldrung of a Nicolai was not exactly to
his taste. Hence we find in the essay a strong aversion for the
mystic and supernatural, and a desire to explain all cultural
phenomena in the light of human intelligence. In the early part
of his Btlckeburg residence, however. Herder's ideas underwent
a tremendous change. So radical, indeed, was the transformation
effected in his general mental attitude that the Preisschrift may
be conveniently considered as marking the end of a definitely
limited period in Herder's life. The mental change referred to
was a break with the older standpoint of "enlightenment" [Auf-
kldrung), which had on the whole, despite Hamann's infiuencx>,
dominated Herder's thoughts, if at times equivocally. He now
(1771-72), very largely under the influence of the mystically
pious Countess of Bflckeburg, leaned toward romanticism, and a
philosophy and theology that did not seek the final explanation
of things in reason. Hence, when the news reached Herder that
he had been awarded the prize, he was anything but elated; the

137

94 Gencnil Linguistics I

whole spirit and tendency of his essay were now quite distasteful to
him, and belonged already for Herder to the dim past. Despite
the change in Herder, congratulations on the winning of the prize
came in from all sides; "the townspeople," he writes, "regard me
as the most celebrated of men because I have now gained the
prize."* These congratulations, as might well be expected under
the circumstances, brought Herder more vexation than satisfaction.
At Easter of the year 1772, Herder's former preceptor, Hamann,
who had not corresponded with his disciple since the latter had
left Riga, and from whom Herder had in the meanwhile become
somewhat estranged, intellectually speaking, wrote a cold and
hostile review of the Ursprung der Sprache in the Kdnigsberger
Journal. Though many ideas developed in the essay had been
largely inspired by suggestions of Hamann himself, nevertheless
Herder's flat denial of the direct agency of God in the invention
of language was by no means to the other's taste. The stand
taken by Hamann is well summarized by Nevinson: "God might
act through nature and the voices of beasts, but from God lan-
guage, as all else, must come, for in God we live and move and
have our being."' Herder felt the sting of criticism all the more
keenly in that he was now largely in sympathy with Hamann's
views, and felt drawn toward his former teacher more powerfully
than ever. Through the mediation of their mutual friend, Hart-
knoch, a reconciliation was effected between the two, Herder
recanting the heresies of which he had been guilty. In a second
and more favorable review, and in a treatise entitled The Last
Words of the Bed-Cross Knight on the Divine or Human Origin
of Language, Hamann clearly shows that the friends of old were
friends again. Perhaps nothing can prove more clearly the
unhealthy element in the mysticism of Herder's Btlckeburg
period than his amazing repudiation of the doctrines he had him-
self so unanswerably demonstrated. For a time he occupied
himself with the thought of adding some words of explanation
and semi-apology to the essay, published by authority of the
Academy in 1772, but nothing came of the plan. Fritz Mauthner,

1 Se« Nevinson, Herder and His Times, p. 185.

2 Ibid., p. 196.

138

One: I he I'lohlini of the Orti^m of Innfiuufie *)5

in the work already referred to, speaks impatiently of Herder's
inconstancy :

Herder bringt sich um jeden KrtxJit, wenn er WMiie Preijischrift nchon
1772 (in einem Brief an Hamuini) als "Schrift cuwh \Vit/.U">l|M«lM" ver-
leugnet; die Denkart dieser Treisschrift kOnm- uud w)lle aiif ihn ao
wenig Einfluss haben, als das Bild, das er jetzt an die Wand na^l<\ Da
ist es denn kein Wunder, wenn Herder in d<T Folgezeit den lielx'n Gott
wieder um die Erfindung der Spracbe iKMniiht.'

Certainly disingenuous is Herder's statement, in the long letter
of explanation addressed to Hamann, that after all he had not
seriously differed from his preceptor, and that in writing for "an
enlightened royal Prussian Academy'" he had been forced to put
on the mask of the "Leibnitzian aesthetic form."* After Herder
had freed himself from the mists of Bftckeburg myfiticism, a
reversal in judgment set in in favor of his earlier comparatively
rationalistic views, so that, when seventeen years later, in 1788-89,
he prepared a second edition of the Preisschrift, he found little
to change in the text, save in the matter of chastening the wild-
ness of the language. We may then safely look upon our Preis-
schrift as the most important and genuine expression of Herder's
views upon the subject of language.

CJoncerning the influence of Herder's prize essay on sub-
sequent linguistic speculation it is difficult to speak deflnitely,
from the very nature of Herder's work. Herder did not, as we
have seen, definitely systematize, nor could his solution of the
problem be considered in any way final; his own subsequent
vacillation shows, perhaps more clearly than anything else, the
unsatisfactory nature of much of the reasoning. Contradictions
even of no small significance and lack of clearness in the terms
used will have been noticed in the course of our ex{X)sition of
Herder's essay; the weak points in it, both wlion judged from
the standpoint of Herder's own time and from that of the |H>st-
Humboldtian and pre-evolutionary view-jxjint of the eixtiefl, are
probably best pointed out by the psychologist and linguistic
philosopher, H. Steinthal, in his Ursprung dcr Sprachv. Setting
aside faults in the essay itself, it is evident that the new vistA-s of

> F. Mauthner, "Zur Sprachwissooschaft" (Vol. II of Kritik drr Sprackr). pp. IT CO
'See Nevinson, op. ctt., p. 197. ilbid.

139

96 General Linguistics I

linguistic thought opened up by the work of Karl Wilhelm von
Humboldt, and the more special labors of Bopp and Grimm,
speedily relegated Herder's treatise to the limbo of things that
were, so that even as early as the period at which Steinthal and
Grimm wrote their works on the origin of language, Herder's
Preisschrift had already become of chiefly historical interest.
The real historic significance, then, of Herder's work would be
shown to lie in the general service it rendered by compelling a
sounder study of the psychologic and historic elements involved
in the investigation of the problem, and perhaps also in the sug-
gestions it gave Humboldt for his far deeper treatment of the
same and closely allied themes.

Perhaps the best testimony which could be offered on the sub-
ject of Herder's more general influence is the following quotation
from Jacob Grimm at the close of his own work on the origin of
language :

Enden kann ich nicht, ohne vorher dem Genius des Mannes zu hul-
digen, der was ihm an Tiefe der Forschung oder Strange der Gelehr-
samkeit abging, durch sinnvoUen Tact, durch reges Gefiihl der Wahrheit
ersetzend, wie manche andere, auch die schwierige Frage nach der
Sprache Urspnmg bereits so erledigt hatte, dass seine ertheilte Antwort
immer noch zutreffend bleibt, wenn sie gleich mit anderen Griinden, als
ihm schon dafur zu Gebote standen, aufzustellen und bestatigen ist.^

On his immediate contemporaries the prize essay doubtless made
a deep impression. To Goethe, who was just at that time under
Herder's personal influence, the author showed the essay while still
in manuscript. Goethe had not thought very much about the
subject, and was inclined to consider it as somewhat superfluous.
"For," he says, "if man was of divine origin, so was language;
and, if man must be regarded in the circle of nature, language
must also be natural. Still, I read the treatise with great pleasure
and to my special edification."*

The extent and even existence of Herder's influence on Hum-
boldt, on the other hand, is a disputed question. It is all the
more important because practically all the later thought on the
philosophy of language (Steinthal, Schleicher, and others) is

' Jacob Grimm, Vber den Ursprung der Sprache (Berlin, 1852), p. 56.
2 See Nevinson, op. cit., p. 163.

140

One: Tin- Prohlcni ofilu- ()rt\^in of l.an^uaiic 97

connected quite directly with Huinljoldt's ideas develoiM'd ju hig
Uher das vergleichende Sjirachstiidium in Bezichxnuj auf die
verschiedenen Epochen dev SpracJirntwickcliuuj, and still more
in the Einleitung in die Kawisprdchc: Uhcr die Verscliirdenheit
des menschlichcn Sprachbaues und iJircn Einjluss auf die gei-
stige Entwickelung der MeJischengeschichte. 8tointlml, himself
an enthusiastic follower and developer of Huniboldtian viows,
most emphatically denies any indebtedness on Humboldt's part
to Herder.'

Haym, Herder's biographer, on the other hand, just as emphati-
cally asserts the perceptible influence of Herder in Humbcjldt'B
writings, and claims that the latter is most decidedly to be con-
sidered as standing on his predecessor's shoulders. He says:

Er [i. e., Humboldt] wiederholt die Gedanken Herder's — er vertioft,
er verfeinert, er bestimmt, er klSrt sie, er denkt das von jenem gleichsam
athemlos Gedachte mit ruhig verweilender Umsicht nmi zweiten Male
nach und durch.^

He goes on to show how, as with Herder, so also with Humboldt,
man is "ein singendes GeschOpf, aber Gedanken mit den TOnen
verbindend";' language is to Humboldt very much as to Herder,
"die natdrliche Entwickelung einer den Menschen als solchen
bezeichnenden Anlage."* To Humboldt the chief task of gen-
eral linguistics is the consideration from a single point of view of
the apparently infinite variety of languages, "und durch alle
Umwandlungen der Geschichte hindurch dem Gauge der goisti-
gen Entwickelung der Menschheit an der Hand der tief in die-
selbe verschlungenen, sie von Stufe zu Stufe begleitenden Sprache
ZTi folgen."* This is evidently little else than a more satisfactory
and scientific formulation of Herder's idea of the gradual growth
of language in concomitance with the growth of Besoinwrihrit. On
the whole, I should be inclined, in view of the greater probability
of the historic continuity of ideas, to side with Haym. A com-
prehensive statement of the position that Herder occupies in the
history of linguistics is given by Lauchert.'

•H. Steinthal, Der Ursprung der Sprache (Rorlin, 1838). p. 12

2 Haym, Herder nach seinem I^ben und scinen W'crken darQCstdlt. p. 40S.

3 Ibid. ilhid. »//>id.

6F. Lauchert, "Die AnschauunKon Herders Qb«r den UrsprunK d«r Spr«cb«»."' £WpA«^
rion. Vol. I, p. 766.

141

98 General Linguistics I

Es geniige, darauf hingewiesen zu haben, dass die neuere Sprach-
philosophie, soweit sie einerseits nicht auf dem Boden des positiven
Christentums und der SchOpfungslehre desselben steht, andrerseits aber
auch noch nicht auf dem der modernen naturalistisch-materialistischen
Weltanschauung, direkt oder indirekt unter Herder's Einfluss steht.

As the last general linguist to discuss language problems from
the standpoint that maintained the existence of a wide, impassable
gulf between man and the lower animals, and stoutly denied any
genetic relationship between animal cries and the rude beginnings
of human speech, should perhaps be mentioned Max Muller. Like
Herder and Humboldt, he saw in language the distinguishing
mark that separated man from the brute world, and was never
tired, to the end of his days, of arguing that this possession of
language was the death-blow to Darwinism. The idea of the
interrelation of language and reason, and of their simultaneous
growth, common to Herder and Humboldt, we find pushed to its
utmost limit by Max MtlUer. So impressed was he by this theory
of the essential identity of language and reason that his slogan
in later days was: "Without reason no language; without language
no reason.'" As is well known, his assertion of this principle
brought on a fruitless logomachy with William Dwight Whitney.
Despite Max MtlUer, however, it seems to me that the path for
future work on the prime problems, more especially the origin, of
language lies in the direction pointed out by evolution. A new
element, the careful and scientific study of sound-reflexes in higher
animals, must now enter into the discussion. Perhaps this, with a
very extended study of all the various existing stocks of languages,
in order to determine the most fundamental properties of lan-
guage, may assist materially in ultimately rendering our problem
more tractable. We should not only try to imagine to what
beginnings the present state of language reaches back, but also
to reconstruct an ideal picture of the evolution of howls and cries,
under the favoring conditions, whatever those were, into less rude
forms of audible expression. Perhaps the ends of the two series
can be bridged over?

Edward Sapie
Columbia University

» See, e. g., title-page of M. MoUer's Science of Thought (New York, 1887).

142

One: r/ic rrohlcni of ilic Origin of iMHiiuu^c
Editorial NdIc

Modem Philology 5 (1907), 109-142. |A rcprmt appeared in: Hiuonogruphta
Lingiiisiica 1 1 ( 19S4). 355-3X3]

The following errors in ihc originally jniblislicd \crsinn have been eorrecled
directly into the text printed here (page relcrenees are lo the oriuinal):
p. 109, 1. 5: sonl ils (correct: sont-ils)

p. 114,1
p. 132, 1
p. 140,1
p. 141,1

30-31: connciissiinces ( 18th-centur\ spelling: connoisstinces)

32: undfort bilden (correct: iind lorlbikleii )

22: when sic (correct: wenn sic)

25: Umwandlnngen (correct: Uniwandlungen)

A further error to be corrected on p. 67 I. 6 concerns the name of ihc secretary
of the Berlin academy. Sapir erroneously writes his name as TourneN: this should
be corrected into Forniey [- Jean-Henri-Samuel l-orme\ ].

Section Two

History. Variety AND Setting u\ Lan(.i age

(1911,1912)

liUroduclioii:
History, Variety and Scllinu ol Laimuaizc

Tlie two texts included in tliis section de;il uiih issues ol geneial. dcscripluc
and historical linguistics. Iit)th articles, published hv Udward Sapir while slill in
his mid-twenties, and shortly after his move to Ottawa, arc the (slighlly) revised
version of papers delivered before an audience of anthropologists.' Hiey Icslify
to Sapir's increasing mastery of various American Indian languages.

In these two texts we find adumbrated the major themes which Sapir was lo
elaborate on in his book Language,^ such as: pri)blems of historical relationship

' For this term, see Language (New York. 1921). p. 221: "Language has a setting. The people that speak il l^
to a race (or a number of races), that is, to a group which is set off by physical characteristics from other vr
Again, language docs not e.xist apart from culture, that is, from the socially inherited assemblage o( pr.i.

and beliefs thai determines the texture of our lives."

- On this period in Sapirs career, see Richard J. Preston, "Reflections on Edward Sapir's Anihr ' "

Canada", Canadian Review of Sociology and Anthropology 17 (I MHO). 3h7-.^74; Stephen <> M
Canadian Winter of Edward Sapir", Historiographia l.inguistica H ( I9S1 ), 6.^-68; HelOne Bernier. "IaI" ... > ■
et la recherche anthropologique au Musee du Canada 191l)-1925". Histonngraphui lingtuMua II i; -•

397-412; William N. Fenton. "Sapir as Museologist and Research Director 1910-1925". m .Nf- ''

Language, Culture, and Personality. Proceedings of the F.dward Sapir Centenary Confrrt >

October 1984). edited by William Cowan, Michael K. Foster, and Konrad Koerner ( Amstcrd.ti.. i ...v ,,

1986). 215-240: Regna l5arnell, Edward Sapir: Linguist. Anthropologist. Humanist (BcTkc\c\. I«N0), 44 Sh

' "The History and Varieties of Human Speech" is based on a lecture delivered on .April I. 1^11 .«( i'-^

University of Pennsylvania Museum: "Language and Environment" was read at the DecemK-r ••"■■ ■■' ■'•

American Anthropological Association in 191 1.

•• Readers familiar with Sapir's linguistic writings will note several passagcN in the Nl I and 1912 papcn ihal are
echoed in Language (New York, 1921 ): e.g..

• "The History and Varieties of Human Speech", p. 46 :: Language, pp. I65-Ih6 (the cumulative cll

linguistic changes): p. 47 and p. 5(S :: Language, p. 202 (English plural nouns): p. 4S ' ■ ■■- "

linguistic relationships): p. 48 and p. 57 :: Language, pp. 191-201 (history of the Cm

51-52 :: Language, pp. 61-64 (grammatical processes): pp. 52-.s4 :: Language r" "^
55-56 :: Language, pp. 183-204 (pht)netic law); p. 57 :: Language, p. 2S and p

affecting infinitive forms); p. 59 :: Language, pp. 205-220 (languages influencii.K

p. 212 (presence of a "dull vowel" in Slavic and Ural-Altaic); p. <>l :: I anguai:e. pill'
ideas of material content through grammatical suffixes appended lo the ni''* ^'^ ">i
101-102 and 97-98 (treatment of reialional concepts in latin and Kxvakuilli

(derivation by final consonant change in I nglish): p. h3 :: / imv"''.*.'**' P- *^^ '"J'""-"-

in Enulish):

• "Language and l-nvironmenl", p. 228 and i^p. 2.W-240 :: language, p ?U ...l .i,..nsl,.P lvt«ccn ...nicnl ..f Un
guage and culture); p. 233 :: Language, pp. 23.^-234 (no correlalit>ii Ivtw

ture); pp. 240-241 :: Language, pp. 231-2.^4 (different rhvlhm oi languam ,: -

guatie): pp. 235-236 :: Language, pp. 228-229 (Hupa. Yurok and Karok: linguislK diwrwly and cultural urniyi.

1 04 General Linguistics J

between languages and remote genetic affiliations; the nature of linguistic chan-
ge, grammatical processes and grammatical techniques as typological parameters,
language and its socio-cultural context."

The first text, "The History and Varieties of Speech," is remarkable for its com-
pactness, and for the vast perspectives it unfolds. Sapir starts from the distinction
between origin and history of language (pp. 45^6); the first theme — which he
had dealt with in his master's thesis (see section I) — is briefly dismissed, whereas
the second forms the central theme of the paper. The history of language consti-
tutes the thematic convergence point for a threefold analysis:

(a) A study of how the linguist builds up knowledge about the past; here Sapir
distinguishes between a philological (documentary) approach and a reconstruc-
tive approach,^ the latter being subdivided into internal and external (or com-
parative) reconstruction;

(b) The analysis of what is constant and what is variable in human language;
Sapir discusses a number of universal conditions or constraints on language,
which properly constitute the nature of language, viz. (/) vocal symbolization,
(//) the use made of a limited set of vocalic and consonantal segments, (///) the
overall presence of a number of grammatical processes, and the basic distinction
between denominating and predicating terms (p. 51). Within the range of these
constraints, a high amount of variation remains possible, and can indeed be
observed world-wide.

(c) The nature of linguistic change; this section constitutes the central part of
Sapir's article, and what Sapir offers us here, is a synoptic treatment of historical

-■' On the topic of (distant) genetic relationships in Sapir's worii, see: Victor GoUa, "Sapir, Kroeber and North
American Linguistic Classification", in New Perspectives in Language, Culture, and Personality. Proceedings of
the Edward Sapir Centenary Conference (Ottawa, 1-3 October 1984), edited by William Cowan, Michael K.
Foster, and Konrad Koerner (Amsterdam/Philadelphia, 1986), 17^0; Michael E. Krauss, "Edward Sapir and
Athabaskan Linguistics, with Preliminary Annotated Bibliography of Sapir's Work on Athabaskan and Na-
Dene", ibid.. 147-190; Ives Goddard. "Sapir's Comparative Method", ibid., 191-214; Marianne Mithun,
"Typology and Deep Genetic Relations in North America", in Reconstructing Languages and Cultures, edited
by Edgar C. Polome and Werner Winter (Berlin/New York, 1992), 91-1 08; Thomas C. Smith Stark, "El metodo
de Sapir para establecer relaciones geneticas remotas", in Reflexiones lingiiisticas y literarias, edited by Rebeca
Barriga Villanueva and Josefina Garcia Fajardo (Mexico, 1992), 17-42; Alan S. Kaye, "Distant Genetic
Relationship and Edward Sapir", 5mnof/cfl 79 (1993), 273-300; Alexis Manaster Ramer, "Sapir's Classifications:
Coahuiltecan" and "Sapir's Classifications: Haida and other Na-Dene Languages", /lAi/Aropo/og/cfl/ Linguistics
38 (1996), 1-38 and 179-215; Regna Darnell, "Indo-European Methodology, Bloomfield's Central Algonquian,
and Sapir's Distant Genetic Relationships", in The Emergence of the Modern Language Sciences. Studies on the
transition from historical-comparative to structural linguistics in honour ofE.EK. Koerner, vol. 2: Methodological
Perspectives and Applications, edited by Sheila Embleton, John E. Joseph and Hans-Josef Niederehe
(Amsterdam/Philadelphia, 1999), 3-16.

'^ See LflAjgwflge (New York, 1921), chapters VILVIII, and IX (historical relationships, broad genetic affiliations;
linguistic change and phonetic law), IV, V and VI (grammatical processes; grammatical concepts and techni-
ques; types of linguistic structure), X (language, race, and culture).

^ It is interesting to note that Sapir makes mention of the method of "relative chronology" (p. 47), which had
its origin in diachronic work by Romance scholars, having been used first by Wilhelm Meyer-Liibke in his
Grammatik der romanischen Sprachen, vol. I: Romanische Lautlehre (Leipzig, 1890) and in his Historische
Grammatik der franzosischen Sprache (Heidelberg, 1908), and later by Elise Richter, Beitrdge zur Geschichte
der Romanismen, vol. I: Chronologische Phonetik des Franzosischen bis zum Ende des 18. Jahrhunderts (Halle,
1934) and Max Kfepinsky, Romanica (Praha, 1952).

Two: llisdnw \ iinctv timl Scitini; i>) Luft^iuif't' 105

linguistics. 1 \c lirsi discusses the general explanatory principles for change in lan-
guage (transmission of language from one generaiit)n lo another, internal lin-
guistic change, i.e. modifications of linguistic structures from i- -nal

influences), while focusing on the uUerplay between phonetic chan»v i;v.

the two basic principles advocated by the Neogrammarians.

Sapir's view of linguistic change is a product-oriented view, not a rulc-onented
view: change is brought about by deviations, modifications m speech: t?
social imitation (and selection) the various indi\idual realizations are ih
brated or uniformi/ed.

In Sapir's discussion o{ ( internal ) change, the Neogrammarian idea i>l j^'
change as the principal factor in language change is maintained, but iu.v ...,,.
has to note Sapir's insistence on the intertwining of phonology and grammar: on
the one hand, phonetic change has an impact on the grammatical type or charac-
ter of a language, and. on the other hand, morphological analogy is seen as pre-
serving (or reorganizing) linguistic structures affected by the destructiNe action
o[' phonetic change (p. 58). Sapir also pt>inis to the role of analogy in language
learning, thus anticipating BIoomfield"s \ ieu on analogy as the basic principle of
synchronic productivity.'

In discussing historical change, Sapir touches upon one i\pe t)f linguistic
variety, viz. variation on the time axis. Geographical variation (and its counter-
part, areal diffusion or uniformization) constitutes a second type. Iliis brinj's
Sapir to raise, at the end of his paper, the problem of the classificatmn ot Ian
guages — a meeting-ground for historical linguistics and general linguistics. Sapir
shows the deficiencies of a genetic classification, and then exannnes the pos-
sibility of a "psychological" classification. Contrary to the "polythetical" stand he
was to adopt in his Liuiguagey Sapir proceeds in an analytical way. discuvsing

^ This is the Neogrammarian view as it is exposed in the classic texibtxik of Hermann Paul. Pnnapten det
Sprach^esthichic (Halle. ISSO: later editions have "Prinzipten").

"* See Leonard nioomtield. l.unt^iuii^c (New York. \9?i?>). p. 27.S; "A uramnialical p.iitcfn (M.nttruc Imv o-o
struclion. or siihslilution) is often called an (nuiloi^v. A regular analogy jHrrmils a spv ■
which he has not heard; we say that he utters them on the nnulofiy of similar forms \^

'" See p. 60: "the linguistic stocks we thus get as our largest units of s|H:ech arc l.x> mmu r.-u% l.< «-mc i* Xhe

simplest possible reduction of the linguistic material to he classified "llic tern

ignate a higher-level grouping of languages, above the more narrow "famiU

speaks of "stock groups." With respect to the American Indian field, Sapir m >

lion into "fifty or more distinct linguistic stocks" (see Jt>hn Wesle\ l'<'\w II In

Mexico". Hiireati af Anuricun lihuoloiiw Scxcnih Annual Kcpurt

rcductit)n of this number, see "A Hird's-cye View of American I ani

4()S [reprinted in //»<• Colltrtfil Works of iilwurd Sapir. vol. V. pp

Languages", i.ncyilopacdia Hriiannua 14th edition ( l'*2''). vol *i I

Edward Sapir. vol. V, pp. 45 1(M|. On Sapir's siv ph\lum J

"Edward Sapir's Six-Unit Classification of American Indi.in I

Essays in the Uislorv of Western l.ini^insiies. editeil by IheiHli'

202-244; Michael K. Foster," ITie Impact o\ Sapir s Six Phylum i

American Indian Culture History: a Bibliographic hssay". Amhropftiogtcni iMtguutns h* ,;•*>••

" See /.((/j,v,'(«;,i,'c (New York. I'):i I. pp 144-lSh

106 General Linguistics I

(/) a psychological classification based on the expression of logical contents,'' a
procedure which he finds unapplicable in typological practice;"
(//) a formal/psychological classification based on the relationship between
forms, contents and processes, which he also considers to be not viable;'^
(/7/) a classification based on the degree of unity which the grammatical processes
bring about between the stem and the increments which express relational con-
cepts.^-

Adopting the latter criterion, Sapir proceeds to a classification into three main
types:'" the isolating type (with Chinese as the classical example), the agglutina-
tive type (exemplified by Turkish),'^ and the inflective (= inflectional) type. As
pointed out by Sapir, the term "polysynthetic," often used to designate a fourth
type, in fact refers to the content of a morphological system, and does not stand
on a par with the terms "isolating," "agglutinative" and "inflective". Further,

'- As is clear from the terminology used ("subject-matter or content", "mere form pure and simple") Sapir is
thinking here of Steinthal's psychologically based classification of language types; see Heymann Steinthal, Die
Classifikation der Sprachen, dargestellt als die Entwicklung der Sprachidee (Berlin, 1850) and Charakteristik der
luiuptsdchlichsten Typen des Sprachbaits (Berlin, 1869).

'3 Sapir thus rejects the typological viability of linguistic characterology as practised by Neo-Humboldtian schol-
ars such as Steinthal; see p. 62; "If, now, it has been shown that no necessary correlation exists between parti-
cular logical concepts and the formal method of their grammatical rendering, and if, furthermore, there can not
even be shown to be a hard and fast line in grammatical treatment between concepts of a derivational and con-
cepts of a more definitely relational character, what becomes of the logical category per se as a criterion of hn-
guistic classification on the basis of form ? Evidently it fails us. Of however great psychological interest it might
be to map out the distribution in various linguistic stocks of logical concepts receiving formal treatment, it is
clear that no satisfactory formal classification of linguistic types would result from such a mapping."

'^ See p. 64; as shown by Sapir, the correlation between forms, contents and grammatical processes is never a
one-to-one correlation.

15 This is the criterion which Sapir was going to label "technique" in his Language (New York, 1921); however,
in his 1921 book (pp. 143-144, and p. 153) Sapir does no longer use "inflective" on a par with "isolating" and
"agglutinative," and instead operates with "fusional" and "symbolic." Note that the formulation of the criterion
in the 1911 article blurs the distinction (made in 1921) between "technique" and "degree of synthesis." On the
basic concepts of Sapir's typology (of languages, but also of cultures), see: Stefano Arduini, Fra cuUiira e lin-
guaggio. Un'interpretazione della tipologia di Edward Sapir (dissertation Pisa, 1984) and "Lenguaje, tipologi'a y
cultura. Edward Sapir", Estudios de Lingidstica de la Universidad de Alicante 5 (1988-89)? 275-290; pFerre
Swiggers," "Synchrony' and 'Diachrony' in Sapir 's Language (192iy\ Neuphilologische Mittedungen 94 (1993),
313-322; Jesiis Pena, "La tipologia morfologica de Sapir", in Scripta in memoriam Manuel Taboada, edited by
Manuel Casado Velarde, Antonio Freire Llamas, Jose E. Lopez Pereira and Jose I. Pascual (A Corufia, 1996),
165-177; Maria Xose Fernandez Casas, "El alcance de la tipologia linguistica en la obra de Edward Sapir",
Verba 27 (2000), 249-287, and "Que entendemos por 'tipo linguistico"? El uso polisemico de este concepto en
la obra de Edward Sapir (1884-1939)", Boletin de la Sociedad Espariola de Historiografia Lingidstica 3 (2002),
79-88, and also her monograph Edward Sapir en la lingidstica actual. Lineas de continuidad en la historia de la
lingidstica (Verba, Anexo 54) (Santiago de Compostela, 2004), pp. 67-120.

'6 In his book Language (New York, 1921) Sapir offers a more qualified and refined classification of language
types (see the table of linguistic types and the subsequent comments there, pp. 149-156).

'^ The examples from Chinese and Turkish are taken from Franz Nikolaus Finck's book Die Haupttypen des
Sprachbaus (Leipzig, 1910). Finck's name is misspelled ("Fisk") in the version published in the Popular Science
Monthly, but not in the version reprinted in the 1912 Annual Report of the Smithsonian Institution.

Two: Hislorv. V(irici\ uml .Scmui^ of 1 iinguiif^e 107

Sapir warns his icadL-r against conclalions established hcl\vccn hnpuiMic ivpcs
and stages in historical (and ciihinal or iniclkciii.il) " development.

In the article "Language and lin iioninent." h.ised on .» paper read in
December 1911. Sapir discusses a topic uhich 1 r.ui/ lioas had dealt with that
same year in an article devoted to the impact ol the environment on physical and

social characteristics, with special reference to the situation of immii

in the United States,'" as well as in his ■Introduction" to the first \..
Handbook of American Indian Lani^uai^es. In his discussion of the relationship
between language and einironnient. Sapir displays his kno\N ledge of the Indo-
European and the American Indian field (there are also sporadic references to
African and Melanesian languages). The wide scope of Sapirs investigation of
the problem — which may have been triggered by the type of audience before
which the paper was read — , as well as the specific linguistic vantage point were
to characterize much of Sapir's *'anthropolinguisiic" writings in the I^Kk. m con
trast to his earlier anthropological publications, focused on the Amerindian con
text, and to his later writings, which testify to a shift towards the study of symbol
ism and of the relationship between society and personality.

Although it was published in an anthropological journal. 'Language and
Environment" is a paper that belongs with Sapir's linguistic publications; its
inclusion in this volume of The Collected Works of Edward Sapir is therefore
fully justified. As a matter of fact, not only does Sapir discuss the issue of pos-
sible linguistic correlates of the sociocultural environment, he also displays in this
paper his acquaintance with data from various languages and language groups,
such as Chinook, Eskimo, Haida, Hupa. Iroquois, Karok. Kwakiull. Maidu.
NahuatI, Nootka, Paiute, Salish, Siouan. Takelma. Tewa. llingit. "^'ana. ^'urok. as
well as Melanesian, Malayan, Mon-Khmer. Chinese, Caucasian languages.
Semitic languages, Hottentot, Ewe, and various European languages (Danish,
French, English, Hungarian, Latvian, Portuguese). Ilie central part of the paper

'** Such correlations had been posited, e.g.. biy August Schleicher [ IS2I - 1S^.S| and the vrhiHtl
guistics," and. shortly before the publication of Sapir's article, by the Russian linguist .Vikol.i| M

'^ Franz Boas, inslabiiily of Human types", in I'lifun on hiurracuil Prohleim ('.■ir.xu.r.i, .,:,.! :.■ :lu

Universal Races Coniiress Helil al the University of l.nmlon. July 2(t-2*J. I^ll (cd. ( i

pp. 99-103: see also Chani;es in Bodily Form of Deseemiants of Ininmr.iniy i^

Congress, Second session. Washington. 191 1; reprinted New ^ork. I9:i

Descendants of Immigrants ". American Anihropolof'isi n >. U i I'M "'i ^

fiir Elhnolugie 45 (1913). 1-22).

-" Franz Boas. "Introduction", in Nanilhook of American hulum lai.
Bureau of American Ethni)logy. Washington. 191 1 ),pp. 1 N3. Ilie inllui r
appears not only from the general topic of the paper, but also from the i^
egories of thought " (p. 236). "modes of thought " (:in M-niiiid m .i l.iiu'
concepts" (p. 236) and "mental stock" (p. 2.^6).

108 General Linguistics I

contains much typological information supplementing that given in the "History
and Varieties" article-' of 1911, and anticipating the comprehensive treatment in
Langiiage.--

The prime importance of "Language and Environment" lies in its methodolo-
gical contribution, which is threefold.

First, Sapir clarifies the notion of "environment." which normally should be
limited to what lies outside the will of man, but which in the present discussion is
used to include physical environment and social (cultural) environment. As Sapir
points out, the physical environment always exerts its impact through the social
prism, which is made up of needs and interests" affecting groups of individuals.
Strictly speaking, the environment cannot influence groups of individuals: it acts
through social forces, and these may be subject to changes caused by the envi-
ronment (pp. 226-227).

In the second place, Sapir makes clear that in the study of the intricate rela-
tionship between language-^ and environment, it is essential to distinguish within
language the lexical content side, the phonetic system, and grammatical form. It
is especially at the level of the (specialized) lexical content that inferences (as to
physical" and cultural-^ environment) can be drawn and indeed have been drawn
(as can be seen from the published record in the field of "linguistic archeolo-
gy"^^). It is also the make-up of the vocabulary of a language that allows, to a cer-
tain extent, to establish correlations between the mind of primitive peoples and
the primitive nature of their language, or better lexicon (as a set of context-
bound words, characterised by a strong "descriptive" orientation; see p. 231). A
further elaboration of diachronic insights to be derived from the character of a
vocabulary can be found in Sapir's Time Perspective in Aboriginal Culture.-^

21 Compare p. 242 the remarks on English with "The History and Varieties of Human Speech", p. 57 and p. 67.

22 Language (New York. 1921), chapters IV, V and VI.

23 See p. 226 and pp. 228-229.

2'' Language is defined as a "complex of symbols" (p. 227).

25 See p. 229.

2^- See p. 232-233.

2'^ See p. 232, where Sapir uses the term "linguistic archeology" with reference to work on Indo-European lan-
guage and culture (e.g.. O. Schrader's work) and to research perspectives for the American Indian field. For this
type of study, the term "linguistic paleontology" has also been used; see Yakov Malkiel, "Linguistic
Paleontology (Geology, Archeology)?", Romance Philology 28 (1975), p. 600. and Richard A. Diebold,
"Paleontology, Hnguistic", in The Encyclopedia of Language and Linguistics (Oxford, 1994), vol. 6, pp.
2906-2913.

28 E. Sapir, Time Perspective in Aboriginal Culture: A Study in Method (Ottawa, 1916) [reprinted in The
Collected Works of Edward Sapir, vol. IV, pp. 31-119].

Two: ffistorv. \ \irici\ mul St'llitifi of LunfiUtiiH' 109

The lliird incihoclolouical conii itniiuui ol ihc jKipcr is a lcsM>n of caution. In
line with Boas's rclicL-nt slatciiiciu on possible corrclalions between language.
race and culture. Sapir warns us auauist premature generalizations'' concerning
the relationship between language and environment. As Sapir ptunts out. no
strong correlations hold between phonetic form and environment.' nor hcl\^ecn
grammatical form " and environment. Whereas some kind of "association** may
be postulated, w ith due reservations, lor "some primitive stage" (p. 240) of a lin-
guistic community, the history of languages and cultures ' is not parallel, since
language and culture ha\e a different evolutionary rhythm, linguistic form being
extremely conser\ alive. "One necessary consequence of this is that the forms of
language will in course o\ time cease to symboli/e those of culture, and this is our
main thesis" (p. 241). in addition, cultural phenomena are much more liable lo
diffusion and to (conscious) adoptit)n or borrt)wing, since they answer immediate
needs. Hie methodological ct)nclusion to be drawn from this is that, hisiorically
speaking, there is a split' of linguistic form ' and culture (p. 241). and. geo-
graphically and typologically speaking, there is no correlation ' between mt)rpho-
logical system and environment (p. 237-238): morphological similarilv can be
observed in extremely diverse environments, and, conversely, within the same
physical and cultural environment, we often find languages with widely diverging
grammatical forms."'

Pierre Swigcjers

-^ Sl'c the opening sentence of the paper and also p. 230.

^' As examples Sapir refers to the scattered distributiDn ol pitcli accent and navtl vimck in Ihc »ofld'» lan-
guages.

^' With "grammatical form" Sapir means morpliology and syntax: morph«)lov;\ is delmcd .t^^ .1. dim- *Mih eani
matical categories and the formal structure oi words (p. 22S: on p. 2^t^ nu>rpholog\ in -

grammatical categories and the formal methods of expressing categories), while ^^'" '» '-^ i--

methods for combinations.

'- It is interesting to note that Sapir also speaks of the 'formls) t)f culture" (p. 241).

^^ Sapir presents this as a "hypothetical explanation" for the failure ii> causaiU con.
guage: the metaphor he uses is that of two men starting i>n a journey in ih-- >> '•"' ^
ging as time goes on (p. 242).

-^ Sapir also speaks of the "formal grmiiulwork" of language (p. 238. p. 241).

'*' Except for those cases where there is grammatical signalling of cultui.i
ical environment; however, as Sapir shows, we are then dealinc «iih 'hi-
with grammatical forms us siuh.

^ See also /.,m,i,'(/,/,i,'.' ( New >ork. U>2I ). p. 227-231.

THE HISTORY AND VARIETIES OF HUMAN SPEPX'H'

IlT Da. KDWARD SAITR

THE CANADIAN OEOUKJICAL BIBVLT

PERHAPS no single feature so markedly ecta off man from the rcnt
of the animal world a.s the gift of speech, which ho alone pos-
sesses. No community of normal human beings, be tlieir advance in
culture ever so slight, has yet been found, or is ever likely to bo found,
who do not communicate among themselves by means of a complex
system of sound 6}"nibols; in other words, who do not make use of a
definitely organized spoken langnage. It is indeed one of the para-
doxes of linguistic science that some of the most complexly organized
languages are spoken by so-called primitive peoples, while, on the otlier
hand, not a few languages of relatively simple structure are found
among peoples of considerable advance in culture. Relatively to the
modern inhabitants of England, to cite but one instance out of an in-
definitely large number, the Eskimos must be considered as rather
limited in cultural development. Yet there is just as little doubt that
in complexity of form the Eskimo language goes far beyond English.
I wish merely to indicate that, however much we may indulge in speak-
ing of primitive man, of a primitive language in the true sense of the
word we find nowhere a trace. It is true that many of the lower ani-
mals, for example birds, communicate by means of various cries, yet
no one will seriously maintain that such cries are comparable to the
conventional words of prcscnt-day human speech ; at best thoy may be
compared to some of our interjections, which, however, falling outride
the regular morphologic and s^mtactic frame of speech, are least typ-
ical of the language of human beings. We can thus safely make the
absolute statement that language is typical of all human communities
of to-day, and of such previous times as we have historical knowle<1ge
of, and that language, aside from reflex cries, is just as untypical of
all non-human forms of animal life. Like all other fonns of human
activity, language must have its history.

Much has been thought and written about the history of language.
Under this term may be included two more or lesa di.-^tinct lines of
inquiry. One may either trace the changes undergone by a particular
language or group of languages for as long a period as tJic evidence at
hand allows, or one may attemj)t to pass beyond the limits of h: ' '
ically recorded or reconstructed sj)eech, to reconstruct the ult;:

'Lecture delivered at the Univcrsitj of Pennsjlvania Muvum, April 1, 1911.

1 12 General Linguistics I

origin of speech in general, and to connect these remote origins by
means of reconstructed lines of development with historically attested
forms of speech. Superficially the latter sort of inquiry is similar in
spirit to the labors of the evolutionary biologist, for in both apparently
heterogeneous masses of material are, by direct chronologic testimony,
inference, analogy and speculation, reduced to an orderly historical
sequence. As a matter of fact, however, the reconstruction of lin-
guistic origins and earliest lines of development is totally different in
kind from biological reconstruction, as we shall see presently.

Taking up the history of language in the sense in which it was first
defined, we find that there are two methods by which we can follow
the gradual changes that a language has undergone. The first and
most obvious method is to study the literary remains of the various
periods of the language of which we have record. It will then be found
that not only the vocabulary, but just as well the phonetics, word
morphology, and syntactic structure of the language tend to change
from one period to another. These changes are always very gradual
and, within a given period of relatively short duration, slight or even
imperceptible in amount. Nevertheless, the cumulative effect of these
slight linguistic changes is, with the lapse of time, so great that the
form of speech current at a given time, when directly compared with
the form of speech of the same language current at a considerably
earlier time, is found to differ from the latter much as it might from
a foreign language. It is true that the rate of change has been found
to be more rapid at some periods of a language than at others, but it
nevertheless always remains true that the changes themselves are not
violent and sudden, but gradual in character. The documentary study
of language history is of course the most valuable and, on the whole,
the most satisfactory. It should not be denied, however, that there
are dangers in its use. Literary monuments do not always accurately
reflect the language of the period; moreover, orthographic conserva-
tism hides the phonetic changes that are constantly taking place. Thus,
there is no doubt that the amount of change that English has under-
gone from the time of Shakespeare to the present is far greater than
a comparison of present-day with Elizabethan orthography would lead
the layman to suppose, so much so that I am quite convinced the great
dramatist would have no little difficulty in making himself under-
stood in Stratford-on-Avon to-day. For some languages a consider-
able amount of documentary historical material is available ; thus, the
literary monuments that enable us to study the history of the English
language succeed each other in a practically uninterrupted series from
the eighth century a.d. to the present time, while the course of develop-
ment of Greek in its various dialects can be more or less accurately
followed from, the ninth century B.C., a conservative date for the
Homeric poems, to the present time.

Two: llistorw V'liriciv mul Sriiins; "I I nn^uage 1 13

For some, in fact for most languages, however, liUrary rnonuujcnta
are either not forthcoming at all or else are restricted to a single
period of short duration. At first eight it would seem that the sci-
entific study of such languages would liave to be limited to purely
descriptive rather than historical data. To a considerable extent thia
is necessarily true, yet an intensive study will always yield at lenat
some, oftentimes a great deal of, information of a historical char-
acter. This historical reconstruction on the basis of purely descrip-
tive data may proceed in two ways. It is obvious that the various
phonetic and grammatical features of a lang\iage at any given time
are of unequal antiquity, for they are the resultants of changes that
have taken place at very different periods; hence it is reasonable
to suppose that internal evidence would, at least within modest
limits, enable one to reconstruct the relative chronology of the lan-
guage. Naturally one must proceed very cautiously in reconstruct-
ing by means of internal evidence, but it is oftentimes surprising how
much the careful and methodically schooled student can accomplish in
this way. Generally speaking, linguistic features that are irregular in
character may be considered as relatively archaic, for they are in the
nature of survivals of features at one time more widely spread. Not
infrequently an inference based on internal evidence can be corrobo-
rated by direct historical testimony. One example will suffice here,
vre have in English a mere sprinkling of noun plurals in -en, such aa
brethren and oxen. One may surmise that nouns such as these are but
the last survivals of a type formerly existing in greater abundance, and
indeed a study of Old English or Anglo-Saxon demonstrates that noun
plurals in -en were originally found in great number but were later
almost entirely replaced by plurals in -s. There is, however, a far
more powerful method of reconstructing linguistic history from de-
scriptive data than internal evidence. This is the comparison of genet-
ically related languages.

In making a survey of the spoken languages of the world, we soon
find that though they differ from each other, they do bo in quite vary-
ing degrees. In some cases the differences are not great enough to
prevent the speakers of the two languages from understanding each
other with a fair degree of ease, under which circumstances we are *pt
to speak of the two forms of speech as dialects of a single language; in
other cases the two languages are not mutually intelligible, but, aa in
the case of English and German, present so many eimilaritios of detail
that a belief in their common origin seems warranted and indecil nec-
essary; in still other cases the two languages are at first glance not at
all similar, but reveal on a closer study so many fundamental traita
in common that there seems just ground for suspec^ting a common
origin. If other languages can be found which serve to lesion the

1 14 General Linguistics I

cliasm between the two, and particularly if it is possible to compare
them in the form in which they existed in earlier periods, this sus-
picion of a common origin may be raised to a practical certainty.
Thus, direct comparison of Russian and German would certainly yield
enough lexical and grammatical similarities to justify one in suspect-
ing them to have diverged from a common source; the proof of such
genetic relationship, however, can not be considered quite satisfactory
until the oldest forms of German speech and Germanic speech generally
have been compared with the oldest forms of Slavic speech and until both
of these have been further compared with other forms of speech, such
as Latin and Greek, that there is reason to believe they are genetically
related to. When such extensive, not infrequently difficult, comparisons
have been effected, complete evidence may often be obtained of what
in the first instance would have been merely suspected. If all the forms
of speech that can be shown to be genetically related are taken to-
gether and carefully compared among themselves, it is obvious that
much information will be inferred as to their earlier undocumented
history; in favorable cases much of the hypothetical form of speech
from which the available forms have diverged may be reconstructed
with a considerable degree of certainty or plausibility. If under the
term history of English we include not only documented but such re-
constructed history as has been referred to, we can say that at least in
main outline it is possible to trace the development of our language
back from the present day to a period antedating at any rate 1500 B.C.
It is important to note that, though the English of to-day bears only
a faint resemblance to the hypothetical reconstructed Indogermanic
speech of say 1500 B.C., there could never have been a moment from
that time to the present when the continuity of the language was
broken. From our present standpoint that bygone speech of 1500 B.C.
was as much English as it was Greek or Sanskrit. The history of the
modem English words foot and its plural feet will illustrate both the
vast difference between the two forms of speech at either end of the
series and the gradual character of the changes that have taken place
within the series. Without here going into the actual evidence on
which the reconstructions are based, I shall merely list the various
forms which each word has had in the course of its history. Starting,
then, with foot — feet, and gradually going back in time, we have
fut—flt, fdt—fet,^ fot—fite, fot—fdte, fot—foti, fot—foti, fot—fdtir,
fot — fotiz, fot — foiis, fot — fotes, fod — fades, and finally pod — podes,
beyond which our evidence does not allow us to go ; the last forms find
their reflex in Sanskiit pad — pddas.

All languages that can be shown to be genetically related, that is,
to have sprung from a common source, form a historic unit to which
the terra linguistic stock or linguistic family is applied. If, now, we

T\v(r nisforv. \ 'arivix ami Sating of l.un^mit^t' 115

were in a jwsition to prove that all known forma of Bpti-ch could be
classified into a single linguistic stock, tlic apparent parallel tlwTe re-
ferred to between linguistic and biological reconstruction would be •
genuine one. As it is, we must content ourselves with operating wiUi
distinct and, as far as we can tell, genetically unrelated linguietu-
stocks. The documentary evidence and the reconstructive evidence
gained by comparison enable us to reduce the bewildering maw of
known languages to a far smaller number of such larger stock group*,
yet the absolute number of these latter groups still remains disquiet-
ingly large. The distribution of linguistic stocks presents great irreg-
ularities. In Europe there are only three such represent<^d : Uic
Indogermanic or Aryan, which embraces nearly all the better known lan-
guages of the continent; the Ural-Altaic, the best known representativca
of which are Finnish, Hungarian and Turkish; and the Basque of
southwestern France and northern Spain. On the other hand, that
part of aboriginal North America which lies north of Mexico alone
embraces fifty or more distinct linguistic stocks. Some stocks, as, for
instance, the Indogermanic just referred to and the Algonkin of North
America, are spread over vast areas and include many peoples or tribes of
varying cultures ; others, such as the Basque and many of the aboriginal
stocks of California, occupy surprisingly small territories. It is {>os-
sible to adopt one of two attitudes towards this phenomenon of the
multiplicity of the largest known genetic speech aggregates. On the
one hand one may assume that the disintegrating effects of gradual
linguistic change have in many cases produced such widely differing
forms of speech as to make their comparison for reconstructive pur-
poses of no avail, in other words, that what appear to us to-day to be
independent linguistic stocks appear such not because they are in fact
historically unrelated, but merely because the evidence of such his-
torical connection has been so obscured by time as to be practically
lost. On the other hand, one may prefer to see in the existence of
mutually independent linguistic stocks evidence of the independent
beginnings and development of human speech at different timcB and
places in the course of the remote history of mankind; there is every
reason to believe that in a similar manner many religious concept* and
other forms of human thought and activity found widely distribut^xl
in time and place have had multiple origins, yet more or \e^9 paral-
lel developments. It is naturally fruitless to attempt to decide be-
tween the monogenctic and polygcnctic standpoints here briefly out-
lined. All that a conservative student will care to do is to shrug hi«
shoulders and to say, "Thus far we can go and no further." It
ehould be said, however, that more intensive study of linguistic data
is from time to time connecting st<x:ks that had hitherto been hxikcd
upon as unrelated. Yet it can hardly be expected that serious research
will ever succeed in reducing the {)re?ent Babel to a pristine unity.

VOL. LXXIX. -4.

1 1 6 General Linguistics I

Although we can not demonstrate a genetic unity of all forms of
human speech, it is ixiteresting to observe that there are several funda-
mental traits that all languages have in common. Perhaps these
fundamental similarities are worthy of greater attention than they
generally receive and may be thought by many to possess a high de-
gree of significance. First of all, we find that in every known lan-
guage use is made of exactly the same organic apparatus for the pro-
duction of speech, that is, the glottal passage in the larynx, the nasal
passages, the tongue, the hard and soft palate, the teeth and the lips.
The fact that we are accustomed to consider all speech as self-evidently
dependent on these organs should not blind us to the importance of
the association. There is, after all, no d priori reason why the com-
munication of ideas should be primarily through sound symbols pro-
duced by the apparatus just defined; it is conceivable that a system
of sound 63Tnbols of noises produced by the hands and feet might have
been developed for the same purpose. As a matter of fact, there are
many systems of thought transference or language in the widest sense
of the word, as a moment's thought will show, that are independent of
the use of the ordinary speech apparatus. The use of writing will occur
to every one as the most striking example among ourselves. Among
primitive peoples we may instance, to cite only a couple of examples
of such subsidiary forms of language, the gesture language of the
Plains Indians of North America and the very highly developed drum
language of several African tribes. From our present point of view it
is significant to note that these and other such non-spoken languages
are either, as in the case of practically all systems of writing, them-
selves more or less dependent on a phonetic system, that is, speech in the
ordinary sense of the word, or else are merely auxiliary systems in-
tended to replace speech only under very special circumstances. The
fact then remains that the primary and universal method of thought
transference among human beings is via a special articulating set of
organs. Much loose talk has been expended by certain ethnologists on
the relatively important place that gesture occupies in the languages
of primitive peoples, and it has even been asserted that several so-
called primitive languages are unintelligible without the use of ges-
ture. The truth, however, is doubtless that the use of gesture is as-
sociated not with primitiveness, but rather with temperament. The
Russian Jew and the Italian, for instance, non-primitive as they are,
make a far more liberal use of gestures accompanying speech than any
of the aborigines of North America.

If we examine in a large way the structure of any given language,
we find that it is further characterized by the use of a definite phonetic
system, that is, the sounds made use of in its words are reducible to a
limited number of consonants and vowels. It does not seem to be true.

Two: Uistorw Variety uml Scifini^ of I .iitifiuai>e 117

certain contradicting statemcnta notwiUistanding, that languages are
to be found in which this phonetic dcfinitcncss is lacking and in which
individual variation of pronunciation takca place praf tirally without
limit. It is of course freely granted that a certain amount of aound
variation exista in every language, but it is important to note that such
variation is always very limited in range and always takes place about
a well-defined center. AH known forms of speech, then, operate with a
definite apparatus of sounds; statements to the contrary will in moet
cases be found to rest either on a faulty perception on the part of the
recorder of sounds unfamiliar to his ear or on his ignorance of regular
Boiind processes peculiar to the language. Naturally the actual phonetic
systems found in- various languages, however much they may resemble
each other in this fundamental trait of definiteness, differ greatly in
content, that is in the sounds actually employed or neglected. Thia ia
inevitable, for the vast number of possible and indeed existing speech
sounds makes an unconscious selection necessary. Even so, however,
it is at least noteworthy with what persistency such simple vowel
sounds as a and t and such consonanta as n and s occur in all parts of
the world.

Even more than in their phonetic systems languages are found to
differ in their morphologies or grammatical structures. Yet also in
this matter of grammatical structure a survey from a broad point of
view discloses the fact that there are certain deep-ljing similarities,
very general and even vague in character, yet significant. To begin
with, we find that each language is characterized by a definite and,
however complex, yet strictly delimited grammatical system. Some
languages exhibit a specific type of morphology with greater clearness
or consistency than others, while some teem with irregularities; yet in
every case the structure tends to be of a definite and consistently car-
ried out type, the grammatical processes employed are quite limited in
number and nearly always clearly developed, and the logical categories
that are selected for grammatical treatment are of a definite sort and
number and expressed in a limited, however large, number of gram-
matical elements. In regard to the actual content of the various mor-
phologies, we find, as already indicated, vast differences, yet here again
it is important to note with what persistence certain fundamental log-
ical categories are reflected in the granmiatical systems of practically
all languages. Chief among these may be considered the clear-cut dis-
tinction everywhere made between denominating and predicating tenna,
that is between subject and predicate, or, roughly speaking, between
substantive and verb. This does not neccbsarily imply that we have in
all cases to deal with an actual difference in plionetic form between
noun and verb, though as a matter of fact such differences are generally
found, but simply that the structure of the sentence is such as to show

118 General Linguistics I

clearly that one member of it is felt by the speaker and hearer to have
a purely denominating oflBce, another a purely predicating one. It may
be objected that in Chinese, for instance, there is no formal distinction
made between noun and verb. True, but the logical distinction of sub-
ject and predicate is reflected in the form of the Chinese sentence, inas-
much as the subject regularly precedes the predicate; thus, while the
same word may be either noun or verb, in any particular sentence it
necessarily is definitely one and not the other. Other fiindamental log-
ical categories will, on a more complete survey, be found to be subject to
grammatical treatment in all or nearly all languages, but thia is not the
place to be anything but merely suggestive. Suffice it to remark on the
wide-spread systematizing of personal relations; the wide-spread devel-
opment of ideas of tense, number and syntactic case relations; and the
clear grammatical expression everywhere or nearly everywhere given to
the largely emotional distinction of declarative, interrogative and im-
perative modes.

Granted that there are certain general fundamental traits of sim-
ilarity in all known languages, the problem arises of how to explain
these similarities. Are they to be explained historically, as survivals
of features deep-rooted in an earliest form of human speech that,
despite the enormous differentiation of language that the lapse of ages
has wrought, have held their own to the present day, or are they to be
explained psychologically as due to the existence of inherent human
mental characteristics that abide regardless of time and race? If the
latter standpoint be preferred, we should be dealing with a phenomenon
of parallel development. It is of course impossible to decide cate-
gorically between the two explanations that have been offered, though
doubtless the majority of students would incline to the psychological
rather than to the historical method. At any rate, it is clear that we
can not strictly infer a monogenetic theory of speech from the funda-
mental traits of similarity that all forms of speech exhibit. Yet even
though these are of psychologic rather than historic interest, it is im-
portant to have demonstrated the existence of a common psychological
substratimi, or perhaps we had better say framework, which is more or
less clearly evident in all languages. This very substratum or frame-
work gives the scientific study of language a coherence and unity quite
regardless of any considerations of genetic relationship of languages.

In spite of the fact that, as we have seen, no tangible evidence can
be brought to bear on the ultimate origin or origins of speech, many
attempts have been made, particularly in the first half of the nineteenth
century, when it was more common for historical and philosophical
problems of extreme difficiilty to be attacked with alacrity, to point out
the way in which human speech originated or at least might have orig-
inated. From the very nature of the case, these attempts could not but

Two: f/istorv. \ (irictv ami Srttin}> of I ani'uasn' 1 19

be deductive in method; hence, however plausible or ingr-nioua in them-
selves, they have at best a merely speculative, not a genuinely ncicntific
interest. We may therefore dispense with anything like a detailed
inquiry into or criticism of these theories. Two of the moflt popular
of them may be respectively termed the onomatopoctic or Round imita-
tive and the exclamatory theories. According to the former, the first
words of speech were onomatopoctic in character, that is, attompta to
imitate by the medium of the Imman organs of speech, the varioos criea
and noises of the animate and inanimate world. Thus, the idea of a
"hawk" would come to be expressed by an imitative vocable based on
the actual screech of that bird ; the idea of a " rock " might be eiprea-MMi
by a combination of sounds intended in a crude way to reproduce the
noise of a rock tumbling down hill or of a rock striking against the butt
of a tree; and so on indefinitely. In course of time, as these imitative
words by repeated use became more definitely fixed in phonetic form,
they would tend to take on more and more the character of conventional
6ound-s}'mbol3, that is of words, properly speaking. The gradual pho-
netic modifications brought on in the further course of time would
finally cause them to lose their original onomatopoctic form. It may
be freely granted that many words, particularly certain nouns and verba
having reference to auditory phenomena, may have originated in this
way; indeed, many languages, among them English, have at various
times, up to and including the present, made use of such onomatopoctic
words. It is difficult, however, to see how the great mass of a vocab-
ulary, let alone a complex system of morphology and sjmtai, could have
arisen from an onomatopoctic source alone. The very fact that
onomatopoctic words of relatively recent origin are found here and
there in sharp contrast to the overwhelmingly larger non-onomatopoetic
portion of the language accentuates, if anything, the difficulty of a gen-
eral explanation of linguistic origins by means of the onomatopoctic
theory.

The exclamatory theory, as its name implies, would find the
earliest form of speech in reflex cries of an emotional character. The.<»e
also, like the hypothetical earliest words of imitative origin, would in
course of time become conventionalized and sooner or later so modified
in phonetic form as no longer to betray their exclamatory origin. The
criticisms urged against the onomatopoctic theory apply with perhapa
even greater force to the exclamatory one. It is, if anything, even more
difficult here than in the former case to see how a small vtx'abulary
founded on reflex cries could develop into such complex linguistic ays-
tems aa we have actually to deal with. It is further significant that
hardly anywhere, if at all, do the interjections play any but an incon-
siderable, almost negligible, part in the lexical or grammatical ma-
chinery of language. An appeal to the languages of primitive peoples

120 General Linguistics I

in order to find in them support for either of the two theories referred
to is of little or no avail. Aside from the fact that their elaborateness
of structure often seriously militates against our accepting them as
evidence for primitive conditions, we do not on the whole find either
the onomatopoetic or exclamatory elements of relatively greater impor-
tance in them than elsewhere. Indeed the layman would be often sur-
prised, not to say disappointed, at the almost total absence of onomato-
poetic traits in many American Indian languages, for instance. In
Chinook and related dialects of the lower course of the Columbia, ono-
matopoesis is developed to a more than usual extent, yet, as though to
emphasize our contention with an apparent paradox, hardly anywhere
is the grammatical mechanism of a subtler, anything but primitive
character. We are forced to conclude that the existence of onomato-
poetic and exclamatory features is as little correlated with relative
primitiveness as we have found the use of gesture to be. As with the
two theories of origin we have thus briefly examined, so it will be
found to be with other theories that have been suggested. They can
not, any of them, derive support from the use of the argument of sur-
vivals in historically known languages; they all reduce themselves to
merely speculative doctrines.

So much for general considerations on language history. Return-
ing to the gradual process of change which has been seen to be charac-
teristic of all speech, we may ask ourselves what is the most central or
basic factor in this never-ceasing flux. Undoubtedly the answer must
be : phonetic change or, to put it somewhat more concretely, minute or
at any rate relatively trivial changes in pronunciation of vowels and
consonants which, having crept in somehow or other, assert themselves
more and more and end by replacing the older pronunciation, which
becomes old-fashioned and finally extinct. In a general way we can
understand why changes in pronunciation should take place in the
course of time by a brief consideration of the process of language learn-
ing. Roughly speaking, we learn to speak our mother-tongue by
imitating the daily speech of those who surround us in our childhood.
On second thoughts, however, it will be seen that the process involved
is not one of direct imitation, but of indirect imitation based on infer-
ence. Any given word is pronounced by a succession of various more
or less complicated adjustments of the speech organs. These adjust-
ments or articulations give rise to definite acoustic effects, effects which,
in their totality, constitute speech. Obviously, if the child's imitative
efforts were direct, it would have to copy as closely as possible the speech
articulations which are the direct source of what it hears. But it is
still more obvious that these speech articulations are largely beyond the
power of observation and hence imitation. It follows that the actual
sounds, not the articulations producing them, are imitated. This

Two: flistorw Vnricty und Scuiny, of Liini^uai^f 121

means that the child is Buhjtxt to a very considerahlc period of rundum
and, of course, wholly involuntary eiperimcnting in tlie production of
6uch articulations as would tend to produce sounda or comhinationa of
sounds approximating more or less closely those the child hears. In
the course of this experimenting many failures are produced, many
partial successes. The articulations producing the former, inasmuch
as they do not give results that match the sounds which it was intended
to imitate, have little or no associative power with these soundfl, hence
do not readily form into habits; on the other hand, articulations that
produce successes or comparative successes will naturally tend to be-
come habitual. It is easy to see that the indirect manner in which
speech articulations are acquired necessitates an element of error, very
slight, it may be, but error nevertheless. The habitual articulationa
that have established themselves in the speech of the child will yield
auditory results that approximate so closely to those used in speech by
its elders, that no need for correction will be felt And yet it is in-
evitable that the sounds, at least some of the sounds, actually pro-
nounced by the child will differ to a minute extent from the correspond-
ing sounds pronounced by these elders. Inasmuch as every word is
composed of a definite number of sounds and as, furthermore, the lan-
guage makes use of only a limited number of sounds, it follows that
corresponding to every sound of the language a definite articulation
will have become habitual in the speech of the child; it follows imme-
diately that the slight phonetic modifications which the child has intro-
duced into the words it uses are consistent and regular. Thus if a
vowel a has assumed a slightly different acoustic shade in one word, it
will have assumed the same shade in all other cases involving the old
a-voWel used by its elders, at any rate in all other cases in whiili tlu?
old a-vowel appears under parallel phonetic circumstances.

Here at the very outset we have illustrated in the individual the reg-
ularity of what have come to be called phonetic laws. The term " pho-
netic law " is justified in so far as a common tendency is to be di.scovcred
in a large number of individual sound changes. It is important, how-
ever, to understand that phonetic law is a purely historic concept, not
one comparable to the laws of natural science. The latter may be paid to
operate regardless of particular times and places, while a phonetic law
is merely a generalized statement of a process that took place in a
restricted area within a definite period of time. The real difficulty in
the understanding of phonetic change in language lies not in the fact
of change itself, nor in the regularity with which such change proceeds
in all cases affected, but, above all, in the fact that phonetic changes are
not merely individual, but social phenomena ; in other words, that the
speech of all the members of a community in a given time and place
undergoes certain regular phonetic changes. Without here attempting

122 General Linguistics I

to go into the details of this process of the transformation of an indi-
vidual phonetic peculiarity into a social one, we will doubtless not be
far wrong in assuming that uniformity is at first brought about by a
process of unconscious imitation, mutual to some extent, among the
younger speakers of a restricted locality, later, perhaps, by the half-
conscious adoption of the new speech peculiarity by speakers of neigh-
boring localities, until, finally, it has spread either over the entire area
in which the language is spoken or over some definite portion of it.
In the former case the historic continuity of the language as a unit is
preserved, in the latter a dialectic peculiarity has asserted itself. In
the course of time other phonetic peculiarities spread that serve to
accentuate the dialectic division. However, the ranges of operation of
the different phonetic laws need not be coterminous, so that a network
of dialectic groupings may develop. At least some of the dialects will
diverge phonetically more and more, until in the end forms of speech
will have developed that deserve to be called distinct languages. It can
not be denied that, particularly after a considerable degree of diverg-
ence has been attained, other than purely phonetic characteristics
develop to accentuate a difference of dialect, but every linguistic student
is aware of the fact that the most easily formulated and, on the whole,
the most characteristic differences between dialects and between lan-
guages of the same genetic group are phonetic in character.

True, some one will say, changes of a purely phonetic character can
be shown to be of importance in the history of language, but what of
changes of a grammatical sort? Are they not of equal or even greater
importance ? Strange as it may seem at first blush, it can be demon-
strated that many, perhaps most, changes in grammatical form are at
last analysis due to the operation of phonetic laws. Inasmuch as these
phonetic laws affect the phonetic form of grammatical elements as well
as of other linguistic material, it follows that such elements may get
to have a new bearing, as it were, brought about by their change in
actual phonetic content; in certain cases, what was originally a single
grammatical element may in this way come to have two distinct forms,
in other cases two originally distinct grammatical elements may come
to have the same phonetic appearance, so that if circumstances are
favorable, the way is paved for confusion and readjustment. Briefly
stated, phonetic change may and often does necessitate a readjustment
of morphologic groupings. It will be well to give an example or two
from the history of the English language. In another connection we
have had occasion to briefly review the history of the words foot and
feet. We saw that there was a time when these words had respectively
the form fot and foti. The final t-vowel of the second word colored,
by a process of assimilation which is generally referred to as " umlaut,"
the of the first syllable and made it '6, later unrounded to e; the final

Two: Historw Vuriclx luul Scititiii <>t luni^iuif^c

t, after being dulled to an e, finally dropped off altogether. 'I'li^ form
foti thus step by step developed into the later fit, which is the normal
Anglo-Saxon form. Note the result. In foti and other words of it«
type the plural is expressed by a distinct sufTn -i, in fii, aa in modem
English feet, and in words of corresponding form it is exprcRfied by a.D
internal change of vowel. Thus an entirely new grammatical feature
in English, as also in quite parallel fashion in German, waa brought
about by a series of purely phonetic changes, in themselves of no gram-
matical significance whatever.

Such grammatical developments on the basis of phonetic changea
have occurred with great frequency in the history of language. In Uic
long run, not only may in this way old grammatical features be lost
and new ones evolved, but the entire morphologic type of the language
may undergo profound modification. A striking example is furnished
again by the history of the English language. It is a well-known
feature of English that absolutely the same word, phonetically speaking,
may often, according to its syntactic employment, be construed as verb
or as noun. Thus we not only love and kiss, but we also give our love
or a kiss, that is, the words love and kiss may be indifferently used to
predicate or to denominate an activity. There are so many examples
in English of the formal, though not syntactic, identity of noun stem
and verb stem that it may well be said that the English language is
on the way to become of a purely analytic or isolating t}'pc, more or
less similar to that of Chinese. And yet the typical Indogermanic
language of earlier times, as represented say by Latin or Greek, always
makes a rigidly formal, not merely syntactic, distinction between these
fundamental parts of speech. If we examine the history of this truly
significant change of type in English, we shall find that it has been
due at last analysis to the operation of merely phonetic laws. The
original Anglo-Saxon form of the infinitive of the verb kiss was cyssan,
while the Anglo-Saxon form of the noun kiss was cyss. The forms in
early middle English times became dulled to kissen and kiss, respect-
ively. Final unaccented -n later regularly dropped off, so that the
infijaitive of the verb came to be kisse. In Chaucer's day the verb and
the noun were still kept apart as kisse and kiss, respectively ; later on,
as a final unaccented -e regularly dropped off, kisse became kiss, so
that there ceased to be any formal difference between the verb and
noun. The history of the Anglo-Saxon verb lufian " to love " and
noun lufu " love " has been quite parallel ; the two finally became con-
fused in a single form luv, modem English love. Once the pace has
been set, so to speak, for an interchange in English between verbal and
nominal use of the same word, the process, by the working of simple
analogy, is made to apply also to cases where in origin we have to deal
with only one part of speech ; thus we may not only have a sick stomach

124 General Linguistics I

but we may stomach an injury (noun becomes verb), and conversely
VI Q may not only write up a person, but he may get a write up (verb
becomes noun). It haa, I hope, become quite clear by this time how
the trivial changes of pronunciation that are necessitated by the very
process of speech acquirement may, in due course of time, profoundly
change the fundamental characteristics of language. So also, if I
may be pardoned the use of a simile, may the slow erosive action of
water, continued through weary ages, profoundly transform the char-
acter of a landscape. If there is one point of historic method rather
than another that the scientific study of language may teach other
historical sciences, it is that changes of the greatest magnitude may
often be traced to phenomena or processes of a minimal magnitude.

On the whole, phonetic change may be said to be a destructive or at
best transforming force in the history of language. Reference has
already been made to the influence of analogy, which may, on the con-
trary, be considered a preservative and creative force. In every lan-
guage the existing morphological groups establish more or less definite
paths of analogy to which all or practically all the lexical material is
subjected; thus a recently acquired verb like to telegraph in English is
handled in strict analogy to the great mass of old verba with their vary-
ing forms. Such forms as he walks and he laughs set the precedent
for he telegraphs, forms like walking and laughing for telegraphing.
Without such clear-cut grooves of analogy, indeed, it would be impos-
sible to learn to speak, a c»rollary of which is that there is a limit to
the extent of grammatical irregularity in any language. When, for
some reason or other, as by the disintegrating action of phonetic laws,
too great irregularity manifests itself in the morphology of the lan-
guage, the force of analogy may assert itself to establish comparative
regularity, that is, forms which belong to ill-defined or sparsely repre-
sented morphologic groups may be replaced by equivalent forms that
follow the analogy of better-defined or more numerously represented
groups. In this way all the noun plurals of English, if we except a
few survivals like feet and oxen, have come to be characterized by a
sufl&xed -s; the analogical power of the old -s plurals was strong enough
to transform aU other plurals, of which Anglo-Saxon possessed several
distinct tjrpes. The great power exerted by analogy is seen in the
persistence with which children, whose minds are naturally unbiased
by tradition, use such forms as foots and he swimmed. Let us not
smile too condescendingly at the use of such forms ; it may not be going
too far to say that there is hardly a word, form, or sound in present-day
English which was not at its first appearance looked upon as incorrect.

The disintegrating influence of phonetic change and the leveling
influence of analogy are perhaps the two main forces that make for
linguistic change. The various influences, however, that one language

Two: Ilistorw Variety and .Srinni: <'t luni^mti^c 125

may exert upon another, generally summed up in the word borrowing,
are also apt to be of importance. As a rule Buch influence is limited
to the taking over or borrowing of certain words of one language by
another, the phonetic form of tlie foreign word almost always adapting
itself to the phonetic system of the borrowing language. Resides thi«
very obvious sort of influence, there are more subtle ways in which one
language may influence another. It is a very noteworthy plienomenon
that the languages of a continuous area, even if genetically unrelated
and however much they may differ among themselves from the point of
view of morphology, tend to have similar phonetic sysiems or, at any
rate, tend to possess certain distinctive phonetic traits in common.
It can not be accidental, for instance, that both the Slavic languages
and some of the neighboring but absolutely unrelated Ural-Altaic lan-
guages (such as the Cheremiss of the Volga region) have in common
a peculiar dull vowel, known in Russian as yeri, and also a set of
palatalized or so-called " soft " consonants alongside a parallel set of
unpalatalized or so-called " hard " consonants. Similarly, we find that
Chinese and Siamese have in common with the unrelated Annamite
and certain other languages of Farther India a system of musical
accent. A third very striking example is afforded by a large number
of American Indian linguistic stocks reaching along the Pacific coafit
from southern Alaska well into California and beyond, which have in
common peculiar voiceless /-sounds and a set of so-called " fortis "
consonants with cracked acoustic effect. It is obvious that in all these
cases of comparatively uniform phonetic areas embracing at the same
time diverse linguistic stocks and types of morphology we must be deal-
ing with some sort of phonetic influence that one language may exert
upon another. It may also be shown, though perhaps less frequently,
that some of the morphologic traits of one language may be adopted
by a neighboring, sometimes quite unrelated, language, or that certain
fundamental grammatical features are spread among several unrelated
linguistic stocks of a continuous area. One example of this sort of
influence will serve for many. The French express the numbers 70,
80 and 90, respectively, by terms meaning 60-10, 4 twenties and 4
twenties 10; these numerals, to which there is no analogy in I^atin,
have been plausibly explained as survivals of a vigesimal method of
counting, that is counting by twenties, the numbers above 20, a method
that would seem to have been borrowed from Gallic, a Celtic language,
and which still survives in Gaelic and other modern Celtic languages.
This example is the more striking as the actual lexical influence which
Celtic has exerted upon French is surprisingly small. So much for the
influence of borrowing on the history of a language.

We may turn now to take up the matter of the varieties of human
speech. One method of classifying the languages of the world has been

126 General Linguistics I

already referred to ; it may be termed the genetic method, inasmuch as
it employs as its criterion of classification the demonstrable relation of
certain languages as divergent forma of some older form of speech.
As we have already seen, the linguistic stocks which we thus get as our
largest units of speech are too numerous to serve as the simplest pos-
sible reduction of the linguistic material to be classified. One naturally
turns, therefore, to a psychological classification, one in which the
classificatory criterion is the fundamental morphological tj'pe to which
a particular language or stock is to be assigned. Such a classification
of morphological types may proceed from different points of view, vary-
ing emphasis being laid on this or that feature of morphology. It is
clear at the outset that we have to distinguish between what we may
call the subject-matter or content of morphology and the mere form
pure and simple. Any grammatical system gives formal expression to
certain modes or categories of thought, but the manner of expression
of these categories or the formal method employed may vary greatly
both for different categories and for different languages. Not infre-
quently the same logical category may be expressed by different formal
methods in the same language. Thus, in English, the negative idea is
expressed by means of three distinct formal methods exemplified by
untruthful, with its use of a prefix un-, which can not occur as a freely
movable word; hopeless, with its use of a suffix -less, which again can
not occur as a freely movable word; and not good, in which the nega-
tive idea is expressed by an element (not) that has enough mobility to
justify its being considered an independent word. We have here, then,
three formal processes illustrated to which may be assigned the terms
prefixing, suffixing and juxtaposing in definite order. While the same
logical category may be grammatically expressed by different formal
methods, it is even more evident that the same general formal method
may be utilized for many different categories of thought. Thus, in
English, the words books and worked use the same method of suffixing
grammatical elements, the one to express the concept of plurality, the
other that of past activity. The words feet and swam, furthermore,
respectively express the same two concepts by the use of an entirely
distinct formal method, that of internal vowel change.

On the whole one finds that it is possible to distinguish between two
groups of grammatically expressed logical categories. One group may
be characterised as derivational; it embraces a range of concepts ex-
pressed by grammatical elements that serve to limit or modify the
signification of the word subjected to grammatical treatment without
seriously affecting its relation to other words in the sentence. Such
merely derivational elements are, in English, prefixes like un-, suffixes
like -less, agentive suffixes like -er in hiker, and numerous others. The
second group of logical concepts and corresponding grammatical ele-

I'wo: Hisiorv. Viiriciv and Si'tiim^ of Litnx^iui^e ill

menta may be cliaracterizcd aa relational ; they not nu-rely 8er?c to gitc
the word affected a new increment of meaning, as ia the case with the
first group, but also assign it a definite syntactic place in the sentence,
defining aa they do its relation to other words of the Bentcnce. Such
a relational grammatical element, in P^nglish, ia the plural ■» Bufhi;
a word, for instance, like hooks differs from its corresponding einguUr
hook not merely in the idea of plurality conveyed by the suffii -s. but
assumes a different grammatical relation to other words in the sentence
— a book is, but books are. Such relational elements are, furthermore,
the case and gender suflBics of nouns and adjectives in Indogermanic
languages; furthermore, the personal endings and tense suffixes of
verbs. On the whole it may be said that derivational elements are of
relatively more concrete signification than the relational ones and tend
to become more thoroughly welded into a word unit with the basic word
or stem to which they are attached or which they affect. This state-
ment, however, is only approximately of general application and is sub-
ject to numerous qualifications. The greatest degree of concrctene^s
of meaning conveyed by derivational elements is probably attained in
many, though by no means all, American Indian languages, where idea*
of largely material content are apt to be expressed by grammatical
means. To this tendency the name of polys^Tithesis has been applied.
Thus in Yana, an Indian language of northern California, such ideas
as up a hill, across a creek, in the fire, to the east, from the south,
immediately, in vain and a host of others are expressed by means of
grammatical suffixes appended to the verb stem; so also in Nootka, an
Indian language of Vancouver Island, so highly special ideas as on the
head, in the hand, on the rocks, on the surface of the water, and many
others, &re similarly expressed as suffixes. It is important to note that,
although the distinction between derivational and relational grammat-
ical elements we have made is clearly reflected in pome way or other in
most languages, they differ a great deal as to what particular logical
concepts are treated as respectively derivational or relational. Such
concepts as those of sex gender, number and tense, which in Indo-
germanic are expressed as relational eleraenta, are in other linguistic
stocks hardly to be separated, as regards their grammatical treatment,
from concepts treated in a clearly derivational manner. On the other
hand, demonstrative ideas, which in most Indogermanic languages
receive no relational syntactic treatment, may, as in the Kwakiutl
language of British Columbia, serve an important relational function,
analogous, say, to the Indogermanic use of gender; just as in I>«tin,
for instance, such a sentence as " I saw the big house " is expressed by
" I-saw house-masculine-objective big-masculine-objective," with a
necessary double reference to the concepts of case relation and gender,
so in Kwakiutl the sentence " I saw the iiousc " would have to be ex-

128 General Linguistics I

pressed by some such sentence as " I-saw-the-objective-near-you house-
visible-near-you," with an analogous necessary double reference to the
demonstrative relations involved. If, now, it has been shown that no
necessary correlation exists between particular logical concepts and the
formal method of their grammatical rendering, smd if, furthermore,
there can not even be shown to be a hard and fast line in grammatical
treatment between concepts of a derivational and concepts of a more
definitely relational character, what becomes of the logical category
per 50 as a criterion of linguistic classification on the basis of form?
Evidently it fails us. Of however great psychological interest it might
be to map out the distribution in various linguistic stocks of logical
concepts receiving formal treatment, it is clear that no satisfactory
formal classification of linguistic types would result from such a
mapping.

Having thus disposed of the subject-matter of linguistic morphol-
ogy as a classificatory criterion, there is left to us the form pure and
simple. Here we are confronted first of all by a number of formal
grammatical methods or processes. These, being less numerous than
the logical categories which they express themselves, and, furthermore,
being on the whole more easily defined and recognized, would seem to
lend themselves more easily to classificatory purposes. The simplest
grammatical process is the juxtaposing of words in a definite order, a
method made use of to perhaps the greatest extent by Chinese, to a
very large extent also by English; the possibilities of the process from
the point of view of grammatical effectiveness may be illustrated by
comparing such an English sentence as " The man killed the bear "
with " The bear killed the man," the actual words and forms being
identical in the two sentences, yet definite case relations being clearly
expressed in both. A somewhat similar process, yet easily enough kept
apart, is compounding, that is, the fusion of two words or independent
stems, into a firm word-unit; the process is particularly well developed
in English, as illustrated by words like railroad and underestimate,
and indeed is found widely spread among the most diverse linguistic
stocks. In some languages, as in the Sioux and Paiute of our own
country, compounding of verb stems is frequent, as illustrated by such
forms as to eat-stand, that is to eat while standing; on the other hand,
in not a few linguistic stocks, as the wide-spread Athabascan stock of
North America and the Semitic languages, compounding as a regular
process is almost or entirely lacking. Perhaps the most commonly
used formal method of all is affixing, that is, the appending of gram-
matical elements to a word or to the body or stem of a word ; the two
most common varieties of aSixing are prefixing and suffixing, examples
of which have been already given from English. Probably the major-
ity of linguistic stocks make use of both prefixes and suffixes, though

Two: History, Variety and Si'ttini; of Limi^uai^c 129

they differ greatly as to the relative importance to bo attacliod to thoso
two clasees of elements. Thus, while both in Indogcrmanic and in the
Bantu languages of Africa prefixes and BufTucs are to be found, we
must note that the greater part of the grammatical machinery of Indo-
germanie is carried on by its suflBxes, while it is the prefiifa that in
Bantu take the lion's share of grammatical work. There are also not
a few linguistic stocks in which suffixing as a process in greatly de-
veloped, while prefixing is entirely unknown; such are Ural-Altaic,
Eskimo, and the Kwakiutl and Nootka languages of British Columbia.
On the other hand, languages in which prefixes are used, but no suf-
fixes, seem to be quite rare. A third variety of affixing, known aa in-
fixing, consists in inserting a grammatical element into the very body
of a stem; though not nearly so wide-spread as either prefixing or
BuflBxing, it is a well-attested linguistic device in Malayan, Siouan,
and elsewhere. Still another wide-spread grammatical process is re-
dupltcation, that is, the repetition of the whole or, generally, only
part of the stem of a word; in Indogermanic we are familiar with this
process in the formation, for instance, of the Greek perfect, while in
many American Indian languages, though in far from all, the process
is used to denote repeated activity. Of a more subtle character than
the grammatical processes briefly reviewed thus far is internal vowel or
consonant change. The former of these has been already exemplified
by the English words feet and swam as contrasted with foot and
swim; it attains perhaps its greatest degree of development in the
Semitic languages. The latter, internal consonant change, is on the
whole a somewhat rare phenomenon, yet finds an illustration in Eng-
lish in at least one group of cases. Beside such nouns as house, mouse,
and teeth we have derived verbs such as to house, mouse around, and
teeth J in other words a certain class of verbs is derived from corre-
sponding nouns by the changing of the final voiceless consonants of
the latter to the corresponding voiced consonants. In several non-
Indogermanic linguistic stocks, as in Takelma of southwestern Oregon
and in Fulbe of the Soudan, such grammatical consonant changes
play a very important part. As the last formal grammatical process
of importance may be mentioned accent, and here we have to distin-
guish between stress accent and musical or pitch accent. An cicellcnt
example of the grammatical use of stress accent is afforded in Eng-
lish by such pairs of words as conflict and conflict, object and object,
the verb being accented on the second syllabic, the noun on the first.
Musical accent is a far more prevalent phonetic characteristic than is
perhaps generally supposed; it is by no means confined to Chinese and
neighboring languages of eastern Asia, but is found just as well in
many languages of Africa and, as has been recently discovcreil by Mr.
J. P. Harrington and the writer, in a few North American Indian

130 General Linguistics I

languages. As a process of definite grammatical significance, how-
ever, musical accent is not so wide-spread. It is found, to give but one
example, in the earlier stages of Indogermanic, as exemplified, among
others, by classical Greek and by Lithuanian.

Having thus briefly reviewed the various grammatical processes
used by different languages, we may ask ourselves whether the map-
ping out of the distribution of these processes would be of more serv-
ice to us in our quest of the main types of language than we have
found the grammatical treatment of logical concepts to be. Here a
difBculty presents itself. If each linguiBtie stock were characteri2ed
by the use of just one or almost entirely one formal process, it would
not be difficult to classify all languages rather satisfactorily on the
basis of form. But there are great differences in this respect. A
minority of linguistic stocks content themselves with a consistent and
thoroughgoing use of one process, as does Eskimo with its suffixing
of grammatical elements, but by far the larger number make use of
80 many that their classification becomes difficult, not to say arbitrary.
Thus in Greek alone every one of the processes named above, excepting
consonant change, can be exemplified. Even if we limit ourselves to a
consideration of grammatical processes employed to express the rela-
tional concepts, we shall find the same difficulty, for the same language
not infrequently makes use of several distinct processes for concepts
of this class.

On a closer study of linguistic morphology, however, we find that
it is possible to look at the matter of form in language from a differ-
ent, at the same time more generalized, point of view than from that
of the formal processes employed themselves. This new point of view
has regard to the inner coherence of the words produced by the opera-
tion of the various grammatical processes, in other words, to the rela-
tive degree of unity which the stem or unmodified word plus its vari-
ous grammatical increments or modifications possesses, emphasis being
particularly laid on the degree of unity which the grammatical proc-
esses bring about between the stem and the increments which express
relational concepts. On the basis of this formal criterion we may
classify languages, at least for the purposes of this paper, into the
three main types of linguistic morphology generally recognized. The
first type is characterized by the use of words which allow of no gram-
matical modification whatever, in other words the so-called isolating
type. In a language of this type all relational concepts are expressed
by means of the one simple device of juxtaposing words in a definite
order, the words themselves remaining unchangeable units that, ac-
cording to their position in the sentence, receive various relational
values. The classical example of such a language is Chinese, an illus-
tration from which will serve as an example of the isolating type of

Two: History, Variety and St'ttini^ < >f l.anintage 1 3 1

ecntence. woo^ (rising from deep tone) pu* (rising from high)
p'd* (sinking from middle) i'd' (high) may be literally tranfllitcd
" I not fear he," meaning " I do not fear him "; woo* " I " a.s Hubjcct
comes first; p'd* "fear" as predicate follows it; pu^ "not," inaamoch
as it limits the range of meariiiig given by the predicate, must precede
it, hence stands between the subject and predicate; finally I'd* "he"
a3 object follows the predicate. If we eichange the positions of
woo' and i'd^ we change their 8}'ntactical bearing; woo' " I" beoomes
" me " as object, while t'd^, which in our first sentence waa beat
translated as " him " now becomes " he " as subject, and the sentence
now takes on the meaning of " he does not fear me."

In the second main type of language, generally known as the ag-
glutinative, the words are not generally unanalyzable entities, as in
Chinese, but consist of a stem or radical portion and one or more gram-
matical elements which partly modify its primary signification, partly
define its relation to other words in the sentence. While these gram-
matical elements are in no sense independent words or capable of being
understood apart from their proper use as subordinate pairts of a
whole, they have, as a rule, their definite signification and are used
with quasi-mechanical regularity whenever it is considered gram-
matically necessary to express the corresponding logical concept; the
result is that the word, though a unit, is a clearly segmented one com-
parable to a mosaic. An example taken from Turkish, a typical ag-
glutinative language, will give some idea of the spirit of the type it
represents. The English sentence " They were converted into the
(true) faith with heart and soul" is rendered in Turkish dzan u
gdnxil-den iman-a gel-ir-ler^ literally translated, " Heart and soul-from
belicf-to come-ing-plural." The case-ending -den " from " is here
appended only to goniil " soul " and not to dzan " heart," though it
applies equally to both; here we see quite clearly that a case-ending is
not indissolubly connected with the noun to which it is appended, but
has a considerable degree of mobility and corresponding transparency
of meaning. The verb form gel-ir-ler, which may be roughly trans-
lated as " they come," is also instructive from our present point of
view; the ending -ler or -lar is quite mechanically used to indicate the
concept of plurality, whether in noun or verb, so that a verb form
" they come," really " come-plural," is to some extent parallel to a
noun form like " books," really " book-plural." Here we Boe clearly
the mechanical regularity with which a logical concept and ita corre-
sponding grammatical element are associated.

In the third, the inflective, type of language, while a word may be
analyzed into a radical portion and a number of subordinate gram-

•The Turkish and Chinese examples aro taken from F. N. Finck'* "Di«
HaupttTpen dee Sprachbaua."

VOL-LXXII. — 5.

132 General Linguistics I

matical elements, it is to be noted that the unity formed by the two is
a very firm one, moreover that there is by no means a mechanical one-
to-one correspondence between concept and grammatical element An
example from Latin, a typical inflective language, will illustrate the
difference between the agglutinative and inflective types. In a sen-
tence like video homines " I see the men/* it is true that the verb form
video may be analyzed into a radical portion vide- and a personal end-
ing -6, also that the noun form homines may be analyzed into a radical
portion homin- and an ending -es which combines the concepts of
plurality with objectivity, that is, a concept of number with one of
case. But, and here comes the significant point, these words, when
stripped of their endings, cease to have even a semblance of meaning,
in other words, the endings are not merely agglutinated on to fully-
formed words, but form firm word-units with the stems to which they
are attached; the absolute or rather subjective form homo, "man," is
quite distinct from the stem homin- which we have obtained by analy-
sis. Moreover, it should be noted that the ending -6 is not mechan-
ically associated with the concept of subjectivity of the first person
singular, as is evidenced by such forms as vidi " I saw " and vidcam
" I may see"; in the ending -es of homines the lack of the mechanical
association I have spoken of is even more pronounced, for not only
are there in Latin many other noun endings which perform the same
function, but the ending does not even express a single concept, but, as
we have seen, a combined one.

The term polysynthetic is often employed to designate a fourth
type of language represented chiefly in aboriginal America, but, as has
been shown in another connection, it refers rather to the content of a
morphologic system than to its form, and hence is not strictly parallel
as a classificatory term to the three we have just examined. As a
matter of fact, there are polysynthetic languages in America which
are at the same time agglutinative, others which are at the same time
inflective.

It should be carefully borne in mind that the terms isolating, agglu-
tinative and inflective make no necessary implications aa to the logical
concepts the language makes use of in its grammatical system, nor is it
possible definitely to associate these three types with particular formal
processes. It is clear, however, that on the whole languages which
make use of word order only for grammatical purposes are isolating in
type, further, that languages that make a liberal grammatical use of
internal vowel or consonant change may be suspected of being inflective.
It was quite customary formerly to look upon the three main types of
morphology as steps in a process of historical development, the isolating
type representing the most primitive form of speech at which it was
possible to arrive, the agglutinative coming next in order as a type

7iv(;. History. Variety and Sfttini^ of Lani^uai^c 133

<'7

evolved from the isolating, and the inflective as the latest and Bo-called
highest t}'pe of all. Further study, however, has ahown that there ia
little to support this theory of evolution of types. The Chinrge lan-
guage, for instance, so far from heing typical of a primitive Btage, as
used to be asserted, has been quite conclusively proven by internal and
comparative evidence to be the resultant of a long prcKcss of fiiraplifica-
tion from an agglutinative type of language. English iticlf, in its
historical afBliations an inflective lang\iage, has ceased to be a cU-ar
example of the inflective type and may pcrliups be said to be an isola-
ting language in the making. Nor should we be too hasty in attaching
values to the various types and, as is too often done even to-day, look
with contempt on the isolating, condescendingly tolerate the agglu-
tinative, and vaunt the superiority of the inflective type. A well-
developed agglutinative language may display a more logical system
than the typically inflective language. And as for myself, I should
not find it ridiculous or even paradoxical if one asserted that the most
perfect linguistic form, at least from the point of view of logic, had
been attained by Chinese, for here we have a language that, with the
simplest possible means at its disposal, can express tlic most technical
or philosophical ideas with absolute lack of ambiguity and with ad-
mirable conciseness and directness.

Editorial Note

Popular Science Monthly 79 (1911 ). 45-67. [Reprinted in: Smithsonian Insiiiulion.
Annual Report 1912, 573-595, and in Selected Readin\^s in Anthropology,
University of California Syllabus Series, no. 101. 202-224)

The following error in the originally published \ersioii has been corrected
directly into the text printed here (the page reference is lo the original):
p. 65. note 2: Fisk's (correct: Finck's)

langua(;k and favihonmiai-

There is n strong tendency to iisciihc m.-my elements of human nil-
ture to the influence of tlie environment in whicli the shnrcrs of that
culture are placed, some even takinj^ the extreme positi-m of rctiurmR
practically all manifestations of human life and ih juj-ht to envimn-
mental influences. I shall not attempt to argue for or aRainst the im-
portance of the influence had by forces of environment on traits of
culture, nor shall I attempt to show in how far the influence of environ-
ment is crossed by that of other factors. To explain any one trait of
human culture as due solely to the force of physical environment,
however, seems to me to rest on a fallacy. Properly speaking, environ-
ment can act directly only on an individual, and in tho.si' ciLses whore
we And that a purely environmental influence is responsible for a com-
munal trait, this common trait must be interpreted as a summation of
distinct processes of environmental influences on individuals. Such,
however, is obviously not the typical foiTn in which we And the forces
of environment at work on human groups. In these it is enough that a
single individual may react directly to his environment and bring the
rest of the group to share consciously or unconsciously in the influence
exerted upon him. Whether even a single individual can be truthfully
said to be capable of environmental influence uncombined with mflu-
ences of another character is doubtful, but we may at least assume the
possibility. The important point remains that in actual society even the
simplest environmental influence is either supported or transformed by
social forces. Hence any attempt to consider even the simplest element
of culture as due solely to the influence of environmcMit nuist be termed
misleading. The social forces which thus transform the purely environ-
mental influences may themselves be looked upon as environ jlTl mental in
character in so far as a given individual is placed in, and therefore
reacts to, a set of social factors. On tlu^ other hand, the social forces
may be looked upon, somewhat metaphorically, :is parallel in their
influence to those of heredity in so far as they are handetl down from
generation to generation. That these traditional .social forces are them-
selves subject to environmental, among other, changes, illustrat<*5 the
complexity of the problem of cultural origins and development. On ihe
whole one does better to employ the term "environment" only when
reference is had to such influences, chiefly physical in charact<»r, as he

136 General Linguistics I

outside the will of man. Yet in speaking of language, which may be
considered a complex of symbols reflecting the whole physical and
social background in which a group of men is placed, it is advantageous
to comprise within the term environment both physical and social
factors. Under physical environment are comprised geographical char-
acters, such as the topography of the country (whether coast, valley,
plain, plateau, or mountain), climate, and amount of rainfall, and what
may be called the economic basis of human life, under which term are
comprised the fauna, flora, and mineral resources of the region. Under
social environment are comprised the various forces of society that
mold the life and thought of each individual. Among the more important
of these social forces are religion, ethical standards, form of political
organization, and art.

According to this classification of environmental influences, we may
expect to find two sets of environmental factors reflected in language,
assuming for the moment that language is materially influenced by the
environmental background of its speakers. Properly speaking, of
course, the physical environment is reflected in language only in so far
as it has been influenced by social factors. The mere existence, for
instance, of a certain type of animal in the physical environment of a
people does not suflfiice to give rise to a linguistic symbol referring to it.
It is necessary that the animal be known by the members of the group
in common and that they have some interest, however slight, in it
before the language of the community is called upon to make reference
to this particular 12281 element of the physical environment. In other words,
so far as language is concerned, all environmental influence reduces at
last analysis to the influence of social environment. Nevertheless it is
practical to keep apart such social influences as proceed more or less
directly from the physical environment, and those that can not be
easily connected with it. Language may be influenced in one of three
ways: in regard to its subject matter or content, i.e., in regard to the
vocabulary; in regard to its phonetic system, i.e., the system of sounds
with which it operates in the building of words; and in regard to its
grammatical form, i.e., in regard to the formal processes and the logical
or psychological classifications made use of in speech. Morphology, or
the formal structure of words, and syntax, or the methods employed in
combining words into larger units or sentences, are the two main as-
pects of grammatical form.

It is the vocabulary of a language that most clearly reflects the physi-
cal and social environment of its speakers. The complete vocabulary of
a language may indeed be looked upon as a complex inventory of all

7'u7>; ffisforv, Viiricix nnn Sfm/ii; >>r Junfiuu^e 137

the ideas, interests, and occupations that take up the nttcntion of the
community, and wcro such a complete thesaurus of the language of a
given tribe at oi-.r disposal, we miRht to a large extent infer the character
of the physical environment and the charnctoristics '' •' '• ,f

the people making use of it. It is not difhcult to find ■
guages wliosc vocabulary thus l)ears the stam|) of li»c physical environ-
ment in which the speakers are placed. This is particuhirly true of the
languages of primitive peoples, for among these culture has not •• ■ ri
such a degree of complexity as to imply practically univcnwil ^.

From this point of view the vocabulary of primitive languages may be
compared to the vocabularies of particular sections of the population
of civilized peoples. The characteristic vocabulary of a coast tnl>e,
such as the Nootka Indians, with its precise terms for many species
of marine animals, vertebrate and invertebrate, might be compared
to the vocabulary- of such European fisher-foIk as the Basques of south-
western France and northern Spain. In contrast to such coxst peoples
may be mentioned the inhabitants of a desert plateau, like the Southern
Paiute of Arizona, Nevada, and Utah. In the |2W| vocabulary of this tribe
we find adequate provision made for many topographical features that
would in some cases seem almost too precise to be of practical value.
Some of the topographical terms of this language that have been col-
lected are: divide, ledge, sand flat, semicircular valley, circular valley
or hollow, spot of level ground in mountains surrounded by ridges,
plain valley surrounded by mountains, plain, desert, knoll, plateau,
canj'on without water, canyon with creek, wash or gutter, gulch, slope
of mountain or canyon wall receiving sunlight, shaded slope of mountain
or canyon wall, rolling country' intersected by several small hill-ridges,
and many others.

In the case of the specialized vocabularies of both Nootka and South-
ern Paiute, it is important to note that it is not merely the fauna or
topographical features of the country' as such that are rcflcctod. but
rather the interest of the people in such environmental features. Were
the Nootka Indians dependent for their food supply primarily on land
hunting and vegetable products, despite their proximity to the sea,
there is little doubt that their vocabular>' would not be a.s [\. - ' 'v
saturated as it is with sea lore. Similarly it is (juite evident .«?

presence in Paiute of such topographical terms as have been listed, that
accurate reference to topography is a necessar>' thing to dwellers in an
inhospitable senii-arid region; so purely practical a need as ■' ' • •• •■
locating a spring might well rocjuire reference to several f« .
topographical detail. How far the interest in the physical environment

138 General Linguistics I

rather than its mere presence affects the character of a vocabulary may
be made apparent by a converse case in English. One who is not a
botanist, or is not particularly interested for purposes of folk medicine
or otherwise in plant lore, would not know how to refer to numberless
plants that make up part of his environment except merely as "weeds,"
whereas an Indian tribe very largely dependent for its food supply on
wild roots, seeds of wild plants, and other vegetable products, might
have precise terms for each and every one of these nondescript weeds.
In many cases distinct terms would even be in use for various condi-
tions of a single plant species, distinct reference being made as to
whether it is raw or cooked, 12301 or of this or that color, or in this or that
stage of growth. In this way special vocabularies having reference to
acorns or camass might be collected from various tribes of California or
Oregon. Another instructive example of how largely interest determines
the character of a vocabulary is afforded by the terms in several Indian
languages for sun and moon. While we find it necessary to distinguish
sun and moon, not a few tribes content themselves with a single word
for both, the exact reference being left to the context. If we complain
that so vague a term fails to do justice to an essential natural difference,
the Indian might well retaliate by pointing to the omnium gatherum
character of our term "weed" as contrasted with his own more precise
plant vocabulary. Everything naturally depends on the point of view
as determined by interest. Bearing this in mind, it becomes evident
that the presence or absence of general terms is to a large extent de-
pendent on the negative or positive character of the interest in the ele-
ments of environment involved. The more necessary a particular cul-
ture finds it to make distinctions within a given range of phenomena, the
less likely the existence of a general term covering the range. On the
other hand, the more indifferent culturally are the elements, the more
likely that they will all be embraced in a single term of general applica-
tion. The case may be summarized, if example can summarize, by saj'ing
that to the layman every animal form that is neither human being,
quadruped, fish, nor bird, is a bug or worm. To this same type of lay-
man the concept and corresponding word "mammal" would, for a con-
verse reason, be quite unfamiliar.

There is an obvious difference between words that are merely words,
incapable of further analysis, and such words as are so evidently sec-
ondary in formation as to yield analysis to even superficial reflection.
A lion is merely a lion, but a mountain-lion suggests something more
than the animal referred to. Where a transparent descriptive term is
in use for a simple concept, it seems fair in most cases to conclude that

Two: llistorw Wiririv ami Sfntni^ oj I.iiniiuage 139

the knowledge of tlie environment ;il element referred to is comparatively
recent, or ;it any rate tliat the present naming haa Uikcn place at a
comparatively recent time. 'I'he destructive agencies of phonetic change
would in the IIMI long run wear down originally descriptive t«'mi.s to more
labels or unanalyzable words pure and simple. I siwak of this niHtf.-r
here because the transparent or untransparent character of a vo^
lary may lead us to infer, if somewhat vaguely, the length of time that
a group of people has been familiar with a particular concept. IVople
who speak of lions have evidently been familiar with that animal for
many generations. Those who speak of mountain lions would seem to
date their knowledge of these from yesterday. The case is even clearer
when we turn to a consideration of place-names. Only the student of
language history is able to analyze such names as Essex, Norfolk, and
Sutton into their component elements as East Saxon, North Folk, and
South Town, while to the lay consciousness these names are etymologi-
cal units as purely as are "butter" and "cheese." The contrast between a
country inhabited by an historically homogeneous group for a long time,
full of etymologically obscure place-names, and a newly settled countn.-
with its Newtowns, Wildwoods, and Mill Creeks, is apparent. Naturally
much depends on the grammatical character of the language it. self;
such highly synthetic forms of speech as are many American Indian
languages seem to lose hold of the descriptive character of their terms
less readily than does English, for instance.

We have just seen that the careful study of a vocabulary leads to
inferences as to the phj'sical and social environment of those who
use the vocabulary; furthermore, that the relatively transparent or
untransparent character of the vocabulary' itself may lead us to infer
as to the degree of familiarit}'' that has been obtained with various
elements of this environment. Several students, notal-)ly Schrader, in
dealing with Indo-Germanic material, have attempted to make a still
more ambitious use of the study of vocabularies of related languages.
By selecting .such words as arc held in common by all, or at least several,
of a group of gcneticall}' related languages, attempts have been made to
gather some idea of the vocabulary of the hypothetical language of
which the forms of sjicoch investigated arc later varieties, and in this
way to get some idea of the range of concepts po.ssessed by the speakers
of the reconstructed |232| language. We are here dealing with a kind of
linguistic archeology. Undoubtedly many students of Indo-CIerrnanic
linguistics have gone altogether too far in their attempts to r« t

culture from comparative linguistic eviflence, but the value oi > . . .■ :.<c
obtained in this way can not be summarily denied, even granted that

140 General Linguistics I

words may linger on long after their original significance has changed.
The only pity is that in comparing languages that have diverged very
considerably from each other, and the reconstructed prototype of which
must therefore point to a remote past, too little material bearing on the
most interesting phases of culture can generally be obtained. We do
not need extended linguistic comparison to convince us that at a re-
mote period in the past people had hands and fathers, though it would
be interesting to discover whether they knew of the use of salt, for in-
stance. Naturally the possibility of secondary borrowing of a word
apparently held in common must always be borne in mind. Yet, on
the whole, adequate knowledge of the phonology and morphology of the
languages concerned will generally enable a careful analyst to keep
apart the native from the borrowed elements. There has been too little
comparative linguistic work done in America as yet to enable one to
point to any considerable body of tangible results of cultural interest
derived from such study, yet there is little doubt that with more inten-
sive study such results will be forthcoming in greater degree. Surely
a thoroughgoing study of Algonkin, Siouan, and Athabascan vocabu-
laries from this point of view will eventually yield much of interest.
As a passing example of significance, I shall merely point out that
Nahua oco-il^ "Pinus tenuifoHa," and Southern Paiute oyo-mp'", "fir,"
point to a Uto-Aztekan stem oko- that has reference to some variety of
pine or fir.

If the characteristic physical environment of a people is to a large
extent reflected in its language, this is true to an even greater extent
of its social environment. A large number, if not most, of the elements
that make up a physical environment are found universally distributed
in time and place, so that there are natural limits set to the variability
of lexical materials in so far as they give expression to concepts derived
from the physical world. A [2331 culture, however, develops in numberless
ways and may reach any degree of complexity. Hence we need not be
surprised to find that the vocabularies of peoples that diff'er widely in
character or degree of culture share this wide difference. There is a
difference between the rich, conceptually ramified vocabulary of a
language like English or French and that of any typical primitive group,
corresponding in large measure to that which obtains between the
complex culture of the English-speaking or French-speaking peoples of
Europe and America with its vast array of specialized interests, and the
relatively simple undifferentiated culture of the primitive group. Such
variability of vocabulary, as reflecting social environment, obtains in
time as well as place; in other words, the stock of culture concepts and

Two: llisiorx. \ 'uricty ami Scttitii^ of liuii'itage 141

therefore also the corresponding; vocaljuhiry become constantly en-
riched and ramified with the increase within a Rroiip of ciiltunil com-
plexity. That a vocabulary should thus to a great dcKrec reflect cultural
complexity is practically self-evident, for a vocabular\', that is, the
subject matter of a language, aims at any given time to serve aa a act
of symbols referring to the culture background of the group. If by
complexity of language is meant the range of interests imphed in ilA
vocabulary, it goes without saying that there is a constant correlation
between complexity of language and culture. If, however, as is more
usual, linguistic complexitj' be used to refer to degree of morphologic
and syntactic development, it is by no means true that such a correla-
tion exists. In fact, one might almost make a case for an inverse correla-
tion and maintain that morphologic development tends to decrease with
increase of cultural complexity. Examples of this tendency are so easy
to find that it is hardly worth our while going into the matter here. It
need merely be pointed out that the history of English and French
shows a constant loss in elaborateness of grammatical structure from
their earliest recorded forms to the present. On the other hand, too
much must not be made of this. The existence of numerous relatively
simple forms of speech among primitive peoples discourages the idea of
any tangible correlation between degree or form of culture and form of
speech.

Is there, then, no element of language but its mere concrete sub-llMl
ject matter or vocabulary that can be shown to have any relation to the
physical and social environment of the speakers? It has sometimes been
claimed that the general character of the phonetic system of a language
is more or less dependent on physical environment, that such communi-
ties as dwell in mountainous regions or under other conditions tcn(hng
to make the struggle for existence a difficult one develop acoustically
harsh forms of speech, while such as are better favored by nature make
use of relatively softer phonetic systems. Such a theory is as easily
disproved as it seems plausible. It is no doubt true that examples m.ty
be adduced of harsh phonetic systems in use among mountaineers, as
for instance those of various languages spoken in the Caucasus; nor is
it difficult to find instances of acoustically pleasant forms of speech in
use among groups that are subjected to a favorable physical environ-
ment. It is just as easy, however, to adduce instances to the contmn*
of both of these. The aboriginal inhabitants of the Northwest Coast of
America found subsistence relatively easy in a countr>' abounding in

many forms of edible marine life; nor can they be said to h" •• ' n

subjected to rigorous climatic conditions; yet in phonetic ; >8

142 General Linguistics I

their languages rival those of the Caucasus. On the other hand, perhaps
no people has ever been subjected to a more forbidding physical en-
vironment than the Eskimos, yet the Eskimo language not only im-
presses one as possessed of a relatively agreeable phonetic system when
compared with the languages of the Northwest Coast, but may even
perhaps be thought to compare favorably with American Indian lan-
guages generally. There are many cases, to be sure, of distinct languages
with comparable phonetic systems spoken over a continuous territory
of fairly uniform physical characteristics, yet in all such cases it can
readily be shown that we are dealing not with the direct influence of
the environment itself, but with psychological factors of a much subtler
character, comparable perhaps to such as operate in the diffusion of
cultural elements. Thus the phonetic systems of Tlingit, Haida, Tsim-
shian, Kwakiutl, and Salish are not similar because belonging to lan-
guages whose speakers are placed in about the same set of environmental
conditions, but merely because these speakers are geographically con-
tiguous to 12351 each other and hence capable of exerting mutual psycholo-
gical influence.

Leaving these general considerations on the lack of correlation be-
tween physical environment and a phonetic system as a whole we
may point to several striking instances, on the one hand, of phonetic
resemblances between languages spoken by groups living in widely
different environments and belonging to widely different cultural
strata, on the other hand, of no less striking phonetic differences that
obtain between languages spoken in adjoining regions of identical or
similar environment and sharing in the same culture. These examples
will serve to emphasize the point already made. The use of pitch accent
as a significant element of speech is found in Chinese and neighboring
languages of southeastern Asia, Ewe and other languages of western
Africa, Hottentot in South Africa, Swedish, Tewa in New Mexico, and
Takelma in southwestern Oregon. In this set of instances we have illus-
strated practically the whole gamut of environmental and cultural
conditions. Nasalized vowels occur not only in French and Portuguese,
but also in Ewe, Iroquois, and Siouan. "Fortis" consonants, i.e., stop
consonants pronounced with simultaneous closure and subsequent re-
lease of glottal cords, are found not only in many languages of America
west of the Rockies, but also in Siouan, and in Georgian and other
languages of the Caucasus. Glottal stops as significant elements of
speech are found not only plentifully illustrated in many, perhaps
most, American Indian languages, but also in Danish and in Lettish,
one of the Letto-Slavic languages of Western Russia. So highly peculiar

Two: Hisiorv. Varicrv and Sciiini: "t Innnua^e 143

sounds as the hoarse lia and stranKuhited-sounding 'ain of Arabic are
found in almost identical form in Xootka. And so on indefinitely. On
the other hand, while the Kni^lish and French may, on tlie whole, be

said to be closely related culturally, there are very htrikinR d-'^ fn

in the phonetic systems made use of by each. TurninK l«> •' »!

America, we find that two such closely related groups of tribes, from a
cultural standpoint, as the Iroquois and nei.[;hl)orinK eastern A' ' -is
speak widely different languages, both phonetically and morpl^
The Yurok, Karok, and Hupa, all three occupying a small u. :M

in northwestern California, form a most intimate cultural unit. Vet
here again we find that the phonetic differences l)etween the lanRua^wi
spoken by these tribes are great, and so on indefinitely again. There
seems nothing for it, then, but to postulate an ab.solute lack of correla-
tion between physical and social environment and phonetic systems,
either in their general acoustic aspect oi' in regard to the distribution
of particular phonetic elements.

One feels inclined to attribute a lack of correlation between phonetic
system and environment to the comparatively accidental character of
a phonetic system in itself; or, to express it somewhat more clearly, to
the fact that phonetic systems may be thought to have a quasi-me-
chanical growth, at no stage subject to con.scious reflection and hence
not likelj' in any way to be dependent on environmental conditions, or,
if so, only in a remotel}' indirect manner. Lingui.stic morpliolog;', on the
other hand, as gi\'ing evidence of certain definite modes of thought
prevalent among the speakers of the language, may be thought to stand
in some .sort of relation to the stock of concepts forming the menial
stock in trade, as it were, of the group. As this stock of concepts, how-
ever, is necessarily determineel by tlu; physical and social environment,
it follows that some sort of correlation between the.se environments and
grammatical .structure might be looked for. And yet the negative evi-
dence is as strong in this case as in the paiallel one just disptk-MMl of.
We may consider the subject matter of morphology as made up of cer-
tain logical or p.sychological categories of thought that receive gnim-
matical treatment and of formal methods of expressing these. The
distinct character of the.se two groups of mor|)hological pi ■ i.i

may be illustrated by pointing out that neighboring langu..^ .y

influence, or at any rate resemble, each other in the one set without
necessary corresponding influence or resemblance in the other. Thus,
the device of reduplication is widespread in .Vmerican In<iian languages.
yet the concepts expressed by this method vary widely, llvrv we deal
with a widespread formal device as such. Conversely, the notion of

144 General Linguistics I

inferential activity, that is, of action, knowledge of which is based on
inference rather than personal authority is also found widely expressed
in American languages, but I237| by means of several distinct formal proc-
esses. Here we deal with a widespread grammatically utilized category
of thought as such.

Now, in rummaging through many languages one finds numerous
instances both of striking similarities in the formal processes of mor-
phology and of striking similarities or identities of concepts receiving
grammatical treatment, similarities and identities that seem to run in
no kind of correspondence to environmental factors. The presence of
vocalic changes in verb or noun stems in Indo-Germanic languages,
Semitic, Takelma, and Yana may be given as an example of the former.
A further example is the presence of the infixation of grammatical
elements in the body of a noun or verb stem in Malayan, Mon-Khmer,
and Siouan. It will be noticed that despite the very characteristic
types of formal processes that I have employed for illustrative purposes
they crop up in markedly distinct environments. A striking example,
on the other hand, of a category of thought of grammatical significance
found irregularly distributed and covering a wide range of environ-
ments, is grammatical gender based on sex. This we find illustrated in
Indo-Germanic, Semitic, Hottentot of South Africa, and Chinook of
the lower Columbia. Other striking examples are the existence of syn-
tactic cases, primarily subjective and objective, in Indo-Germanic,
Semitic, and Ute; and the distinction between exclusive and inclusive
duality or plurality of the first person found in Kwakiutl, Shoshonean,
Iroquois, Hottentot, and Melanesian.

The complementary evidence for such lack of correlation as we have
been speaking of is afforded by instances of morphologic differences
found in neighboring languages in use among peoples subjected to
practically the same set of environmental influences, physical and
social. A few pertinent examples will suffice. The Chinook and Salish
tribes of the lower Columbia and west coast of Washington form a
cultural unit set in a homogeneous physical environment, yet far-
reaching morphologic differences obtain between the languages of the
two groups of tribes. The Salish languages make a superabundant use
of reduplication for various grammatical purposes, whereas in Chinook
reduplication, though occurring in a limited sense, has no grammatical
significance. On the other hand, the system of sex gender rigidly carried
out in the |238| noun and verb system of Chinook is shared by the Coast
Salish dialects only in so far as prenominal articles are found to express
distinctions of gender, while the interior Salish languages lack even

Two: Hisiorw \'urict\ ntul Stfitnii "I I iiniiua^e 145

this feature entirely. Perhaps an even more striking insUnce of radical
morphological dissimilarity in neighboring languages of a single cul-
ture area is afforded by Yana and iMaidu, spoken in north central
California. Maidu makes use of a large numl)er of grammatical prefixea
and employs reduplication for grammatical purpo.ses to at Icaat some
extent. Yana knows nothing of either prefi.xes or re<luplication. On the
other hand, Maidu lacks such characteristic Yana features as the differ-
ence in form between the men's and women's language, and the em-
ployment of several hundreds of grammatical sufRxes, some of them
expressing such concrete verbal force as to warrant their being inter-
preted rather as verb stems in secondary position than as suffixes proper.
To turn to the Old World, we find that Hungarian differs from the
neighboring Indo-Germanic languages in its lack of sex gender and in
its employment of the principle of vocalic harmony, a feature which,
though primarily phonetic in character, nevertheless has an important
grammatical bearing.

In some respects the establishment of failure of phonetic and mor-
phologic characteristics of a language to stand in any sort of relation to
the environment in which it is spoken seems disappointing. Can it be,
after all, that the formal groundwork of a language is no indication
whatsoever of the cultural complex that it expresses in its subject
matter? If we look more sharply, we shall find in certain cases that at
least some elements that go to make up a cultural complex are em-
bodied in grammatical form. This is true particularly of .synthetic
languages operating with a large number of prefixes or suffixes of rela-
tively concrete significance. The use in Kwakiutl and Xootka, for in-
stance, of local suffixes defining activities as taking place on the beach,
rocks, or sea, in cases where in most languages it would be far more
idiomatic to omit all such reference, evidently points to the nature of
the physical environment and economic interests connected therewith
among these Indians. Similarly, when we find that such ideas as those
of bu3'ing, giving a feast of some kind of food, giving a potlatch for
some person,and|239| asking for a particular gift at a girl's pul)erty ceremony,
are expressed in Nootka by means of grammatical suffixes, we are led
to infer that each of these acts is a highly typical one in the life of the
tribe, and hence constitute important elements in its culture. This
type of correlation may be further exemplified by the use in Kwa-
kiutl, Nootka, and Salish of distinct series of numerals for various
classes of objects, a feature which is pushed to its greatest length,
perhaps, in Tsimshian. This grammatical peculiarity at le.ast sug-
gests definite methods of counting, and would seem to emphasize the

146 General Linguistics I

concept of property, which we know to be so highly developed among
the West Coast Indians. Adopting such comparatively obvious ex-
amples as our cue, one might go on indefinitely and seize upon any
grammatical peculiarity with a view to interpreting it in terms of
culture or physical environment. Thus, one might infer a different
social attitude toward woman in those cases where sex gender is made
grammatical use of. It needs but this last potential example to show
to what flights of fancy this mode of argumentation would lead one.
If we examine the more legitimate instances of cultural-grammatical
correlation, we shall find that it is not, after all, the grammatical form
as such with which we operate, but merely the content of that form;
in other words, the correlation turns out to be, at last analysis, merely
one of environment and vocabulary, with which we have already
become familiar. The main interest morphologically in Nootka suffixes
of the class illustrated lies in the fact that certain elements used to
verbify nouns are suffixed to noun stems. This is a psychological fact
which can not well be correlated with any fact of culture or physical
environment that we know of. The particular manner in which a noun
is verbified, or the degree of concreteness of meaning conveyed by the
suffix, are matters of relative indifference to a linguist.

We seem, then, perhaps reluctantly, forced to admit that, apart from
the reflection of environment in the vocabulary of a language, there is
nothing in the language itself that can be shown to be directly asso-
ciated with environment. One wonders why, if such be the case, so
large a number of distinct phonetic systems and types of linguistic
morphology are found in various parts of the world. Perhaps the whole
problem of the relation between culture and 1240] environment generally,
on the one hand, and language, on the other, may be furthered some-
what by a consideration simply of the rate of change or development of
both. Linguistic features are necessarily less capable of rising into the
consciousness of the speakers than traits of culture. Without here
attempting to go into an analysis of this psychological difference be-
tween the two sets of phenomena, it would seem to follow that changes
in culture are the result, to at least a considerable extent, of conscious
processes or of processes more easily made conscious, whereas those of
language are to be explained, if explained at all, as due to the more
minute action of psychological factors beyond the control of will or
reflection. If this be true, and there seems every reason to believe that
it is, we must conclude that cultural change and linguistic change
do not move along parallel lines and hence do not tend to stand in
a close causal relation. This point of view makes it quite legitimate

Two: Iflsforv, Xariciy and Si'ttiny; of I iin\iuage 147

to grant, if necessary, the existence at some primitive stage in the
past of a more definite association between environment and hnguistic
form than can now be posited anywhere, for the different character and
rate of change in linguistic and cultural phenomena, conditioned by the
veiy nature of those phenomena, would in the long run very materially
disturb and ultimately entirely eliminate such an association.

We may conceive, somewhat schematically, the development of cul-
ture and language to have taken place as follows: A primitive group,
among whom even the beginnings of culture and language are a* yet
hardly in evidence, may nevertheless be supposed to behave in uccortl-
ance with a fairly definite group psychology, determined, we will sup-
pose, partly by race mind, partly by physical environment. On the basis
of this group psychology, whatever tendencies it may possess, a lan-
guage and a culture will slowly develop. As both of these arc directly
determined, to begin with, by fundamental factors of race and physical
environment, they will parallel each other somewhat closely, so that
the forms of cultural activity will be reflected in the grammatical system
of the language. In other words, not only will the words themselves of a
language serve as symbols of detached cultural elements, as is true of
languages at all periods of development, but we may suppose the IXII
grammatical categories and processes themselves to symbolize corre-
sponding types of thought and activity of cultural significance. To
some extent culture and language may then be conceived of as in a
constant state of interaction and definite association for a considerable
lapse of time. This state of correlation, however, can not continue
in(lcfinitel)^ With gradual change of group p.sychology and physical
environment more or less i)rofoun(l changes must be efTectcil in the
form and content of both language and culture. Language and culture,
however, are obviously not the direct expression of racial psychology*
and physical environment, but depend for their existence and con-
tinuance primarily on tlu^ foiccs of tradition, llcnce, despite necess;ir>*
modifications in either with the lapse of time, a con.s^rvati\*e tendency
will always make itself felt as a check to those tendencies that make foi
change. And here we come to the crux of the matter, ('ultural elements,
as more definitely serving the ininiediale needs of society and entering
more clearly into consciousness, will not only change more rapidly than
those of language, but the form itself of culture, giving each element
its relative significance, will be continually shaping itself anew. Lin-
guistic elements, on the other hand, while they may and do readily
change in themselves, do not so easily lend themselves to regroujiings,
owing to the subcon.seious character of grammatical <*Ia.ssilication. A

148 General Linguistics I

grammatical system as such tends to persist indefinitely. In other
words, the conservative tendency makes itself felt more profoundly in
the formal groundwork of language than in that of culture. One neces-
sary consequence of this is that the forms of language will in course of
time cease to symbolize those of culture, and this is our main thesis.
Another consequence is that the forms of language may be thought to
more accurately reflect those of a remotely past stage of culture than
the present ones of culture itself. It is not claimed that a stage is ever
reached at which language and culture stand in no sort of relation to
each other, but simply that the relative rates of change of the two
differ so materially as to make it practically impossible to detect the
relationship.

Though the forms of language may not change as rapidly as those of
culture, it is doubtless true that an unusual rate of cultural change is
accompanied by a corresponding accelerated rate of 1242] change in language.
If this point of view be pushed to its legitimate conclusion, we must be
led to believe that rapidly increasing complexity of culture necessitates
correspondingly, though not equally rapid, changes in linguistic form
and content. This view is the direct opposite of the one generally held
with respect to the greater conservatism of language in civilized com-
munities than among primitive peoples. To be sure, the tendency to
rapid linguistic change with increasingly rapid complexity of culture
may be checked by one of the most important elements of an advanced
culture itself, namely, the use of a secondary set of language symbols
necessarily possessing greater conservatism than the primarily spoken
set of symbols and exerting a conservative influence on the latter. I
refer to the use of writing. In spite of this, however, it seems to me that
the apparent paradox that we have arrived at contains a liberal ele-
ment of truth. I am not inclined to consider it an accident that the rapid
development of culture in western Europe during the last 2000 years
has been synchronous with what seems to be unusually rapid changes in
language. Though it is impossible to prove the matter definitely, I
am inclined to doubt whether many languages of primitive peoples
have undergone as rapid modification in a corresponding period of time
as has the English language.

We have no time at our disposal to go more fully into this purely
hypothetical explanation of our failure to bring environment and lan-
guage into causal relation, but a metaphor may help us to grasp it.
Two men start on a journey on condition that each shift for himself,
depending on his own resources, yet traveling in the same general di-
rection. For a considerable time the two men, both as yet unwearied,

Two: History, Wiriciv ami Scnini: of Idni^uaf^e 149

will keep pretty well together. In course of time, however, the varying
degrees of physical strength, resourcefulness, ability to orient oneself.
and many other factors, will begin to manifest themselves. The actual
course traveled by each in reference to the other and to the course
originally planned will diverge more and more, while the absolute
distance between the two will also tend to l)ecome greater and greater.
And so with many sets of historic sequences which, at one time causally
associated, tend in course of time to diverge.

Editorial Note

American Anthropologist n.s. 14 (1912), 226-242. [ Reprinted in: Idward Sapir.
Selected Writings in Language, Culture, and Personality. I^dilcd by David (i.
Mandclbaum. Berkeley: University of California Press, 1949, pp. 89-103). Read
before the American Anthropological Association. Washington. DC . December

28,1911.

Section Three

Theoretical, Descriptive and His k )ki( a
Linguistics ( 1 923 - 1 929 )

Inlroduclion:
Theorelical, Descriptive and Hisioncal 1 Jimuisiics.

1923-1929

The papers in this seelion date I'rom the end ol S.ipii s si;i\ m « Miawa to his
first years in Chieago. They testify to Sapirs intelleelual deveh)pment as a gen
eral hnguist, and to his attempt at situating language in the encompassing study
of human behaviour and social structure. The texts included here consist of four
reviews, an encyclopedia article, and live journal articles, two of which \\crc
published in the then recently created journal of the linguistic S(K-ietv of
America, Language. Some of the papers thus coincide with the aulonomi/ation
of linguistics as an academic discipline in the I IS. and with the recognition of the
study of language as a (social) science on its own.

The major thematic Hnes running through almost all of the papers of this sec-
tion are the concept of patterning in language — Sapir's "Sound Patterns m
Language" (1925) marks a crucial date here — . the emphasis laid on linguistic
symbolization and the symbolism inherent in the linguistic material (.m issue
with respect to which Sapir's reading and reviewing of Ogden and Richards's //»<•
Meaning of Meaning (1923) is of high relevance), and the preoccupation with
defining the place of linguistics as a science.

Sapir's "Sound Patterns in Language," a classic article which Sapir regularK
referred to in his later publications, shows that the study of phonetics transcends
the domain of physicalist description, and necessarily includes the study of the
''psychology of a language." The sounds of a language belong as such ti> a defi-
nite system, defined by its proper functionalits and its specific range of variation.
Materially (phonically) similar sounds across languages will differ functionally
and systemically. Sapir takes up here insights from historical-comparative lin-
guistics (more specifically in the field oH phonology) and descripti\e anthrt^poio-
gical linguistics (e.g., Boas's views on 'allernating sounds'), hut integrates them in
a general structural view of language: the theoretical concept griiunding this \ieNK
is that oi patterning, the language-immanent organization of structural relation-
ships. It is precisely in terms of systemic patterning that a distinction can be made
between phones and phonemes: distinct sounds are not necessariK distinctive
sounds, nor do similar sounds necessarily belong to the same phonemic unit.
Whereas in his Language (1921 ). Sapir did not use the notion o\ phoneme, here
he makes a clear distinction between phonetics and [thoneinics (p. 40).'I'he struc-
tural description of a language involves decisions on alignment (p. 41 ). which can
only be taken with respect to patterns (pp. 41-42). Sapir ends this paper with draw-
ing the implications for sound change: the (phonemic) patterns of a language
define a range of variation but they also pio\ide an orientation for the changes
that can affect the system.

154 General Linguistics I

In "The Grammarian and his Language" (1924) Sapir addresses the issue of
patterns in language, discussing it in general terms ("language as form;" cf. his
book Language. 1921, chapters IV and V), and combining it with recent insights
into linguistic symbolization. Undoubtedly, his reading of Ogden and Richards's
The Meaning of Meaning (1923) had made him aware of the complex symbolic
function of language. His short review, entitled "An Approach to Symbolism,"
reflects the deep interest he had taken in this "original" book, which opened up
vistas for new sciences (p. 573), and which, while showing the relevance of lan-
guage for philosophers and psychologists, at the same time dismissed the tradi-
tional philosophical approach to language.' Largely subscribing to the "relativistic"'
approach of Ogden and Richards, who pointed out the pervasive (and also delu-
sive) role played by words in habitual thinking, Sapir somehow deplores their
neglect of language form as symbolic on itself^ (an issue which is explored in his
1929 paper "A Study in Phonetic Symbolism"). Sapir's short, artfully written
review of Ogden and Richards's book gives only a dim reflection of the deep
impact the work seems to have had on his linguistic thinking. It allowed him to
integrate a non-reductionist form of psychology in his general approach of lan-
guage(s). As is clear from the article "The Grammarian and his Language," Sapir
could hardly feel intellectual affinity with the behaviouristic psychology which
was then flourishing in the United States. This form of psychology, in which lan-
guage is defined as "subvocal laryngeating" ("The Grammarian and his
Language," p. 150; see also the article "Philology"), was, in Sapir's view, a poor
ally to linguistics. In his paper Sapir deplores the lack of general interest taken by
Americans in linguistics, and in language as a structure. He attributes this to an
overly rationalistic, pragmaticist attitude and to a general lack of culture (pp.
150-151). Moreover, linguistics as an autonomous science hardly seems to appeal
to Americans, who expect to find in linguistics answers to questions of a larger
interest (such as the relation of language to culture, or the relation of language
to psychology): on both counts, they are likely to be disappointed. In his paper,
published in the American Mercury and clearly written for a larger audience,
Sapir sets himself the difficult task of replacing the view of the grammarian as a
"pedant" by that of the grammarian (or hnguist) interested in studying the for-
mal completeness of language (made visible in the variety of languages as formal
systems). The emphasis here is on language as grammatical form, not as vocabu-
lary (p. 151); this form serves as a frame of reference, as a method or approach to
experience. Using an analogy with mathematics (see also Sapir's 1931
"Conceptual Categories in Primitive Languages," included in section VI), Sapir

On the impact of Ogden and Richards's work on linguistic theory, see Terrence Gordon, "C.K. Ogden, E.
Sapir, L. Bioomfield and the Geometry of Semantics", in History and Historiography of Linguistics. Papers from
the Fourth International Conference on the History of the Language Sciences (Trier, 24-28 August, 1987), edited
by Hans-Josef Niederehe and E.F.K. Koerner (Amsterdam/Philadelphia, 1990), 821-832.

^ An approach which Sapir also welcomes in the writings of Fritz Mauthner [1849-1923].

^ "Is It not a highly significant fact none the less, that its form is so essentially of symbolic pattern" (p. 573).

Three: Descriptive and llistonml I .in\^uisius 155

defines language as a matrix lor ihc expression ol iIiduuIu (p. 152): different lan-
guages show different formal -leehniques " (p. 152). correspond to a specific
form-feeling (pp. 152-153) of their speakers, and eonstilule diflerenl trames of
reference. Taking up the idea of (uneonseious) pallermng. Sapir sees hnguistic
forms as providing an unconsciously elaborated t)rienlalion in the experience of
reality. Language form provides us with an "intuitive" feeling for relalions.
though not in an absolutely uniform way: drawing the attention of philosopher.
lo the work of Ogden and Richards (p. 154). Sapir warns us against prect)nceived
ideas and unwarranted extrapolations. The study of language is recommended as
a relativistic eye-opener: "Perhaps the best way to get behind our thought pro-
cesses and to eliminate from them all the accidents or irrelevances due lo their
linguistic garb is to plunge into the study of exotic modes of expression. Al any
rate, I know of no better way to kill spurious "entities"" (p. 154). As a symbolic
system, language is related to experience, but not in a deternnnistic wav: there are
many facets to experience, and language forms allow for divergent analvses of
what could be superficially described as "the same experience." Just like in chap-
ter V of his Language (New York, 1921 ), Sapir shows (pp. 154-155) how the same
event ("the stone falls") can be categorized in manifold ways, in relation to the
set of formal statements available in each language, llie study of language thus
opens fascinating perspectives for approaching what Sapir calls the "relativiiy of
the form of thought" or relativity of concepts (p. 155). A passing reference is
made to physical and psychological relali\ ily.

Much in line with this 1924 article is the paper "Language as a Form of Human
Behavior,"^ in which Sapir seems to envisage a dialogue, or rappriK'hement with
psychology, philosophy and sociology. As Sapir notes at the beginning i>f this
paper, the fundamental problems of linguistics have to be related to the study
of human behaviour in general (p. 42l).TIie central part o{ the paper is taken up
by the presentation of the essential characteristics of language, ilituuc! .is the
arrangement of all the elements of experience (p. 425).'I"hese are

(a) the completeness of its formal development (or "grammar:" cl. p. 42.^):

(b) its status as a specific system of behaviour: although language is all-per\asivc
in human behaviour it is also an aulononK)Us. unconscious svstem ot behaviour:
it incorporates naturally acquired knowledge (p. 423). which cm be made expli-
cit in statements expressing linguistic kncnvledge;

(c) the indirect character of its symbolic nature; this feature correspon.is i.. tin
arbitrariness of linguistic signs (p. 424):

(d) the universality of language as a human fact, coupled Nsith the inlinilc"
variety of words, forms and eonslruetions across languages; this eharaelerislic

'*This paper can be seen as occupving a midwav pi)siiu)n (eluoni>li>i;ically and inlcllcclualK %pcaking) bclwccn
Sapir's more narrowly -liniiinstic" approach lo lanjiua.uc (as in his \^\\ paper " tlic Hislor> and Nanclic^ of
Human Speech" |reprinled here in section \\\. and in his Lunniuif^c ol 1^21). and his bnvadcr vKu>logical
approach to language (as we find it in his article •l^anguage" of 193.^ |rcpnnlcd in section VI of lh» volume)).

156 General Linguistics I

allows Sapir to oppose the ground-plan of language to the overt forms of lan-
guages;

(e) the joining of a denotative and an expressive dimension in language: on the
one hand, language is an abstract classification of reality, while on the other hand,
it constitutes the locus of the most individual expressions (p. 426, p. 431).'

As noted by Sapir, the latter characteristic is also a source of misunderstand-
ings: the signs of natural languages carry with them a number of not strictly con-
trollable connotations (p. 432); such connotations do not occur in artificially con-
structed languages. An international language for communication — the need for
which Sapir stresses at the end of his paper" — can be constructed as an objective
language of reference.

In the second part of the paper Sapir considers the origin of language.^ He pro-
poses the hypothesis that language originated as conventionalized gesture (pp.
426-427): auditory gestures became conventionalized as spoken language. This
process involved a functional shift: from secondary symbolization to primary
symbolization (pp. 428^29). The origin of speech thus involved two stages: a
stage of gestural communication, and a stage of secondary referentialization (p.
430). In Sapir's view this hypothesis accounts for two facts:

(a) the iconicity of the material shape of language (p. 429; see also the 1929 article
"A Study in Phonetic Symbolism");

(b) the universal development of language as a symbolic system; the explanation
of the latter fact should be sought in the disponibility of speech-unspecific organs
for producing linguistic signs.^

Some of the issues discussed in the three articles "Sound Patterns in
Language," "The Grammarian and his Language," and "Language as a Form of
Human Behavior," are also touched upon in Sapir's reviews included in this sec-
tion. As noted above, the review of Ogden and Richards's The Meaning of
Meaning highlights the symbolic function of language, and the complex psychic
and social role played by linguistic signs in our approach to reality.

The review of Kent's Language and Philology (1923), published in 1928, con-
tains an implicit criticism of the linguistic myopia of Classicists and Indo-
European scholars, and corrects the exaggerated view of Enghsh as an "analyt-
ical," untypical Indo-European language: as Sapir shows, a more balanced typolo-
gical view (based on the typological theory proposed in Language, 1921, chapter
VI) places Enghsh within the group of fusional and mixed-relational languages,
showing the same "patterning" (p. 85) as ("synthetic") Indo-European languages
like Sanskrit or Latin.

^ See also Language (New York, 1921), pp. 39^2.
" See also the papers reprinted in sections IV and V.

See also the papers reprinted in sections I and II.
^ See also Language (New York, 1921), pp. 7-8.

Three: Descriptive und /lisioriml I ini'tiistics 157

The ct|iially short review or jesperseii's Munkimi. \uiinn ami huintdiutl from
a I.ini^uistie l\>ini oj \'ie\v (1^>25) suiiimari/cs. or simply mcnlions Jcspcrscn's
views on noliiMis siieh as hini^iiui^e, speech, (huh'ci. and on ihc need for an inter-
national auxihary language.' Sapir deplores ihc lack of originality of Jcspcrscn's
booklet, but praises the authors 'common sense." He crilici/cs Jcspcrscn for dis-
missing the Saussurean distinction between lan^ue (language: the global pattern
underlying the linguistic behaviour of a group) and parole (speech; the limcand-
place-bound linguistic behaviour of individuals in a group), and righlK points »)Ut
that Jespersen's (positivistic) dismissal is contradicted by his own writings on the
history of English, and would lead to a lully atomistic description of cultural
(including linguistic) phenomena: it carried to their logical conclusion.
Jespersen's strictures would demolish the study of all cultural patterns and con-
demn the social scientist to the interminable listing of individual events" (p. 498).

The review oi the collecti\e Nolume edited in 1^24 b\ Meillet and (iihen.
as well as the article "Philology" bring us back to .Sapir's education and early
work as a philologist (in the sense of "practitioner of historical-comparative lin-
guistics"). In his review of Les hingiies du monde — a work written by "linguistic
specialists" (p. 373),'" Sapir shows his familiarity with the state of research on the
world's languages, and with ambitious, monogenetic reconstructit)ns like those of
the Italian scholar Alfredo Trombetti [1(S66- 192^)]. While approving of the (pho-
netically unavoidable) division of labour necessitated by a survey of the world's
languages, and while noting a few merits of the volume (such as the unified treat-
ment of Hamitic-Semitic, or the use of the term "Sino-Tibetan")/' Sapir regrets
some serious omissions (Siberian and Andaman languages), and the disparity
(both in coverage and in [more or less] systematic treatment) of the separate
chapters, and shows the need for a combined perspectise. that of the lypologis!
(here Sapir recommends Franz Nikolaus Finck's work. Pie SpriichMiimnw drs
Erdkreises) and the comparatist. Sapir mentions the possibility of including a
structural sketch of Amerindian, African or Polynesian language families, but his
most serious criticism concerns the treatment of American Indian languages, a
domain too vast and too c(Miiple\ to be assigned to a single scholar (m the pres-
ent case a specialist of South American languages, vi/. Paul Ri\et)

The article "Philology" of the 13th edition of the Encyclopaedia Hniannica^' is
mainly devoted to historical-com|-»arati\e linguistics and to a survey of the

Sec also the papers in sections !\ and \
'"nic work IS also rclcrrcd to in the article "PhiloUigy" Ircprintcd m this section)

' ' Note that in Lan^iuigc (Now York. 1^21 ), p. 155 and p 164. Sapir had used IndoChtnc^r. Ulcr he u»cd Suto-

Tibeiun and Siniiic (see the article "Philology").

'- On this encyclopedia article hy Sapir. see Yakov Malkiel. "Sapirs Panoramic Vic^ of Rivni Xilvauc* in

Linguistics", in (icncrul uiul Amcniulian I ihnnliUf'ui.Muw In Kcmcmhrinuc of Stanley Vr»» -• ' ^ir>

Ritchie Key and Henry M. Hoenigswald (Berlin/Ncw York. l''.s«ii s<J |o4 In ihc ..mi
|1929-1932]. the article "Philology" was as,signcd lo Otto Jcspcr>cn

1 58 General Linguistics I

world's languages; this justifies the traditional heading "Philology" under which
it appears in the Encyclopaedia Bhtannica. Sapir opens the article with a short
statement on the progress of general linguistics (or "linguistic science"), referring
to the manuals of Otto Jespersen, Joseph Vendryes and his own Language, all
published in 1921-1922.'^

Sapir then shows the relevance of linguistics for, and its ties with psychology,
philosophy, sociology and anthropology. In the paragraph on psychology he men-
tions, without much sympathy,'^ behaviourism, but he welcomes the work of J.R.
Kantor and of Gestalt^sy choXogy in general (which gives primary importance to
systemic patterning)."

The paragraph on philosophy puts in evidence the work of Ogden and
Richards, and stresses the methodological importance of adopting a relativistic
stand'" (illustrated in the appended paragraph on "Forms of speech"). Along the
same lines, Sapir refers, in the paragraph on sociology and anthropology, to the
cognate linguistic-anthropological work of Mahnowski, showing the role of lan-
guage as a "delimiter and index of social groups," as the medium of symbolic
socialization. The crucial role of field work and text collection'^ is stressed by
Sapir and is illustrated with a reference to Boas and Westermann.

The paragraph on sociology and anthropology prompts the transition to the
study of language in its historical context. Primary evidence of this is found in the
synchronic relics of place names (and ethnic names), a topic for linguistic folklore
or paleontology (Sapir does not expHcitly refer to the latter type of research).

The deeply historical nature of language forms the subject matter of historical
linguistics, which Sapir presents to the readers of the Encyclopaedia Britannica in
a few paragraphs, dealing with internal factors of change, consistency of change,
external contacts; this provides an occasion to discuss some of his preferred
themes, such as resistance to change (attested in Athabaskan),'** drift, ''^ the role of
bilingualism,^" and "convergences" in development.^' The link between language

1 "^

^^ See my "Note sur la linguistique generale en 1921-1922", Beitrdge zur Geschichte der Sprachwissenschaft 1
(1991), 185-191.

^^ See also the 1924 paper "The Grammarian and his Language" [reprinted in this section].

^^ On Sapir's reception of G^i7a/rpsychology, see Michael Cain, "Edward Sapir and Gestalt Psychology",
Anthropological Linguistics 22 (1980), 141-150; and the rejoinder by Stephen Murray, "Sapir's Gestalt",
Anthropological Linguistics 23 (1981), 8-12.

^" See also "The Grammarian and his Language" (1924), and the encyclopedia article "Language" (1933)
[reprinted here in section VI].

1 7

See Regna Darnell, "Franz Boas, Edward Sapir and the Americanist Text Tradition", Historiographia

Linguistica 17 (1990), 129-144.

1^ See Language (New York, 1921), p. 209.

19 See Language, o.c, chapter VII, especially pp. 160-163. 165-166, 172-174, 182-184.

^" See the article "Language" of 1933 [reprinted here in section VI].

On this notion, see Language (New York, 1921), chapter IX (especially p. 213) and Antoine Meillet,

"Convergence des developpements linguistiques", Revue philosophique 85 (1918), 97-110 [reprinted in A.

Meillet, Linguistique historique et linguistique generale, vol. I [Paris, 1921], pp. 61-75].

Three: Dcscnpiivc ami llistornal l.tni;uistn \ |SV

history mikI hinguage geography is made iii the paragraphs on language families"
and on newly discmered hinguages ( Tocharian ' and llittile * in ihe Indo-
Furopean field).

In between these hisit>rieally slanted paragraphs there is a paragraph on mor-
phologieal l\pology (suniniai i/iim ihc three typologieal parameters — types of
coneepts. teehnicjue and degree ol synthesis - used bv Sapir m his I .an^ua\'t)^
and on phoneties (Sapir stresses the iniporlanee o! phonelies for field work.
which crucially hinges on the quality ol the field worker's ear; he also adds a few
remarks on the geographical distribution ol phonetic and lonal ivpes).

"A Study in Phonetic Symbolism" explores. \n an experimental was. a iheorel-
ical issue touched upon in "Language as a I orni ol Human Behavior" and m
Sapir's review of Ogden and Richards. This paper deals with the expressive
dimension of language (coexisting with the arbitrary or referential s\mb«)lic
dimension). TTie paper reports on an experiment conducted by Sapir and aiming
at revealing the "symbolic suggestiveness of sound contrasts" (p. 227) or intuition
of unsocialized symbolisms (p. 2^)). The experiment involves psycholt>gical
correlations between acoustic properties of sounds and material properties of
supposed referents. The set-up o{ the experinieiii is described, and the methtnlo-
logical pitfalls are clearly defined (avoidance of association with actual words:
avoidance of self-induced systematizations). The twci major conclusions of the
study are (a) the demonstration that on the range u — /, a has a greater [X)tential
magnitude symbolism than / (irrespective of the native language of the subject).
(b) on the other hand, the linking of the perception o{ phonetic symbolic weight
to the phonemic patterning of the subject's native language.

For this "unconscious" expressive, translinguistic symbolism (p. 235. pp.
238-239), Sapir sees two factors (which may interact): an acoustic and a kin-
aesthesic one. The role of both factors is bricHy discussed (pp. 23.^ 23h). The final
section of the paper reports on one part of the experiment calling for further
exploration, viz. the factor of individual variation in the perception v)f phonetic
symbolism and its association with referential properties.

Although this paper has obvious links with Sapirs work on p.illernmg of
sounds, and although it illustrates a psychological approach to language which

-'- Sapir alludes to the possihility of Iransccndiim the gc-nclic ciasMficaliunN vi
lars; he also seems to refer with approval lo Hermann Mt)ller's atlempl lo link li

2-^ Tocharian is riiihlly identified as having Iwo dialects (now commonly referred to m TiKhAnan A and
Tocharian B). For Sapir's study of tocharian. see the relevant papers rcpnnlcd m volume II of l^f Ittlhttrd

Works of Edwaril Sapir.

•^■* Sapir also mentions non Indo-liuropean Hattic Colder Hiltite iunguage") and Ihe minor AnAli^tan lanttua-
ges Lycian, I.ydian and ( arian

"^^ See /.(;/;,i,'(/«i;r ( New ^ork. \^)2\). chapter \l In Ihe article ■•Phil.»logv" Sapir c\plioll\ r< r'*

attempt at establishing a parallelism hetween morpholoijical t>|x-s and cultural evolution n-

ticism. see l.tirtfiua^e. p. 2.^4).

160 General Linguistics I

Sapir was clearly heading for from the late 1920s on, the topic was never taken
up later by Sapir in a comprehensive general-linguistic study. ^^

The very concise paper "The Status of Linguistics as a Science" (1929) consti-
tutes Sapir's second paper published in the journal Language. It offers a balan-
ced synthesis of his linguistic work in the 1920s (especially of his publications in
the period covered in this section, 1923-1929). Sapir's starting point is that 19th-
and early 20th-century linguistics had acquired scientific status in the form of
historical-comparative grammar, which rests on two basic notions, sound laws
and analogical levelling. As noted by Sapir, these concepts had been fruitfully
applied in the field of Indo-European and Semitic languages, and more recently
in the field of African and American Indian'^ languages. The ultimate explanation
for the principles of historical-comparative linguistics would have to be sought in
sociology and psychology.-^

Sapir then proceeds to show that linguistics is connected with other disciplines,
and that it fulfills a central role in the study of social behaviour. This is due to the
dialectic relationship between language and social reality: on the one hand, our
cultural patterns are "indexed" in the language (p. 209),''^ and on the other hand,
language is the symbolic guide to social reahty (p. 209). The notion of "pattern-
ing" (or "configuration"), at the linguistic and cultural level, is recurrent through-
out the paper (see especially pp. 212-214). In a central (and often quoted)
passage of the paper we find an adumbration of the Sapir -Whorf hypothesis^":
"Human beings do not live in the objective world alone, nor alone in the world
of social activity as ordinarily understood, but are very much at the mercy of the

Oft

■^° It is a significant fact that in his article "SymboHsm" written for the Encyclopaedia of the Social Sciences,

vol. 14 (New York, 1934), pp. 492^95 [reprinted in The Collected Works of Edward Sapir, vol. Ill, pp. 319-325],

Sapir hardly discusses phonetic symbolism.

References are made to Leonard Bloomfield's work on Algonquian languages and Sapir 's own work on
Athabaskan; cf. Sapir's 1931 paper "The Concept of Phonetic Law as Tested in Primitive Languages by Leonard
Bloomfield" [reprinted in section VI].

-70

^° See also Antoine Meillet, "L'etat actuel des etudes de linguistique generale", Revue des idees 3 (1906)

296-308 [reprinted in A. Meillet, Linguistique historique et linguistique generale, vol. I [Paris, 1921], pp. 1-18].

^^ Sapir illustrates this with reference to domains of central interest to the historical linguist: linguistic paleon-
tology and history of techniques (p. 210).

^" On the antecedents and the posterity of the hypothesis, see John E. Joseph, "The Immediate Sources of
the 'Sapir - Whorf Hypothesis'", Historiographia Linguistica 23 (1996), 365^04; E.F.K. Koerner, "Towards a
Full 'Pedigree' of the Sapir - Whorf Hypothesis: From Locke to Lucy", in Explorations in Linguistic Relativity,
edited by Martin Putz and Marjolijn Verspoor (Amsterdam/Philadelphia, 2000), 1-24. On Whorf's view of
linguistic relativity, see Penny Lee, The Whorf Theory Complex: a Critical Reconstruction (Amsterdam/
Philadelphia, 1996). For an interesting linguistic-anthropological reformulation of the linguistic relativity
hypothesis see John A. Lucy, Language Diversity and Thought (Cambridge, 1992) and Grammatical Categories
and Cognition (Cambridge, 1992), and the volume Rethinking Linguistic Relativity, edited by John Gumperz
and Stephen Levinson (Cambridge, 1996).

Three: Descriptive and Historical Linf^uiMics 161

particular laiimumc which has hcconic llic nicdmm dI expression l».i .. ■ ---tv.

It is quite an illusion to iiiiauinc that owe adjusts to reality essenli.ilh u }»c

use of hinguagc and that language is merely an ineidenlal means of solving ipc-
cific problems ol communication ox retlcclion. Ilie fact of tlu he

"real world* is to a large extent unconsciously buill up on the i .f

the group. No \\\o languages are ever sufficiently similar to be
representing the same social reality. Hie worlds in which different s.-. >c

are distinct worlds, not merely the same world with ditterent lalxls ,w 1

We see and hear and otherwise experience very largelv as we do b. I»c

language habits of our commuiiil\ predispose certain eht>iccs of interpretation **
(p. 2(W, p. 210).

ITiis paper, like most of Sapir's \\o\\ m the years WZ-S-iy."^,^. throws a bridge
between linguistics, sociology and psychology: linguistics is shown to be a neces-
sary foundation for the sociologist (p. 2 10). who has to study its role in sinrial svm-
bolization, and for the psychologist (p. 211 ). who has to study patterns of beha-
viour. As in the encyclopedia article "Philology" Sapir notes the relevance of
Ge^ra/rpsychology for (structural) linguistics (see pp.21 1-212). Also. the relation-
ship of linguistics to philosopin is brought into the discussion: linguists can
prevent philosophers from making naive generali/ations, and from lormulaimg
metaphysical abstractions on the basis of one's "own speech": the linguist "by
the very nature of his subject matter" "the most relativist [student)" of human
behaviour (p. 212) — should make the philosopher aware of the relatiMts i>f lin-
guistic symbolization,'- especially since our symbols are subject to eNa(X)ralion
(cf. pp. 21 1-212). Linguistics, a key science Iim understanding social behaviour.
connects with various other disciplines; it is, however, an autonomous science.
which studies self-contained patterns (p. 212): this is a theme which Sapir repeat-
edly dealt with in the 1920s, more specifically under the heading "the formal
completeness of language."

At the end of the paper Sapir addresses the ciuesiion wliat kind *>i sckiuc im
guistics is.'' Whereas some o{ its aspects (e.g., phonetic description) beltuig within
the natural sciences, and other pertain to biology (when defined as the science of
the "free" development of natural organisms), there can be no doubt f»»r S.ipir
that linguistics, which studies language as a cultural and social product. Iv i-n.-s.
to the social sciences; within this field, it has direct relevance for |v»>cl

• ' Sapir refers ti) hchiiviourislic psycholi>i;y in iix ,ipplii..itu>n lo i.iiivii.iv;i m.
(as a siibstitiilivc slinuilus, see p. 21 1) was later elaKiraleil up«»n hv I con.inl Ml
York. 11)33). pp. 2.^24. l.V>-144.

■^- See also Sapir's review (if Oiiikn aiul Kiehanls s I !„■ V/..//iu<i, ■' \/r.i/M/is' ( I'C^I. inchldcd in a

^^ Tliis problem had been insighilully discussed al the end of the l«»ih ccniun »-» '^ •'•••«-'' !'«'"«♦• ^'*^«*

Henry |!S5()- 1^)07] in his Aninumius Itni^msiuiiu's (Pans. IH*>M

1 62 General Linguistics I

sociology, and anthropology. And Sapir admonishes linguists to integrate their
work within the "interpretation of human conduct in general" (p. 214). As he
observes, linguistics is endowed with a crucial role^^ in the elaboration of a gener-
al methodology for the social sciences.

Pierre Swiggers

One should note that this is partly owing to the fact that its object, language, shows regularities similar to
those observed in the natural sciences (p. 213).

AN AI^IM<(),\( 11 K ) S^ \lh( )1 ls\l

1 572a I A MACHINE is something which a man introduces It) his fclhm-mcn in

order to make things easier, more agreeahle. or more worlh u ' ' ' ' '.if
himself, or for some third, generally iinsti|nilated. p.irls. Ii is m n a

new maehiiie is broiighi up lor nur acceptance, that wc have been sadly hanuMcd
up to that point by the necessity of putting in more time and energy in a given

pursuit, such as striking a light, moving from A to Ii. or disco\ering «m;' <>n

about something, than is advisable in the nature ol things; that the m.. . a

humble slave who would like to cut down this serious expenditure oi lime and
energy; and that, accepting the machine's services, we at once proceed lo read
Shakespeare and to make other explorations into the higher life. Bui ihe aula-
mobile, a labour-saving coiilrixance of obscure intention, msists on the cross-
country spin, on getting itself exhibited, and on divers attentions not menln>ned
in the bond. It saves us five minutes in order that it may diclale the schedule
for five hours. It translates our regret that Shakespeare is inaccessible lo-day into
the impossibility of touching Shakespeare for another month at the least. 'Ilic
tyranny of incidental ser\ ices should be the one obsession of social reformers

Of all insidious machines. wt)rds are the most insidious. Like the humblest o(
kitchen help they worm themselves \nU) tuir good-natured, patronizing confi-
dence and have us at their mercy before we realize that their almost indispcnsiile
usefulness has grooved our minds inti) an infinite tracery of habit. We Ivgm by
coining or adapting words for such symbolic uses as the shifting needs and con-
veniences of custom require. Ihc old need and the old convenience may be left
behind for good and all. but the WDrds which once gave them a habitation we do
not readily relinquish. They tend to remain as landmarks in a vast but finite and
wellnigh intlexible world of symbols, housing new needs and new conveniences.
enlarging or contracting their hospitality, yet always mysteriously themselves
Their hvpnotized creators have no recourse but to pronounce them sanctuaries
and to look anxiously \'ov the di\init\ that must dwell in each of them. Who has
not asked himself the agonizing question. "W hat di>es this word rcallv meany

|572b| Every intelligent person kni>ws that words delude as much as they help.
Many a heated argument, many a difference of philosophical attitude seems lo
resolve itself into \ariously preferred emphases on this or that facet of a word's
customarv surface-range o{ significance. I 'nfortunately for rigorous thinking, the
significance is only in part a coldly symbi>lic reference lo the wt»rld ol cxivri*
ence;more often than not. it also embodies emoti\e elements that have no place
in the objectively verifiable context of things. And yet few accept with or

and conviction the nt)torious failure of a given universe of speechs\int . in-

guage, to correspond lo the universe of phenomena, physical and menial. Il is

164 General Linguistics I

distressing to have two remorseless and even humorous Enghsh thinkers' dis-
cover for us not only sixteen types of aesthetic theory based on as many kinds of
definition of the beautiful, but no less than sixteen appreciably distinct ways of
understanding the term ''meaning."

Messrs. Ogden and Richards are no mere sophists, no clever hair-sphtters. It is
doubtful if the essential limitations of speech have ever been more vividly, yet
sympathetically, realized than in their radical study of symbolism. They make it
clear, as no philologist has ever quite made it clear, why an understanding of the
nature of speech is a philosophic essential, why every epistemology and every
system of logic that does not subject speech, its necessary expressive medium, to
a searching critique is built upon the sands, is sooner or later snared in the irrel-
evances of the medium. Philosophers and psychologists, most of them, have had
Httle patience with the ways of speech. They have either dismissed it as a by-pro-
duct of human behaviour, as an adventitious code that only grammarians need be
seriously interested in, or they have seen in it but a conveniently externalizing
expression, an adequate symbolic complement, of a mental life that is open to
direct observation. They have been either blindly disdainful or blindful trustful.
Profounder insights into the normative influence of speech are not absent from
philosophic and hnguistic literature — see Fritz Mauthner's httle-known "Kritik
der Sprachwissenschaft" — but they have been slightly regarded. "The Meaning
of Meaning" is written from the angle of the logician and the psychologist rather
than from that of the linguist. It seems more than usually significant, therefore,
that the writers have gone so fully into the linguistic factors which are involved
in the puzzhng processes known as thought and interpretation.

The originality of "The Meaning of Meaning" lies chiefly in this, that it refuses
to see a special relation between symbol and "referent" or thing (event) symbol-
ized; further, that it looks upon thinking as the interpreting of "signs," which
interpreting is merely the psychological reaction to the "sign" in the light of past
and present experience. A "door" may be a thing thought of or referred to, what
the authors call a "referent," but it may also be an indication of some other thing
or some event or some attitude that has been or is hnked with it in a context,
physical or psychological or both. In the latter case the "door" (not merely the
written or heard symbol "door," but the thought of the door, whether imaged or
not) becomes a "sign" or natural symbol for a "referent," such as house or opening
or banging or entry into the dining room or whatever else its particular context
and direction of reference lead us to. Symbols, as ordinarily understood, are the
"signs" of thoughts or references (sign-interpretations) and are "causally" re-
lated, in psychological contexts, to these references somewhat as the sign-inter-
pretations themselves are related, again "causally," to the "referents." The relation

1 "'The Meaning of Meaning: a Study of the Influence of Language upon Thought and of the Science of
SymboHsm." C. K. Ogden and I. A. Richards. New York: Harcourt, Brace and Co. $3.75

Three: Descriptive <//;</ Hi\torieal I in^uistics l(iS

Ix'lw ecu a sMiibol (sa\.lhc word "dDor") and a rcfcrcnl (say.a dtK)r or ihis |573m|
door) is merely imputed, even fictitious. Ihc ihouuhi of Mcssrv Ogdcn and
Richards is as simple as it is difficult to urasp. It looks a\say not onl\ from the uni-
\ersals of the realist, hut from the more innocent "concepls" (alvsiracled shorl-
hand references) o\ the ciMiceplualisl ami orthodox linguist as well. It pins its
faith to the closest possible psychological scrutmy of experienced contexts and
feels its way with the canniest of "canons of svmboli/alion."

New sciences are adumbrated in this book. I "lie y are a general theory of signs
(a psychological approach to the problems of epistemology):a theory of symbol-
ism; and, as the most importaiil special development of a general theor> of s\m-
bolism.a broader theor\ o{ language than the philologists hase \et .ittempted. In
an admirable chapter on "Symbol Situations" the writers make it abundantly
clear that language is on\\ in part a coherent system o{ symbolic reference. To a
far greater extent than is generally realized language serves also -a -ffeclive and
\olitional purposes. Perhaps a criticism may be ventured at this [-Hunt. It is true
that the function o{ language is not in practice a purely symbolic or referential
one. but is it not a highly significant lacl. none the less, that its Utrni is so essen-
tially of symbolic pattern ? Most students o{ language, aside from some\^hat naive
teleologists like Professor Jespersen. are inclined to be more interested in the
form than in the function of speech, but. as Messrs. Ogden and Richards might
reflect, that is perhaps their private weakness. In any event, the psychtilogy of the
varying, yet eventually equivalent, forms o{ linguistic expression is a fascinating
subject. Little of real importance seems \el to ha\e been said about it.

ED\N ^ IK

Editorial Note
The Freeman 7 (1923), 572-573.

THE GRAMMARIAN AND HIS LAXCirAGE*

The normal man of intelligence has somethinR of a ronlempt for
linguistic studies, convinced as he is that nothinR can well be more
useless. Such minor usefulness as he concedes to them is of a purely
instrumental nature. French is worth studying because there are French
books which are worth reading. Greek is W(jrth studying -if it is —
because a few plays and a few passages of verse, written in that curiouB
and extinct vernacular, have still the power to disturb our hearts— if
indeed they have. For the rest, there are excellent translations.

Now it is a notorious fact that the linguist is not necessarily deeply
interested in the abiding things that language has done for us. He
handles languages very much as the zoologist handles dogs. The zo-
ologist examines the dog carefully, then he dissects him in order to
examine him still more carefully, and fmally, noting reseml)lances
between him and his cousins, the wolf and the fox, and differences 1k?-
tween him and his more distant relations like the cat and the l)ear,
he assigns him his place in the evolutionary scheme of animated na-
ture, and has done. Only as a polite visitor, not as a zoologist, is he even
mildly interested in Towzer's sweet parlor tricks, however fully ho
may recognize the fact that these tricks could never have evolved
unless the dog had evolved first. To return to the philologist and the
layman by whom he is judged, it is a precisely parallel indifference to the
beauty wrought by the instrument which nettles the judge. And yet
the cases are not altogether parallel. When Towzer has {performed his
tricks and when Porto has saved the drowning man's life, they n^lap.se,
it is true, into the status of mere dog — but even the zoologist's dog is of
interest to all of us. But when Achilles has bewailed the death of his
beloved Patroclus and Clytaemncstra has done her worst, what are we
to do with the Greek aorists that arc left on our hands.' There is a
traditional mode of procedure which arranges them into patterns. It is
called grammar. The man who is in charge of grammar and is called a
grammarian is regarded by all plain men as a frigid and dehumaniz(vl
pedant.

It is not difficult to understand the very pallid status of linguistics
in America. The purely instrumental u.sefulness of language study is
recognized, of course, but there is not and cannot be in this country
that daily concern with foreign modes of expression so natural on the
continent of Europe, whore a number of languages jostle each other in

* American Mercury, 1 (1921): 140-1.35.

168 General Linguistics I

everyday life. In the absence of a strong practical motive for linguistic
pursuits the remoter, more theoretical, motives are hardly given the
opportunity to flower. But it would be a profound mistake to ascribe
our current indifference to philological matters entirely to the fact that
English alone does well enough for all practical purposes. There is
something about language itself, or rather about linguistic differences,
that offends the American spirit. That spirit is rationalistic to the very
marrow of its bone. Consciously, if not unconsciously, we are inclined
to impatience with any object or idea or system of things which cannot
give a four-square reckoning of itself in terms of reason and purpose.
We can see this spirit pervading our whole scientific outlook. |150| If psy-
chology and sociology are popular sciences in America today, that is
mainly due to the prevailing feeling that they are convertible into the
cash value of effective education, effective advertising, and social
betterment. Even here, there is, to an American, something immoral
about a psychological truth which will not do pedagogical duty, some-
thing wasteful about a sociological item which can be neither applied
nor condemned. If we apply the rationalistic test to language, it is found
singularly wanting. After all, language is merely a level to get thoughts
"across." Our business instinct tells us that the multiplication of levers,
all busy on the same job, is poor economy. Thus one way of "spitting
it out" is as good as another. If other nationalities find themselves using
other levers, that is their affair. The fact of language, in other words, is
an unavoidable irrelevance, not a problem to intrigue the inquiring
mind.

There are two ways, it seems, to give linguistics its requisite dignity
as a science. It may be treated as history or it may be studied descrip-
tively and comparatively as form. Neither point of view augurs well
for the arousing of American interest. History has always to be some-
thing else before it is taken seriously. Otherwise it is "mere" history.
If we could show that certain general linguistic changes are correlated
with stages of cultural evolution, we would come appreciably nearer
securing linguistics a hearing, but the slow modifications that eat into
the substance and the form of speech and that gradually remold it
entirely do not seem to run parallel to any scheme of cultural evolution
yet proposed. Since "biological" or evolutionary history is the only
kind of history for which we have a genuine respect, the history of
language is left out in the cold as another one of those unnecessary
sequences of events which German erudition is in the habit of worry-
ing about.

But before pinning our faith to linguistics as an exploration into

ihrcc: Descriptive uml llistumul / im;iii\tic\ 169

form, we might cast an appealing glance at the psychologist, for he is
likely to prove a useful ally. He has himself looked into the subject of
language, which he finds to be a kind of "behavior," a rather siwcial-
ized type of functional adaptation, yet not so specialized but that it
may be declared to be a series of laryngeal habits. We may go even
further, if we select the right kind of psychologist to help us, and have
thought put in its place as a merely "subvocal laryngeating." If these
psychological contributions to the nature of speech do not altogether
explain the Greek aorists bequeathed to us by classical poets, they arc
at any rate very flattering to philology. Unfortunately the philologist
cannot linger long with the psychologist's rough and ready mechani.srns.
These may make shift for an introduction to his science, but his real
problems are such as few psychologists have clearly envisage<l, though
it is not unlikely that psychology may have much to say about them
when it has gained strength and delicacy. The psychological problem
which most interests the linguist is the inner structure of language, in
terms of unconscious psychic processes, not that of the individual's
adaptation to this traditionally conserved structure. It g(x*s without
saying, however, that the two problems are not independent of each
other.

To say in so many words that the noblest task of linguistics is to
understand languages as form rather than as function or as historical
process is not to say that it can be understood as form alone. The formal
configuration of speech at any particular time and place is the result of
a long and complex historical development, which, in turn, is unintelli-
gible without constant reference to functional factors. P\^rm is even more
liable to be stigmatized as "mere" than the historical process which
shapes it. For our characteristically pragmatic American attitude forms
in themselves seem to have little or no reality, and it is for this |I5I| reason
that we so often fail to divine them or to realize into what new patterns
ideas and institutions are balancing themselves or tending to do so.
Now it is very probable that the poise which goes with culture is largely
due to the habitual appreciation of the formal outlines and the formal
intricacies of experience. Where life is tentative and experimental,
where ideas and sentiments are constantly protruding gaunt elbows
out of an inherited stock of meagre, inflexible patterns instead of gra-
ciously bending them to their own uses, form is necessarily felt as a
burden and a tyranny instead of the gentle embrace it should l>e. Per-
haps it is not too much to say that the lack of culture in America is in
some way responsible for the unpopularity of linguistic studies, for
these demand at one and the same time an intense appreciation of a

170 General I.ini^uistics I

given form of expression and a readiness to accept a great variety of

possible forms.

The outstanding fact about any language is its formal completeness.
This is as true of a primitive language, like Eskimo or Hottentot, as of
the carefully recorded and standardized languages of our great cultures.
By "formal completeness" I mean a profoundly significant peculiarity
which is easily overlooked. Each language has a well defined and ex-
clusive phonetic system with which it carries on its work and, more
than that, all of its expressions, from the most habitual to the merely
potential, are fitted into a deft tracery of prepared forms from which
there is no escape. These forms establish a definite relational feeling or
attitude towards all possible contents of expression and, through them,
towards all possible contents of experience, in so far, of course, as ex-
perience is capable of expression in linguistic terms. To put this matter
of the formal completeness of speech in somewhat different words, we
may say that a language is so constructed that no matter what any
speaker of it may desire to communicate, no matter how original or
bizarre his idea or his fancy, the language is prepared to do his work.
He will never need to create new forms or to force upon his language a
new formal orientation — unless, poor man, he is haunted by the form-
feeling of another language and is subtly driven to the unconscious dis-
tortion of the one speech-system on the analogy of the other. The world
of linguistic forms, held within the framework of a given language, is a
complete system of reference, very much as a number system is a com-
plete system of quantitative reference or as a set of geometrical axes of
coordinates is a complete system of reference to all points of a given
space. The mathematical analogy is by no means as fanciful as it ap-
pears to be. To pass from one language to another is psychologically
parallel to passing from one geometrical system of reference to another.
The environing world which is referred to is the same for either lan-
guage; the world of points is the same in either frame of reference. But
the formal method of approach to the expressed item of experience, as
to the given point of space, is so different that the resulting feeling of
orientation can be the same neither in the two languages nor in the
two frames of reference. Entirely distinct, or at least measurably dis-
tinct, formal adjustments have to be made and these differences have
their psychological correlates.

Formal completeness has nothing to do with the richness or the
poverty of the vocabulary. It is sometimes convenient or, for practical
reasons, necessary for the speakers of a language to borrow words from
foreign sources as the range of their experience widens. They may

Three: Descriptive ami lliMonml I iniiuiMus 171

extend the meanings of words which they already poesess, create new
words out of native resources on the analogy of existing terms, or take
over from another people terms to apply to the new conceptions which
they arc introducing. None of these processes affects the form of the
language, any more than the enriching (if a certain portion of space hy
the introduction of new objects afTects the geometrical form of that
region as defined by an accepted mode of reference. It would l>c absurd
to say that Kant's |I52| "Critique of Pure Reason" could be rcndere<l forth-
with into the unfamiliar accents of Eskimo or Hottentot, and yet it
would be absurd in but a secondary degree. What is really meant is
that the culture of these primitive folk has not advanced to the point
where it is of interest to them to form abstract conceptions of a philo-
sophical order. But it is not absurd to say that there is nothing in the
formal peculiarities of Hottentot or of Eskimo which would obscure the
clarity or hide the depth of Kant's thought — indeed, it may l)e sus[KTtr<i
that the highly synthetic and periodic structure of Eskimo would more
easily bear the weight of Kant's terminology than his native German.
Further, to move to a more positive vantage point, it is not absurd to
say that both Hottentot and Eskimo possess all the formal apparatus
that is required to serve as matrix for the expression of Kant's thought.
If these languages have not the requisite Kantian vocabulary', it is not
the languages that are to be blamed but the Eskimo and the Hottentots
themselves. The languages as such are quite hospitable to the addition
of a philosophic load to their lexical stock-in-trade.

The unsophisticated natives, having no occasion to speculate on the
nature of causation, have probably no word that adequately translates
our philosophic term "causation," but this shortcoming is purely and
simply a matter of vocabulary and of no interest wliatever from the
standpoint of linguistic form. From this standpoint the term "causa-
tion" is merely one out of an indefinite number of examples illustrating
a certain pattern of expression. Linguistically — in other words, as
regards form-feeling — "causation" is merely a particular way of ex-
pressing the notion of "act of causing," the idea of a certain type of
action conceived of as a thing, an entity. Xow the form-feeling of such
a word as "causation" is perfectly familiar to Eskimo and to hundretls
of other primitive languages. They have no dithculty in expressing the
idea of a certain activity, say "laugh" or "speak" or "run," in terms of
an entity, say "laughter" or "speech" or "running." If the particular
language under consideration cannot readily adapt itself to this type of
expression, what it can do is to resolve all contexts in which such forms

J 72 General Linguistics I

are used in other languages into other formal patterns that eventually
do the same work. Hence, "laughter is pleasurable," "it is pleasant to
laugh," "one laughs with pleasure," and so on ad infinitum, are func-
tionally equivalent expressions, but they canalize into entirely distinct
form-feelings. All languages are set to do all the symbolic and expres-
sive work that language is good for, either actually or potentially. The
formal technique of this work is the secret of each language.

It is very important to get some notion of the nature of this form-
feeling, which is implicit in all language, however bewilderingly at
variance its actual manifestations may be in different types of speech.
There are many knotty problems here — and curiously elusive ones —
that it will require the combined resources of the linguist, the logician,
the psychologist, and the critical philosopher to clear up for us. There
is one important matter that we must now dispose of. If the Eskimo
and the Hottentot have no adequate notion of what we mean by causa-
tion, does it follow that their languages are incapable of expressing the
causative relation? Certainly not. In English, in German, and in Greek
we have certain formal linguistic devices for passing from the primary
act or state to its causative correspondent, e.g., English to fall, to fell,
"to cause to fall"; vride, to widen; German hangen, "to hang, be sus-
pended"; hdngen, "to hang, cause to be suspended"; Greek phero,
"to carry"; phoreo, "to cause to carry." Now this ability to feel and
express the causative relation is by no manner of means dependent
on an ability to conceive of causality as such. The latter ability is
conscious and intellectual in character; it is laborious, like most con-
scious processes, and it is late in developing. The former ability is un-
conscious and nonintellectual in character, exercises itself 11531 with great
rapidity and with the utmost ease, and develops early in the life of the
race and of the individual. We have therefore no theoretical difficulty in
finding that conceptions and relations which primitive folk are quite
unable to master on the conscious plane are being unconsciously ex-
pressed in their languages — and, frequently, with the utmost nicety.
As a matter of fact, the causative relation, which is expressed only
fragmentarily in our modern European languages, is in many primitive
languages rendered with an absolutely philosophic relentlessness. In
Nootka, an Indian language of Vancouver Island, there is no verb or
verb form which has not its precise causative counterpart.

Needless to say, I have chosen the concept of causality solely for the
sake of illustration, not because I attach an especial linguistic impor-
tance to it. Every language, we may conclude, possesses a complete and

////■('(•; Dcscripitvc mul UiMorunl l.ini;ui\iics 173

psychologically satisfying formal orientation, but thia orienUtion is
only felt in the unconscious of its speakers— is not actually, that is,
consciously, known by thorn.

Our current psychology does not seem altogether adequate to explain
the formation and transmission of such submerged formal systems aa arc
disclosed to us in the languages of the world. It is usual to say that
isolated linguistic responses are learned early in life and that, as these
harden into fixed habits, formally analogous responses are made, when
the need arises, in a purely mechanical manner, 8{)ecific prece<lcnt«
pointing the way to new responses. We are sometimes told that these
analogous responses are largely the result of reflection on the utility
of the earlier ones, directly learned from the social environment. Such
methods of approach see nothing in the problem of linguistic form
beyond what is involved in the more and more accurate control of a
certain set of muscles towards a desired end, say the hammering of a
nail. I can only believe that explanations of this type are seriously
incomplete and that they fail to do justice to a certain innate striving
for formal elaboration and expression and to an unconscious patterning
of sets of related elements of experience.

The kind of mental processes that I am now referring to are, of course,
of that compelling and little understood sort for which the name "in-
tuition" has been suggested. Here is a field which psychology has barely
touched but which it cannot ignore indefinitely. It is preci.sely l>er.iuse
psychologists have not greatly ventured into these difficult readies
that they have so little of interest to offer in explanation of all those
types of mental activity which lead to the problem of form, such as
language, music, and mathematics. We have every reason to surmise
that languages are the cultural deposits, as it were, of a vast and self-
completing network of psychic processes which still remain to be clearly
defined for us. Probably most linguists are convinced that the language-
learning process, particularly the acquisition of a feeling for the formal
set of the language, is very largely unconscious and involves mecha-
nisms that are quite distinct in character from either sensation or re-
flection. There is doubtless something deeper about our feeling for
form than even the majority of art theorists have divined, and it is not
unreasonable to suppose that, as psychological analysis becomes m«rc
refined, one of the greatest values of linguistic study will be in the un-
expected light it may throw on the psychology of intuition, this "in-
tuition" being perhaps nothing more nor less than the "feeling" for
relations.
There is no doubt that the critical study of language may also be of

174 General Linguistics I

the most curious and unexpected helpfulness to philosophy. Few phil-
osophers have deigned to look into the morphologies of primitive lan-
guages nor have they given the structural peculiarities of their own
speech more than a passing and perfunctory attention. When one has
the riddle of the universe on his hands, such pursuits seem trivial
enough, yet when it begins to be |154| suspected that at least some solutions
of the great riddle are elaborately roundabout applications of the rules
of Latin or German or English grammar, the triviality of linguistic
analysis becomes less certain. To a far greater extent than the philos-
opher has realized, he is likely to become the dupe of his speech-forms,
which is equivalent to saying that the mould of his thought, which is
typically a linguistic mould, is apt to be projected into his conception of
the world. Thus innocent linguistic categories may take on the formi-
dable appearance of cosmic absolutes. If only, therefore, to save himself
from philosophic verbalism, it would be well for the philosopher to
look critically to the linguistic foundations and limitations of his
thought. He would then be spared the humiliating discovery that many
new ideas, many apparently brilliant philosophic conceptions, are
little more than rearrangements of familiar words in formally satisfy-
ing patterns. In their recently published work on "The Meaning of
Meaning" Messrs. Ogden and Richards have done philosophy a signal
service in indicating how readily the most hardheaded thinkers have
allowed themselves to be cajoled by the formal slant of their habitual
mode of expression. Perhaps the best way to get behind our thought
processes and to eliminate from them all the accidents or irrelevances
due to their linguistic garb is to plunge into the study of exotic modes
of expression. At any rate, I know of no better way to kill spurious
"entities."

This brings us to the nature of language as a symbolic system, a
method of referring to all possible types of experience. The natural or,
at any rate, the naive thing is to assume that when we wish to com-
municate a certain idea or impression, we make something like a rough
and rapid inventory of the objective elements and relations involved in
it, that such an inventory or analysis is quite inevitable, and that our
linguistic task consists merely of the finding of the particular' words and
groupings of words that correspond to the terms of the objective an-
alysis. Thus, when we observe an object of the type that we call a
"stone" moving through space towards the earth, we involuntarily
analyze the phenomenon into two concrete notions, that of a stone and
that of an act of falling, and, relating these two notions to each other
by certain formal methods proper to English, we declare that "the

Ihrcc: Descriptive mul Histnrunl I ini;iii\iics 175

stone falls." We assume, naively enough, that this is about the only
analysis that can properly be made. And yet, if we look into the way
that other languages take to express this ver>' simple kind of impression,
we soon realize how much may be added to, subtracte<i from, or re-
arranged in our own form of expression without materially altering our
report of the physical fact.

In German and in French we are compelled to assign "stone" to a
gender category— perhaps the Freudians can tell us why this object is
masculine in the one language, feminine in the other; in Chippewa we
cannot express ourselves without bringing in the apparently irrelevant
fact that a stone is an inanimate object. If we find gender beside the
point, the Russians may wonder why we consider it necessar>' to specify
in every case whether a stone, or any other object for that matter, is
conceived in a definite or an indefinite manner, why the difference be-
tween "the stone" and "a stone" matters. "Stone falls" is good enough
for Lenin, as it was good enough for Cicero. And if we find barbarous
the neglect of the distinction as to definiteness, the Kwakiutl Indian
of British Columbia may sympathize with us but wonder why we do
not go a step further and indicate in some way whether the stone is
visible or invisible to the speaker at the moment of speaking and
whether it is nearest to the speaker, the person addressed, or some third
party. "That would no doubt sound fine in Kwakiutl, but we are too
busy!" And yet w^e insist on expressing the singularity of the falling
object, where the Kwakiutl Indian, differing from the Chip}>owa, can
generalize and make a statement which would apply |155| equally well to
one or several stones. Moreover, he need not specify the time of the
fall. The Chinese get on with a minimum of explicit fc.rmal statement
and content themselves with a frugal "stone fall."

These differences of analysis, one may object, are merely formal;
they do not invalidate the necessity of the fundamental concrete an-
alysis of the situation into "stone" and what the stone dcx's, which in
this case is "fall." But this necessity, which we feel so strongly, is an
illusion. In the Nootka language the combined impression of a stone
falling is quite differently analyzed. The stone need not be sfv^cjfically
referred to, but a single word, a verb form, may be used which is in
practice not essentially more ambiguous than our English sentence.
This verb form consists of two main elements, the first indicating gen-
eral movement or position of a stone or stonelike object, while the sec-
ond refers to downward direction. We can get some hint of the ffH'ling
of the Nootka word if we assume the existence of an intransitive verb
"to stone," referring to the position or movement of a atonelike object.

I7(, General Linguistics I

Then our sentence, "The stone falls," may be reassembled into some-
thing like "It stones down." In this type of expression the thing-quality
of the stone is implied in the generalized verbal element "to stone,"
while the specific kind of motion which is given us in experience when a
stone falls is conceived as separable into a generalized notion of the
movement of a class of objects and a more specific one of direction. In
other words, while Nootka has no difficulty whatever in describing the
fall of a stone, it has no verb that truly corresponds to our "fall."

It would be possible to go on indefinitely with such examples of in-
commensurable analyses of experience in different languages. The
upshot of it all would be to make very real to us a kind of relativity
that is generally hidden from us by our naive acceptance of fixed habits
of speech as guides to an objective understanding of the nature of ex-
perience. This is the relativit}' of concepts or, as it might be called, the
relativity of the form of thought. It is not so difficult to grasp as the
physical relativity of Einstein nor is it as disturbing to our sense of
security as the psychological relativity of Jung, which is barely be-
ginning to be understood, but it is perhaps more readily evaded than
these. For its understanding the comparative data of linguistics are a
sine qua non. It is the appreciation of the relativity of the form of
thought which results from linguistic study that is perhaps the most
liberalizing thing about it. What fetters the mind and benumbs the
spirit is ever the dogged acceptance of absolutes.

To a certain type of mind linguistics has also that profoundly serene
and satisfying quality which inheres in mathematics and in music and
which may be described as the creation out of simple elements of a self-
contained universe of forms. Linguistics has neither the sweep nor the
instrumental power of mathematics, nor has it the universal aesthetic
appeal of music. But under its crabbed, technical, appearance there lies
hidden the same classical spirit, the same freedom in restraint, which
animates mathematics and music at their purest. This spirit is antag-
onistic to the romanticism which is rampant in America today and
which debauches so much of our science with its frenetic desire.

Editorial Note

American Mercury 1 (1924), 149-155. [Reprinted in: Edward Sapir, Selected
Writings in Language, Culture, and Personality. Edited by David G. Mandelbaum.
Berkeley: University of California Press, 1949, pp. 150-159]

Three: Dcwripiivc uiui HiMitruiil I iniiutstics 177

Li's Laiii^ucs du Monde, par un uroupc Jc linguislcs sous la direction Uc A.
Mhiii.it ct Marcel Cohen (Collcciion 1 muuisiiquc publico par la Socidld
<de> Linguisliqiic dc Paris. X\'l: Paris. 1 ibiairic Ancicnnc Ivdouard Champion.
1924. 81 1 pp., KS plalcs ol maps)

|373j To the tireless energies dI Prof. A. Meillet. the distinguished Indo-Euro-
peanist, we owe this admirable review ot the languages ol the \M)rld. Ihc work
could hardly have been undertaken except as here pl.inned and carried i»ul. that
is. b\ a number of linguistic specialists. Ii is true that works of a similar nature,
such as Friedrich Miiller's Gnindriss dcr Sprdchwisscnsc/uifi and .A Iroinbelti's
Elenwnti di CJloitoloiiia, have been written by individual scholars, but. on the
whole, it was probably wise to sacrifice something of unity of treatment to the
greater authoritativeness that was bound to result from a division of labor

The parts into which the book falls are: an Introduction, by A. Meillet: Indo-
European, by J. Vendryes; Mamito-Semitic, by Marcel (\)hen (it is with great
satisfaction thai one sees a conservative book ol this ivpe recognizing the funda-
mental points of accord that have long been pointed out between Semitic and
"Hamitic" to the point of frankly uniting them into a single genetic group);
Finno-Ugrian and Samoyed, by A. Sauvageot; Turkish. Mongol, and lungusic
Languages, by J. Deny; Japanese; Corean; Ainu: "Hyperborean" Languages (i.e.
Chukchee. Yukagir, and Gilyak). all four by S. Llissee\: Special languages t>f the
Ancient Near East (particularly Sumerian. I'lamite. Hiitite. l.\dian. and iJrus-
can), by C. Autran; Basque, by George Lacombc: Ni>rthcrn Caucasic Languages,
by N. Troubetzkoy; Southern Caucasic Languages, by A. Meillet: I)ra\idian. by
Jules Bloch; Sino-Tibetan, by J. Przyluski (this term is much to be preferred to the
misleading "Indo-Chinese" that has been current: "Sinic" is perhaps even better);
Austroasiatic Languages (Mon-Khmer. Annamiic. and Munda). by J. Pry/luski;
Malayo-Polynesian.by Gabriel Ferrand (Papuan Languages, which do not pro|X'rly
belong here, are brielly treated at the end of this section ): Australian Language\ by
A. Meillet; Languages of the Soudan and of Guinea, by Maurice Delafosso; Bantu.
by Miss L. Homburger; Bushman and Hottentot, by Miss L. Homburger: |374|
and American Languages, by P. Rivet. Iliere is an adequate ecjuipment of biblio-
graphies and maps and an excellent index. All in all. the book is an achievement
and no serious student of general linguistics or descriptive anthro|xWogy can afford
to do without it. That it will need to be replaced by ani)thcr work of similar sco(V
in a few decades goes without saying (certain oi its paragraphs became antiquated
in the writing !) but for the present it is indispensable.

Just because this work is so precious lor the linguist it will not seem ungracu>us
if we point out certain shortciMnings. In the lirsl place a number ol impt>itaiit lan-
guages have slipped out from under ihe specialists, llie editors and their staff will
be chagrined to discover that the Andaman group, which includes a considerable
number of quite distinct dialects or languages, and the isolated Siberian group to
which belong "Yenissei Ostyak" {\o be carefully distinguished from the Ugro-
Finnic"Ostvak"and from the "Oslvak" dialect of Samoyed) and Kott are entire-

1 7j^ General Linguistics I

Iv omitted. Both ol these isolated families are treated in considerable detail in
irombetti's Elementi and both are of crucial importance for the early linguistic
history of Asia.Trombetti produces some evidence, by no means to be despised,
which tends to connect the Ycnissei Ostyak group with Sino-Tibetan. A mere
glance at F. N. Finck's useful little Sprachstdmme des Erdkreises would have in-
sured at least a mention of the two groups. A more excusable omission is that of
Zandawe, a language recently discovered in east central Africa and showing
unmistakable resemblances to the Bushman and Hottentot languages far to the
south (see Trombetti).The historical importance of this language is obvious.

A second and probably more serious criticism is the lack of a consistent plan
in the treatment of the various sections. Mechanical uniformity was rightly rejec-
ted by the editors, but they have gone to the opposite extreme. As it is, certain
languages or groups of languages receive an altogether disproportionate share of
attention. In some sections a good deal of useful information is given on the mor-
phology of the languages listed, in others there is considerable detail of a biblio-
graphical and geographical nature but no vitalizing hints as to the nature of the
languages themselves, in still others a vast field is dismissed with a few perfunc-
[375|tory remarks and a shrug of the shoulders. The editors cannot honestly
retort that they have had to omit all grammatical discussion where none is given
in the book because of the scantiness of the data. As a matter of fact, the descrip-
tive material available in many such cases is of a very high order of merit. There
would have been no more essential difficulty, for instance, in giving some ele-
mentary idea of Algonkin or Siouan or Athabaskan or Maya structure than of
Hottentot or Polynesian structure and such indications would have added
immeasurably to the value of the work, which now hovers uncertainly between
the geographical listing of groups and sub-groups and the morphological discus-
sion of languages. The ideal method would probably have been to combine the
two, as in the admirable section on Hamito-Semitic, which could well have spa-
red, on the other hand, a great deal of its rather irrelevant historical detail.

One other point. It was cruel to assign the vast field of American Indian lan-
guages to a single specialist. No one person living today could even begin to get
his bearings in it, let alone do justice to it. It might have been necessary for the
editors to go outside of France and to secure the cooperation of at least one spe-
cialist for North America north of Mexico and another for Mexico and Central
America, leaving the South American field in the hands of M. Rivet, who is
obviously the one best quahfied to handle it. If it was the intention of the editors
to show how well an essentially international task could be carried out with the
splendid resources of French scholarship alone, all we can say is that they must
be congratulated on coming as near solving an impossible task as it was reason-
ably possible to do.

Ottawa, Ont. E, Sapir.

Editorial Note
Modern Language Notes 40 (1925), 373-375.

SOUND PATTERNS IN LANGUAGE

Edward Sapiu
Victoria Museum, Ottawa

There used to be and to some extent still is a fcolinK atnonR linguists
that the psychology of a language is more particularly concerned with
its grammatical features, but that its sounds and its phonetic procos.sed
belong to a grosser physiological substratum. Thus, we sometimes
hear it said that such phonetic processes as the palatalizing of a vowel
by a following i or other front vowel ("umlaut") or the series of shifts
in the manner of articulating the old Indo-European stopped consonants
which have become celebrated under the name of "Grinmi's Law" are
merely mechanical processes, consummated by the organs of spe«*ch and
by the nerves that control them as a set of shifts in relatively simple
sensorimotor habits. It is my purpose in this paper, as briefly as
may be, to indicate that the sounds and sound processes of speech
cannot be properly understood in such simple, mechanical terms.

Perhaps the best way to pose the problem of the psychology of
speech sounds is to compare an actual speech sound with an identical
or similar one not used in a linguistic context. It will become evident
almost at once that it is a great fallacy to think of the articulation of
a speech sound as a motor habit that is merely intended to bring alx>ut
a directly significant result. A good example of superficially similar
sounds is the wh of such a word as when, as generally jjronounred in
America (i.e., voiceless w or, perhaps more accurately analyzeil, a,^pira-
tion plus voiceless w plus voiced i^'-glide), and the sound made in blow-
ing out a candle, with which it has often been compared. We are
not at the present moment greatly interested in whether thc^^e two
articulations are really identical or, at the least, very similar. L«*t us
assume that a typically pronounced wh is identical with the sound
that results from the expulsion of breath through pursetl lips when a
candle is blown out. We shall assume identity of both articulation
and quality of perception. Does this identity amount to a psycho-
logical identity of the two processes? Obviously not. It is worth
pointing out, in what may seem pedantic detail, wherein they ilifier.

180 General Linguistics I

1. The candle-blowing sound is a physical by-product of a directly
functional act, the extinguishing of the candle by means of a peculiar
method of producing a current of air. So far as normal human interest
is concerned, this sound serves merely as a sign of the blowing out, or
attempted blowing out, itself. We can abbreviate our record of the
facts a little and say that the production of the candle-blowing sound
is a directly functional act. On the other hand, the articulation of
the lyA-sound in such a word as when has no direct functional value;
it is merely a link in the construction of a symbol, the articulated or
perceived word when, which in turn assumes a function, symbolic at
that, only when it is experienced in certain linguistic contexts, such
as the saying or hearing of a sentence like When are you coming? In
brief, the candle-blowing wh means business; the speech sound wh is
stored-up play which can eventually fall in line in a game that merely
refers to business. Still more briefly, the former is practice; the
latter, art.

2. Each act of blowing out a candle is functionally equivalent, more
or less, to every other such act; hence the candle-blowing wh is, in the
first instance, a sign for an act of single function. The speech sound
wh has no singleness, or rather primary singleness, of reference. It is a
counter in a considerable variety of functional symbols, e.g. when,
whiskey, wheel. A series of candle-blowing sounds has a natural func-
tional and contextual coherence. A series of ly/i-sounds as employed
in actual speech has no such coherence; e.g., the series wh{en), wh(iskey),
wh(eel) is non-significant.

3. Every typical human reaction has a certain range of variation
and, properly speaking, no such reaction can be understood except as
a series of variants distributed about a norm or type. Now the candle-
blowing wh and the speech sound wh are norms or types of entirely
distinct series of variants.

First, as to acoustic quality. Owing to the fact that the blowing out
of a candle is a purely functional act, its variability is limited by the
function alone. But, obviously, it is possible to blow out a candle in
a great number of ways. One may purse the lips greatly or only a little ;
the lower lip, or the upper lip, or neither may protrude; the articulation
may be quite impure and accompanied by synchronous articulations,
such as a x-hke (velar spirant) or sA-like sound. None of these and
other variations reaches over into a class of reactions that differs at all
materially from the typical candle-blowing wh. The variation of wh
as speech sound is very much more restricted. A when pronounced,

Three: Descriptive mul Ihstoru ttl I tn^uisiies 181

for instance, with a wh in whicli the linvcr lip protrudiMi or with a xvh
that was contaminated with a .s/j-sound would h<? frit a.s diHtinctly
"off color." It could be tolerated only as a joke or a jxTsonal .sj>eech
defect. But the variability of wh in lanj^uai^e is not only Ie.Hs wide
than in candle-blowing, it is also different in tendency. The latter
sound varies chiefly along the line of exact i^laee (or [)lace.s) of articula-
tion, the former chiefly along the line of voicing. Psycholoi^ically uh
of when and similar words is related to the w of well and Kiriiilar
words. There is a strong tendency to minimize the a.spiration and to
voice the labial. The gamut of variations, therefore, run.s roughly
from hW (I use W for voiceless w) to w. Needless to say, there is
no tendency to voicing in the candle-blowing vh, for such a tendency
would contradict the very purpose of the reaction, which is to releiisc
a strong and unhampered current of air.

Second, as to intensity. It is clear that in this respect the two
series of variations differ markedly. The normal intensity of the
candle-blowing sound is greater than that of the linguistic uh; this
intensity, moreover, is very much more variable, depending as it does
on the muscular tone of the blower, the size of the flame to be extin-
guished, and other factors. All in all, it is clear that the resemblance
of the two t6'/i-sounds is really due to an intercrossing of two absolutely
independent series, as of two independent lines in space that have one
point in common.

4. The speech sound wh has a large number of associations with
other sounds in symbolically significant sound-groups, e.g. uh-e-n,
wh-i-s-k-ey, wh-ee-l. The candle-blowing sound has no sound as.>*o-
ciations with which it habitually coheres.

5. We now come to the most essential point of difference. The
speech sound wh is one of a definitely limit^xl number of .^^ound.'^ (e.g.
wh, s, t, I, i, and so on) which, while differing qualitatively from one
another rather more than does wh from its candle-blowing e<juivalent,
nevertheless belong together in a definite system of symlHilically
utilizable counters. Each member of this .system is not only character-
ized by a distinctive and slightly variable articulation and a corre-
sponding acoustic image, but also— ar^/ this is crucial — by a psycho-
logical aloofness from all the other members of the system. The
relational gaps between the .sounds of a language are just as neres,>iary
to the psychological definition of these sounds as the articulations
and acoustic images which are customarily used to ilefine them. \

182 General Linguistics I

sound that is not unconsciously felt as "placed"^ with reference to
otlier sounds is no more a true element of speech than a lifting of the
foot is a dance step unless it can be "placed" with reference to other
movements that help to define the dance. Needless to say, the candle-
blowing sound forms no part of any such system of sounds. It is not
spaced off from nor related to other sounds — say the sound of humming
and the sound of clearing one's throat^ — which form with it a set of
mutually necessary indices.

It should be sufficiently clear from this one example — and there are
of course plenty of analogous ones, such as m versus the sound of
humming or an indefinite series of timbre-varying groans versus a
set of vowels — how little the notion of speech sound is explicable in
simple sensorimotor terms and how truly a complex psychology of
association and pattern is implicit in the utterance of the simplest
consonant or vowel. It follows at once that the psychology of phonetic
processes is unintelligible unless the general patterning of speech
sounds is recognized. This patterning has two phases. We have been
at particular pains to see that the sounds used by a language form a
self-contained system which makes it impossible to identify any of
them with a non-linguistic sound produced by the "organs of speech,"
no matter how great is the articulatory and acoustic resemblance
between the two. In view of the utterly distinct psychological back-
grounds of the two classes of sound production it may even be seriously
doubted whether the innervation of speech-sound articulation is ever
actually the same type of physiological fact as the innervation of
"identical" articulations that have no Hnguistic context. But it is not
enough to pattern off all speech sounds as such against other sounds
produced by the "organs of speech." There is a second phase of
sound patterning which is more elusive and of correspondingly greater
significance for the linguist. This is the inner configuration of the
sound system of a language, the intuitive "placing" of the sounds with
reference to one another. To this we must now turn.

Mechanical and other detached methods of studying the phonetic
elements of speech are, of course, of considerable value, but they have
sometimes the undesirable effect of obscuring the essential facts of
speech-sound psychology. Too often an undue importance is attached
to minute sound discriminations as such; and too often phoneticians

1 This word has, of course, nothing to do here with "place of articulation."
One may feel, for instance, that sound A is to sound B as sound X is to sound Y
without having the remotest idea how and where any of them is produced.

Three: Dcscnpiivc ami HiMormil 1 ni^utsucs 183

do not realize that it is not enough to know tliat a certain sound occurs
in a lan<z;uage, but that one must ascertain if the sound is a typical
form or one of the points in its sound pattern, or is merely a variant
of such a form. There are two types of variation that tend to ol*curc
the distinctiveness of the difTerent points in the [)h<)nf'tic pattern of a
lanp;uag;e. One of these is inciividual variation, it is true that no
two individuals have precisely the same pronunciation of a lanRuaRe,
but it is equally true that they aim to make the same sound discrimina-
tions, so that, if the qualitative difTerences of the sounds that mako up
A's pattern from those that make up B's are perceptible to a minul4»
analysis, the relations that obtain between the elements in the two
patterns are the same. In other words, the patterns are the sainc
pattern. A's s, for instance, may differ (piite markedly from B's «,
but if each individual keeps his 8 equally distinct from such points id
the pattern as th (of think) and sh and if there is a one to one cor-
respondence between the distribution of A's s and that of B's, then
the difference of pronunciation is of little or no interest for tlie phonetic
psychology of the language. We may go a step further. Ix't u.s
s>Tnbolize A's and B's pronunciations of s, th, and sh as follows:

A: th s sh
B: thi Si shi

This diagram is intended to convey the fact that B's s is a lisped s
which is not identical with his interdental th, but stands nearer
objectively to this sound than to A's s; similarly, B's sh is acoustically
somewhat closer to A's s than to his sh. Obviously we cannot discover
B's phonetic pattern by identifying his sounds with their nearest
analogues in A's pronunciation, i.e. setting thi = th, si = variant of
thj shi = s. If we do this, as we are quite likely to do if we are olv
sesscd, like so many linguists, by the desire to apply an absolute and
universal phonetic system to all languages, we get the ft)llowing pattern
analj^'sis:

A: th s sh

E: thi si shi —

which is as psychologically perverse as it is '•objectively" accurate.
Of course the true pattern analysis is:

A: th s sh

B: thi .«?, .«•/»;

184 General Linguistics I

for the objective relations between sounds are only a first approxima-
tion to the psychological relations which constitute the true phonetic
pattern. The size of the objective differences th — s, s — sh, thi — si,
Si — shi, th — Si, si — s, s — shi, and shi — sh does not correspond to the
psychological "spacing" of the phonemes th, s, and sh in the phonetic
pattern which is common to A and B.

The second type of variation is common to all normal speakers of
the language and is dependent on the phonetic conditions in which
the fundamental sound ("point of the pattern") occur^. In most
languages, what is felt by the speakers to be the "same" sound has
perceptibly different forms as these conditions vary. Thus, in (Amer-
ican) English there is a perceptible difference in the length of the vowel
a of bad and bat, the a-vowel illustrated by these words being long or
half-long before voiced consonants and all continuants, whether voiced
or unvoiced, but short before voiceless stops. In fact, the vocalic
alternation of bad and bat is quantitatively parallel to such alternations
as bead and beat, fade and fate. The alternations are governed by
mechanical considerations that have only a subsidiary relevance for
the phonetic pattern. They take care of themselves, as it were, and
it is not always easy to convince natives of their objective reality,
however sensitive they may be to violations of the unconscious rule
in the speech of foreigners. It is very necessary to understand that
it is not because the objective difference is too slight to be readily
perceptible that such variations as the quantitative alternations in
bad and bat, bead and beat, fade and fate stand outside of the proper
phonetic pattern of the language (e.g., are not psychologically parallel
to such qualitative-quantitative alternations as bid and bead, fed and
fade, or to such quantitative alternations as German Schlaf and schlaff,
Latin ara and ara), but that the objective difference is felt to be slight
precisely because it corresponds to nothing significant in the inner
structure of the phonetic pattern. In matters of this kind, objective
estimates of similarity or difference, based either on specific linguistic
habits or on a generalized phonetic system, are utterly fallacious. As
a matter of fact, the mechanical English vocalic relation bad: bat
would in many languages be quite marked enough to indicate a rela-
tion of distinct points of the pattern, while the English pattern rela-
tion -^ -d, which seems so self-evidently real to us, has in not a few
other languages either no reality at all or only a mechanical, condi-
tional one. In Upper Chinook, for instance, t: d exists objectively
but not psychologically; one says, e.g., inat 'across,' but inad before

Three: Drsinpiivc und lliMorn ul l.iUf^uiMuw 185

words boiiinning with a vowel, aiul the two forms of the final cr»n.Honanl
are undoubtedly felt to be the "same" sound in exactly the same mime
in which the English vowels of bad and bat are felt by uh to l>o identical
phonetic elements. The Upper Chinook d exists only an a mechaiiiraJ
variant of t; hence this alternation is not tin* saine p«ycholf>gically a/<
the Sanskrit sandhi variation -t: -d.

Individual variations and such conditional variations as wp have
discussed once cleared out of the way, we arrive at the genuine pattern
of speech sounds. After what we have siiid, it almost j;o<\s without
saying that two languages, A and B, may have identical .sounds but
utterly distinct phonetic patterns; or they may have mutually in-
compatible phonetic systems, from the articulat<jry and acoustic stand-
point, but identical or similar patterns. The following schematic
examples and subjoined comments will make this clear. Souiuis which
do not properly belong to the pattern or, rather, are variants within
points of the pattern are put in parentheses. Long vowels are desig-
nated as a'; 77 is ug of sing: 6 and 5 are voiceless and voiced interdental
spirants; x and 7 are voiceless and voiced guttural spirant.s; ' is glottal
stop; ' denotes aspirated release; c and o are open e and u.

(e)

(o)

(a)

(*')

{e)

(0-)

' h w y I m n (tj)

p t k

V' V k'

(b) (d) (g)

f d s X

(v) (5) (z) (7)

but B: a e e i u

(a) (€-) (e) (i-) (u) (u, ^^j

(') h (iv) (y) (/) m n r,

It t

;p') in

ik')

b d

(!) io) ^

(:r)

V b z

235 General Lini^iiistics I

We will assume for A and B certain conditional variants which are
all of types that may be abundantly illustrated from actual languages.
ForA:

1. e occurs only as palatalized form of a when following y or i.
In many Indian languages, e.g., ye = ya.

2. e is dropped from z-position when this vowel is final. Cf. such
mechanical alternations as Eskimo -e: -i-t.

3. is dropped from w-position when this vowel is final. Cf. 2.

4. 5 occurs only as labialized form of a after w or u. Cf. 1. (In
Yahi, e.g., wowi 'house' is objectively correct, but psychologically
wrong. It can easily be shown that this word is really wawi and
"feels" like a rhyme to such phonetic groups as lawi and bawi; short
in an open syllable is an anomaly, but o' is typical for all Yana dialects,
including Yahi.)

5. 77 is merely n assimilated to following k, as in Indo-European.

6. 6, d, g, V, z, 8, 7 are voiced forms of p, t, k, f, s, 6, x respectively
when these consonants occur between vowels before the accent (cf.
Upper Chinook wa'pul 'night': wabu'lmax 'nights'). As the voiced
consonants can arise in no other way, they are not felt by the speakers
of A as specifically distinct from the voiceless consonants. They feel
sharply the difference between p and p', as do Chinese, Takelma, Yana,
and a host of other languages, but are not aware of the alternation
p: h.

And for B:

1. Long vowels can arise only when the syllable is open and stressed.
Such alternations as ma'la: u"-mala are not felt as involving any but
stress differences. In A, ma'la and viala are as distinct as Latin
"apples" and "bad" (fem.).

2. ' is not an organic consonant, but, as in North German, an attack
of initial vowels, hence 'a- is felt to be merely a-. In A, however, as
in Semitic, Nootka, Kwakiutl, Haida, and a great many other lan-
guages, such initials as 'a- are felt to be equivalent to such consonant
+ vowel groups as via- or sa-. Plere is a type of pattern difference
which even experienced linguists do not always succeed in making
clear.

3. w and y are merel}^ semi-vocalic developments of u and i. Cf.
French oui and hier. In A, w and y are organically distinct con-
sonants. Here again linguists often blindly follow the phonetic feel-
ing of their own language instead of clearly ascertaining the behavior

Time: Dcsciiptivf and lhsu»ru nl I in^uiMus 187

of the language invostipalfd. Tlio difTcrmco, o.^., Ix-twi-on aua and
men is a real one for some languages, a phantom for othir^

•1. / arises merely as dissimilated variant of n.

5. p', t', k' are merely p, t, k with i)reath release, chaructoriutic of B
at the end of a word, e.g. ap-a: «;/. This sort of alternation is < •
mon in aboriginal America. It is the rever.-^e of the Mngli.'^h li.'. •
tame with aspirated t (t'e'hfi) but hate with una.spirated, or vrry wr
aspirated, release (heH).

G. /, d, and x similarly arise from the unvoicing of final v, 6 and 7;
e.g., av-a: of. 2 and s also alternate in this way, but there i.s a true »
besides. From the point of view of B, .v in such phonems at? sa and asa
is an utterly distinct sound, or rather point in the phonetic pattern,
from the objective!}^ identical as which alternates with az-n}

The true or intuitively felt phonetic systems (patterns) of A and B,
therefore, are:

^ If B ever develops an orthography, it is likely to fall into the lial)it of writing
az for the pronounced as in cases of typ>e az-a: as, hut as in cases of type aa-o: a$.
Philologists not convinced of the reality of phonetic patterns h.** hrn* • ■•■I

will then be able to "prove" from internal evidence that the change (»f « ■_ , i-

cal V, z, 5, 7 to -/, -s, -d, -X did not take place until after the language wna reduced
to writing, because otherwise it would be "impossible" to explain why -a should
be written -z when there was a sign for s ready to hand and why signs should not
have come into use for/, 0, and x. As soon as one realizes, ht.wevor, that '■ideal
sounds," which are constructed from one's intuitive feeling of the significant rela-
tions between the objective sounds, are more "real" to a naive speaker ihjui the
objective sounds themselves, such internal evidence loses iiiurh of its force
The example of s in B was purposely chosen to illustrate an intcre.Hting phen'-n^-
non, the crossing in a single objective phoneme of a true element of the ph.
pattern with a secondary form of another such element. In H, e g , ohjec:... .
is a pool of cases of "true s" and "pseudo-.s." Many interesting and subtle euuu-
ples could be given of psychological difference where there i.s objortiv.
or similarity so close as to be interpreted by the reconier as identity 1
an Athabaskan language with significant pitch diflcrcnces, there is a true n
tone and a pseudo-middle tone which results from the lowering of a high 1«m.. ;••
the middle position because of certain mechanical rules of tone sandhi. I doubl
very much if the intuitive psychology of these two middle tones is •
There are, of course, analogous traps for the unwary in Chincac. H.-v "^

Chinese kindly formalized for us their intuitive feeling a^xjut lb© e«arnlial
tone analysis of their language, it is exceedingly dt.ul»tful if «»ur <' ' " .!
ears and kymograjjhs wouUi have succeeded in discovering the exact j ^

of Chinese tone.

138 General Linguistics I

A: n I u

a' t* u'

w y I m n

6
e

s
i

B:a€ e i u o o

h m n 7]

p t k

b d g

V 8 z y

which show the two languages to be very much more different pho-
netically than they at first seemed to be.

The converse case is worth plotting too. C and D are languages
which have hardly any sounds in common but their patterns show a
remarkable one to one correspondence. Thus:

C: a e i

w y I m n

t
d

f s

q (velar k)

9 (velar g)

X (velar x)

r m ij

7 (velar 7)

h (laryngeal h)

* As in French jour.

* Bilabial v, as in Spanish.
' As in German ich.

Three: Descriptive (irul llisutrual l.m^uisiics 189

Languages C and D have far loss suporficial siniihirity m thoir mund
systems than have A and B, but it is obvious at a ^hmrc thai tlirir
patterns are built on very much more similar lines. If wo alluwcil
ourselves to speculate genetically, we might suspect, on goneral prin-
ciples, that the phonetic similaritios b('t\v«'(>n A ami li, which wr will
suppose to be contiguous languages, are due to historical contact , but
that the deeper pattern resemblance between C and I) is an index of
genetic relationship. It goes without saying that in the cr>mplcx
world of actual linguistic history wo do not often find the i'
facts working out along such neatly schematic linos, but it i

expedient to schematize here so that the pattern concept might emerge
with greater clarity.

An examination of the patterns of C and I) shows that there is still
a crucial point that we have touched on only by implication. We
must now make this clear. We have arranged the sounds of C and D
in such a way as to suggest an equivalence of "orientation" of any
one sound of one system with some sound of the other. In comparing
the systems of A and B we did not commit ourselves to specific o<juiva-
lences. We did not wish to imply, for instance, that A's s wa.s or was
not "oriented" in the same way as B's, did or did not occupy the same
relative place in A's pattern as in B's. But here we tlo wish to imply
not merely that, e.g., C's y corresponds to J)'s // or C"s h to I)'.»< h,
which one would be inclined to grant on general phonetic groundji,
but also that, e.g., C's w corresponds to D's v while ("s b corresponda
to D's /3. On general principles such pattern alignments a.'? the
latter are unexpected, to say the least, for bilabial ^i ros«>nibles «r rather
more than dentolabial v does. Why, then, not allow /3 to occupy the
position we have assigned to y? Again, why shouM D's; Ih» supix*^.^!
to correspond to C's y when it is merely the voiced form of .s? Should
it not rather be placed under s precisely as, in C's sj-ntom, 6 i.s placi^I
under /;? Naturally, there is no reason wliy the intuitive patti^rn
alignment of sounds in a given language should not be iilontical willi
their natural phonetic arrangement and, one need hardly say, it is
aknost universally true that, e.g., the vowels form l>oth a natural rtnd
a pattern group as against the consonants, that such stopiKx! sounds
as p, t, k form both a natural antl a i)attorn group Jt«< <»p|Kised to thr
equally coherent group b, d, g (provided, of course, the languaKo |m»vsoj*-
ses these two scries of stoi)ped consonants). And yet it is nuwt im-
portant to emphasize the fact, strange but indubitable, that a pattern
alignment does not need to corrosi)ond exactly to the more obvious

I9() General Linguistics I

phonetic one. It is most certainly true that, however likely it is that
at last analysis patternings of sounds are based on natural classifica-
tions, the pattern feeling, once established, may come to have a linguis-
tic reality over and above, though perhaps never entirely at variance
with, such classifications. We are not here concerned with the histor-
ical reasons for such phonetic vagaries. The fact is that, even from a
purely descriptive standpoint, it is not nonsense to say that, e.g., the
s or IV of one linguistic pattern is not necessarily the same thing as the
s or w of another.

It is time to escape from a possible charge of phonetic metaphysics
and to face the question, ''How can a sound be assigned a 'place' in a
phonetic pattern over and above its natural classification on organic
and acoustic grounds?" The answer is simple. "A 'place' is in-
tuitively found for a sound (which is here thought of as a true 'point
in the pattern,' not a mere conditional variant) in such a system be-
cause of a general feeling of its phonetic relationship resulting from
all the specific phonetic relationships (such as parallelism, contrast,
combination, imperviousness to combination, and so on) to all other
sounds." These relationships may, or may not, involve morphological
processes (e.g., the fact that in English we have morphological alterna-
tions like wife: wives, sheath: to sheathe, breath: to breathe, 7nouse: to
mouse helps to give the sounds /, 6, s an intuitive pattern relation to
their voiced correlates v, b, z which is specifically different from the
theoretically analogous relation p, t, k: b, d, g; in English, / is nearer
to V than p is to 6, but in German this is certainly not true).

An example or two of English sound-patterning will help us to fix
our thoughts. P, t, and k belong together in a coherent set because,
among other reasons: 1, they may occur initially, medially, or finally;
2, they may be preceded by s in all positions (e.g. spoon: cusp, star:
hoist; scum: ask); 3, they may be followed by r initially and medially;

4, they may be preceded by s and followed by r initially and medially;

5, each has a voiced correspondent (6, d, g); 6, unlike such sounds as/
and 6, they cannot alternate significantly with their voiced correspon-
dents; 7, they have no tendency to be closely associated, either pho-
netically or morphologically, with corresponding spirants (p:f and t:d
are not intuitively correct for English; contrast Old Irish and Hebrew
t:d, k:x, which were intuitively felt relations — Old Irish and Hebrew
6 and x were absolutely different types of sounds, psychologically,
from English 6 and German x). These are merely a few of the rela-
tions which help to give p, t, k their pattern place in English.

ihrcc: Pcscripiivc luiil Hisionciil I iniiuisfics

A second example is 7; of ,sing. In spile of wlial phonotician-H tell
us about this sound {b:tn as d:n as g'.rj), no nnivc KnKlisli-.s|M'Hkit»K
person can be made to feci in liis bones that it I>e|()np8 to a sintde
series with 7)i and //. Psychulo^ieally it cannot be nroup<'d with thfin
because, unhke them, it is not a freely movable consonant (tl!'--- '••
no words beginninp:; with 17). It still /er/.s- like 77^7, howovcr littU- it
like it. The relation antiand = sink: sing is psych()lo;;icully oa well
as historically correct. Orthography is by no means solely resiKinsible
for the "ng feeling" of 77. Cases like -rjg- in fingrr and nngrr do not
disprove the reality of this feeling, for there is in Knglish a pattern
equivalence of -rig-'--V ^^^^ -nd-:-}id. What cases likr .'^iriger with
-rj- indicate is not so much a pattern difTerence -rjg-'.-Tj-, which is not
to be construed as analogous to -?id-:-n- (e.g. u-indow: winnow), a.s an
analogical treatment of medial elements in terms of their final form
{singer: sing like cutter : cut) .^

To return to our phonetic patterns for C and D, we can now brttcr
understand why it is possible to consider a sibilant like j as le.vs closely
related in pattern to its voiceless form s than to such a set of voiced
continuants as v, r, 7n, t). We might find, for instance, that i never
alternates with j, but that there are cases of s'.b analogous to cii.scs of
/:/3 and x'.y; that ava, aja, ara alternate with au, ai, ar; that com-
binations like -al3d, -adg, -ayd are possible, but that coml^inations
of type -ajd and -avd are unthinkable; that v- and /- are possible initials,
like r-, jn-, and rj-, but that (3-, 8-, y-, 7- are not allowed. The product
of such and possibly other sound relations would induce a feeling
that j belongs with v, r, m, 77; that it is relatetl to i; and that it has
nothing to do with such spirants as i and 5. In other words, it "feels"
like the y of many other languages, and, as // itself is alxsent in D, we
can go so far as to say that j occupies a "place in the pattern" that
belongs to tj elsewhere.

In this paper I do not wish to go into the complex ami tanpUxl

« Incidentally, if our theory is correct, such a. form as singer I •
conscious analysis into a word of absolute significance sing and n ■
dent agentive element -er, which is appended not to a stem, an ah^tracted rmdiral
element, hut to a true word. Hence sing: siJiger is not psychologically :•• ' - •''
to such Latin forms as can-: can-tnr. It wouUi jihnost st'om th:it ihr Kn>;
tence on the absoluteness of its significant words tended at the same t«:
many of its derivative suffixes a secondary. revitaUzod reality, -rr, for .
might ahuost be construed as a "word" which occurs only as the second elentcnt
of a compound, cf. -man in words like Inngshnrntum. An Prof \. Bloonifjr'.d
points out to me, the agentive -€r contnusts with the cinparativo -<r. whuh
allows the adjective to keep its radical ionu in -ng- (e.f.. toru; with -f; longrr
with -7/7-).

192 General Linguistics I

problems of the nature and generality of sound changes in language.
All that I wish to point out here is that it is obviously not immaterial
to understand how a sound patterns if we are to understand its history.
Of course, it is true that mechanical sound changes may bring about
serious readjustments of phonetic pattern and may even create new
configurations within the pattern (in Modern Central Tibetan, e.g.,
we have 6-, d-, g-\ B'-, D'-, G'-,'' while in classical Tibetan we have, as
correspondents, 7nb-, nd-, rig-: b-, d-, g-; mb-, nd-, rjQ- are here to be
morphologically analyzed as nasal prefix + b-, d-, g-). But it is
equally true that the pattern feeling acts as a hindrance of, or stimulus
to, certain sound changes and that it is not permissible to look for
universally valid sound changes under like articulatory conditions.
Certain typical mechanical tendencies there are (e.g. nb > mb or
-az> -as or tya> tsa), but a complete theory of sound change has to
take constant account of the orientation of sounds in our sense. Let
one example do for many. We do not in English feel that d is to be
found in the neighborhood, as it were, of s, but that it is very close to
5 In Spanish, 6 is not far from s, but is not at all close to 5.^ Is it
not therefore more than an accident that nowhere in Germanic does 6
become s or proceed from s, while in certain Spanish dialects, as so
frequently elsewhere, 6 passes into s (in Athabaskan 6 often proceeds
from s)? In English 6 tends to be vulgarized to ^ as 5 tends to be
vulgarized to d, never to s; similarly. Old Norse 6 has become t in
Swedish and Danish. Such facts are impressive. They cannot be
explained on simple mechanical principles.

Phonetic patterning helps also to explain why people find it difficult
to pronounce certain foreign sounds which they possess in their own
language. Thus, a Nootka Indian in pronouncing English words with
77 or I invariably substitutes n for each of these sounds. Yet he is
able to pronounce both rj and I. He does not use these sounds in
prose discourse, but rj is very common in the chants and I is often
substituted for n in songs. His feeling for the stylistic character
of 77 and for the n-l equivalence prevents him from "hearing" English
7] and I correctly. Here again we see that a speech sound is not
merely an articulation or an acoustic image, but material for sjon-
bolic expression in an appropriate linguistic context. Very instruc-
tive is our attitude towards the English sounds j, 77, and ts. All

"> B, D,G represent intermediate stops, "tonlose Medien." In this series they
are followed by aspiration.

8 The slight objective differences between English and Spanish 6 and 8 are of
course not great enough to force a different patterning. Such a view would be
putting the cart before the horse.

Three: Descriptive anil llisutrical l.iitf'uisucs 193

three of those sounds are faniiUar to us (e.g. azure, sing, haU). Nt»i.«-
occurs initially. For all that the attempt to pronounce tliem initi-
ally in foreign words is not reacted to in the .same way. »j<i- :irid ha-
are naively felt to be incredible, not so ja-, which in ea.sily actjuinti
without replacement by flja- or sa-. Why is this'.' rja- in incnHliblo
because there is no itiba-, 7ula-, r]{g)(i- series in Knirlish. ttn- in in-
credible because there is no psa-, tsa-, ksa-, .series in HnRJi.Mh; -U is
always morphologically analyzablc into -I -f -■<, hence no feeling
develops for is as a simple phoneme despite the fact that its |)h<)i:
parallel ts (ch of church) is found in all positions.' But j.i- i> ;
difficult, say in learning French, because its articulation and p«'r.- ; -
tion have been mastered by implication in the daily use of our plionetic
pattern. This is obvious from a glance at the formula:

-j- -z- -5- -V-

— 2- 8- V-
which is buttressed by:

.§. .s- ~d- -/-

s- s- d- f-

Is it not evident that the English speaker's pattern has all but lau;'.!il
him j- before he himself has ever used or heard an actual j-?

There are those who are so convinced of the adequacy of purely
objective methods of studying speech sounds that they do not hesitate
to insert phonetic graphs into the body of their descriptive gramm.'^ir-i.
This is to confuse linguistic structure with a particular meth(Kl of
studying linguistic phenomena. If it is justifiable in a grammaticil
work to describe the vocalic system of a language in terms of k\in<)grai>h
records,^" it is also proper to insert anecdotes into the mnrpholoj^v to
show how certain modes or cases happened to come in hantly. And
a painter might as well be allowed to transfer to his ranvjus his unro-
vised palette! The whole aim and spirit of this paper ha.s U^n to
show that phonetic phenomena are not physical [ihenomena /k-
however necessary in the preliminary stages of inductive linjfui
research it may be to get at the phonetic facts by way of their phN-^iicui
embodiment. The present discussion is really a sfx^cial illn
of the necessity of getting behind the .sense data of any typ* ■-. •
pression in order to grasp the intuitively felt and comnjunicjilod foniis
which alone give significance to such expression.

•Obviously we need not expect -ts and -ts t.. .irvolop nnalo([ou«ly even if •
and h do.

1" Needless to say, such recordH arc in place in atudics explicitly d«vo(«d to
experimental phonetics.

1 1)4 General Linguistics I

Editorial Note

Lanmtui^e 1 (1925), 37-51. [Reprinted in: Edward Sapir, Selected Writings in
Liingiiage, Culture, and Personality. Edited by David G. Mandelbaum. Berkeley:
University of California Press, 1949, pp. 33^5; in Readings in Linguistics: The
development of descriptive linguistics in America, 1925-56. Edited by Martin Joos,
New York: American Council of Learned Societies, 1957 [various reprints], pp.
19-25; and in V. Becker Makkai ed.. Phonological Theory: Evolution and Current
Practice, New York: Holt, Rinehart & Winston, 1972, pp. 13-21]

Correction:
p. 40. 1. 1 1. read: phonemes (not: phonems)

Three: Descriptive (ind Historiml I in^uistics 195

11121 PHILOLOGY —Since h)l() there has been a reniarkahle rcMval ot inter-
est in linguistic science. ,\ luimher ol hnizuisis have turned aside frimi their
specialist activities and concerned theniselves with the restatement of funda-
mental principles. Among these works may be mentioned Otto Jespersen.
Language, Its Nature, Development and Origin, and I he I'hilosophv of (irammar.
J. Vendryes, Le Langage, hunxhuiion l.inguistitpie a lllistmrc and I:. Sapir's
Language: an Introduction to the Study of Speech. Hiese vsriters approach the
study of language from notably distinct viewpoints, so that their books supple-
ment rather than parallel each other. Jespersen is mainly concerned with the
more important modern languages of Europe and takes into consideration a
good many matters of practical interest ordinarily neglected b\ the scientific lin-
guist. Vendryes writes from the standpoint of the Indo-L.uropeanisi and stresses
the comparative and historical points of \iew. Sapir. a specialist in American
Indian languages, is chiefly concerned with formal and psychological fundamen-
tals, and uses freely examples taken from the languages of primiii\e peoplev

Psychology. — It is very noteworthy that other social and humanistic sciences
than linguistics itself have taken a serious interest in ihc data of language.
Psychology, which had been disposed to neglect language behaviour, has begun
to analyse it more carefully in terms of stimulus and response; of habit: of adap-
tive function. A notable contribution to the understanding of language as a par-
ticular type of behaviour is J. R. Kantor's paper on An Analysis of Psvchidogical
Language Data, in which the peculiar characteristics of speech, whether commu-
nicative or expressive, are sought in its indirect nature as a response, the 'adaptive
stimulus" being responded to not directly but in the form of a reference, while a
secondary stimulus, generally the person spoken to. is substituli\ely reacted
to. J. B. Watson, the extreme exponent of behaviourism, sees in language merely
a series of highly specialised laryngeal habits, and goes so far as to identify lan-
guage with thinking by interpreting the latter type of behaviour as implicit or
"sub-vocal" laryngeating. Such ultra-beha\iourislic interpretations of language
are not likely to meet with the approval o[ the linguists ihcmseUes. but they may
have a certain value in accustoming us to approach the stud\ of language habits
without necessary reference to the logicians' world of "concepts." Hiere is reason
to believe that the kind of psychology {see Psv(U()ic)(iY) which will prove of the
greatest value to linguistic science is the Gestalt psychologic (configurative psy-
chology), which is still in its elementary stages. In this type o\ thinking the empha-
sis on behaviour is placed on the tola! form or configuration ot a sequence ot acts
viewed as a system, lb apply conligurati\e ps\cholog\ to language one may say
that no linguistic act, however elementary, can be looked upon as a mere res-
ponse, nor can even the simplest speech articul;iiu>n be uiulcrstood m terms ol
muscular and nervous adjustments alone.

Language always implies a particular kind ol selective organisation: no speech
sound, as Sapir has shown in a paper. Sound Patterns m Language, is intelligible
as a habit without reference to the complete system o\ sounds characteristic of a
given language, the individual sound being defined not merely as an articulation,
but as a point in a pattern, with the other points of which it has intuitively felt

1 96 General Linguistics I

relations (see Phonetics). For the more complex levels of linguistic organisation
the pattern point of view is more obviously in place.

Pliilosophw —Even more fruitful for linguistics than psychology has been
the work of certain philosophers. As the relativity of all knowledge and all
experience to the habitual symbolism by means of which they are expressed has
become more and more clearly understood, philosophy has begun to take a very
lively interest in the relation between language and thought, in the nature of the
symbolic process exemplified in, but not exhausted by, language, and in allied
problems of meaning, reference and classification of experience. Philosophers of
standing, such as Cassirer, Delafosse and Ogden and Richards, [1131 have been
occupying themselves with linguistic problems as never before. New viewpoints
have been arrived at which are of capital interest for both philosophy and lin-
guistics. In an important work entitled The Meaning of Meaning Ogden and
Richards have carefully explored the nature of the symbolic and referential pro-
cess involved in the use of language and have classified the concept of "meaning"
itself Many problems that have occupied the attention of philosophers and logi-
cians are shown to be not essential problems but pseudo-problems that arise
from the almost unavoidable temptation to read an absolute validity into lin-
guistic terms that are really devoid of meaning when they are disconnected from
a more or less arbitrarily defined context.

Forms of Speech. — The study of forms of speech that are very different from
those that most of us are accustomed to — say English, French or Latin — disclo-
ses the possibility of markedly distinct analyses of experience where one might
naively suppose that our customary analysis via speech is resident, as it were, in
the nature of things. Categories that are carefully developed in one language are
but weakly developed, or not at all, in another. Even the elements of sensible
experience, whether conceived of as thing (say "tree"), quality (say "blue") or
action (say "give"), are not necessarily taken as equivalent ranges of reference
("concepts") in different languages, but may be included in, or distributed
among, respectively different ranges. It may even happen that what in one lan-
guage is a definite experiential concept, with an unambiguous mode of reference
to it. finds not even a partial expression in another, but is left entirely to the impli-
cations of a given context. Such a language may be said to have no "word" or
other element for the "idea" in question, from which, however, it need not in the
least follow that it is incapable of satisfactorily conveying the total psychic
sequence ("thought") or unit of communication in which the "idea," in another
language, figures as an essential element. Linguistic expressions of this kind are
naturally of the greatest value for our conception of the nature of "reality" and
of our symbolised attitude toward it, and certain philosophers have not been slow
to turn to language for this reason.

That m the process of thought linguistic mechanisms play an important part is
beyond doubt. Material rich in interest for the philosopher, the psychologist, the
philologist and the sociologist is afforded by the modes of classification in use in
various language families, such as classificatory prefixes and generic determina-
tives.

Three: Descriptive atul Hisiorn ul 1 tniiuiMux 197

Sociology andAni/uopology. — Sociologisls and anlhropt)U)gisis icx) have inlcr-
ested themselves in linguistic phenomena as indispensable lo an undcrsianding

of social behaviour. Especial mention may be made of H Mahntmskis s|>ccula-

tions on the linguistic expression o{ prmutive peoples. Ihe value ol lan-'i is a

delimiter and index of social groups is being increasingly recognised : .lo-

gists. One may wonder, indeed, if there is any set of siKJal habils thai is more
cohesive or more disrupting than language habits. Not only has laii ' an

object of study in its own right, but it has shown ilsell mcreasmr the

greatest instrumental value in all historical studies. Hie best kind of ethnological
and folk-loristic material is that which is secured m the form of native lexK for
material of this kind is not open to the charge of misrepresentation of the native
point of view. Accordingly, we find that the text metht)d of studying the cultures
of exotic peoples, whether lettered or not, has been growing in favour, in spiic of
the obvious difficulties of the method, implying as it does a preliminars siudv of
the exotic language itself. Indeed, much of the most competent and aulhonlalivc
ethnological information that we possess has been obtained as material ancillary
to linguistic studies. One may instance, for America, the KwakiutI publications of
Franz Boas and, for Africa, the Shilluk texts of D. Westermann.

Place Names. — In another sense, too, linguistic researches ha\e been of great
assistance, and that is in enabling us to make inferences as to the history and pre-
history of various peoples. The value of place names, e.i^.. in ihe study of the for-
mer distributions of the various ethnic elements that go lo make up the present
population of England, is well understood (see Folki^orh). Much of the prehis-
tory of Europe and Western Asia is being rewritten with the aid of a profounder
study of place names. The non-Hellenic character of hundreds of well-known
Greek place names, such as Athenai, Korinthos and Tiryns, is interesting and
important in connection with our rapidly increasing kni>wledge of the prc-
Hellenic or Minoan civilisation of the Aegean and o\ the mainland of (ireece.
The Etruscan problem, too, has been furthered by a study oi the recorded place
and personal names of Etruria that are clearly not of Italic origin. Iliese show so
many resemblances to names recorded frc^n western Asia Minor (e.g., Lydia.
Lycia) that there is now less hesitation than before lo credit the lestimonv of
Herodotus, who derives thelurrhenoi (I-lrusci) from Asia Minor.

General Tendencies. — We may point out a luiinbci of tendencies in recent lin-
guistic thinking. First of all, there is a growing reali.sation ihat the lite ol l.f

is similar in all parts of the world, regardless of the race or cultural de\cli^, . .

of the speakers of the language, and that the rate of linguistic change is not
seriously dependent on the presence or absence ot unting. Ilie sup|x>sed con-
servative power of a system of writing, it is now generallv avireed. is altogether
mythical. It is an illusion to imagine, for instance, that ( hmesc as a s|>oken lan-
guage was hindered from the normal rate of change bccau.se of an early hlerary
fixation. Not only have the modern dialects a notablv different viK-abular>-
from classical Chinese but their pronunciatiini has been so modified that it is
clear that present-day Mandarin or Cantonese would be quite as unintelligible lo
Confucius as present-day Spanish to Cicert).

198 General Linguistics I

Inwnuil Factors in Change. — The most important single factor making for an
increased or retarded rate of linguistic change would seem to be the formal set
of the language itself. It is significant, for instance, that the Semitic languages
have changed very much less in the last 3,000 years than have the Indo-European
languages, llie Athabaskan (Dene) languages of North America, spoken by
unlettered tribes which had, for the most part, reached but a very primitive level
of culture, have tended to resist morphological change because of a certain for-
Fnal equilibrium, despite the complexity of their grammatical structure, hence
such widely separated languages as Navaho (New Mexico and Arizona), Hupa
(Northern California) and Chipewyan (Mackenzie Valley) differ probably less
than French and Italian {see Archaeology: Central America).

Consistency of Change. — One of the most impressive things about linguistic
change within a given genetic group is the relative consistency of its direction
over a long period of time. Well-known examples of this principle are the pro-
gressive simplification of the case system in all branches of the Indo-European
family and the ever-increasing tendency to isolation in the structure of the Sinitic
(Indo-Chinese) languages. It is very remarkable, too, that in many cases it can
be shown that related languages have undergone similar developments inde-
pendently of each other and at very different periods. "Umlaut," for instance,
seems to have developed independently in West Germanic and in Scandinavian,
just as certain tonal developments in modern Tibetan (central dialects) are signi-
ficantly parallel to the far older tonal developments of the earliest Chinese. It
is difficult to explain these paralleHsms except on the assumption that a given
formal set implies a certain liability to modification in one rather than in another
direction.

External Contacts. — Important, yet less important than the inner "drift" of a
language, is the tendency to change as a result of external contact. Not only lex-
ical, but also far-reaching phonetic influences may be ascribed to the contact of
unrelated languages. Although Annamite seems to be basically a Mon-Khmer
language, therefore originally toneless, it has acquired a complicated tone mecha-
nism which is, in principle, identical with that of the neighbouring Tai languages
(Siamese, Shan, Laos, Tho). Morphological features, too, may be freely diffused,
though the evidence for this is less convincing in most cases than for the spread
of words and of phonetic elements and tendencies. Often it is difficult to say
whether a morphological [114] parallelism is due to historic contact or to genetic
relationship or to independent development.

The effects of bilingualism deserve attention, especially in those areas where
permanent contact is established between two different families of speech, as in
India, where Indo-Aryan languages are spoken in addition to Dravidian or
Austnc or Indo-Chinese tongues. The principles of the growth of linguae francae
have also been examined, but much yet remains to be done. Internal social devel-
opments as well as exterior relations induce linguistic features of interest and of
philological value in themselves. The phenomena recorded by anthropologists of
secret speech, of slang, of special modes of speech associated with social divi-
sions, women, age-grades, priests and kings bear witness to the importance and

Three: Descriptive ami fli\i,>ru nl I infiuistuw 199

validity ot linguisiics in ihc sukK oI social auurciiatcs and their a.rpt.i.iit- fii.im
testations.

Morphology. — llic niorpliolouical classilKation t)t lanr ich

to be desired. It is abundanlK clear that the tr.idilional cl ...,,..,, ... ..j-v^ol

speech into isolating, agglutinative, inllective and polysynlhetic is impcrlccl It t%
difficult to find easily applicable criteria of general linguistic structure. Perhaps
the safest plan is to adopt a number of distinct points of vic\\ and i " S a
given language from each of these. .Sapir suggests ihc possibility of i; . ...de-
pendent types of classification, based respectivly on degree of synlhcsiv on ihe
predominant "technique"' (manner and degree of wekhng of elements into uni-
ties), and on the conceptual expressions. From the first pt)int of vieu larv-

may be described as analytic (Chinese, Ewe), weakly synthetic ( I .nglish. 1 1.
synthetic (Latin, Turkish, Japanese) and polysynthetic (Hskimo. Algonquin). Ihc
classification based on "technique" is '\n\o isolating (Chinese. Annamite), .u'ulu-
tinali\e (Turkish. Nootka, Bantu), fusional (Latin. Irench. Yokuts) and "ssmhol-
ic". i.e.. with the characteristic employment of various types of internal change.
such as vocalic and consonantal change, reduplication and differences of stress
and pitch (Semitic. Shilluk). A conceptual classification would distinguish be-
tween languages which can and those which cannot freely derive words from basic
elements and, further, between those which express the fundamental relational
(syntactic) concepts as such and those which need adventitious concepts like gen-
der or number to bring out these necessary relational ideas. Iliere would, there-
fore, be theoretically four conceptual types of languages — pure-relational, non-
deriving languages (Chinese), pure-relational deriving languages (Pt)lynesian.
Haida), mixed-relational non-deriving languages (but meagrelv represented) and
mixed-relational deriving languages (Latin. Semitic. Algonkin).

Phonetics. — Much progress has been made in phonetic research {see Pho-
netics). A vast number of new sounds have been discovered and whole classes of
articulation come to light from time to time whose existence could not readilv
have been foretold by an a priori phonetic analysis. Ilie help derived b\ linguists
from objective methods of investigation (various types of recording apparatus)
has been welcome but. none the less, disappointing on the whole. A well-lraincd
ear can readily make and classify sound diflcrences which a k\nu>graph diK-s not
materially help us to understand. As our know ledge oi phonetics grows, wc real-
ise that sounds and phonetic discriminations originally believed to have a quite
restricted distribution are really rather widely distributed. Ihe Hi>llenlol and
Bushman clicks, partly borrowed by certain Bantu languages, seem to be confined
to South Africa. But the curious glottalised consonants and voicelevs laterals of
Western American languages are found as well in the Caucasus and in a number
of African languages. Again, lone differences or signilicanl elements in the v^ord
are by no means confined to Chinese and related languages in Eastern Asia. Wc
now know that pitch languages are exceedingly common. Most of c m

languages that are not Semitic or Hamitic are pitch languages (Sudan. - . ...lu.
Hottentot. Bushman), while in aborigm.il America a very considerable number
of languages have been IouikI lo recognise tt)nal differences (fc|f.. Athabaskan.

200 General Linguistics I

Tlinyil, Achomawi, Takelma, Tewa, Mohave, Mixteczapotec, Chinantec, Otomi
and many others.)

Genetic Relations. — For a long time Hnguists hesitated to look for genetic
connections between the groups and isolated languages that had been establish-
ed. Iliere is an increasing tendency now to make larger syntheses and to suggest
as at least probable, if not entirely demonstrable, relations that at first blush seem
farfetched. Among serious linguists Alfredo Trombetti is perhaps the only one
who has ventured to commit himself to the theory of linguistic monogenesis and
has actually attempted {see his Elementi di Glottologia) to show in what manner
the various groups of languages that are generally recognised are related to each
other. Tlie most important general survey of languages published in recent years,
Les Langues dii Monde, edited by A. Meillet and Marcel Cohen, is very conser-
vative in the matter of genetic theories, but even in this a number of syntheses
are allowed that would not have passed muster a few decades ago. There is no
doubt that as our comparative knowledge becomes more profound we shall be
enabled to extend our genetic groups with safety.

North and South American Languages. — We can only glance at some of the
new genetic theories. The incredibly complicated linguistic picture of North and
South America, which has long been proverbial among linguists is Hkely to become
very appreciably simplified. R. B. Dixon, A. L. Kroeber, P. Radin, Sapir, <J.> R.
Swanton and P. Rivet are among those who have sought to bring order out of the
linguistic chaos that still largely prevails in America. For the groups of languages
spoken north of Mexico and in part in Mexico Sapir suggests a greatly simplified
classification into six genetic groups, Eskimo, Nadene (Haida, Tlingit-
Athabaskan), Hokan-Siouan, Algonquin-Wakashan, Penutian and Aztek-
Tanoan. These groups, aside from Eskimo, embrace languages, however, which
differ vastly more than the Indo-European languages. In South America Rivet
has connected various groups of languages hitherto beheved to be unrelated. He
has also attempted to prove a relationship between certain Fuegian languages
and the languages of Australia, and between the Hokan languages of California
and Polynesian.

The Austric and Australian Languages. — More likely to prove sound than
these latter theories is Father W. Schmidt's hnguistic synthesis which passes under
the name of "Austric" and which includes the vast group of languages — Mon-
Khmer (Mon-Khmer proper; Khasi; Proto-Malaccan, Nicobarese; Munda) and
Malayo-Polynesian.

Analysis of the Australian languages by Father Schmidt seems to disclose the
existence of a definite distinction between the northern and southern groups. The
northern group is resolved into two main divisions with a third division of an
intermediate nature. There are marked differences in the southern group, some
of which are regarded as approaching Tasmanian (now extinct) and as therefore
markedly primitive. To a very large extent hnguistic differences in this area,
claims Father Schmidt, can be correlated with variations of, or differences in,
other cultural features such as social organisation.'

• Anthropos, vol. 8 (1912), p. 260 and p. 463; vol. 9 (1913), p. 526.

Three: l^cscripdvc ,in,l lli\u>ru id I inf(Uislics 201

African Lungiuiges. —In Aliica wc iiias iiolc as im|X)rlant ihc discovcn' oC
a language spoken in the casl-ccnlral portion of ihc a)nimcnt: Sandawc,
which is unniislakahly related to Hottentot, spoken far m the south Hiis lend*
an unexpected colour to theories lonu jircvalent as to the Mamilie .ifdhalmn c»f
Hottentot. The sharp hue ol deniarcalion which used to he dra\sn Jvlx^een Ihc
Bantu and Sudanese languages is giving way. liverylhing points to ihc fact that
the Sudanese huiguages are a worn-down form of a hinguage. oi ' ' in-

guages, which was genetically related to Bantu. A nuinhcr ol Sud.u :^c»

possess class prefixes (or suffixes) that are strt)nuly reminiscent of the Bantu
prefixes, while bare survivals ol these elements persist m a great many other
Sudanese languages.

The publication of Sir H. H. .Johnstons woik on the Bantu languages was an
event of first class scientific importance. It summarises and surveys with master*
ly power the sweep of this family of languages throughout Africa and arLUies that
"the Bantu family was finally moulded by some non-negro inet)mers ot |II5|
possibly Hamitic affinities, akin at any rate in ph\sique and culture, if not m lan-
guage, to the dynastic Egyptians, the (lallas and perhaps most of all li> those
'Ethiopians' of mixed Egyptian and Negro-Nubian stock that, down to I.CJOO
years ago, inhabited the Nile basin south of Wadi-Halfa and north of Kordofan."

Indo-European, Semitic and Uaniitic Languages. — Hermann Moller's very
systematic and detailed attempt to connect Indo-European and Semitic (the
relationship of which to Hamitic is now generally recognised) seems not to have
been cordially received by either the Indo-Europeanists or the Semiiists. hut
to the general linguist who studies his works his demonstration seems highly
suggestive, not to say convincing. Less solid seems to be the attempt of N. Marr.
a Russian linguist, to establish a "<J>aphetic'" group of languages, consisting of
the Caucasic languages, Basque, and that large number c>{ still \ery imperfectly
known languages which preceded the Indo-fiuropean group in southern Europe
and western Asia (Etruscan. Minoan. "Asianic" [early Hittite)).

More and more it is becoming evident that the linguistic cartograph> of the
Near East and of Europe was a complicated one in remote times. ITie discovery
and partial decipherment of a series of cuneiform •Hittite" inscriptions from the
second millennium B.C. proves a number oi interesting things that a language
closely related to Sanskrit and known from quoted terms was spoken in the
neighbourhood of the Hittite country; that the language o\ the Hittite rulers had
unmistakable Indo-Iuiropean features, but was not typicalK. or even mainly.
Indo-European in character; and thai an okler llillite language was quite unre-
lated to this. Just what relalionship. if any. the non Indo-European elements of
these Hittite languages bear to 'Asianie" (l.ycian. I ydian. Cariar
and the Caucasic, remains to be discoNereil ( see Ak( iiai tmn.^ : \\

Discovery of Tokharian. —Perhaps the most interesting linguistic dwcovcry
that has been made of recent years in the domain of Indol urojvan phs
apart from the "pseudo-Hittite" documents just referred to. is the presc;..-
two dialects of a language, generally termed lokharian m Chinese Turkistan. as
late as the 7lh centurv ad lokharian is quite distinct from any other known

202 General Linguistics I

Indo-European language and has thrown Hght on a number of points of Indo-
European grammar. In certain phonetic respects it agrees more closely with
Greek and Latin than with the Indo-lranian languages that were geographically
nearer to it.

Pisueha Languages. — Problems of an interesting nature are raised by the evi-
dence published (1919) in vol. 8, part 2, of the Linguistic Survey of India by Sir
George Grierson, who describes the Dardic or Pisacha languages as not posses-
sing all the characteristics either of Indo-Aryan or of Eranian. They exhibit
almost unaltered and in common use words which in India are hardly found
except in Vedic Sanskrit. The wild, mountainous country in which these lan-
guages are found has not attracted the conqueror. An interesting feature is the
survival of words from Burushaski, a form of speech which has not yet been
satisfactorily related to any other language group.

Karen Languages. — Examination of the languages spoken in Burma, an
area not covered by the Linguistic Survey, has justified the view that the Karen
group of languages constitute a new family of languages which exhibits features
resembling those of the Chin and Sak languages, even of some of the sub-
Himalayan dialects. Sir George Grierson suggests the possibihty of a widespread
pre-Tibeto-Burman population, which was absorbed, with parts of its language,
by the later Tibeto-Burman immigrants.

Other Families. — The Man family is similarly regarded as distinct but it is a
newcomer from Southern China, whence further evidence may be available to
identify its main relationship. There are thus in the confines of the Indian Empire
language stocks of world-wide distribution and languages — the Dravidian ton-
gues — Burushaski, Karen, Man and Andamanese, which survive in isolation. It
has been surmised that Dravidian languages may be related to Sumerian or to
Basque, or to a common prototype, but conclusive evidence has not yet been put
forward in proof of these hypotheses.

Bibliography. — Bureau of American Ethnology, Bulletin 40, pt. 1 (1912); and
pt. 2 (1922); Sir H. H. Johnston, A comparative study of the Bantu and semi-Bantu
Languages (1919); J. Vendryes, Le Langage (1921); E. Sapir, Language (1921); C.
K. Ogden and I. A. Richards, The Meaning of Meaning (1923); O. Jespersen,
Language, its Nature, Development, and Origin (1922); Sir G. A. Grierson,
Lingidstic Survey of India, vol. 8-11 (1919-22); India Census Report (1921).

Editorial Note

Encyclopaedia Britannica, 13th ed., Supplementary Volumes, vol. 3 (1926),
112-115. [Reset after the originally published version; editorial changes are
indicated with <>]

Mankind, Nation and Individutil from a I infiuisiu Point of View. BY OTTO
JHSPl-RSFN. Ciimhriclgc. Massacluisctts: ll.ir\.irci University Press (H.

Aschchoug c^ Co.), 192S. Pp. 221

1 498 1 Ihc eleven ehapleis ol this very readable and commcndaNy untcchnical
book were origiiiallN delivered as a series i)! leelures for ihc Norv^ciiian Instilulc
for Comparali\e Researeh in Human C'ullure. ilic volume forms ihc li»urth in
Series A of the publications of this Institute, its title is pcrha|-)> ltK> ambitious for
what is included between its covers, w hich consists of a discussion of ihc concepts
"speech" and ■"language." remarks on the actual influence on l.r
indi\idual as such, two chapters on the relation between dialect
language," an account o{ what constitutes good usage in speech, examples of soci-
ally determined linguistic diflerences. and some interesting material on slang and
other vagaries or eccentricities of language. In its concluding chapter. Jes|x:rscn
stresses what is universally valid in human speech, cutting across all the bewilder-
ing variety of phonetic and morphological expression in the languages of the
world. He casts a prophetic glance at the (or rather, an) international auxiliary
language — Jespersen's interest in Ido. an offshoot of I^speranto. is well kni>ssn —
but does not enlarge upon this somewhat contentious subject.

There is little that is new in the book, nor can the presentation be said to be
characterized by any noteworthy originality of point of view. But it is all worth
while, and it is all pervaded by Jespersen's common sense and good practical
judgment. One may make some demur, however, to his unsympathetic dismissal
of the distinction that certain linguists, like De Saussure and Harold I-. Palmer.
have made of "speech" [499] and "language." According to these, "speech" is the
totality of articulatory and perceptual phenomena that take place when given
individuals indulge in language behavior at a given lime and place. "Language."
on the other hand, is society's abstracted pattern-whole of such behavior, all
purely individual variations being dismissed as irrelevant. Jespersen's criticisms
of this useful distinction are obvious but unsound, it seems ti> the reviewer. A cer-
tain class of phenomena cannot be shown to be illusory, as Jespersen apfXMrs to
think, merely because it is unthinkable in terms of actual experience except as a
mode of abstraction of another, more empirically ascertained, class. If carried lo
their logical conclusion, Jespersen's strictures would demolish the study of all cul-
tural patterns and condemn the social scientist to the intermmable lisiini! of indi-
vidual events. Needless lo say, Jespersen merely overstates the consequences of a
characteristically "extraverted" spirit of linguistic research. His own exccllcnl
work in the history of the English language shows that he mstmcliNeK and wise-
ly recognizes a distinctit>n that he is theorelicalb at a l'<^- '" \ 'Iid-.r.

UnIVLRSII^ Ol ("IIK A(i() ' ^ '^^

Editorial Note
American Journal of Socudo^y 32 ( 1^2(0. 4^AS -4v'v'.

LANGUAGE AS A FORM OF HUMAN BEHAVIOR*

EDWARD SAPIR

So long as the facts of speech are discussed only by students
professionally interested in language its peculiar characteristics
are very likely to be overlooked. It is not surprising, therefore, that
philologists and teachers of particular languages are apt to have
vaguer and cruder notions of the fundamental characteristics of
human speech than such unspecialized students of human behavior
as concern themselves with this most human of all reactive sys-
tems. It is clear that if v^e are ever to relate the fundamental prob-
lems of linguistic science to those of human behavior in general —
m other words, to psychology — we must learn to see language as
possessed of certain essential characteristics apart from those of
particular languages that we may happen to be familiar with and
as rooted in some general soil of behavior that gives birth to other
than strictly linguistic forms of expression.

If we take a bird's-eye view of the languages of the world we
find that there are certain things that characterize them as a whole
and that tend to mark them off somewhat from other forms of cul-
tural behavior. In the first place we are struck by the marvelous
completeness of formal development of each and every language
that we have knowledge of. Popular opinion to the contrary, there
is no known language, whether spoken by a culturally advanced

* An address delivered before the joint session of the National Council of Teach-
ers of English and the English Language Congress at Philadelphia, November, 1936.

Three: Descriptive und llisii>rual I in\^iusucs 205

422

group or by one of the more backward peoples of the world, whic h
has not its perfectly definite phonetic system, its clearly develofxxi
and often very complex system of forms, and its established rules
and idioms of usage. In other words, there Ls no such thing as a
language that has not its well-defined grammar, though it goes with-
out saying that this grammatical system is never consciously known
or capable of communication by the untutored folk. We may fur-
ther say of all normal forms of human speech that they have the
appearance of systems of behavior that are rather definitely marked
off from all other organizations within the general range of conduct .
When we deal with such flexible categories of socialized behavior af
religion or art or government or education, it is difficult to draw
convincing lines. A given reaction may be placed now here, now
there, but there seems never to be any reasonable doubt as to the
formal identity of a speech act. This is all the more remarkable as
speech is not in actual life handled as a self-satisfying reactive sys-
tem. It comes into use and consciousness merely as a general lever
for all types of expression. There is no pure linguistic conduct.
There is only conduct in which the need is felt for communication
or expression and in which this need is solved with the help of an
elaborate hnguistic apparatus, always ready to hand.

We are learning to attribute more and more importance to the
part played by the unconscious mind in the development of our
habits. We realize more clearly than the psychologists and philoso-
phers of the past have realized that the consciously controlled typi-
of behavior is always a secondary type, and that all intellectual
formulations are little more than a conscious control, in set tcrnv.,
of knowledge that is already possessed in an unconscious or intui-
tive form. It is characteristic of the more important types of so-
cialized conduct that they tend to draw the attention of individuals
and societies to themselves and thus become an object of conscious
formulation — in other words, of scientific study. It does not follow,
however, that a scientific statement of facts, say of religion or other
aspects of social conduct, is more accurate or more delicately
nuanced than the intuitive knowledge and control which preceded
it. One may have, and often does have, an astonishing fineness of
intuitive control without being in the least possessed of an rvrn

-)Q^ General Linguistics I

423

elementary knowledge of the reactive system as such. Hence it is
not surprising that statements made by sophisticated, no less than
by more naive, people about their own forms of speech and about
the nature of their linguistic responses are often -wide of the mark-
It is well known that Plato indulged in extraordinarily childish
speculations on the nature of the Greek language.

We must, therefore, carefully refrain from making the common
mistake of confusing the intuitively felt complexities of language
with an intellectually controlled statement of them. If we do not
learn to make this important distinction we shut ourselves of! from
the possibility of understanding the marvelous formal intricacy of
so many of the languages of primitive man, and — ^what is of more
general interest — of understanding how the child is able to ac-
quire as firm and accurate and extended a control of linguistic hab-
its as he does. Lx)ng before the child knows the difference between
a singular and a plural, long before he has any notion that there is
such a thing as tense, he has learned to use the categories of number
and tense with perfect freedom and comfort. With the development
of civilization, language has tended more and more to rise into con-
sciousness as a subject of speculation and study; but it still remains
characteristic of language as a whole that of all the great systems
of social patterning it is probably the one which is most definiiely
unconscious in its operation. But the unconsciousness of the speech
processes is not to be interpreted in any vague or mystic sense. All
iliai we mean is that our intuitive sense of the relations bct\vecn
form? of expression is keen enough to enable us to acquire a full
coiitrol of speech habits without the necessity of an intellectual
overhauling. The clarity and certainty of this intuitive process is
best evidenced by the fact that when natives who have never
learned the grammar of their o^-n language come in contact with a
held investigator who endeavors to discover this grammar by in-
fluclive methods of inquiry, the natives often develop — and in a
•^lifTL time — a ver>' real comprehension of what the inquirer is
aficr. They can often help him with the statement of e.xphcit rules
and with observations that could never have occurred to them in
the normal course of their daily Hfe.

The next and most obvious of the general features of speech is

Three: Descriptive ami Ihsioneut I in^'utsucs Jlj,

424

its indirect or symbolic character. No normal sp<'och utterance re-
ally means what it biologically seenis to mean. In other words if
I move the lips or tongue or glottal cords in this or that fashion it
is not because these movements are of any direct use in my adjust-
ment to the environment. It is because by means of these move-
ments I am able to eflect articulations that are perceived as arbi-
trary sequences of sounds conveying more or less definite notions
which stand in no intelligible relation to the articulations them-
selves. If I move my tongue in order to lick a piece of candy I
carry out a movement which has direct significance for the desired
end, but if I put my tongue in the position needed to articulate the
sound "1" or a given vowel, the act is in no way useful except in so
far as society has tacitly decided that these movements are capable
of symbolic interpretation. It is highly important to realize that
symbolic systems, which are systems of indirect function, are nut
in theory confined to speech. One may have a pantomimic symbolic
system or any other kind of symbolic system developed by some
part of the organism, but it remains strikingly true that no otin-r
human type of symbolic behavior compares for a moment in com-
pleteness or antiquity or universality with the symbolism known
as speech. It is further important to observe that there seem to be
no appreciable differences in the languages of the world in regard
to their relatively symbolic character. The sound sequences that
are used as symbols of reference by the Hottentot or the Eskimo
are to all intents and purposes as arbitrary as those used by a \\v d
em Englishman or Frenchman or German.

This brings us to what is, in some respects, the most suq)ri^ing
fact about language: that, universal as it is, it is at the same time
the most variable of all human institutions in the actual det:iil of
its overt expression. All types of reUgious behef , all systems of dec-
orative art, all kinship systems, all methods of organizing society
have certain unmistakable things in common so far as their actual
content is concerned; but when it comes to speech, which is more
deeply rooted in human society than any one of these systems of
activity, we find that its actual content in terms of words, linguistic
forms, and methods of articulation is almost infinitely variable.
We must conclude from this that what is fundamental! v character-

90^ General Linguistics I

425

istic of speech at any given time and place is not its overt form,
which can only be understood as the resultant of a very complicated
series of historical circumstances, but its ground plan. Languages
do not profoundly matter. It is the habit of language as such that
is of such tremendous importance for humanity. The best proof of
this is the ease with which human beings learn to pass from one
language to another and the ease with which they transfer from one
symbolic medium to ancJther — from oral speech to writing, from
writing to the telegraph code, and so on indefinitely. It comes to
this: that language cannot be adequately defined as a set of phy-
sical habits, but must be understood rather as an arrangement, for
purposes of communication and expression, of all the elements of
experience in accordance with a complex system of intuitively felt
relations that can choose any perceptible tokens or symbols they
have a mind to.

So far we have been assuming that the task of language is a
purely denotive one; that when we say "horse," for instance, our
sole purpose is to convey a reference to a class of well-known ani-
mals. If we look more closely at what actually happens in human
speech, however, we are impressed by the fact that the denotive
function of speech is always compounded with certain expressive
factors which we are in the habit of leaving out of account in our
formal designations of hnguistic processes, but which are alwaj'^s
])resent in the actual hfe of language. It is imp)ossible to pro-
nounce even so indifferent a word as "horse" without a lesser or
greater show of interest, without some change of emotion. This
ex]->re5siveness may relate to our personal attitude toward horses in
general or to a particular horse that has been called up as an image
in our minds. Or very frequently, too, the expressiveness of artic-
ulation may have nothing to do with the animal but may relate to
our attitude toward the person that we are speaking to or thinking
of, or to our own general state of mind. In extreme cases our pro-
nunciation of the word "horse" may be infinitely more significant
as revealing a distracted state of mind or an intense interest in the
I'crson spoken to than as a pure denotive symbol. It becomes clear
that in the course of our speech activities we are really doing two
ratlier distinct things, though these are never to be completely sun-

Three: Descriptive and llistoriml I ini^uistics li)*^)

426

dered except by a process of abstraction. On the one hand, we are
using words as algebraic symbols for classes of ideas to which we
refer certain individual references that we wish to make. On the
other hand, we are giving some hint of our attitude toward these
ideas or toward some of the things or persons present in the back-
ground of the speech act. It is because we are alert to the endless
conflict between the denotive and the expressive aspects of speech
that we can safely interpret many words or phrases or statements
in a sense that is completely at variance with their supposed mean-
ing. If a man slaps his friend on the back with the remark that he
is a "rascally dog," he knows very well that his words will not be
understood to mean what they seem to mean. Again, measured
words of praise may be so uttered as to amount to an insult. All
these remarks are commonplaces, but their importance is perhaps
not sufficiently understood by students of language. We have been
too successfully schooled by the grammarians and the systematiz-
ers of speech phenomena to have much charity for these apparent U-
dubious or secondary uses of speech. But what if it turned out, in
the upshot, that these usages are not as secondary as they seem.
that we have reason to believe that in the remote past, when lan-
guage was beginning to emerge as a symbolic system, its use was
even more expressive than at present? What if language were
merely conventionalized gesture — using the word "gesture" in its
widest possible sense as indicating any form of expression?

A parallel from an entirely imconnected type of human be-
havior may be useful at this point. If I am very much angered by
something that one has said or done, I may so forget myself as to
lunge forward with the obvious intention of hitting him. My reac-
tion would be directly expressive in a functional sense. Now, I may
inhibit the reaction without entirely destroying its form. There
may be enough of the gesture left to make it obvious to the by-
standers that I had had the unpulse to strike. This abbreviated or
rudimentary gesture will then be felt as in some sort a symbol of
my attitude. But it is an individual symbol which I have created,
at it were, on the spur of the moment, and which is not to be inter-
preted in accordance with an artificial social code. But suppose.
further, that this gesture becomes accepted by society at large as a

f]o General Linguistics I

427

pantomimic symbol of the striking act, then we have what amounts
to all intents and purposes to an unemotional, denotive symbol of
the notion of striking. It would then be possible for human beings
to use such a gesture for purposes of reference only. They would
not necessarily feel the impulse to strike when they used the pan-
tomimic gesture for it. Now it is, of course, perfectly obvious that
rudimentary symbolisms of this sort are constantly being created
by individuals and that they tend to become more or less charac-
teristic of our varying social groups, but they have never succeeded
in elevating themselves to the status of universal symbolisms of ref-
erence. W^at if speech were, at last analysis, a highly convention-
alized system of just such expressive gestures which by constant
use had lost their original expressive content, having been rubbed
down to the status of purely referential symbols?

In order clearly to understand the point of view that we are
trying to develop it is necessary to take a glance at human respon-
siveness in general. We shall then be able to put speech in its
proper perspective, to see it, not as the extraordinarily isolated and
peculiar thing that it at first seems to be, but as a highly evolved
product of a type of activity that is far more general in scope. In
the case of the arrested gesture of striking that we discussed before
there was an obvious resemblance between the activity that result-
ed from the impulse to strike and the eventually pantomimic sym-
bol built up out of it. But a more inclusive observation soon dis-
closes the fact that symbols need not arise in quite so simple a
manner.

We are in the habit of looking at all forms of human activity
from a functional point of view. A man moves his jaws when en-
gaged in the act of eating in order to prepare the food for swallow-
ing. He moves his limbs in order to walk. He strains his eyes in
order to see more clearly. He lifts his voice in order to reach the
ears of a distant person. And so on indefinitely. But these obvi-
ously functional acts by no means constitute the sum total of hu-
man behavior; it becomes necessary to inquire if a large portion of
this behavior may not be interpreted in other than strictly func-
tional terms. We may return, for purposes of illustration, to the
example that we gave before. It is perfectly true that the act of

Three: Descriptive and Hi^: ■'■.. ; / in^:u:\::L ^ .11

428

lunging forward and of beginning to use the hands is the central
fact in the response of the angr>' individual, and is, for that reason
likely to be the only one that engages our attention. But it is true
to say that the whole body piarticipates in the response in some
form or other. The eyes, the brows, the toes, and any number of
other organs or systems of organs are in%'olved in the angry re-
sponse and can in a sense be said to help along the central response
by a kind of s>inpathetic imitation. In other words, we have reason
to believe that the particular p)art of the body that carries out a
given act is alwav'S accompanied by other segments of the organ-
ism, which lend their consent, as it were, in s>-mbolic form. It is
essentially artificial, for instance, to think of the expression of won-
der as bound up with the staring e>'es alone. There will be some-
thing about the tension of the hands and about the whole set of the
organism that is in some manner corroborative of what the e>''es
disclose. If, for one reason or another, the central expressive organ
or s>'5tem of organs is inhibited, the energy of the impulse may
concentrate in some other part of the organism, so that the sec-
ondary expressive svTnbol may become th^ symbol par excellence
The more highly evolved we become as sociahzed individuals whose
business it is to inhibit many of our expressive impulses, the more
likely it is that these impulses £nd lodgment in parts of the organ-
ism that were not primarily designed, so to speak, for such expres-
sions. The facts that I allude to are familiar enough, though they
are probably rarely seen in their true generahty. If I cannot box
a man on the ear, I can at least double my fist behind my back or
quietly stick out my tongue at him when he is not looking. These
are merely overemphasized s>'mbols of a t>-pe which are present in
embryo in the natural consummation of the striking gesture. The
tongue darts forward as a functionally ineffective but symbolically
relieving substitute for the striking act itself.

Let us now return to speech. There is no reason to bcheve that
speech articulations are different from any other expressions of the
human body. At moments of intense excitement, when the current
patterns of society fall away from us, the articula.tory apparatus is
very likely to regress into its primitive condition, and we produce
all kinds of involuntary sounds that are highly expressive of our

212 General Linguistics I

429

emotions or impulses. In ordinary life these emotions and impulses
merely color speech, which has come to be an essentially denotive
system. But we may surmise that in the remote past speech was
neither more nor less than a series of auditory gestures, directly
symboliziDg various types of adjustment to the environment. These
auditory gestures would have arisen, then, as substitutive symbols
for activities or other expressions which it was not possible or con-
venient to effect at the time. Thus, one's natural impulse in describ-
ing a large object is to move the arms in a more or less circular
fashion suggesting the notion of bigness. According to the theory
that we are now developing, other parts of the organism would be
expressing the notion of bigness in other and more indirect ways.
The speech apparatus would help along in the expressive symbol-
ism by parting the lips, dropping the tongue as low as possible,
rounding the lips, sinking the pitch of the voice, and in other ways.
If anyone is inclined to doubt the reality of such involuntary
symbolisms in speech, let him try the following experiment, which
I have myself tried a number of times with practically 100 per cent
success. Let him tell a number of people, or a class, that there are
three imaginary words: "la" (rhyming with "pa"), "law," and
*^ee," all meaning "table," but with a connotation of difference of
size. Let them then tell which of these three hypothetical words
indicates the big table, which the little table, and which the mid-
dle-sized table. I think it will be found that the normal English-
speaking person, or French-speaking person, for that matter, will
think of "lee" as symbolizing the small table, "law" the big table,
and "la" the middle-sized table or table par excellence. This simply
means that even at this late day we have not lost the feeling for the
gesture significance of sounds and combinations of sounds. If we
examine these three imaginary words more closely from the pho-
netic point of view, we find that the responses are well justified.
The pronunciation of the vowel "ee" of "lee" is such as to demand
a shortening of the distance between the tongue and the roof of the
mouth, as contrasted with the pronunciation of the "a" of "la," in
which the tongue articulates lower and farther back. Similarly, in
"law" the tongue drops stiU more, the lower jaw also drops a little,
and there is a slight tendency toward lip rounding. These phonetic

Three: Descriptive uml Hisiorunl I tnt^ui.sitcs 213

430

modifications amount to gestures, the "ee" gesture being in effect
an expressive symbol for smallness, the "aw" gesture a similarly
expressive symbol for bigness.

All that is claimed here is that if we see sound articulations as
highly variable and intuitively intelligible symbolic gestures, we
have all we need to explain the phenomenon we call speech. The
precise how and when and where of the historical development of
speech is of course another matter. We must assume that speech
expressiveness was highly variable, in an individual sense, to begin
with; that only very slowly and painfully did groups of individuals
come to look upon certain of these symbols as possessing a fixed
symbohc significance; that once an expressive symbolism had been
fixed by social habit, it could lose its expressive content and take
on a denotive one, as in the case of the pantomimic gesture we dis-
cussed before; and that, finally, owing to the changes, unconscious
and ceaseless, which set in to blur the original outlines of any social
pattern of conduct, the actual sounds used for any symbol of ref-
erence would in course of time depart so widely from their original
form as to obscure the whole mechanism of gesture symbolism
which gave rise to the speech process in the first place. It should be
carefully noted that this theory of the nature and development of
speech is only superficially similar to the older interjectional and
onomatopoetic theories of speech which used to be current in lin
guistic circles. If our view of the nature of speech is correct, it
follows that the denotive function of speech, which we now con-
ceive to be its primary function, is in actual fact a secondary one.
just as it is a secondary fact that one shakes one's fist at an enemy.
not in order to initiate the act of striking, but to indicate in a {pic-
turesquely symbolic manner that one has no use for him and might
punch him soundly under appropriate circumstances. It may b«"
suggested at this point that students of speech psychology could
hardly do better than work out experiments intended to test to
what extent imaginary speech forms can be constructed that arc
capable of intuitive interpretation in a primary symbolic sense.
We see that language, in short, has embodied two distinct strata
of symbohc expression, a primary gesture syTnbolism and a sec-
ondary referential one which has largely swamped the former. ^^ e

-, J ^ General Linguistics I

431

caji, therefore, see why it is that all normal speech involves a con-
flict between, or intertwining of, two distinct strata of expression
or types of mechanism, one of which has to do with direct expres-
sion, the other with indirect or denotive expression.

It is legitimate to ask why language, a system of articulated
and perceived sounds, is the one kind of symbolic system that all
human beings have developed in the course of cultural evolution.
I believe that the answer is not difficult to find. Most of the availa-
ble parts of the human organism are in constant demand for di-
rectly functional purposes. This was even more true in the earliest
stages of human development than today. The muscles of the
hands and feet were too busily engaged in grasping, striking, walk-
ing, running, climbing, and other directly useful activities to make
it possible for a successful symbolism to develop with their aid
alone. The case was different with the so-called "organs of speech."
The primary purpose of these organs, which include the larynx
(more particularly the glottal cords), the nose, the palate, the
tongue, the teeth, and the hps, is respiration, smelling, and the
handling of food, of which the first and the last are by far the most
important. Respiration is so nearly automatic a type of behavior
that any symbolisms that might be secondarily worked out with the
help of the larynx, the nose, and the mouth would interfere with it
to only a slight extent. The chewing and swallowing of food is far
less automatic, but occurs so seldom in actual practice that the or-
gans needed for these important acts are most of the time left free
for secondary activities. In other words, they are in a strategic
position for the development of just such secondary symbolisms
as we have shown Vv'ere vrithin the reach of all other parts of the
organism as well. If tliis is true, speech is what it is, not be-
cause there is a mystic connection between articulated sounds and
thought, or the process of symbolization, but merely because the
general tendency of the human body to develop symbolic modes of
expression over and above the directly functional ones found fa-
vorable soil in tlie tract known as the "organs of speech.'' An indi-
vidual who does not need to earn his living by the constant use of
his fingers for primary purposes may develop great virtuosity as a
pianist. In precisel}' the same way the organs of speech developed

Three: Dcscripiivc mid llistoriKil I ini^uisiics 215

432

special social virtuosity as symbol execuLints because they were
idle enough of the time to make such a luxury px)ssiblc.

The two aspects of speech, expressive and referential, are rare-
ly seen in their purity. In the workaday world they are constantly
intertwining their functions in countless compromises, and it is this
highly variable process of compromise that is so Lirgely resp<jnsible
for the misunderstandings and clashes of human contact. If words
really meant what we say they mean, there should be little room
for misunderstandings; but it is of course only too true that they
rarely mean quite what in our moments of intellectual isolation we
claim as their due significance, but that they convey thousands of
connotations over and above this ostensible meaning of theirs.
Now, it is clear that with the growth of the power of analysis there
is an ever increasing demand for the development of a perfectly
objective and unemotional set of symbols that can stay put and
mean exactly what they are supposed to mean — no more, no less.
The more exact a system of thought becomes, the more impatient it
is apt to be with the rough-and-ready symboUsms of normal speech.
Hence it is not surprising that the mathematical disciplines have
been driven to invent a great many special symbols which can be
defined with complete accuracy and which will allow no room for
expressive modifications. A plus sign that left ciny room for doubt
cLS to its meaning would be useless. But language as ordinarily
handled by society is soaked with overtones or connotations that
are nicely felt by the members of the particular society that makes
use of one of its specific forms, but that it is difficult to convey to
outsiders. One may know the vocabulary and the grammar of a
foreign language ever so well, but one is not likely to use it with that
unconscious appeal to the expressive values that attach to those
words and forms unless one has spent a considerable time among
the people who use it or unless one as a child is very much more
impressionable to these values than we adults normally are.

Language would be a poorer thing than it is if it were a deno-
tive system alone. But we must not be so sentimental as to over-
look the equally obvious fact that it is precisely the expressiveness
of particular languages that makes any one of them a misleading or
even a dangerous tool for the problem of sheer reference. 1 ho

^l^ General Linguistics I

433

necessity of evolving a complicated denotive symbolism that is
absolutely, or so far as may be, devoid of expressive values is not
very keenly felt by normal human beings, but it is more than likely
that as time goes on this purely speculative need will become more
and more imperative. International complications, for instance,
must be avoided at all costs, and there is perhaps no one device
which would do more to eliminate the subtle misunderstandings
which arise from the use of language as we ordinarily handle it
than a system of symbols which are as cold-blooded as those used
by the mathematicians, but which are inclusive enough to provide
for every possible kind of communication. We must, then, conceive
the essential task of an international language to be not so much
the choosing of this or that particular form of speech, but the crea-
tion of a system of symbols of maximal simplicity and of absolute
lack of ambiguity. The task of constructing such a system is far
more difficult than it seems to be at first blush. Esperanto and other
systems that have been suggested are to an amazing extent little
more than translations into new terms of old habits of reference,
the expressive "plus" being always included. One of the great
tasks of the future may be the creation of such an objective lan-
guage of reference from which every possible nuance of individual
or social expressiveness has been removed. Perhaps such a system
is philosophically inconceivable; but the history of mathematical
and other scientific symbolisms shows clearly that the tendency has
been toward the creation of just such a system. It goes without say-
ing, however, that a purely denotive language need not and is not in
the least likely to supersede the actual languages in use today, with
their bewildering flexibility of individual and social expressiveness.
These are likely to remain for incalculable periods to come.

Editorial Note
The English Journal 16 (1927), 421-433.

Language and I'hilology. By Roliiiui ( i. Kcni. Bosion: Marshall Jones ("omp.iin
(1923). Pp.vii+ 174'.

[83j Professor Kcnls liiik hook. l.;inuuauc ;iik1 I'lnlology. uhich is one of ihc
series entitled Our Debt lo ( ircece and Rome, edited by Professors (i. D. Had/sits
and D. M. Robinson, is an excellent and most readable statement of the extent
of our linguistic indebtedness to the classical laneuaues. Hie title is obviously,
but inoffensively, misleading, for by "Language" is meant Presenlday Iru-lish
Particularly in its Written 1 brm". and "Philology" means "Latin and Greek

How powerfully Ljiglish leans on these languages has perhaps never l")c(orc
been made so evident. Our vocabulary, our apparatus of prefixes and suffixevour
alphabet are eloquent of the far-reaching cultural influence exerted at various
times by the classical tradition. The most frenzied purist can no more successfully
de-Hellenize or de-Latinize our everyday Hnglish speech than a reformer can oust
the decimal system of notation and put a duodecimal one in its place, or than a
bolshevistic biologist can persuade us to give up the charming ritual of our meals
and revert to the more elemental law of 'bite when hungry".

Professor Kent writes in just that simple, patient, well-documented style which
is needed to make a somewhat technical array of facts intelligible and interesting
to the lay public. Here and there he fits the words he discusses into their back-
ground of use and in this way gives his discourse a liveliness — at limes e\en a
jauntiness — which is surely not native to lists of words as such. Only seldom docs
he seem to fall a victim to the temptation of saying merely pretty things, as when
certain words composed of Latin elements are said to be "as truly part of our debt
to the Latin language as though they had fallen trippingly from the lips of Cicert^
against Catiline". O tempora, O mores!

A very significant passage occurs at the end o{ the chapter on ( irammatical
Studies (128-138). It reads (138):

Notwithstanding these differences between Latin and modern English, the oldest
form of English, namely Anglo-Saxon, was a highly inflected language sery simi-
lar to Latin in forms and in syntax; and the essentials of case in nouns, of person
and number in verbs, of the use of the subjuncti\e nu)od. and oi the various
agreements between different members of the sentence, still abide in Fuiglish. and
are rarely well understood except by those who know tlieiii in their I alin aspect.

There is such a thing as seeing English thiougii l.aiin and ( Ireek eyes, but there
is also the even more insidious danger of exaggerating the degree of fundamental
structural difference between English and its more liigliK inllected prtMoiNpesand
relatives. Analogies that it has been somewhat fashionable lo point oul between
English and such thoroughgoing analytic languages as Chinese are superficial a!

<' The volumes of the Scries Our Dcbl lo ("ircccc aiul Rome .in puM.slu-.l n..\v h\ Mcxsrv I onemaav Cifccn.
and Company. New York City. C. K.>.

21X General Linguistics I

best. It IS not a qiicsiii)n ol how complex is English morphology as compared with
that of Anglo-Saxon or Latin but of what are the basic lines of its patterning, and
these arc as undeniably Indo-European and inflective'— or, as I should prefer to
sav, -fusional"— in technique and "mixed-relationar' in principle as are those of
Sanskrit itself. Just as it is more significant to compare the structural principles of
a humble frame house with those of a magnificent mansion of Occidental type than
to dilate on its similarity, as regards economy of means, to an Indian tepee or an
liskimo snow house, so too English should be seen with an eye which [86] has
learned to follow the more involved lines of Latin and Greek and Anglo-Saxon
structure.

llie contents of the book are as follows:

I. Introduction (3-7); IL Language Relationship and Behavior (8-13); IILThe
Greek Language (14-18); IV. The Latin Language (19-25); V. The English Lan-
guase (26-38); VI. Statistics and Examples (39-57); VII. Our Present-Day Vocabu-
iary^58-76); VIII. Prefixes (77-90); IX. Suffixes (91-108); X. Words and Forms
(109-127); XI. Grammatical Studies (128-138); XII. Grammatical Terminology
(139-143); XIII. The Alphabet and Writing (144-155); Conclusion: Latinless
English (156-158); Notes (161-172); Bibliography (173-174).

University of Chicago Edward Sapir

Editorial Note
The Classical Weekly 21 (1928), 85-86.

THE STATUS OF LINGUISTICS AS A SCIENCE*

E. Sapiu

University of Chicago

(The long tried methods of Indo-Euro[)ean linguistics have proved them-
selves by the success with which they have been applied to other (ielda,
for instance Central Algonkian and Athabaskan. An increasing interest
in linguistics may be noted among workers in anthropolog>', culture his-
tory, sociology, psychology, and philosophy. For all of them linguistics
is of basic importance: its data and methods show better than those of
any other discipline dealing with socialized behavior the possibility of a
truly scientific study of society. Linguists should, on the other hand, be-
come aware of what their science may mean for the interpretation of hu-
man conduct in general.)

Linguistics may be said to have begun its scientific career with the
comparative study and reconstruction of the Indo-European languages.
In the course of their detailed researches Indo-European linguists have
gradually developed a technique which is probably more nearly perfect
than that of any other science deahng with man's institutions. Many
of the formulations of comparative Indo-European linguistics have a
neatness and a regularity which recall the formulae, or the so-called
laws, of natural science. Historical and comparative linguistics has
been built up chiefly on the basis of the hypothesis that sound changes
are regular and that most morphological readjustments in language fol-
low as by-products in the wake of these regular phonetic development^^.
There are many who would be disposed to deny the psychological nece.s-
sity of the regularity of sound change, but it remains true, as a matter of
actual hnguistic experience, that faith in such regularity has Uh'u the
most successful approach to the historic problems of language. \\ hy
such regularities should be found and why it is necessary to assiiinc regu-
larity of sound change are questions that the averag(^ linguist is perhaps
unable to answer satisfactorily. But it does not follow that he can ex-
pect to improve his methods by discarding well tested hy|X)thes<\s and

* Read at a joint meeting of the Linquistic Socikty ok Amkkka , (he .Vmkricam
Anthropological Association, and Sections TI and L of the .Amkiucan .\s.m<h'i-
ATiON for the Advancement of iSciENcE, New York City, I)eceinl)er 2S, 19iS.

207

220 General Linguistics I

208

throwing the field open to all manner of psychological and sociological
explanations that do not immediately tie up with what we actually know
about the historical behavior of language. A psychological and a socio-
logical interpretation of the kind of regularity in linguistic change with
which students of language have long been familiar are indeed desirable
and even necessary. But neither psychology nor sociology is in a posi-
tion to tell linguistics what kinds of historical formulations the linguist is
to make. At best these disciplines can but urge the linguist to concern
himself in a more vital manner than heretofore with the problem of see-
ing linguistic history in the larger framework of human behavior in the
individual and in society.

The methods developed by the Indo-Europeanists have been applied
with marked success to other groups of languages. It is abundantly
clear that they apply just as rigorously to the unwritten primitive lan-
guages of Africa and America as to the better known forms of speech of
the more sophisticated peoples. It is probably in the languages of these
more cultured peoples that the fundamental regularity of linguistic
processes has been most often crossed by the operation of such conflict-
ing tendencies as borrowing from other languages, dialectic blending, and
social differentiations of speech. The more we devote ourselves to the
comparative study of the languages of a primitive linguistic stock, the
more clearly we realize that phonetic law and analogical leveling are the
only satisfactory key to the unravelling of the development of dialects
and languages from a common base. Professor Leonard Bloomfield's
experiences with Central Algonkian and my own with Athabaskan leave
nothing to be desired in this respect and are a complete answer to those
who find it difficult to accept the large scale regularity of the operation
of all those unconscious linguistic forces which in their totality give us
regular phonetic change and morphological readjustment on the basis of
such change. It is not merely theoretically possible to predict the cor-
rectness of specific forms among unlettered peoples on the basis of such
phonetic laws as have been worked out for them — such predictions are
already on record in considerable number. There can be no doubt that
the methods first developed in the field of Indo-European linguistics
are destined to play a consistently important role in the study of all other
groups of languages, and that it is through them and through their
gradual extension that we can hope to arrive at significant historical
inferences as to the remoter relations between groups of languages that
show few superficial signs of a common origin.

It is the main purpose of this paper, however, not to insist on what

Three: Descriptive and IliMoruul I ini'uistics 221

ao9

linguistics has already accomplished, but rather to point out some of
the connections between linguistics and other scientific disciplin««.s, and
above all to raise the question in wliat sense linguistics cnn be cillrd n.
'science'.

The value of linguistics for anthropology and culture history \\i\n long
been recognized. As linguistic research has proceeded, language ha8
proved useful as a tool in the sciences of man and has itself re(juired and
obtained a great deal of light from the rest of these sciences. It ia
difficult for a modern linguist to confine himself to his traditional subject
matter. Unless he is somewhat unimaginative, he cannot but shan; in
some or all of the mutual interests which tie up linguistics with anthro-
pology and culture history, with sociology, with psychology', with philos-
ophy, and, more remotely, with physics and physiology.

Language is becoming increasingly valuable as a guide to the scientific
study of a given culture. In a sense, the network of cultural {)attern3
of a civilization is indexed in the language which expresses that civiliza-
tion. It is an illusion to think that we can understand the significant
outlines of a culture through sheer observation and without the guide of
the linguistic symbolism which makes these outlines significant and
intelligible to society. Some day the attempt to master a primitive
culture without the help of the language of its society will seem as ama-
teurish as the labors of a historian who cannot handle the original docu-
ments of the civilization which he is describing.

Language is a guide to 'social reality'. Though language is not
ordinarily thought of as of essential interest to the students of social
science, it powerfully conditions all our thinking about social problems
and processes. Human beings do not live in the objective world alone,
nor alone in the world of social activity as ordinarily understood, but
are very much at the mercy of the particular language which has Ix'come
the medium of expression for their society. It is quite an illusion to
imagine that one adjusts to reality essentially without the use of lan-
guage and that language is merely an incidental means of solving
specific problems of communication or reflection. The fact of the
matter is that the 'real world' is to a large extent unconsciously built
up on the language habits of the group. No two languages an' ever
sufficiently similar to be considered as representing the same social
reahty. The worlds in which different societies live are distinct worlds,
not merely the same world with different labels attached.

The understanding of a simi)le poem, for instance, involves not merely
an understanding of the single words in their average significance, but

222 General Linguistics I

210

a full comprehension of the whole life of the community as it is mirrored
in the words, or as it is suggested by their overtones. Even compara-
tively simple acts of perception are very much more at the mercy of the
social patterns called words than we might suppose. If one draws some
dozen lines, for instance, of different shapes, one perceives them as
divisible into such categories as 'straight', 'crooked', 'curved', 'zigzag'
because of the classificatory suggestiveness of the linguistic terms them-
selves. We see and hear and otherwise experience very largely as we do
because the language habits of our community predispose certain
choices of interpretation.

For the more fundamental problems of the student of human culture,
therefore, a knowledge of linguistic mechanisms and historical develop-
ments is certain to become more and more important as our analysis of
social behavior becomes more refined. From this standpoint we may
think of language as the symbolic guide to culture. In another sense too
linguistics is of great assistance in the study of cultural phenomena.
Many cultural objects and ideas have been diffused in connection with
their terminology, so that a study of the distribution of culturally
significant terms often throws unexpected light on the history of inven-
tions and ideas. This type of research, already fruitful in European
and Asiatic culture history, is destined to be of great assistance in the
reconstruction of primitive cultures.

The value of linguistics for sociology in the narrower sense of the
word is just as real as for the anthropological theorist. Sociologists
are necessarily interested in the technique of conamunication between
human beings. From this standpoint language faciUtation and language
barriers are of the utmost importance and must be studied in their inter-
play with a host of other factors that make for ease or difficulty of trans-
mission of ideas and patterns of behavior. Furthermore, the sociologist
is necessarily interested in the symbolic significance, in a social sense, of
the linguistic differences which appear in any large community. Cor-
rectness of speech or what might be called 'social style' in speech is of far
more than aesthetic or grammatical interest. Peculiar modes of pro-
nunciation, characteristic turns of phrase, slangy forms of speech,
occupational terminologies of all sorts — these are so many symbols of
the manifold ways in which society arranges itself and are of crucial
miportance for the understanding of the development of individual and
social attitudes. Yet it will not be possible for a social student to eval-
uate such phenomena unless he has very clear notions of the linguistic
background against which social symbolisms of a linguistic sort are to be
estimated.

Three: Dcscripiivc ami Uistorndl I ini^uisiics 223

211

It is very encouraging that the psychologist has Ix^en concorninR him-
self more and more with linguistic data. So far it is doubtful if he htm
been able to contribute very much to the understandmg of lunKuan*-*
behavior beyond what the linguist has himself been ablo to fonnulatc
on the basis of his data. But the feeling is growing rapidly, and juHtly,
that the psychological explanations of the linguists th('ms<'lve.s need to
be restated in more general terms, so that purely linguistic facts may be
seen as specialized forms of symbolic behavior. The psychologists have
perhaps too narrowly concerned themselves with the simple psycho-
physical bases of speech and have not penetrated very deep)ly into the
study of its symbolic nature. This is probably due to the fact that
psychologists in general are as yet too little aware of the fundamental
importance of symbolism in behavior. It is not unlikely that it is
precisely in the field of symbohsm that linguistic forms and processes
will contribute most to the enrichment of psychology.

All activities may be thought of as either definitely functional in the
immediate sense, or as symbolic, or as a blend of the two. Thus, if I
shove open a door in order to enter a house, the significance of the act
lies precisely in its allowing me to make an easy entry. But if I 'knock
at the door', a Httle reflection shows that the knock in itself does not
open the door for me. It serves merely as a sign that somebody is to
come to open it for me. To knock on the door is a substitute for the
more primitive act of shoving it open of one's own accord. Wv have
here the rudiments of what might be called language. A vast number of
acts are language acts in this crude sense. That is, they are not of im-
portance to us because of the work they immediately do, but because*
they serve as mediating signs of other more important acts. A primi-
tive sign has some objective resemblance to what it takes the place of or
points to. Thus, knocking at the door has a definite relation to intended
activity upon the door itself. Some signs become abbreviated forms of
functional activities which can be used for reference. Thus, shaking
one's fist at a person is an abbreviated and relatively harmless way of
actually punching him. If such a gesture becomes sufficiently expres-
sive to society to constitute in some sort the equivak'nt of an abuse or a
threat, it may be looked on as a symbol in the proper s<'nse of tlie word.

SjTTibols of this sort are primary in that the resemblance of the
symbol to what it stands for is still fairly evident. As time goes on,
symbols become so completely changed in form as to Io.se all outward
connection with what they stand for. Thus, there is no resemblfvnec
between a piece of bunting colored red, white, and blue, imd tin* United

224 General Linguistics I

212

States of America, — itself a complex and not easily definable notion.
The flag may therefore be looked upon as a secondary or referential
symbol. The way to understand language psychologically, it seems, is
to see it as the most complicated example of such a secondary or referen-
tial set of symbols that society has evolved. It may be that originally the
primal cries or other types of symbols developed by man had some con-
nection with certain emotions or attitudes or notions. But a connection
is no longer directly traceable between words, or combinations of words,
and what they refer to.

Linguistics is at once one of the most difficult and one of the most
fundamental fields of inquiry. It is probable that a really fruitful
integration of linguistic and psychological studies lies still in the future.
We may suspect that linguistics is destined to have a very special value
for configurative psychology ('Gestalt psychology'), for, of all forms of
culture, it seems that language is that one which develops its fundamen-
tal patterns with relatively the most complete detachment from other
types of cultural patterning. Linguistics may thus hope to become some-
thing of a guide to the understanding of the 'psychological geography' of
culture in the large. In ordinary life the basic symbolisms of behavior
are densely overlaid by cross-functional patterns of a bewildering vari-
ety. It is because every isolated act in human behavior is the meeting
point of many distinct configurations that it is so difficult for most of
us to arrive at the notion of contextual and non-contextual form in
behavior. Linguistics would seem to have a very peculiar value for
configurative studies because the patterning of language is to a very ap-
preciable extent self-contained and not significantly at the mercy of
intercrossing patterns of a non-linguistic type.

It is very notable that philosophy in recent years has concerned itself
with problems of language as never before. The time is long past when
grammatical forms and processes can be naively translated by philoso-
phers into metaphysical entities. The philosopher needs to understand
language if only to protect himself against his own language habits, and
so it is not surprising that philosophy, in attempting to free logic from
the trammels of grammar and to understand knowledge and the meaning
of symbolism, is compelled to make a prehminary critique of the linguis-
tic process itself. Linguists should be in an excellent position to assist
in the process of making clear to ourselves the implications of our terms
and linguistic procedures. Of all students of human behavior, the lin-
guist should by the very nature of his subject matter be the most relativ-
ist in feeling, .the least taken in by the forms of his own speech.

Three: nrscriprivc and Historical I .in\>uistks 225

213

A word as to tho relation between linguistics and the natural hcicnccK.
Students of linguistics have been greatly indebted for their t4.'chnical
equipment to the natural sciences, particularly physics and physioloK\'.
Phonetics, a necessary prerequisite for all exact work in linguiaticn, ih
impossible without some grounding in acoustics and the phyMolog^- of
the speech organs. It is particularly those students of Innguage who are
more interested in the realistic details of actual speecli behavior m the
individual than in the sociahzed patterns of language who must have
constant recourse to the natural sciences. But it is far from unlikely
that the accumulated experience of linguistic research may {)rovide more
than one valuable hint for the setting up of problems of research to acous-
tics and physiology themselves.

All in all, it is clear that the interest in language has in recent years
been transcending the strictly linguistic circles. This is inevitable, for
an understanding of language mechanisms is necessary for the study of
both historical problems and problems of human behavior. One can
only hope that linguists will become increasingly aware of the signifi-
cance of their subject in the general field of science and will not stand
aloof behind a tradition that threatens to become scholastic when not
vitalized by interests which lie beyond the formal interest in language
itself.

Where, finally, does linguistics stand as a science? Does it belong to
the natural sciences, with biology, or to the social sciences? There seem
to be two facts which are responsible for the persistent tendency to view
linguistic data from a biological point of view. In the first place, there is
the obvious fact that the actual technique of language behavior involves
very specific adjustments of a physiological sort. In the second place,
the regularity and typicality of linguistic processes leads to a quiisi-
romantic feehng of contrast with the apparently free and undetermined
behavior of human beings studied from the standpoint of culture. Hut
the regularity of sound change is only superficially analogous to a bio-
logical automatism. It is precisely because language is as strictly
sociahzed a type of human behavior as anything else in culture and yet
betrays in its outlines and tendencies such regularities jis only the
natural scientist is in the habit of formulating, that linguistics is of
strategic importance for the methodology of social science. Hehind
the apparent lawlessness of social phenomena there is a regularity of
configuration and tendency which is just as real as the regularity of ph>-F-
ical processes in a mechanical world, though it is a reguhirity of infi-
nitely less apparent rigidity and of another mode of apprehension on our

226 General Linguistics I

214

part. Language is primarily a cultural or social product and must be
understood as such. Its regularity and formal development rest on
considerations of a biological and psychological nature, to be sure. But
this regularity and our underlying unconsciousness of its typical forms
do not make of linguistics a mere adjunct to either biology or psychology.
Better than any other social science, linguistics shows by its data and
methods, necessarily more easily defined than the data and methods of
any other type of discipline dealing with socialized behavior, the
possibility of a truly scientific study of society which does not ape the
methods nor attempt to adopt unrevised the concepts of the natural
sciences. It is peculiarly important that linguists, who are often ac-
cused, and accused justly, of failure to look beyond the pretty patterns
of their subject matter, should become aware of what their science may
mean for the interpretation of human conduct in general. Whether
they like it or not, they must become increasingly concerned with the
many anthropological, sociological, and psychological problems which
invade the field of language.

Editorial Note

Language 5 (1929), 207-214. [Reprinted in: Edward Sapir, Selected Writings in
Language, Culture, and Personality. Edited by David G. Mandelbaum. Berkeley:
University of California Press, 1949, pp. 160-166]

The following error in the originally published version has been corrected
directly into the text printed here (the page reference is to the original):
p. 210, 1. 23: values (correct: value)

A STUDY IN PHONETIC SYMBOLISM

BY EDWARD SAPIR
Univerjity of Chicago *

The symbolism of language is, or may be, twofold. By
far the greater portion of its recognized content and structure
is symbolic in a purely referential sense; in other words, the
meaningful combinations of vowels and consonants (words,
significant parts of words, and word groupings) derive their
functional significance from the arbitrary associations between
them and their meanings established by various societies in
the course of an uncontrollably long period of historical
development. That these associations are essentially arbi-
trary or conventional may be seen at once by considering
such a proportion as

phonetic entity *boy*: idea (or reference) *boy*
B» phonetic entity 'man': idea (or reference)
'man.*
In passing from the notion of 'boy* to that of 'man* we
experience a definite feeling of relationship between the two
notions, that of increase in size and age. But the purely
phonetic relationship of 'boy* : 'man* takes no account of
this. So far as the referential symbolism of language is
concerned, the words 'boy* and 'man* are discrete, incom-
parable phonetic entities, the sound-group b-o-y having no
more to do with the sound-group m-a-n, in a possible scale
of evaluated phonetic variants, than any randomly selected
pair of sound-groups, say 'run' and 'bad,' have to do with
each other.

This completely dissociated type of symbolism is of course
familiar; it is of the very essence of linguistic form. But

* Publication of the Behavior Research Fund, the Institute for Juvenile Reiearch,
Chicago (Herman M. Adler, Director), Ser. B, No. 132. For valuable lugKeitioni ia
the preparation of this paper I am indebted to Professor H. A. Carr, Univcnity of
Chicago.

225

228 General Linguistics I

226

there are other types of linguistic expression that suggest a
more fundamental, a psychologically primary, sort of sym-
bolism. ^ As examples may be given the interrogative tone
in such a spoken sentence as "You say he's dead?" in com-
parison with the simple declarative tone of the corresponding
"You say he's dead"; further, the emphatically diminutive
f^ of teeny as contrasted with the normal i of tiny. In both
of these examples the phonetic difference is undoubtedly felt
as somehow directly expressive of the difference of meaning
in a sense in which the contrast between say *boy' and *man'
is not. We may call this type of symbolism ^expressive'
as contrasted with the merely 'referential' symbolism which
was first spoken of. It goes without saying that in actual
speech referential and expressive symbolisms are pooled in a
single expressive stream, the socialization of the tendency to
expressive symbolism being far less extreme, in the great
majority of languages, than of the tendency to fix references
as such.

Wc may legitimately ask if there are, in the speech of a
considerable percentage of normal individuals, certain prefer-
ential tendencies to expressive symbolism not only in the
field of speech dynamics (stress, pitch, and varying quantities),
but also in the field of phonetic material as ordinarily under-
stood. Can It be shown, In other words, that symbolisms
tend to work themselves out in vocalic and consonantal
contrasts and scales in spite of the arbitrary allocations of
these same vowels and consonants in the strictly socialized
field of reference.'' The present paper is a preliminary report
of certain aspects of a study, still in progress, intended to
probe into any such latent symbolisms as may be thought to
exist. The field of inquiry is vast and difficult to chart and
I cannot hope to have guarded against all the possible fallacies
of interpretation. For the present I have limited myself to
the meaning contrast Marge* : * small' as oflfering the most
likely chance of arriving at relatively tangible results.

The main object of the study is to ascertain if there

* For the two symbolic layers in speech, as in all expression, sec E. Sapir, Language
as a form of human behavior, Engl, J., 1927, 16, 421-433.

Three: Descriptive <///</ Hisu,rual / iniitiisdcs 229

227

tends to be a feeling of the symbolic magnitude value of
certain differences in vowels and consonants, regardless of the
particular associations due to the presence of these vowels and
consonants in meaningful words in the language of the
speaker. The results so far obtained seem to go far in
demonstrating the reality of such feelings, whatever may be
their cause. It has also become very clear that individuals
differ a good deal in the matter of sensitiveness to the symbolic
suggestiveness of special sound contrasts.

A number of distinct schedules have been devised and
applied in the research. In the early stages of the work
the various types of sound difference were studied inde-
pendently. For instance, the contrast between the vowel
a and the vowel i (the phonetic or continental values are
intended) was illustrated in every one of sixty pairs of
stimulus words, the subject being requested to indicate in
each case which of the two in themselves meaningless words
meant the larger and which the smaller variety of an arbi-
trarily selected meaning. For example, the meaningless
words mal and mil were pronounced in that order and given
the arbitrary meaning 'table.' The subject decided whether
mal seemed to symbolize a large or a small table as contrasted
with the word mil.

In the first experiments schedules of sixty stimulus word-pairs were used, each of
which was divided into two sections. The first thirty word-pairs involved only such
sounds as the subject, an English-speaking person, would be familiar with, the lecond
set of thirty word-pairs, while still illustrating the same phonetic contrast at the
first thirty, say that of a to t, also involved sounds that the subject was not familiar
with. Each of the two sets of thirty was further subdivided into functional groups:
nouns, verbs with reference to large or small subject of verb, adjectives with reference
to large or small things, verbs with reference to large or small object of verb, and
verbs with reference to intense or normal degree of activity. It is important to note
that the words were so selected as to avoid associations with meaningful words and
it was the special purpose of the second set of thirty word-pairs to remove the subject
still further from the intercurrent influence of meaningful linguistic associations.

If the results obtained from a considerable number of individuals can be relied
upon as symptomatic, the influence of accidental, meaningful linguistic asiociatioaa is
less than might have been supposed, for the. percentage of responses in fa\x>r of one
of the two vowels as symbolizing the large object tended to be little less, if at all, in
the second set of word-pairs than in the first. For example, Subj. IK. found that
of the first thirty word-pairs illustrating a contrast between the vowels a and 1 twenty-
two examples of a "naturally " carried with them the connotation " large," five examples

230 General Linguistics I

228

of 1 carried this connotation, and three word-pairs were responded to indiflPcrently.
The effective score in favor of a as the vowel inherently Symbolizing a large rather
than a small reference was 22/27 or 81 per cent. In the second set of thirty word-
pairs illustrating the same vocalic contrast, 21 of the words involving the vowel a
were said to connote the large reference, 5 with the vowel 1 connoted the small reference,
and 4 were indifferent. Here the effective score in favor of the symbolic value of the
vowel a as large by contrast with i is 21/26 or, again, 81 per cent. In the case of the
vowel contrast a to f (with the short value of the French e, as in etf) IK's effective
score in favor of the a vowel as connoting the larger reference was 24/29 or 83 per cent
for the first 30 word-pairs, 73 per cent for the second 30 word-pairs.

The essential points that seemed to appear from these
first experiments with individuals were; (i) that vocalic and
consonantal contrasts tended with many, indeed with most,
individuals to have a definite symbolic feeling-significance
that seemed to have little relation to the associative values
of actual words, (2) that it made surprisingly little difference
whether the phonetic contrast was contained in a phonetically
"possible" or a phonetically "impossible" context and (3)
that the certainty of the symbolic distinction tended to vary
with the nature of the phonetic contrast. The last point,
which is important, will be discussed later on in this report.
These earlier experiments with individuals, though re-
vealing, were felt as the work proceeded to be deficient in
one important respect, namely, that the simple nature of the
vocalic or consonantal contrast in a set of word-pairs might
be expected to lead to a too ready systematization of responses
on the part of the subject. In other words, the average
subject could not help noticing after responding to a few
stimuli that a certain consistency in the responses would
naturally be expected, and that if the vowel a, for example,
as contrasted with e or i, is felt satisfactorily to symbolize the
larger of two objects, all other examples of word-pairs illus-
trating the same vocalic contrast should be dealt with in
the same manner. The primary purpose of the experiment,
however, was to elicit spontaneous feelings of symbolic con-
trast, unrevised by any judgment as to consistency of re-
sponse. For this reason a further and, It is believed, much
more efficient experiment was devised consisting of 100 word-
pairs involving every type of phonetic contrast that was
investigated. These hundred word-pairs were not arranged

Three: Descriptive ami HiMonml I in^imucs 231

229

in any logical order, nor was the order of the contrasted
phonetic elements in any particular entry necessarily the
same as in another entry involving the same contrast. In
the table that was finally adopted the first word-pair illus-
trated the contrast between a and 1, the second the contrast
between e and fl, the third the contrast between z and /,
and so on through the list. The contrast between a and :
was illustrated not only in Entry i but also in Entries 41,
81, and 87. In this way, it was hoped, systematization on
the part of the subject was necessarily hindered, if not entirely
blocked, and the responses actually obtained may be looked
upon as normally spontaneous feeling judgments following in
the wake of an initial suggestion as to preferred class of
symbolic response {i.e. variations in magnitude).

For this second experiment 500 subjects were employed, most of them students
of the University of Chicago High School. The subjects were eventually analyzed
into the following groups; 6 cases of n-year-old children, 30 of 12 years, 86 of 13
years, 94 of 14 years, 124 of 15 years, 81 of 16 years, 33 of 17 years, 10 of 18 years,
21 University of Chicago students, 8 adults who were not students and 7 Chinese.
The subjects were provided with forms in which there were blank spaces for each of
the entries, and they were carefully instructed to check off the first of the two stimulus
words announced by the investigator as to whether it symbolized the larger or the
smaller reference. If the response was indifFcrcnt, no check was to be entered in either
the large or the small column. Very little difficulty was experienced in explaining
the conditions of the experiment, which seemed to be enjoyed by the great majority
of the subjects as a rather interesting game. It is believed that the results obtained
are as reliable as material of this kind can be, every precaution having been taken to
arrange conditions favoring simple and unambiguous resj^onses and only the investi-
gator himself pronouncing the stimulus words, in order that all confusion due to slight
variations of pronunciation might be avoided.

The phonetic contrasts may be classified on phonetic and
acoustic grounds into five main groups. There arc also two
minor groups which are of lesser interest. In the first group
the contrasting vowels belong to the scries a, a, (, ^, 1. The
pronunciation of these vowels, as of all other vowels, was
quantitatively uniform in a given pair in order that the in-
dependent symbolic suggestivcness of quantity difTcrences as
such be ruled out of consideration where quality alone was
being studied. The phonetic values of these vowels were
respectively those of a of German Mann (a), a of English
hat (d), e of English met (c), e of French ete (e), i of French

232 General Linguistics I

230

fini (1). It will be observed that the phonetic contrast Is
gradually lessened within the scale as one moves from a to t.
Thus, a to i affords the greatest objective contrast, a to i or
^ to <r a lesser contrast, c to i or fl to € a still lesser one, and
a to a or a to € or < to <f or <f to i a minimal contrast. In other
words, on purely objective phonetic grounds, one might
imagine that the responses would tend to be further removed
from a purely random or 50-50 distribution the greater the
contrast between the vowels. It was therefore of great
interest to determine not only whether there were preferred
symbolisms, but also whether the varying percentages of
response bore a fairly close relation to objective differences in
the sounds themselves as determined on phonetic and acoustic
grounds.

The second group of word-pairs illustrates the contrast
between vowels on the scale a, J, 0, u^ i.e. a scale with pro-
gressive lip-rounding. The third group illustrates contrasts
between rounded back vowels (w, 0, 0) and unrounded front
vowels (i, e, «, a). In the fourth group of word-pairs there
was illustrated the contrast between voiced and voiceless
consonants, e.g. between z and J, v and f, b and p. The fifth
group illustrates the contrast between stopped consonants and
spirants or fricatives, e.g. between / and p, x (ch of German
Bach) and k.

It would be quite impossible to report on all the details
of the experiment in this place. I shall content myself with
giving two selected tables. The first shows the distribution
of responses for the word-pairs illustrating the contrast
between a and i, classified according to the groups of subjects
(11-18 yrs, university students, adults and Chinese).

Table I

Percentage of responses showing preference for a vs. 1

TO SYMBOLIZE ' LARGE*

Entry Obs. 6 30 86 94 124 8i 33 10 21 8 7

no. Age II 12 13 14 IS 16 17 18 Univ. Adits. Chin.

* 83.3 86.7 90.6 92.3 83.1 84.0 78.8 80.0 85.0 lOO.O lOO.O

41 loo.o 70.0 82.7 78.0 76.4 71.6 69.7 50.0 95.2 loo.o 85.7

J* 83.3 93.3 74.7 72.2 8i.8 80.0 77.4 loo.o 70.0 85.7 85.7

^7 83.3 83.3 84.1 86.0 91.8 86.1 72.7 80.0 90.0 loo.o 42.9

Ave... 87.5 83.3 83.0 82.1 83.3 80.4 74.6i 77.S Ss.oJ 964 78.6

Three: Descriptive ami l/isiorn ul I iniiuiMus 233

231

It will be observed that the percentage of responses in
favor of a vs. : ranges all the way from about 75 per cent to
about 96 per cent. For the largest group of subjects, the
124 fifteen-year-olds, the percentage is as high as 83, while
the small number of il-year-olds reach the figure 87.5. It is
obvious that, regardless of infinite differences of an individual
nature as to the general symbolic value of this phonetic
contrast or as to its specific value in particular cases, English-
speaking society does, for some reason or other, feel that of
these two vowels, a, by and large, is possessed of a greater
potential magnitude symbolism than the contrasted vowel i.
The same feeling seems to be illustrated by the small number
of Chinese cases. Furthermore, within the English-speaking
community there seems little reason to believe that there is a
significant growth in the firmness of the symbolic feeling after
the age of 11. The case of the eight adults is not really
significant because they consisted of high school teachers of
English who answered the forms at the same time as their
classes. They would naturally have a more self-conscious
attitude toward the problem of sound symbolism than indi-
viduals selected at random. In other words, however these
symbolisms are fixed, it is probable that they arc so fixed at a
rather early age and that familiarity with literature is not
likely to count as a heavy factor in the situation. These
general considerations are borne out by all the other findings,
and it is of particular interest to note that the Chinese
evidence is nearly always in the same general direction as
that of the English-speaking subjects. Further work needs
to be done on responses of this kind from younger children
and from other groups of foreigners before the age and
language factors can be properly evaluated or dismissed as
irrelevant.

The second table is an attempt to show the differential
symbolic value of the vocalic contrasts in the a to :" scries.
Four age-groups (13-16), involving 385 subjects, are repre-
sented in this table. It was found in comparing the responses
to the different vocalic pairs that they tended to arrange
themselves roughly into four distinct groups (A, B, C, D).

234 General Linguistics I

232

tr> J?" do Q I

^,^

t<^

00 00

«>

c§

vo' fi

v^S

/ — \

/— V

^"^/"^

/— >

,,-v

c^-

a a

•

»w'

t< N

l-l

>»— 'V-**

^-^

^— '

'••

■••

•»• W

■

'••

«*

■ •

• •

«•

••

• •

* «3

•

>— 1

•

l-H

■•*

M Cf< CT^

xr\

^ '

•J

•*

<->

<->.

^— \

♦J

a
♦J

M-

•>♦•

«^

M f* 1-

^w'

^ka'

V_^>_^V«^

Vw'

■••

•m <« w

*•*

'^J

«•> q i;j

t^O

•«^«^

00 to

'«^ tA

f^ M

(

» 00 00 00

t^r>-

vo^o

./^^y—^

,^-^

^-s

,,-i-^

<~N/— N

0..

aS.

Ck.

«J

« >«

♦-•

«->

4-<

♦J -

C>oo

bo
<

M U

Three: Dcscnpiivc <///,/ lli\i,,nciil I

infiui.Mitw 235

233

In the first group, typically illustrated by the contrast
between a and i and a and 1, the percentage of a response in
favor of the vowel nearer a of the scale ranged from 80 per
cent upward. The second group of responses was found to
be somewhat set off from the preceding one by a marked
decrease in the percentage of responses favoring the vowel
toward a of the scale. This group is typically illustrated by
the contrast between a and €, the percentage in favor of the
Marger' vowel running from about 73 per cent to 78 per cent.
The third group, illustrated by the typical contrast ^ to t,
ranges from about 60 per cent to 70 per cent. The last
group, that of minimal psychological contrast in the a to t
set, runs below 60 per cent in favor of the vowel toward a of
the scale.

The table has been arranged chiefly from the point of
view of the internal 'hiatus' between the percentages of
response within each age-group. It is noteworthy that the
'configurated distribution' of the responses runs fairly parallel
in the four age groups both as to the stepwise discriminations
which seem to be felt by many of the subjects and as to the
actual order of the specific vocalic contrasts when evaluated
by means of percentages in favor of the vowel toward a of
the scale. Naturally, the reality and normal limits of these
stepwise discriminations need to be tested by a careful
examination of the individual records, supplemented by
further experiments.

On the whole, it will be observed that the symbolic discrimi-
nations run encouragingly parallel to the objective ones based
on phonetic considerations. This may mean that the chances
of the responses being to a high degree determined by actual
word associations of the language of the subject are slim,
the meanings of words not being distributed, so far as known,
according to any principle of sound values as such ; and, further,
that we are really dealing with a measurably independent
psychological factor that for want of a better term may be
called 'phonetic symbolism.'

One vocalic contrast, however, falls out of the expected
picture. This is the a to f set, which is starred in the tabic.

>36 General Linguistics I

234

Though the a vowel is judged prevailingly Marge' as con-
trasted with ^, there seems to be present some factor of
hesitation which lessens the value of the contrast. If we
go by objective distances between vowels, the ^ to <f contrast,
being a '3-step' one, should have fallen into Group A, instead
of which it actually either comes last in Group B or falls
even as low as Group C. I believe that a very interesting
and sufficient reason can be given for this curious fact.
The short vowel ^, as in French ete^ is not native to the
English language. Subjects hearing the vowel ^, when pro-
nounced in the proximity of a, which is acoustically far
re-moved from it, would tend not to hear what was actually
pronounced, but to project the characteristic long V- vowel'
familiar to us in such words as raise or lake. In other words,
the qualitative symbolism would tend to receive a revision
in the opposite sense because of an intercurrent quantitative
symbolism. This example is suggestive as illustrating the
importance of the linguistic factor vs. the merely phonetic
one, though not in the sense in which the term 'linguistic
factor' is ordinarily understood. What skews the picture
here is probably not the associative power of particular
English words but the phonetic configuration of English as
such.' That even this configuration, however, is of limited
importance in interpreting the experiment is shown by the
fact that in word-pairs illustrating the contrast e to i, e to ^,
the acoustic nearness of the two vowels prevents the un-
consciously imputed quantitative interference from making
itself felt in the symbolic response.

These and many other similar results need interpretation.
One's first temptation is to look about for some peculiarity
of English speech, some distribution of sounds in actual
words, that would make the results we have secured in-
telligible. A simple associational explanation, however, is
not likely to prove tenable. The weighting of the responses
is altogether too much in accordance with an absolute phonetic

• For the significance in language of 'sound patterns' or 'phonetic configurations'
as distinct from sounds as such, see E. Sapir, Sound patterns in language, Language,
1925, I, 37-51-

Thrt'c: DcscrifUix c <///</ I lisioncul 1 iniiutslus 237

«15

scale to make it possible in the long run to avoid at least some
use of ^natural' or 'expressive,' as contrasted with socially
fixed verbal, symbolism as an explanation. It is difficult to
resist the conclusion that in some way a significant proportion
of normal people feel that, other things being equal, a word
with the vowel a is likely to symbolize something larger than
a similar word with the vowel :, or ^, or «, or a. To put it
roughly, certain vowels and certain consonants 'sound bigger*
than others. It would be an important check to amass a
large number of randomly distributed meaningful words, to
classify into the two groups of Marge' and 'small' those which
could be so classified without serious difficulty, and to see
if in sets in which equal numbers of phonetically contrasted
words are found the meaning classes were or were not corre-
lated with the sound classes and to see further, if they are so
correlated, if the distributions are of the same nature as
those studied in the experiments.

The reason for this unconscious symbolism, the factor of
linguistic interference being set aside for the present, may be
acoustic or kinesthetic or a combination of both. It is
possible that the inherent 'volume' of certain vowels is
greater than that of others and that this factor alone is
sufficient to explain the results of the experiment. On the
other hand, it should be noted that one may unconsciously
feel that the tongue position for one vowel is symbolically
'large' as contrasted with the tongue position for another.
In the case of : the tongue is high up toward the roof of the
mouth and articulates pretty well forward. In other words,
the vibrating column of air is passing through a narrow
resonance chamber. In the case of a the tongue is very
considerably lowered in comparison, and also retracted. In
other words, the vibrating column of air is now passing
through a much wider resonance chamber. This kinestiictic
explanation is just as simple as the acoustic one and really
means no more than that a spatially extended gesture is
symbolic of a larger reference than a spatially restricted
gesture. In discussing some of the results with the children
themselves, who seemed very much interested in the rationale

238 General Linguistics I

236

of the experiment, the impression was gained that the subjects
differed somewhat in the psychological basis of the symbolism,
some being apparently swayed entirely by the acoustic factor,
others by the acoustic factor only or mainly insofar as it was
itself supported by the kinesthetic factor.

The tabulated results, of which we have given a brief sample, have the dis-
advantage of drowning out significant individual variations. For a preliminary report
such a method of presentation is at least suggestive; but it would be important to
know to what extent individuals differ significantly in their ability to feel symbolism
in sound contrasts. The schedules need to be gone over from the point of view of
working out individual indices of 'symbolic sensitiveness' to sounds.

Meanwhile a third experiment, intended to bring out
individual idiosyncrasies, was carried out with a number of
selected subjects, chiefly adults. The results were interesting.

In this experiment an artificial 'word' was taken as a starting point and assigned
an arbitrary meaning by either the investigator or the subject. The subject was
asked to hold on to this arbitrary meaning and to try to establish as firn\ an association
as possible between the imaginary word and its given meaning. Some phonetic
clement in the word, a vowel or a consonant, was then changed and the subject aslccd
to say what difTerencc of meaning seemed naturally to result. The answer was to be
spontaneous, unintellcctuali/.ed. The process was kept on for as long a period as
seemed worth while, the saturation point of meaningful and interested responses
being reached very soon in some cases, very late in others. In the case of certain
individuals more than 50 distinct 'words' were found to build up a constellated system
in which the meanings were rather obviously the results of certain intuitively felt
symbolic relations between the varied sounds. In the case of other individuals actual
word associations tended to creep in, but on the whole there was surprisingly little
evidence of this factor. The subjects were found to differ a great deal in their ability
to hold on without effort to a constellation once formed and to fit new meanings into
It consistently with the symbolisms expressed in previous responses. Some would
give identically the same response for a stimulus word that had been — so it was claimed
— forgotten as such. In its imaginary, constellated context it evoked a consistent
response. Others lost their moorings ver>' rapidly. It is hoped to discuss these
interesting variations of sensitivity to sound symbolism, i.e. to the pjotential meaning-
fulness of relations in sound sets, in the final report of these investigations.

In the present purely preliminary report we can do little
more than give a few examples of the responses of two of
the subjects, KP and JS. The word mila was arbitrarily
defined as 'brook' by the subject KP. Fifty-three responses
were obtained from her, starting with this nucleus. The
following excerpts from the material will be illuminating.

1. mila: "Brook."

2. mila: "Smaller brook."

Three: Di'scripiivc mul /hsinrntil l.irmnisrics 239

^37

3. mila: "Larger brook; nearer a river; swifter; no longer thought of as pin of the

meadow landscape."

4. mela: "Larger, not so flowing; large lake like Lake Superior."

5. mela: "Little lake for fishing."

6. mrla: "Smaller brook than m'tla, larger than mila."

7. m^a: "Larger than brook. Perhaps water running through a ravine; mixed up

with the scene."

8. mlna: "Water trickling down in a ravine through the rocks scattered on the side."

9. mena: "A little larger but still diminutive. Water travels through a gravel

10. mina: "Deep, narrow, swiftly moving stream rushing through a cut in the rocks."

14. mini: "Tiny but swift stream spurting out of the rocks like a jet of water."

18. mula: "Fairly large, rather rambling brook at night."

19. mola: "Ocean at night."

20. m3la: "Ocean in the daytime."

21. mala: "Bright ocean."

2^ mila: (21st response after 2): "A little brook." (The jump back to the earlier
response was made at once, without hesitation.)

14*. fntni: (isth response after 14): "Spurt of water from the rocks, small but swift."

48. wila: "Can be wet, but water is more like dew on wet grass after rain. Belongs

to the same set."

49. wela: "Wet trees after a rain. No feeling of a body of water. General dampness,

a 'larger' feeling than wila."

Not all subjects by any means were as responsive as KP;
but a surprising number showed a very definite tendency
toward the constellating of sound symbolisms. A few re-
sponses from JS, based on the same stimulus word, will be
interesting for purposes of comparison. The meaning 'brook'
was assigned by the investigator and accepted as satisfactory
by the subject.

1. mila: "Brook."

2. mela: "Seems to sort of broaden out. Brook got much calmer."

3. mila: "Got to chattering again; smaller brook; stones visible, which make the

noise."

^40 General Linguistics I

238

4. m'tla: "Brook gcta stagnant with rushes growing in it. The rushes hold the water

back so it forms pools. The flow is in the middle; relatively stagnant at the
cdgtrs."

5. mala: "Almost like a lake. An uninteresting lake."

6. mala: "More color in it. May have been shallow before; now has greater depth

of color, greener shadows; still a lake."

7. m^la: "Pools taken out at the side from 4. Regains a little of its chattering. Sort

of tiny. Less cheerful and chirpy than 3. No great difference as to size
between 7 and 3. Merely has a deeper note."

G. mala (4th response after 6): "Nice broad pool with all nice colors in it. Shadows
and water rich green, as of tree shadow in pool."

l». m'lla (nth response after 1): "Rather nice chattering brook.'

17. m'llt: "A little splash of water. Tiny stream hit a rock and spattered out io all

directions."

18. mili: "Water has gone. A bit of rather dense woods with lots of moisture.

Water not evident, but obviously somewhere. You don't sec water but you
know it is there. Rather soggy to walk around."

6*. mala (23d response after 6*, with much material in between that was definitely
removed from suggestions of 6): "Quick sweep of water view over a lake. Not
just a pond. A few islands, but they look like dots. The sun is setting.
There are nice black shadows this side of the island. The scenery is darkest
where I am. I am interested in the distant brightness."

A comparison of these excerpts from the two schedules
shows certain interesting resemblances and differences. Both
subjects constellate their responses; but KP does so more
rigidly, * geometrically,* as it were. With JS the underlying
* geometry' of response is enriched by imaginative overtones.
Incidentally, it will be observed by the attentive reader, a
considerable number of the responses here quoted from the
third series of experiments check some of the magnitude
symbolisms independently obtained from the first and second.
This is true of most of the schedules in this set and is significant
because neither magnitude variations nor any other class of
variations in the responses had been suggested.

It is believed that studies of this type are of value in
showing the tendency of symbolisms to constellate in accord-

Three: Descriptive and llistonenl I in^uistics 241

239

ance with an unconscious or intuitive logic which is not
necessarily based on experience with the stimuli in their
normal, functional aspect. In the realm of articulate sounds
to take a specific type of perceptive field, it is believed that
the experiments here referred to give cumulative evidence for
the belief that unsocialized symbolisms tend to work them-
selves out rather definitely, and that the influence of specific,
functional language factors need not be invoked to explain
these symbolisms.

(Manuscript received September 12, 1928)

Editorial Note

Journal of Experimental Psychology 12 (1929), 225-239. [Reprinted in: I-dward
^2li^\x, Selected Writings in Language, Culture, and Personality. Edited bv Da\id (J.
Mandelbaum. Berkeley: University of California Press, 1949, pp. 61-72]

Section Four

The Problem of an International Ai'xii.iakv
Language (1925-1933)

Inlroduclion:
The Problem of an Inlernalional /XlimImia Laiimumc

The texts in this section cover a reiali\cly short pciUKl ( \^)2> \^>\^) m s.jpir s
scholarly career, during which he puhhshed on the problem of the choice and the
construction of an inlcrnalit)nal auxiliary language. In 1^;25 Sapir met Mrs Alice
Morris, the driving force behind the International Auxiiiarv Language Association
(I ALA). Morris was interested in ct)n\incing leading linguists —and she uas
indeed successful in obtaining the commitment of Otto Jespersen and I dsvard
Sapir — to support the movement and to contribute to the conceptual founda-
tions of the project. Sapir's extensive background in general linguistics and lan-
guage typology made him the ideal person for exploring the possibilii) ot a univer-
sally valid, minimal grammar,' which would also be psychologically well groun-
ded. Sapir's work on grammatical processes and on grammatical concepts, as
well as his growing interest in Chinese,' were to be major assets in the undertak-
ing.

It seems that Sapir lost no time in writing a "Memorandum on the Problem of
an International Auxiliary Language": in a letter of March 26, \*^)2> Alice Morris
approved the substance of a first draft, and recommended Sapir to get signatures
from other linguists. Eventually the paper appeared with the signatures of Sapir.
Bloomfield, Boas, Gerig and Krapp. The "Memorandum" appeared in volume
XVI of The Romanic Review, a journal published by Columbia University Press.
and edited by John L. Gerig, who was one of the co-signers. Iliis paper — clearK
written by Sapir alone' — starts out from the thesis that linguistics has an auto-
nomous status,'' but that linguists can serve the practical goals of an international
language project, especially since the adherents to the international language

' See now A Linguist's Life. An English iranslaiion of Olio Jespersen s autobiography with noir\ pholos and a
bibliography. Edited by Arne Juul. Hans F. Nielsen and Jorgen t:rik Nielsen (Odensc. IW.S). csp. pp. 220-222,
225-226 concerning Jespersens involvement with the I ALA.

2 As noted by Regna Darnell, in her book Edward Sapir: Linguist, Anthropologist, Humanul (Berkeley. I*W»)).

p. 272, Morris was interested in Sapir's idea of constructing a kind of universal prammar or "world

which would contain the essential formal framework allowing for the expression of even kind
Darnell qut)tes from the correspondence between Sapir and Morris, fri>m which it appear^ ■
was thinking of adopting categories from very diverse languages, and (b) thai he felt that the i

guage should be "simple, natural. Ilexible, self-creative, and incidentally, logical sMih .i minim.

machinery ";cf. the ideas put forward in the "Memorandum".

-^ E. Sapir. Language (New York, 1^21 ), pp. .59-85 and Sh-126.

■* See "The History and Varieties of Human Speech" (1911) (reprinted in section I '
Language, ac. pp.'ftfi, 7(),7.\8().8.V,S4,96 97. 101. 1 IK-1 19. 134 \M\ l.SO, 1.54 155.:'
two publications with Hsii Isan Hwa in Journal of Ameruan Lolklore M^ (1923).
Tales') and 31-35 ("Humor o{ the Chinese F'olk") [reprinted in ///«• Collected Wor-
pp. 799-809].

■^ See the intratextual references to "the writer" (p. 244. p. 245).

^ See also the papers "The Grammarian and his l^inguage" (1924) and "The Slalus of UngunUo u • Sacoce"

(1929) [reprinted here in section III|.

246 General Linguistics I

movement are not sufficiently acquainted with linguistic diversity. The paper
testifies to Sapir's intimate knowledge of American Indian languages (including
the Chinook jargon, a trade language^), and his familiarity with the structure of
Chinese.' The paper also reflects Sapir's attachment to concepts" and techniques
used in his Language (1921): the notion of grammatical concepts (p. 247),'" the
distinction between factual concepts and relational concepts (p. 250)," and the
notation used for derivational concepts (p. 248).''

The general principles put forward in the "Memorandum" are that the (desi-
red) international language should be characterized by simplicity, economy of
categories, and flexibility. More concretely, the first part '^ "General principles"
specifies that the international auxihary language should have an "accessible"
phonetic and grammatical structure, psychological (conceptual) simplicity, and
should be easily convertible into the world's major languages (English, French,
German, Japanese, Chinese are mentioned, p. 255), as well as be made suited for
secondary transpositions such as writing and radio transmission. In the second
part of the paper, some applications of the general principles are outlined: avoid-
ance of suprasegmental complexities, such as tones and length, absence of
inflection, and construction of a unified vocabulary (p. 252, with the suggestion to
base the vocabulary on Peano's Latino sine flexione).^^ Conversion to major
extant languages will be achieved if the international language is maximally ana-

^ Chinook jargon (Chinook Pidgin or Chinook Wawa) is a trade language used in the Northwest of the United
States and in British Columbia; it is based on a simplification of the phonological and grammatical structure of
Chinook; see Allan R. Taylor, "Indian lingua francas", in Charles A. Ferguson - Shirley Brice Heath (eds.).
Language in the USA (Cambridge/New York, 1981), pp. 175-199, and Sarah Grey Thomason - Terrence
Kaufman, Language Contact, Creolization, and Genetic Linguistics (Berkeley, 1988), pp. 256-263.

** See "Memorandum on the Problem of an International Auxiliary Language", pp. 244, 248, 249 ("We are like-
ly to find that it is helped, rather than hindered, by the unassuming simplicity of such languages as Chinese.
Much of our seeming subtlety in expression is really verbiage"; cf. Language, o.c.,p. 102: "An intelligent and sen-
sitive Chinaman, accustomed as he is to cut to the very bone of linguistic form, might well say of the Latin sen-
tence, 'How pedantically imaginative !' "), and p. 250.

'^ Note also the expression "grooves of thought" (p. 249), which Sapir used in his Language (New York, 1921),
p. 232 (and compare also p. 14 there).

'0 See Language, o.c, pp. 86-94, 104-105, 109-113.

1' See Language, o.c, pp. 86-87, 89-93, 98-102, 106-107 (instead of "factual concepts," Sapir uses there the term
"concrete concepts"; "relational concepts" are subdivided into "concrete relational" and "pure relational con-
cepts").

'2 See Language, o.c, pp. 87-88, 92, 106, 109-1 1 1 .

'-* The text is divided into four parts; "General Principles", "Certain Applications of the General Principles".
"Suggestions for Research", "Affiliation with Scientific Bodies".

1'* See p. 252; "In a wider historical sense too Latino sine Flexione has a great advantage. It is worth remember-
ing that Latin has a practically unbroken history as the international language of West European civilization.
Of late centuries this tradition has become rather threadbare but it has never died out completely. The various
proposals submitted in this memorandum are perhaps best synthetized by taking Peano's Latino as a basis and
simplifying it still further in the direction of a thoroughly analytic language, minimizing, so far as possible, the
use of derivational suffixes." The Italian scholar Giuseppe Peano [1858-1932], professor of mathematics at the
University of Turin, had constructed the "interlanguage" Latino sine flexione, which was based on Latin(ate),
and. to a lesser extent, Germanic and Slavic lexical bases.

Four: The Problem oj an hiicnuuionul Ai(\iluir\ I .iinf^utiiie 247

lytical. and iniiiinialK dcri\ alioiial ( to a\oKl dcrivalii)n Sapir suggests two slratc-
gies: compound Icxilicalion and lexical concrclization).

Sapir's major concern in Ihc "Memorandum" is with the ps\cti()h)gic.il ac\- p
lability o\ the international lanuuage that has to be consirucled; he iherel"!^
deems it necessary to jirepare the project throuuh acoustic-arliculalory" research
and psycholinguistic research, which WDuld anticipate possible psychological
resistance and rejection.

The "Memorandum" paper appeared in \^)2^. but does not seem to have arou-
sed widespread international interest. It seems that the ideas put lorward in
the paper could not convince scholars of the (urgent) need ol an international
language: most probably, the principles outlined were telt to be generally st)und.
but extremely abstract. Also, one should not torget that the movement for an
international auxiliary language had by then a rather long tradition'' — in Wll
the Belgian scholar Jules Meysmans had coined the term "interlinguistics" as a
designation for this field of applied research"* — , and that most of the ideas put
forward in the "Memorandum" had already been expressed by major liuropean
linguists, such as Hugo Schuchardt'' and Jan Baudouin de Courtenay.'"

'^ In his Language, o.c, pp. 207-210. Sapir had already pointed out the importance of psychological resistance
to borrowing (of words) from other languages: such psychological resistance would a fortiori apply m the case
of the adoption of a new language.

'^ In 1939 Nikolaj S. Trubetzkoy devoted a tiy now classic article to the problem of constructing a ph>':
ically acceptable international language: "Wie soil das Lautsystem einer kunstlichen internaii<>:
Hilfssprache beschaffen sein '.'". Iravuiix dit Cercle lingiiisiiquc de Prague 2 (19.^9). .s-21.

'^ In 1860 August-'ITieodor von Grimm wriUe a foundatii)n-laying "t'ri)gramm /ur Bildung einer allgcmemen
Sprache". published in Die Weltsprache (Bamberg. 1S.S7). pp. 8-1.^: in l.Sh.l (imsto ik-ilavitis followed up with
his book Pensieri sopra una lingua universale e su ahiini argonienii analoglu (Venice. ISh.^). lietween IS'"
!<S90 numerous books and articles appeared on theoretical and practical aspects of the problem of an ir
tional language, and various auxiliary languages were developed (some of which became ver\ successlu
as V\)lapuk and Esperanto). In l<S89 the first catalogue of "interlinguistic" literature was published (A
iiber die Sonderstellung der welfsprachlichen l.iurauir in den Rdumen de\ HiuligeMerhemu.seiinn. l.eip/iL'
The best surveys on the early period of the international language movement are I.ouis Couturat I
Leau. Hisuiire de la langue universelle (Paris. 1903. second ed. 1907) (reprinted: Mildeshemi Ne» > •
and Werner Fraustiidter. Die iniernaiionale Hilfssprache. Line kurze iieschuhie der Weltsprach H,
(Husum. 1910); for later developments see Albert L. Ciuerard. A Short History of the International l.ar.
Movement (London. 1922). Henry Jacob. A Plantu-d Auxiliary Language (London. 1947). Ric Rercer //
del lingua internatioiuil (Morges, 1972, 2 vols.). Alessandrt) Bausani. Le lingue invcntaie. l.ir
Linguaggi secreii. Linguaggi universali (Rome, 1974) and the useful anthology of texts ed::
Uaupenthal. Planspraehen (Darmstadt. 1976). which contains the Clerman translation (pp. UJ 147) ul the
"Memorandum."

"* See Jules Meysmans, "Une science nouvelle". Lingua iniernaiionale I ( 191 1-12). 14-16. On present U.is p '
spectives of interlinguistics, see the various contributions in Klaus Schubert Dan Maxwell (cJv i.
Interlinguistics. Aspects of the Science of Planned Languages (Berlin/New ^■i>rk. 19S9)

'*' See especially the following publications by Hugo Schuchardi: Weltsprache and \S>

Meyer (Strassburg. l.S94):"Die Wahl einer ("lemeinsprache". Heilage :ur .Mlgeniemen /-

"Bericht iiber die auf Schaffung einer kunstlichen mternalionalen Hilfssprache genthlctc \-

Almanach der Akademie der ivissenschaften in Wien (1904). 2SI 296; "/.ur F-'raec drr i

Gemeinsprache ". Heilage zur Allgemeinen /.eitiing 107 (1907), 2.S9-261. On Schuchardi

for an international auxiliarv language, and on his extensive corresp«indence wiih interli

ing articles by Pierre Swiggers and Herman Seldeslachts: '-line lettre de 1 e\ ."sOerha a M

pos de la creation dune iangue auxiliaire Internationale". Orhis M< ( I99.S). 21.'» 22.V '/u ^

der interlinguistischen Bewegung: Das Zeugnis Heinrich von Mans/ynys ". Orhu W { I99.S). 224-12^. Unt kn-

248 General Linguistics I

In the early 1930s there was a upsurge of interest in the question of an inter-
national auxiliary language, as can be seen from the rapid succession of major
publications by O. Jespersenr' D. Szilagyir and W.E. Collinson.-' Perhaps the best
testimony is the fact that the question was put on the program of the second
international conference of hnguists, held in Geneva, in August 1931. For the
section devoted to the construction of an international language, Sapir sent in
a paper —in fact a condensed version of a longer manuscript''—, which was
published in 1933, in the proceedings of the conference.'' In "The Case for [a]
Constructed International Language,"'' Sapir adopts a more propagandistic view
than in the 1925 "Memorandum." This explains the emphasis laid on the practical
and intellectual (broadly humanistic and cognitive) advantages of an interna-
tional language, the strong rebuttal of three criticisms generally formulated against

re de Paul Chappellier a Hugo Schuchardt a propos de la creation d'une langue auxiliaire Internationale", Orbis
39 (1996-97). 163-166; "Philosophe et linguiste devant le choix d'une langue Internationale: Albert Schinz et
Hugo Schuchardt". Orbis 38 (1996-97), 167-173; "Die Kontakte zwischen Josef Weisbart und Hugo Schuchardt
hinsichtlich der Plansprachenproblematik", Orbis 39 (1996-97), 175-179; "Zu Couturats und Schuchardts
Beschiiftigung mit der Frage einer internationalen Hilfssprache", Orbis 40 (1998), 179-184; "Die
Auseinanderselzung zwischen Albert Ludwig und Hugo Schuchardt hinsichtlich der Schaffung einer kunstli-
chen internationalen Hilfssprache", Orbis 40 (1998), 185-190; "Schuchardts Beschaftigung mit dem Volapuk:
ein Zeugnis aus dem Briefwechsel", Orbis 40 (1998), 191-195. Schuchardt, in his writings on an international
auxiliary language, criticizes some misconceptions concerning the function of an international language and
refutes the organicist reactions to it (viz. the criticism of the artificial and non-natural character of an interna-
tional auxiliary language).

-•^ See especially Jan Baudouin de Courtenay, "Zur Kritik der kunstlichen Weltsprachen", Annalen der
Naturphilosophie 6 (1907), 385-433; this text is a masterful reply to the brochure published by the
Neogrammarians Karl Brugmann and August Leskien, Zur Kritik der kunstlichen Weltsprachen (Strassburg,
1907).

-' Otto Jespersen, "A New Science: Interlinguistics", Psyche 11 (1930-31), 57-67, reprinted (under the title
"Interlinguistics") in Herbert N. Shenton - Edward Sapir - Otto Jespersen, International Communication. A
symposium on the language problem (London, 1931), 95-120 [see also note 29]. Two years earlier Jespersen had
published his booklet An International Language (New York, 1928).

22 Denes Szilagyi, "Versus interlinguistica", Schola et Vita 6 (1931), 97-120.

23 William Edward Collinson, "International Languages", The Year's Work in Modern Language Studies 2
(1932), XI-XVIII. William Edward Collinson [1889-1969], professor of Germanic philology at the University of
Liverpool, published extensively on the problem of an international language; see, e.g., his books Esperanto and
Its Critics (Edinburgh, 1924) and La homa lingvo (Berlin, 1927) and his articles "Discussion: The case for
Esperanto", Modern Languages 13 (1931-32), 109-112 and "The Structure of Esperanto compared with that of
Some National Languages", Transactions of the Philological Society (1931-32), 77-79. He also participated with
Sapir in the lALA-sponsored research on semantic categories, publishing a work (with Alice V. Morris) on
Indication (Baltimore, 1937), and commenting on Sapir's work on "Totality" and "Grading" [see section V].

2"* An 8-page typescript, of which a carbon copy is in Philip Sapir's archives, contains a somewhat more exten-
sive text than the one published in the Actes. The typescript bears the title "The Case for a constructed inter-
national language," and a handwritten subtitle "Resume". Within the text there are handwritten deletions, cor-
rections (mainly of typographical errors) and a restricted number of changes and additions, all in Edward
Sapir's hand. Of another typescript version (of 3 1/4 page) a carbon copy also survives; this version was sent to
Alice v. Morris. Sapir sent a corrected version of the "Resume" text to Albert Sechehaye, the secretary of the
second international conference of linguists.

25 Actes du Deuxieme Congres international de linguistes, Geneve 25 - 29 aoiit 1931 (Paris, 1933).

26 Apparently the proofs of the published contribution were not (re)read by Sapir: the typescript versions have
the indefinite article in their title ("The Case for a ...") and have the correct form "particularly" towards the end
of the text (here "particulary"). Also, the printed 1933 text ends with a comma, a clear misprint for a period.

Four: The Prohlc/n of an /nicniiiiionul Auxilmrv l.ani^ua^e 249

international language(s),and the stronu rejection dI an •Drganicisl" view on lan-
guage.-' This 1931 "resume" offers few concrete proposals, since Sapir limits his
considerations to the necessity of the conslriielion of a -'highlv efficient and maxi-
mally simple international language," based on a calibrated "slock of words" and
"grammatical techniques."-** From this short paper we can retain, however, the
definition of the international language project as an attempt to ensure transna-
tional communication on spheres of interest which have universal significance: its
goal is to elaborate a consciously regularized and adaptable svstem (m contrast
with the unconscious nature, the "local" integration and the more personallv and
societally bound domains of a "mother tongue").

In 1931 Sapir published two other papers on the question of an international
auxiliary language. The paper published in Psyche-" under the title * llie 1- unction
of an International Auxiliary Language" was written for a scholarly audience,*
and testifies to the increasing need for, as well as to the growing scientific inter-
est in an international auxiliary language (p. 1 10 and p. 121 ). Sapir points out two
main directions, viz. the use of a constructed language, or the adoption of (a sim-
plified form of) an established language, and suggests that the "modern world"
may need a full-fledged constructed international language. What Sapir does in
this paper, is the following:

(1) First, he discusses the general requirements which an international auxiliary
language must satisfy: it should be analytic, simple, regular, but also creati\e. re-
fined, and adapted to the modern mind: "What is needed above all is a language
that is as simple, as regular, as logical, as rich, and as creative as possible: a lan-
guage which starts with a minimum of demands on the learning capacity of the
normal individual and can do the maximum amount of work: w hich is to ser\ e as
a sort of logical touchstone to all national languages and as the standard medium
of translation. It must, ideally, be as superior to any accepted language as the
mathematical method of expressing quantities and relations between quantities
is to the more lumbering methods of expressing these quantities and relations in
verbal form. This is undoubtedly an ideal which can never be reached, but ideals
are not meant to be reached: they merely indicate the direction of mo\enient"( p.
113):

-^ For a similar criticism of the organicist view, see Hugo Schuctiardt. "licricht ubcr die aiif Schaffiing cincr
kiinstlichen intcrnationalen Hiifssprache gerichlete Bevvegung",fl.f. (see note l')|.

-^ Here the term is used as a synonym for grammatical "processes" in general: in /.«//iv"</.c«' (Nov* \oil. \^2\)
Sapir used the term technique to refer to the grammatical processes relatmg to the ways of combining (i»r of nui
combining) a cxincrete concept with a relational concept (this includes the range from isolahon lo agglulinaliiHi
and fusion, with possible "symbolic" expression).

29 The papers published by Herbert N. Shenton in /•v\(7i<- 1 1 :1 ( 1930-31). 6-20. by Otto Jcspcrscn in P\vche\\3
(193(K31), 57-67 and bv Edward Sapir in Psyche 1 1:4 ( 1930-31 ). 3 I. S were also jointly published in Kxik-form
(see the reference in note 21). Sapirs paper is reprinted here after the version published in the Srlcttni Mn/mift
(1949).

■^" In this paper (as well as in its shortened version "Wanted: a World 1 ancuage") Sapir us<s ih.- ictm murim
guist(s)."

250 General Linguistics I

(2) lb counteract negative reactions (based on nationalism and intellectual or
affective myopia) to the international language movement, he then proceeds to
an exercise in dcmystification, showing the illusions and the false ideas one
has/can have about one's native language and pointing out that behind the appar-
ent simplicity, there often lies great complexity." Sapir's discussion of asymmet-
ries in English- and in French'- foreshadows B.L. Whorf s work on ''overt" and
"covert" ' categories (incidentally, Sapir uses the terms "overt form," "overt sim-
pHcity" and "to cover up");

(3) Finally, Sapir refutes the label of "inferiority" which is erroneously attached
to constructed languages, and he shows the logical and psychological advantages
of a constructed language, and insists on the (intellectual and linguistic) freedom
allowed by it. The paper ends with a strong plea for an open and liberal human-
istic education.

The paper "Wanted: a World Language," '' published the same year in The
American Mercury, is aimed at a large audience; it is based'" on the longer paper
pubhshed in Psyche, from which various passages are reproduced (including the
more technical discussion of English formal categories).

Pierre Swiggers

31 Sapir speaks of "a perfect hornet's nest of bizarre and arbitrary usages."

32 Sapir discusses two cases: noun-derivation (zero-derivation, derivation with -ing. with -th. or with a Latinate
formation, such as obedience) and verb phrases with put or get.

33 For French Sapir takes the case of the multiple values of the reflexive voice.

^■^ See Benjamin Lee Whorf, Language, Thought, and Reality. Selected Writings of Benjamin Lee Whorf. Edited
and with an introduction by John B. Carroll (Cambridge [Mass.], 1956), esp. pp. 69-70, 88-89, 113, 132. On
Whorfs views on language and the categorization of experience, see Penny Lee. The Whorf Theory Complex: a
Critical Reconstruction (Amsterdam/Philadelphia, 1996).

35 The title was reused in 1969 by Mario Pei for his brochure Wanted: a World Language (New York. 1969).

3'' Apart from a rather different paragraphing and the deletion of a few repetitive sentences, the version publish-
ed m The American Mercury differs from the one published in Psyche (and its 1931 and 1949 reprints) by having
five longer passages deleted from it: the second paragraph on the purpose of the paper "The Function of
an International Auxiliary Language," part of the more "technical" paragraph on temporal expressions, part of
the paragraph on symbolic systems (such as used in mathematics and symbolic logic), the passage on the
"Chinaman's and the Indian's indifference to the vested interests of Europe," and part of the last but one
paragraph, where Sapir warns us against the danger of international language doctrinarism.

MEMORANDUM ON THE PROBLEM OF AN INTER-
NATIONAL AUXILIARY LANGUAGE

THE following memorandum is offered from the point of view
of one who is greatly interested in linguistic study on its
own account. Like many, perhaps most, linguistic students he
has until recently been only moderately interested, if at all, in
the various proposals that have been made from time to time to
create an International Language that might be used alongside
the many national languages already in use. Within the last
year or two, however, the increasingly practical nature of the
problem has been borne in upon him as well as the reasonable
possibility of its solution. It has seemed to him, however, that
a wider acquaintance with linguistic phenomena than most of
those who are interested in the International Language move-
ments presumably possess would have enabled them to evolve
far simpler and more readily acquired auxiliary languages than
those which have actually been proposed. The writer is par-
ticularly Impressed by the needless adherence to the irrelevances
and elegances of our western European languages and he wonders
why a language like Chinese, which has produced a poetic and
philosophic literature of the greatest subtlety, can do without
cases, modes, tenses, and a complex system of derivations when
an international language like Esperanto, which is supposed to
be a carefully thought out and ideally simple means of communi-
cation. Indulges In all kinds of linguistic luxuries. In the
following remarks the attempt has been made to cut to the bone
of what Is necessary in practical communication that does not
aim to ape the literary graces of English or French. The needs
of aliens who have not grown up in our Occidental civilization
are particularly borne In mind. It seems not unreasonable to
proceed on the assumption that it is worth while to consider these
needs and to try to learn something from the structure of
languages simpler than Italian or Spanish or any of their Inter-
national derivatives.

244

252 General Linguistics I

245

But the writer feels strongly that ruthless simplicity is not
the only thing to consider. A great deal of useful energy has
already been expended on International Language work and this
energy and its results must be utilized. Moreover, the move-
ment is mainly in the hands of the Occidental world, and it is
very possible that a maximum of theoretical simplicity would
present certain unforeseen psychological difficulties. The writer
is very far indeed from wishing to put forward radically new
proposals. They would be utterly futile. Yet he hopes that
some of the points raised in this memorandum may assist in
simplifying the International Language problem, whatever basis
(Esperanto, Ido, Latino sine Flexione, or other) be ultimately
adopted, and make clear the need for experimental research
before responsible bodies commit themselves to any one form of
International Language.

The memorandum is divided into four parts: A, General
Principles; B, Certain Applications of the General Principles;
C, Suggestions for Research; D, Affiliation with Scientific
Bodies.

A. General Principles

An International Auxiliary Language should have the
following characteristics:

1. It should have no sounds (vowels and consonants) that
cause serious difficulty to large bodies of speakers.

2. It should have the simplest grammatical structure that is
consonant with effectiveness. It should not merely have a
structure that is theoretically simple, logical, and regular, but
that is most easily assimilated, on psychological grounds, by the
greatest number of diverse peoples.

3. It should be so constructed as to be readily convertible
into any of the major languages now in use. And, conversely,
it should be able to render the essential meaning, without danger
of ambiguity, of a text composed in any of these languages.

4. It should have considerable flexibility of structure, so that
any speaker may not too greatly impair its intelligibility if he
bends it involuntarily to constructions familiar to him in his
own language. There should be some opportunity for alter-

Four: The rrohlcni of an Inicrndiionul Ainiliurv I Ainj^iuii^f 253

246

nation of expression, such as the option of usinp or not using
elements indicating certain concepts, like numl)er or tense.

5. It should be built as far as possible out of materials that
are familiar to the speakers of West European languages.

Of lesser importance, yet worthy of consideration, are the
following principles:

6. The International Language should, so far as possible, be
a logical development of international linguistic habits that have
been formed in the past.

7. It should be capable of expression in shorthand with the
utmost ease.

8. Its phonetic system should be such as to make it intelli-
gible with a minimum of ambiguity on the telephone, phono-
graph, and by radio.

B. Certain Applications of the General Principles

The numeration corresponds to A.

I. (a) There are certain consonants which should be avoided

because found in relatively few languages, such as th of English

thin and th of English then, ch of German Bach and ch of German

ich. But this is not all. There are certain consonants which,

while not exactly uncommon, are not generally found or are often

not found if certain other consonants that resemble them are in

existence in the language. Thus, neither w nor v of English

wine and vine is an uncommon sound, but there are many

languages which do not possess both. Hence to recognize both

w and V is to invite confusion in such languages. Either w or v

should be explicitly recognized and the other considered an

alternative pronunciation. Thus, if w is adopted, Poles and

Italians may legitimately pronounce it r; if r is adopted,

Chinese and Arabic speakers may legitimately pronounce it w.

Had this principle been adopted in Esperanto, we should not

have both s and }, nor both c and c; 5 and s, for instance, would

have been considered as merely variant pronunciations of one

sibilant consonant. In fact, the whole group of sibilants— 5, },

c, c — might with advantage have been reduced to one, s. There

is no doubt that a careful survey of the whole phonetic field

would suggest a simplified consonant system that would make

254 General Linguistics 1

247

the learning of Esperanto, Ido, or other International Language
very much easier.

(6) Consonant combinations should be avoided as far as
possible, but this principle would probably need to be checked
considerably by other considerations. By simplifying too much,
we might in many cases lose the very real advantages of im-
mediate recognition of words and of historical continuity.

(c) The vocalic system should be cut down to a minimum.
The series a, e, i, 0, u is plenty; a, t, u alone would have very
distinct ethnic and acoustic advantages, but would probably so
distort the appearance of words as to introduce new difficulties.
No prosodic peculiarities, such as differences between long and
short vowels, differences of stress accent, or differences of tone,
should be recognized, as the habits of different languages are too
various and inflexible on such points as these.

2. (a) There seems to be no need to insist on the specific
expression of certain grammatical concepts that most of us are
accustomed to. The usefulness of tense distinctions is greatly
overestimated, for instance. Even in English there is no gram-
matical difference between present and future in cases like "I'm
working" and "I'm working tomorrow." Word particles can
always be appended if it is necessary to convey the idea of tense.
Such complexities as the three Esperanto tenses with their
symbolic vowels and attached participles are quite uncalled for,
add nothing to clarity of thought, make for pedantry in ex-
pression, and greatly increase the difficulty of learning the
language. In general, neat symbolisms of expression are more
attractive on paper than they are either necessary or desirable
in practice. What applies to tense applies also to gender, case,
mode, probably number, and several other categories. "Yester-
day he kill several cow" is quite as adequate as "yesterday
he killed several cows." Certain rules of order of words, impli-
cation as to concepts not definitely expressed, and optional use
of "empty" words to define case relation, tense, and other gram-
matical ideas, could be very easily worked out and would prove
astonishingly effective. There are many cases where "ambi-
guity" is a real advantage. "He kill man" might be looked
upon as a blanket statement for "he kills a (the) man," "he

Four: The /'rohlcni of dn Inicrnaiiomil Au\iliar\ I uniiua^v 255

248

kills (the) men," "lie killed . . . ," and so on, precisely as in
Chinese. In real usage it is most instructive to sec how little
ambiguity such bare and simple propositions contain, because
there is always a context. Moreover, we are often driven to
greater defmiteness of expression than we are actually aiming at;
we are the dupes of our forms. For instance, in a legal clause
like "Any person or persons who has or have knowledge of, or
who has or have had knowledge of, such act or acts" there is
obviously a labored attempt to express the generalized "Person
who know (about) such act." Where number and tense have to
be particularized, one can always add qualifying elements, thus,
"One person who know such act," "Several person who did
know such several act," and soon. (Of course "one," "several,"
and "did " are here used merely as approximate English counters
for whatever appropriate terms may actually be adopted.) In
short, the ideal of effective simplicity is attained by a completely
analytic language, one in which the whole machinery of formal
grammar is reduced to carefully defined word order and to the
optional use of "empty" independent words (like "several,"
"did," "of"). Inflection is reduced to zero. This is the ideal
that English has been slowly evolving towards for centuries and
that Chinese attained many centuries ago after passing through
a more synthetic prehistoric phase. (The simplicity of Chinese
grammar is not a primitive trait, but is at the polar extreme
from "primitiveness.")

{h) In the expression of derivative ideas (place, instrument,
adjectival, and many others) there is room for great simplifi-
cation. The international languages that have been suggested
seem to make it a matter of pride to have a great many deriving
affixes and to luxuriate in the endless possibilities of coining new
words, whenever wanted, by means of the derivational appa-
ratus. A far better economy of material would seem to demand
that derivation be either eliminated or reduced to a minimum.
Psychologically, it is quite false to imagine that the memorizing
of a series of derivative words of type A -f x, B + x, C -f x,
D -f x {e.g., bak-er, farm-er, cutt-er, press-cr) reduces to the
memorizing of the root words A, B, C, D {e.g., bake, farm, cut,
press) plus the memorizing of a deriving affix x {e.g., -er) of given

756 General Linguistics I

249

function {e.g., "one who . . .")• As a matter of fact, such
derivative words have to be learned as units, though the memo-
rizing- of them is naturally less laborious than of words absolutely
unrelated to words already mastered. There seem to be two
ways of simplifying the problem of derivative formations. One
is to compound independent words, e.g., bake man or bake person
for baker, cold time for winter, make strong for strengthen, more old
for older. In a sense such compound expressions have to be
learned as units too, but there seems to be a very real psycho-
logical advantage in having every element in the language
independently expressive. The speakers and readers of such a
language come to feel that in a comparatively short time they
have memorized everything there is to know and that they have
a free, creative use of the language after that. The second
method is intertwined with the first. It consists in a simplifi-
cation of the form pattern. A rigorous thinking out of the true
content of a sentence as contrasted with its purely formal
convolutions often reveals the humiliating fact that it could have
been expressed with half the apparatus. Abstract nouns, in
particular, are not nearly so useful or necessary as is generally
assumed. There should, perhaps, be some provision for their
formation, perhaps by means of some indefinite noun like thing
or matter or way {e.g., wise way for wisdom), but the real point is
that they can often be easily avoided, and with a gain in vivid-
ness. Thus, there is nothing in the sentence "The wisdom of
old age chills youth" which is not as adequately expressed, and
with a more intuitive impact, in such a sentence as "Wise old
person make cold (to) young person" (or "The wise old make
cold the young"). In other words, we must not too lightly
assume that the grooves of thought which we are accustomed to
in our European languages are the easiest or most natural in a
universal sense. It may be worth our while to get into the
habit of simplifying the pattern of our thought. We are likely
to find that it is helped, rather than hindered, by the unassuming
simplicity of such languages as Chinese. Much of our seeming
subtlety in expression is really verbiage.

Four. The Prohlcni of tin Itifcrtiniiomil Auxiluii \ I .ani^uii^t' 257

250

as is suggested here (a, h). This would not necessarily prevent
one from aiming towards the gradual realization of an analytic
ideal of linguistic expression. The main point at present would
seem to be to introduce the possibility of far greater flexibility of
individual expression. If, for instance, it gradually became
apparent that a more Chinese-like use of Esperanto or Ido or
Latino or Romanal, whether by Chinese or others, had certain
definite advantages, there is no reason why such use might not
gradually grow in favor at the expense of the uses already
standardized. It seems a mistake to legislate too rigidly at the
present time on points of grammar. Attention should be concen-
trated rather on the formation of a universally accepted mini-
mum vocabulary, sufficient for ordinary purposes.

3. If a language is too synthetic, translation from it or into
it is necessarily more difficult than if it is analytic in structure.
If two languages, one of which is to be translated into the other,
are very different in structure, each must be analyzed, con-
sciously or unconsciously, into the concepts, both factual and
relational, which are expressed in it, so that the equivalences of
the two languages may be discovered or constructed. In dealing
with a thoroughly analytic language this task of mutual ac-
commodation is appreciably lightened because the conceptual
analysis has been made by at least one of the two languages
itself. The more analytic a language is, the more easily does it
serve as a circulating medium for all others. It should be
carefully borne in mind that tests of the efficiency of Esperanto,
for instance, as an expressive equivalent of French or Spanish or
Italian or German do not really prove the adequacy of Esperanto
as a universal "circulating medium" for the simple reason that
Esperanto is modeled on these very languages. A Frenchman
or a Spaniard is heavily biased in its favor, in advance of any
knowledge he may have of it, where a Chinaman or Japanese or
other non-European is not nearly so greatly impressed by its
simplicity or its ready equivalence to his own language. Uni-
versal adequacy does not mean a readiness to provide word for
word translations of other languages, but simply ease in reflecting
their essential meaning.

4. The importance of grammatical flexibility or choice has

-,<^^ General Linguistics I

251

been shown in 2. We cannot hope to reduce the linguistic
psychology of all speakers to one level. Hence we need a
language of structural "lowest terms." Many people may feel
that a certain poverty results, but this poverty, if such it really
be, is likely to make for an increase of true mutual understanding.
It is remarkable what excellent work can be accomplished by so
unpretentious a lingua franca as the Chinook Jargon, which has
been, and to a large extent still is, used between the whites and
Indians of the Pacific Coast and between various Indian tribes
of this region that speak mutually unintelligible languages.
This Jargon, which has not a large basic vocabulary and is built
on strictly analytic lines, is not merely a trade language but has
developed such adroitness with its seemingly slender means that
long religious and political harangues can be and are delivered in
it. The Indians themselves, who speak perhaps the most
complexly synthetic languages that are to be found anywhere,
seem to have no notion that the Jargon is an "imperfect"
language but consider it a perfectly adequate medium for inter-
tribal communication. A vast part of our vocabulary is dedi-
cated to feeling rather than to meaning and is of no use for
scientific, business or other practical work.

5. There is no theoretical reason why an Auxiliary Inter-
national Language should not be made out of whole cloth, as it
were, but the practical advantages of using known material are
too obvious to be insisted upon. It is perhaps unfortunate that
Esperanto is built out of such historically diverse elements as
French, Latin, Greek, English, and German, though the history
of the English language is abundant testimony of the practical
possibilities of combining words of different origins into new
syntheses. There is a certain incongruity that results which
affects some people much more unpleasantly than others, and
undoubtedly this feature has done a gi'eat deal to prevent
Esperanto from spreading as rapidly as it might have. The
Romanal idea of a historically unified vocabulary is psycho-
logically sounder, because such a vocabulary canalizes easily
with systems of word associations that are widely prevalent.
Latino sine Flexione too is psychologically sounder than
Esperanto or Ido. It has all the advantages of these of being

Four: f'lir Problem of iin hiwrndiiomil Au\iliiir\ I iintiuage 259

252

built out of generally known materials and the important further
advantage of not forcing violently new associations. A vast
number of people have a fair smattering of the Latin vocabulary
but an imperfect memory of the rules of Latin grammar. A
language which capitalizes both this knowledge and this igno-
rance is really in a psychologically impregnable position.

6. In a wider historical sense too Latino sine Flexione has a
great advantage. It is worth remembering that Latin has a
practically unbroken history as the international language of
West European civilization. Of late centuries this tradition has
become rather threadbare but it has never died out completely.
The various proposals submitted in this memorandum are per-
haps best synthetized by taking Peano's Latino as a basis and
simplifying it still further in the direction of a thoroughly
analytic language, minimizing, so far as possible, the use of
derivational suffixes. One of the incidental advantages of
Latino sine Flexione is that it can serve as a useful stepping stone
towards the learning of Latin itself.

7. The requirements of a shorthand of maximum ease
emphasize once more the importance of a very simple scheme of
consonants and vowels. It is worth working for a stenographic
system that is so simple and transparent, so rapid even without
abbreviations, that the International Language can be directly
learned, written, and printed in it. This may ultimately prove
to be an important economic asset. If the phonetic system of
the International Language is simple enough, the labor of learning
and using a good shorthand system would be appreciably less
than that of learning and using longhand.

8. Experience seems to show that certain sound difTerenccs
that seem clear enough in ordinary speech tend to be minimized
or obscured in mechanical transmission. Examples are the
vowels e : i and : 11 and the consonant pairs p : b, I : d, k : g,
s : z, f : V. It might be found advantageous to level these pairs
of sounds and to consider them as one each. If one had the
option of constructing an ideal universal phonetic system, he
would probably limit himself to:

(a) 3 vowels — a, i, u.

(b) 8 consonants — p, t, k, s, I, m, ;?, v.

260 General Linguistics I

253

Such a system (built out of syllables of type a, t, u: pa, pi, pu:
la, li, hi; 9 x 3 or 27 basic syllables) would be absurdly easy to
learn to write in shorthand and would provide more than enough
basic vocables for even the most elaborate vocabulary. If we
limit ourselves to words of one, two, and three syllables, a simple
calculation shows that this system gives us the means of forming:

9X3= 27 (monosyllabic)
27 X 27 = 729 (disyllabic)
27 X 27 X 27 = 19,683 (trisyllabic)

20,439 basic words.

A language built out of such materials could be acoustically
perceived at once without the slightest real danger of ambiguity,
could be pronounced accurately at once (allowing for such
optional pronunciations as r for / and w for v) by every person on
the globe, and could be learnt as a shorthand orthography in an
hour by any person of normal intelligence. The great disad-
vantage of so simplified a system is, of course, that it would so
distort the Latino, say, or Romanal or Ido words as to impair the
historic usefulness of their vocabularies. But Oriental and other
exotic habits of speech might gradually suggest or even force a
compromise with it.

C. Suggestions for Research
Certain kinds of experimental work may now be suggested.
These are intended to substantiate or, possibly, disprove some of
the points made in the preceding paragraphs. The numeration
corresponds to that of A and B.

I. Experiments could be undertaken to test the relative ease
with which various sounds are heard and sound difTcrcnccs are
perceived. As many distinct nationalities as possible should be
represented. The test words should be nonsense words, so that
the helping or hindering influence of actual word associations
may be avoided. Another set of experiments would test the
ability of diff'erent nationalities to pronounce various sounds.
If it is found, for instance, that the acoustic and articulatory
distinction between 5 and z, or / and r, causes real embarrassment

Four: The Problem oj iin Inicrnuiiotuil .\u\iliiir\ l.ini^ua^e 261

to large and important populations, there is good reason to
eliminate the distinction in an international language or, if this
cannot be done, so to tinker with the vocabulary as to minimize
the danger of too many words occurring which differ only in such
sound distinctions as cause trouble.

2. Experiments could be undertaken to ascertain with what
ease people of various nationalities can learn to understand, in
writing and as spoken, a highly simplified recasting of their own
language along the analytical lines that we have laid down.
How readily, for instance, after the rules for the simplified form
of their language have been carefully explained, can Germans get
at the meaning of sentences like "er tat gebe zwei Pferd zu ich
{or mich) " for "er hat mir zwei Pferde gegeben"? Next, with
what ease can they learn to compose in such a broken-down form
of their language? These two sets of experiments would attempt
to discover how readily the average person can learn to think in
a completely analytic mould without complicating the problem
by the necessity of memorizing a stock of unfamiliar words.

After this, other sets of experiments could be designed to test
the ability of various people to learn to understand, in writing
and as spoken, and to compose in a constructed analytical
language based on Esperanto or Romanal or Latino sine
Flexione. Compare with their ability to do the same for
Esperanto or Ido or Latino sine Flexione as actually used.
Direct comparisons, however, should not be made after too brief
a period of experimentation, for a highly analytic language, built
on Chinese lines, is likely to be unconsciously resisted on
emotional grounds as "ridiculous" or "too childish" for a while.
After a short period of resistance, however, the advantages of
such a language are likely to sink in at a rapid rate.

3. After the more tentative experiments, chiefly with isolated
sentences, recommended in 2, more elaborate tests should be
made in translating from and to the suggested analytic language
(using native, Esperanto, and Latin material). Then compare
with similar translation experiments in actual Esperanto and
Latino sine Flexione. Esperantists and accomplished Latin
scholars are probably best excluded as subjects from these cxpx^ri-
ments. It would be worth while getting personal estimates from

95? General linguistics I

255

individuals of different nationalities as to the relative ease and
adequacy of translation in the different groups of cases, also some
indication of the emotional attitude (readiness to acceptance,
irrational dislike) of those experimented on.

4. Check or control experiments might be valuable. Selected
business or scientific texts in, say, English, French, German,
Japanese, and Chinese might be translated by those speaking
these languages into other accepted languages, into Esperanto
(by an Esperantist), into Ido (by an Idist), into Latino sine
Flexione, and into some form or forms of thoroughly analytic
languages. These translations could then be retranslated both
directly and also via a third language into their originals or a
third language by other individuals and compared with the
original texts to see if the essential meaning has not been lost in
the processes of translation.

5. It might be worth preparing a questionnaire intended to
throw light on the psychological attitude of different people
towards the question of an international language with a homo-
geneous or with a mixed vocabulary.

6. It might be worth making an effort to cooperate with
Peano to see if a universally satisfactory form of Latino sine
Flexione might not be agreed upon, in which simplification of the
language is pushed even further, as many as possible of the
derivatives being dispensed with.

7. One or more of the existing shorthand systems might be
adapted to various forms of International Auxiliary Language.
Speed, ease of writing, and legibility could be tested. In par-
ticular, it would be worth while finding out if a shorthand system,
when applied to a language of maximum phonetic simplicity,
could be made universally legible when applied with no more
than average care, instead of merely legible to the writer himself.

8. The telephone and radio people might be asked to test out
the possible advantages in transmission of an extremely simple
and unambiguous phonetic system as compared with the ones
used by Esperanto, Latino sine Flexione, or actually spoken
languages. Inasmuch as these experiments would be designed
to test the unambiguous transmissibility of sounds and sound
combinations as such rather than of languages, it might be

Four: The Prohlcni <>t nn Itiicrnittumal Aiixiliarx I.an^iui^e 263

256

advisable to use nonsense material built out of the respective
phonetic systems. If the highly simplified phonetic system sug-
gested in B proves to have very decided advantages from the
point of view of C 7, 8, it becomes a rather more than academic
matter whether or not the phonetic system of the International
Auxiliary Language be left as at present used in Esperanto or
Latino sine Flexione.

D. Affiliation with Scientific Bodies
It would be advantageous to have the auxiliary language
movement get into as close touch as possible with the various
scientific bodies that are Interested in linguistic research, so that
eventually they may give the movement active sympathy and
collective backing. No doubt many of the members of such
Societies are at present uninterested In the problem. But It
should be possible to get an Important nucleus of membership In
each Society interested, which may then draw the attention of
the Society as a whole to the Importance of the problem and
Invite general discussion.

Signed by Edward Sapir

Leonard Bloomfield
Franz Boas
John L. Gerig
George Philip Kr/\pp

LditDrial Note

/"//(' Ronumic Review Id (l^->25). 244 2.'Sh. |(icini;in tiaiishition in: Rcinhard
Haupciithal (Hrsg.), Plansprae/ien. Heiiriii^e :iir Iniirliniiuistik. Dannsladt:
Wisscnschaftlichc Biichizcscllschafl. l^)7h. j^p. 133 I47|

)h4 General Linguistics I

THE FUNCTION OF AN INTERNATIONAL
AUXILIARY LANGUAGE

As TO THE theoretical desirability of an international auxiliary lan-
guage there can be little difference of opinion. As to just what factors in
the solution of the problem should be allowed to weigh most heavily
there is room for every possible difference of opinion, and so it is not
surprising that interlinguists are far from having reached complete
agreement as to either method or content. So far as the advocates of a
constructed international language are concerned, it is rather to be
wondered at how much in common their proposals actually have, both
in vocabulary and in general spirit of procedure. The crucial differences
of opinion lie not so much between one constructed language and an-
other as between the idea of a constructed language and that of an
already well-established national one, whether in its traditional, au-
thorized form or in some simplified form of it. It is not uncommon to
hear it said by those who stand somewhat outside the international
language question that some such regular system as Esperanto is
theoretically desirable but that it is of little use to work for it because
English is already de facto the international language of modern times —
if not altogether at the moment, then in the immediate future — that
English is simple enough and regular enough to satisfy all practical
requirements, and that the precise form of it as an international lan-
guage may well be left to historical and psychological factors that one
need not worry about in advance. This point of view has a certain
pleasing plausibility about it but, like so many things that seem plausible
and effortless, it may none the less embody a number of fallacies.

It is the purpose of this paper to try to clarify the fundamental
question of what is to be expected of an international auxiliary lan-
guage, and whether the explicit and tacit requirements can be better
satisfied by a constructed language or by a national language, including
some simplified version of it. I believe that much of the difficulty in the
international language question lies precisely in lack of clarity as to
these fundamental functions.

There are two considerations, often intermingled in practice, which
arouse the thought of an international language. The first is the purely
practical problem of facilitating the growing need for international

[110]

Four I he riohlcm <>t iin Intcimiiional Auxiliarx Iiin^iiafie 265

111

communication in its most elementary sense. A firm, for instance, that
docs business in many countries of the world is driven to spend an
enormous amount of time, labour, and money in providing for transla-
tion services. From a i)uroly technological point of view, all this is
sheer waste, and while one accepts the necessity of going to all the
linguistic trouble that the expansion of trade demands, one docs so
with something like a shrug of the shoulder. One speaks of a 'necessary
evil.' Again, at an international scientific meeting one is invariably
dissappointed to find that the primary difficulty of communicating
with foreign scientists because of differences of language habits makes
it not so easy to exchange ideas of moment as one had fantasied might
be the case before setting sail. Here again one speaks of a 'necessary
evil,' and comforts oneself with the reflection that if the scientific
ideas which it was not too eas}' to follow at the meeting are of moment
they will, sooner or later, be presented in cold print, so that nothing
is essentially lost. One can always congratulate oneself on having
had an interesting time and on having made some charming personal
contacts. Such examples can, of course, be multiplied ad infinitum.
Too much is not made, as a rule, of any specific difficulty in linguistic
communication, but the cumulative effect of these difficulties is stu-
pendous in magnitude. Sooner or later one chafes and begins to wonder
whether the evil is as 'necessary' as tradition would have it. Impa-
tience translates itself into a desire to have something immediate done
about it all, and, as is generally the case with impatience, resolves
itself in the easiest way that lies ready to hand. Why not push English,
for instance, which is already spoken over a larger area than any other
language of modern times, and which shows every sign of spreading in
the world of commerce and travel? The consideration which gives rise
to reflections of this sort, grounded in impatience as it is, looks for no
more worthy solution of the difficulty than a sort of minimum language,
a lingua franca of the modern world. Those who argue in this spirit
invariably pride themselves on being 'practical,' and, like all 'practical'
people, they are apt to argue without their host.

The opposed consideration is not as easy to state and can be so stated
as to seem to be identical with the first. It should be put in something
like the following form: An international auxiliary language should
serve as a broad base for every type of international understanding,
which means, of course, in the last analysis, for every type of expression
of the human spirit which is of more than local interest, which in turn
can be restated so as to include any and all human interests. The ex-
igencies of trade or travel are from this point of view merely some of

266 General Liniiuistics I

112

the more obvious symptoms of the internationalizing of the human
mind, and it would be a mistake to expect too little of an organ of inter-
national expression. But this is not all. The modern mind tends to be
more and more critical and analytical in spirit, hence it must devise for
itself an engine of expression which is logically defensible at every point
and which tends to correspond to the rigorous spirit of modern science.
This does not mean that a constructed international language is ex-
pected to have the perfection of mathematical symbolism, but it must
be progressively felt as moving in that direction. Perhaps the speakers
of a national language are under profound illusions as to the logical
character of its structure. Perhaps they confuse the comfort of habit
with logical necessity. If this is so — and I do not see how it can be
seriously doubted that it is — it must mean that in the long run the
modern spirit will not rest satisfied with an international language that
merely extends the imperfections and provincialisms of one language
at the expense of all others.

These two opposing considerations seem to me to be the primary
ones. They may be rephrased as "what can be done right now" and
"what should be done in the long run," There are also other considera-
tions that are of importance, and among them perhaps the most ob-
vious is the attitude of people toward the spread or imposition of any
national language which is not their own. The psychology of a language
which, in one way or another, is imposed upon one because of factors
beyond one's control, is very different from the psychology of a lan-
guage that one accepts of one's free will. In a sense, every form of
expression is imposed upon one by social factors, one's own language
above all. But it is the thought or illusion of freedom that is the im-
portant thing, not the fact of it. The modern world is confronted by the
difficulty of reconciling internationalism with its persistent and tighten-
ing nationalisms. More and more, unsolicited gifts from without are
likely to be received with unconscious resentment. Only that can be
freely accepted which is in some sense a creation of all. A common
creation demands a common sacrifice, and perhaps not the least potent
argument in favour of a constructed international language is the
fact that it is equally foreign, or apparently so, to the traditions of all
nationalities. The common difficulty gives it an impersonal character
and silences the resentment that is born of rivalry. English, once ac-
cepted as an international language, is no more secure than French has
proved to be as the one and only accepted language of diplomacy or as
Latin has proved to be as the international language of science. Both
French and Latin are involved wuth nationalistic and religious implica-

Four: The Prohlcni of dn hilftnuiioniil Auxilinrv I iini^iiage Ittl

113

tions which could not be entirely shaken off, and so, while they seemed
for a long time to have solved the international languaKe problem up to
a certain point, they did not really do so in spirit. English would prob-
ably fare no better, and it is even likely that the tradition of trade,
finance, and superficial practicality in general that attaches to English
may, in the long run, prove more of a hindrance than a help to the
unreserved acceptance of English as an adequate means of international
expression. One must beware of an over-emphasis on the word 'aux-
iliary.' It is pcrfecti}' true that for untold generations to come an inter-
national language must be auxiliary, must not attempt to set itself up
against the man}' languages of the folk, but it must for all that be a free
powerful expression of its own, capable of all work that may reasonably
be expected of language and protected by the powerful negative fact
that it cannot be interpreted as the symbol of any localism or nation-
ality.

Whether or not some national language, say, English, or a constructed
language, say Esperanto, is to win out in the immediate future, does not
depend primarily on conscious forces that can be manipulated, Init on
many obscure and impersonal political, economic and social deter-
minants. One can only hope that one senses the more significant of these
determinants and helps along with such efforts as one can master. Even
if it be assumed for the sake of argument that English is to spread as an
auxiliary language over the whole world, it does not in the least follow
that the international language problem is disposed of. English, or
some simplified version of it, may spread for certain immediate and
practical purposes, yet the deeper needs of the modern world may not
be satisfied by it and we may still have to deal with a conflict between
an English that has won a too easy triumph and a constructed language
that has such obvious advantages of structure that it may gradually
displace its national rival.

What is needed above all is a language that is as simple, as regular,
as logical, as rich, and as creatiA'e as possible; a language which starts
with a minimum of demands on the learning capacity of the normal
individual and can do the maximum amount of work; which is to serve
as a sort of logical touchstone to all national languages and as the
standard medium of translation. It must, ideally, be as superior to any
accepted language as the mathematical method of expressing quantities
and relations between quantities is to the more lumbering methods of
expressing these quantities and relations in verbal form. This is un-
doubtedly an ideal which can never be reached, but ideals are not meant
to be reached: they merely indicate the direction of movement.

268 General Linguistics I

114

I spoke before about the illusions that the average man has about
the nature of his own language. It will help to clarify matters if we
take a look at English from the standpoint of simplicity, regularity,
logic, richness, and creativeness. We may begin with simplicity. It is
true that English is not as complex in its formal structure as is German
or Latin, but this does not dispose of the matter. The fact that a be-
ginner in English has not many paradigms to learn gives him a feeling
of absence of difficulty, but he soon learns to his cost that this is only
a feeling, that in sober fact the very absence of explicit guide-posts to
structure leads him into all sorts of quandaries. A few examples will
be useful. One of the glories of English simplicity is the possibility of
using the same word as noun and verb. We speak, for instance, of
"having cut the meat" and of "a cut of meat." We not only "kick a
person," but "give him a kick." One may either "ride horseback" or
"take a ride." At first blush this looks like a most engaging rule but a
little examination convinces us that the supposed simplicity of word-
building is a mirage. In the first place, in what sense may a verb be
used as a noun? In the case of "taking a ride" or "giving a kick" the
noun evidently indicates the act itself. In the case of "having a cut on
the head" or "eating a cut of meat," it just as clearly does not indicate
the act itself but the result of the act, and these two examples do not
even illustrate the same kind of result, for in the former case the cut is
conceived of as the wound that results from cutting, whereas in the
latter case it refers to the portion of meat which is loosened by the act
of cutting. Anyone who takes the trouble to examine these examples
carefully will soon see that behind a superficial appearance of simplicity
there is concealed a perfect hornet's nest of bizarre and arbitrary'' usages.
To those of us who speak English from the earliest years of our child-
hood these difficulties do not readily appear. To one who comes to
English from a language which possesses a totally different structure
such facts as these are disconcerting. But there is a second difficulty
with the rule, or tendency, which allows us to use the unmodified
verb as a noun. Not only is the function of the noun obscure, but in a
great many cases we cannot use it at all, or the usage is curiously re-
stricted. We can "give a person a shove" or "a push," but we cannot
"give him a move" nor "a drop" (in the sense of causing him to drop).
We can "give one help," but we "give obedience," not "obey." A com-
plete examination, in short, of all cases in which the verb functions as a
noun would disclose two exceedingly cheerless facts: that there is a
considerable number of distinct senses in which the verb may be so
employed, though no rule can be given as to which of these possible

Four: The Proh/cni ofun luicniutiomil Auxilmrx I (ini^iia^e 269

llf)

senses is the proper ono in any particular case or whether only one or
more than one such moauinp is possible; and that in many cases no such
nouns may be formed at all, but that either nouns of an entirely difTer-
ent formation must be used or else that they are not possible at all.
We thus have to set up such rather cranky-looking configurations as

to hclp:help = to obey : obedience
= to grow: growth
= to drownidrowning,

a set-up which is further complicated b}' the fact that such a word
as 'drowning' not only corresponds to such words as 'help' and 'growth,'
but also to such words as 'helping' and 'growing.' The precise disen-
tanglement of all these relations and the obtaining of any tiling like
assurance in the use of the words is a task of no small difficulty. Where,
then, is the simplicity with which we started? It is obviously a phantom.
The English-speaking person covers up the difficulty for himself by
speaking vaguely of idioms. The real point is that behind the vagaries
of idiomatic usage there are perfectly clear-cut logical relations which
are only weakly brought out in the overt form of English. The sim-
plicity of English in its formal aspect is, therefore, really a pseudo-
simplicity or a masked complexity.

Another example of apparent, but only apparent, simplicit}' in
English is the use of such vague verbs as 'to put' and 'to get.' To us
the verb 'put' is a very simple matter, both in form and in use. Ac-
tually it is an amazingly difficult word to learn to use and no rules can
be given either for its employment or for its avoidance. 'To put at
rest' gives us an impression of simplicity because of the overt simplicity
of the structure, but here again the simplicity is an illusion. 'To put at
rest' really means 'to cause to lest,' and its apparent analogy to such
constructions as 'to put it at a great distance,' so far from helping
thought, reall}' hinders it, for the formal analogy is not paralleled by a
conceptual one. 'To put out of danger' is formally analog(His to 'to
put out of school,' but here too the analogy is utterly misleading, unless,
indeed, one defines school as a form of danger. If we were to define
'put' as a kind of causative operator, we should get into trouble, for it
cannot be safely used as such in all cases. In such a sentence as "The
ship put to sea," for example, there is no implied causative relation. If
English cannot give the foreigner clear rules for the employment of
verbs as nouns or for such apparently simple verbs as 'put,' what
advantage is derived by him from the merely negative fact that he
has not much formal giammar to learn in these cases? He may well

-,-7^ General Linguistics I

116

feel that the apparent simplicity of English is purchased at the price
of a bewildering obscurity. He may even feel that the mastery of
English usage is, in the long run, much more difficult than the appli-
cation of a fairly large number of rules for the formation of words, so
long as these rules are unambiguous.

English has no monopoly of pseudo-simplicity. French and German
illustrate the misleading character of apparent grammatical simplicity
just as well. One example from French will serve our purpose. There is
no doubt that the French speaker feels that he has in the reflexive verb
a perfectly simple and, on the whole, unambiguous form of expression.
A logical analysis of reflexive usages in French shows, however, that
this simplicity is an illusion and that, so far from helping the foreigner,
it is more calculated to bother him. In some cases the French reflexive
is a true reflexive; that is, it indicates that the subject of the sentence
is the same as the object. An example of a reflexive verb of this sort
would be se tuer, 'to kill oneself.' To French feeling this sort of verb is
doubtless identical with the type illustrated by s'amuser. Logically,
however, one does not 'amuse oneself' in the sense in which one 'kills
oneself.' The possibility of translating 'to amuse oneself into 'to have
a good time' and the impossibility of translating 'to kill oneself into
'to have a bad time killing,' or something of that sort, at once shows the
weakness of the analogy. Logically, of course, s'amuser is not a true
reflexive at all, but merely an intransitive verb of the same general
type as 'to rejoice' or 'to laugh' or 'to play.' Furthermore, the French
verb se baiire gives the Frenchman precisely the same formal feeling as
se tuer and s'amuser. Actually it is a reciprocal verb which may be trans-
lated as 'to strike one another' and, therefore, 'to fight.' Finally, in such
a verb as s'^tendre, 'to extend' or 'to stretch,' the Frenchman distinctly
feels the reflexive force, the stretching of the road, for instance, being
conceived of as a self-stretching of the road, as though the road took
itself and lengthened itself out. This type of verb may be called a
pseudo-reflexive, or a non-agentive active verb, the point being that
the action, while of a type that is generally brought about by an out-
side agency, is conceived of as taking place without definite agency. In
English, verbs of this kind are regularly used without the reflexive,
as in 'the road stretches,' 'the string breaks,' 'the rag tears,' 'the bag
bursts,' which are the non-agentive correspondents of such usages as
'he stretches the rubber band,' 'he breaks the string,' 'he tears the rag,'
'he bursts the balloon.' It should be clear that a linguistic usage, such as
the French reflexive, which throws together four such logically distinct
categories as the true reflexive, the simple intransitive, the reciprocal.

Four: The Problem of nn Inicrnniionul An\iluir\ / (inf^iia^e 271

117

and the non-agentive active, purchases simplicity at a considerable
price. For the Frenchman such usage is convenient enough and no
ambiguity seems to result. But for the outsider, who comes to French
with a different alignment of forms in his mind, the simplicity that is
offered is puzzling and treacherous.

These examples of the lack of simplicity in English and French, all
appearances to the contrary, could be multiplied almost without limit
and apply to all national languages. In fact, one may go so far as to say
that it is precisely the apparent simplicity of structure which is sug-
gested by the formal simplicity of many languages which is responsible
for much slovenliness in thought, and even for the creation of imaginary
problems in philosophy. What has been said of simplicity applies equally
to regularity and logic, as some of our examples have already indicated.
No important national language, at least in the Occidental world, has
complete regularity of grammatical structure, nor is there a single
logical category w^hich is adequately and consistently handled in terms
of linguistic symbolism. It is well known that the tense sj-'stems of
French, English and German teem with logical inconsistencies as they
are actually used. Many categories which are of great logical and psycho-
logical importance are so haltingly expressed that it takes a good deal
of effort to prove to the average man that they exist at all. A good
example of such a category is that of 'aspect,' in the technical sense of
the word. Few English-speaking people see such a locution as 'to burst
into tears' or 'to burst out laughing' as much more than an idiomatic
oddit3^ As a matter of fact, English is here trying to express, as best it
can, an intuition of the 'momentaneous aspect'; in other words, of
activity seen as a point in contrast to activity seen as a line. Logically
and psychologically, nearly every activity can be thought of as either
point-like or line-like in character, and there are, of course, many ex-
pressions in English which definitely point to the one or to the other,
but the treatment of these intuitions is fragmentary and illogical
throughout.

A standard international language should not only be simple, regular,
and logical, but also rich and creative. Richness is a difficult and sub-
jective concept. It would, of course, be hopeless to attempt to crowd
into an international language all those local overtones of meaning
which are so dear to the heart of the nationalist. There is a growing
fund of common experience and sentiment which will have to be ex-
pressed in an international language, and it would be strange if the
basic fund of meanings would not grow in richness with the interac-
tions of human beings who make use of the international medium.

272 General Linguistics I

118

The supposed inferiority of a constructed language to a national one on
the score of richness of connotation is, of course, no criticism of the idea
of a constructed language. All that the criticism means is that the con-
structed language has not been in long-continued use. As a matter of
of fact, a national language which spreads beyond its own confines very
quickly loses much of its original richness of content and is in no better
case than a constructed language.

More important is the question of creativeness. Here there are many
illusions. All languages, even the most primitive, have very real powers
of creating new words and combinations of words as they are needed,
but the theoretical possibilities of creation are in most of these national
languages which are of importance for the international language
question thwarted by all sorts of irrelevant factors that would not
apply to a constructed language. English, for instance, has a great
many formal resources at its disposal which it seems unable to use
adequately; for instance, there is no reason why the suffix -ness should
not be used to make up an unlimited number of words indicating qual-
it3% such as 'smallness' and 'opaqueness," yet we know that only a
limited number of such forms is possible. One says 'width,' not 'wide-
ness'; 'beauty,' not 'beautifulness.' In the same way, such locutions as
'to give a kick' and 'to give a slap' might be supposed to serve as models
for the creation of an unlimited number of momentaneous verbs, yet
the possibilities of extending this form of usage are strictly limited.
The truth is that sentiment and precedent prevent the national lan-
guage, with its accepted tradition, from doing all it might do, and the
logically possible formations of all kinds which would be felt as awk-
ward or daring in English, or even in German, could be accepted as the
merest matters of course in an international language that was not tied
to the dictates of irrational usage.

We see, then, that no national language really corresponds in spirit
to the analytic and creative spirit of modern times. National lan-
guages are all huge systems of vested interests which sullenly resist
critical inquiry. It may shock the traditionalist to be told that we arc
rapidly getting to the point where our national languages are almost
more of a hindrance than a help to clear thinking; yet how true this is
is significantly illustrated by the necessity that mathematics and
symbolic logic have been under of developing their own systems of
symbolism. There is a perfectly obvious objection that is often raised
at this point. We are told that normal human expression does not crave
any such accuracy as is attained by these rigorous disciplines. True,
but it is not a question of remodeling language in the spirit of mathe-

Four: The I'lohlcni of nn hucrnmioiuil .\u\iluii\ lAin^uage 273

119

matics and symbolic logic, but merely of giving it the structural means
whereby it may refine itself in as economical and unambiguous a manner
as possible.

It is likely that tlie foundations of a truly adequate form of inter-
national language have already been laid in Esperanto and other pro-
posed international auxiliary languages, but it is doubtful if the exacting
ideal that we have sketched is attained by any one of them, or is likely
to be attained for some time to come. It is, therefore, highly desirable
that along with the practical labour of getting wider recognition of the
international language idea, there go hand in hand comparative re-
searches which aim to lay bare the logical structures that are inade-
quately s3''mbolized in our present-day languages, in order that we may
see more clearly than we have yet been able to see just how much of
psychological insight and logical rigour have been and can be expressed
in linguistic form. One of the most ambitious and important tasks that
can be undertaken is the attempt to work out the relation between
logic and usage in a number of national and constructed languages, in
order that the eventual problem of adequately symbolizing thought
may be seen as the problem it still is. No doubt it will be impossible,
for a long time to come, to give a definite answer to all of the questions
that are raised, but it is something to rai.se and define the Cjuestions.

I have emphasized the logical advantages of a constructed inter-
national language, but it is important not to neglect the psychological
ones. The attitude of independence toward a constructed language
which all national speakers must adopt is really a great advantage,
because it tends to make man .see himself as the master of language
instead of its obedient .ser\-ant. A common allegiance to form of ex-
pression that is identified with no single national unit is likely to prove
one of the most potent symbols of the freedom of the human spirit
that the world has 3'et known. As the Oriental peoples become of more
and more importance in the modern world, the air of .sanctity that
attaches to English or German or French is likely to seem le.^s and less
a thing to be taken for granted, and it is not at all unlikely that the
eventual triumph of the international language movement will owe
much to the Chinaman's and the Indian's indifTerence to the vested
interests of Europe, though the actual stock of basic words in any
practical international language is almost certain to be bixscd on the
common European fund. A further psychological advantage of a con-
structed language has been often referred to by those who have had
experience with such language? as Esperanto. This is the removal of
fear in the public use of a language other than one's native tongue.

->74 General Linguistics I

120

The use of the wrong gender in French or any minor violence to English
idiom is construed as a sin of etiquette, and everyone knows how par-
alyzing on freedom of expression is the fear of coimnitting the slightest
breach of etiquette. Who knows to what extent the discreet utterances of
foreign visitors are really due to their wise unwillingness to take too
many chances with the vagaries of a foreign language? It is, of course,
not the language as such which is sinned against, but the conventions of
fitness which are in the minds of the natives who act as custodians of the
language. Expression in a constructed language has no such fears as
these to reckon with. Errors in Esperanto speech are not sins or breaches
of etiquette; they are merely trivialities to the extent that they do not
actually misrepresent the meaning of the speaker, and as such they may
be ignored.

In the educational world there is a great deal of discontent with the
teaching of classical and modern languages. It is no secret that the
fruits of language study are in no sort of relation to the labour spent
on teaching and learning them. Who has not the uncomfortable feeling
that there is something intellectually dishonest about a course of study
that goes in for a half-hearted tinkering with, say, Latin and two
modern languages, with a net result that is more or less microscopic in
value? A feeling is growing that the study of foreign languages should
be relegated to the class of technical specialties and that the efforts of
educators should be directed rather toward deepening the conceptual
language sense of students in order that, thus equipped, they may as
occasion arises be in a better position to learn what national languages
they may happen to need. A well-constructed international language is
much more easily learned than a national language, sharpens one's
insight into the logical structure of expression in a way that none of
these does, and puts one in possession of a great deal of lexical material
which can be turned to account in the analysis of both the speaker's
language and of most others that he is likely to want to learn. Certain
beginnings have already been made toward the adoption of international
language study as a means toward general language worlc. Time alone
can tell whether this movement is a fruitful one, but it is certainly an
aspect of the international language question that is worth thinking
about, particularly in America, with its growing impatience of the
largely useless teaching of Latin, French, German, and Spanish in the
high schools. The international language movement has had, up to the
present time, a somewhat cliquish or esoteric air. It now looks as though
it might take on the characteristics of an international Open Forum.
The increasing degree to which hnguists, mathematicians and scientists

Four: T/w I^rohlctn of dn Inicrnutiotuil Auxiluirx I (ini^ua^e 275

121

have been thinking about the problem is a sign that promises well
for the future. It is a good thing that the idea of an international hin-
guage is no longer presented in merely ideaUstic terms, but is more
and more taking on the aspect of a practical or technological problem
and of an exercise in the cleaning up of the thought process. Intelli-
gent men should not allow themselves to become international lan-
guage doctrinaires. They should do all they can to keep the problem
experimental, welcoming criticism at every point and trusting to the
gradual emergence of an international language that is a fit medium
for the modern spirit.

Tlie spirit of logical analysis should in practice blend with the prac-
tical pressure for the adoption of some form of international language,
but it should not allow itself to be stampeded by it. It would be ex-
ceedingly unfortunate if an international language, wliether Esperanto
or English or some form of simplified English, were looked upon as
thenceforth sacred and inviolate. No solution of the international
language problem should be looked upon as more than a beginning
toward the gradual evolution, in the light of experience and at the hand
of all civilized humanity, of an international language which is as rich
as any now known to us, is far more creative in its possibilities, and is
infmitcly simpler, niore regular, and more logical than any one of them.

Editorial Note

Psyche 11:4 (1930-31), 3-15. (Also published in: Herbert N. Shcnton - Hdward
Sapir - Otto Jcspcrsen. Inlcrndtio/uil Conuuiiniaiiion: A Symposium on the
luuii^udi^c f^rohlcm. London: Kcgan Paul. Trench. Trubncr c'Ci (\)., 1^31 . pp. (t^ 94;
reprinted in Edward Sapir, Selected Wrilitii^s in Luni^uui^e, C'ullure, and
Persomdity. Edited by David (i. Mandclbaum. Berkeley: University of California
Press, 1949, pp. 110-121] Reprinted here after the compactly printed version in
the Selected Writings.

WANTED: A WORLD LANGUAGE

BY EDWARD SAPIR

As TO the theoretical desirabihty of an
international auxiliary language
'" there can be little difference of
opinion. But as to just what factors in the
solution of the problem should be allowed
to weigh most heavily there is room for
every possible difference of opinion, and so
it is not surprising that interlinguists are
far from having reached complete agree-
ment. The crucial differences lie not so
much between one constructed language
and another as between the idea of a con-
structed language and that of an already
established national one, whether in its tra-
ditional form or in some simplified form.
It is not uncommon to hear it said by
those who stand somewhat outside the
movement that some such regular system
as Esperanto is theoretically desirable, but
that it is of little use to work for it be-
cause English is already de facto the inter-
national language of modern times — if not
altogether at the moment, then in the im-
mediate future—, that English is simple
enough and regular enough to satisfy all
practical requirements, and that the pre-
cise form of it as an international language
may well be left to historical and psycho-
logical factors that one need not worry
about in advance. This point of view has a
certain pleasing plausibility about it but,
like so many diings that seem plausible
and effortless, it may none the less embody
a number of fallacies.

There are two considerations, often in-
termingled in practice, which arouse the

thought of an international language. The
first is the purely practical problem of fa-
cilitating the growing need for interna-
tional communication in its most elemen-
tary sense. A firm, for instance, that docs
business in many countries is driven to
spend an enormous amount of time, labor,
and money in providing a translation serv-
ice. All this is sheer waste. Again, at an
international scientific meeting; one is al-
ways disappointed to find that the dif-
ficulty of communicating with foreign
scientists makes it much harder to ex-
change ideas than one had fancied might
be the case before setting sail. Such
examples might be multiplied ad inpni-
tum. Sooner or later one chafes and be-
gins to wonder whether the evil is as
necessary as tradition would have it. Impa-
tience translates itself into a desire to have
something immediate done about it, and,
as is generally the case with impatience,
resolves itself in the easiest way that Hes
to hand. Why not push English, for in-
stance, which is already spoken over a
larger area than any other language of
modern times and which shows every sign
of spreading .^^ But reflections of this sort,
grounded in impatience as they are, look
for no more worthy solution of the diffi-
culty than a sort of minimum language, a
lingua franca of the modern world. Those
who argue in this spirit invariablv pride
themselves on being "practical," and. like
all "practical" people, they are apt to argue
without their host.

Four: The Frohlcfn of dn fnicrndiioruil Auxiliurv Lani^utifie

The opposed consideration mny be put
in something hkc the following form: An
intcrnaiionnl auxiUary language should
serve as a broad base for every type of
international understanding, which means
for every type of expression of the human
spirit which is of more than local interest.
The exigencies of trade or travel are from
this point of view merely some of the
more obvious symptoms of the interna-
tionalizing of the human mind, and it
would be a mistake to ask too little of an
organ of international expression.

But this is not all. The modern mind
tends to be more and more critical and
analyical in spirit, hence it must devise for
itself an engine of expression which is logi-
cally defensible at every point and which
tends to correspond to the rigorous spirit
of modern science. This does not mean that
a constructed international language is ex-
pected to have the perfection of mathe-
matical symbolism, but it must be pro-
gressively felt as moving in that direction.
Perhaps the speakers of a national lan-
guage are under profound illusions as to
the logical character of its structure. Per-
haps they confuse the comfort of habit
with logical necessity. If this is so — and I
do not see how it can be seriously doubted
that it is — it must mean that in the long
run the modern spirit will not rest satisfied
with an international language that merely
extends the imperfections and provin-
cialisms of one language at the expense of
all others.

There are also other considerations that
are of importance, and among them per-
haps the most obvious is the attitude of
people toward the spread or imposition of
any national language which is not their
own. The psychology of a language which,
in one way or another, is imposed upon
one because of factors beyond one's control
is very different from the psychology of a

language that one accepts of one's free
will. In a sense, every form of expression
is imposed upon one by social factors, one's
own language above all. But it is the
thought or illusion of freedom that is the
important thing, not the fact of it.

The modern world is confronted by the
difhculiy of reconciling internationalism
with its persistent and tightening national-
isms. More and more, unsolicited gifts
from without are likely to be received with
unconscious resentment. Only that can be
freely accepted which is in some sense a
creation of all. A common creation de-
mands a common sacrifice, and perhaps
not the least potent argument in favor of a
constructed international language is the
fact that it is equally foreign, or apparently
so, to the traditions of all nationalities. The
common difficulty gives it an impersonal
character and silences the resentment that
is born of rivalry.

English, as an international language, is
no more secure than French has proved to
be as the accepted language of diplomacy,
or as Latin has proved to be as the interna-
tional language of science. Both French
and Latin are involved with nationalistic
and religious implications which could not
be entirely shaken off, and so, while they
seemed for a time to have solved the inter-
national language problem up to a certain
point, they did not really do so in spirit.
English would probably fare no better,
and it is even likely that the tradition of
superficial practicality that attaches to it
may, in the long run, prove more of a hin-
drance than a help to its acceptance.

One must beware of an over-emphasis
on the word "auxiliary." It is pcrfccdv true
that for generations to come an interna-
tional language must be auxiliary', must
not attempt to set itself up against the
many languages of the folk, but it must
for all that be a free and powerful cxprcs-

278

General Linguistics I

104

sion of its own, capable of all work that
may reasonably be expected of language
and protected by the powerful negative
fact that it cannot be interpreted as the
symbol of any nationality.

Even if it be assumed for the sake of
argument that English is to spread as an
auxiliary language over the whole world,
it docs not follow that the international
language problem is disposed of. English,
or some simplified version of it, may
spread for certain immediate and practical
purposes, yet the deeper needs of the mod-
ern world may not be satisfied by it and
wc may still have to deal with a conflict
between an English that has won a too
easy triumph and a constructed language
that has such obvious advantages of struc-
ture that it may gradually displace its na-
tional rival.

What is needed above all is a language
that is as simple, as regular, as logical, as
rich, and as creative as possible; a language
which starts with a minimum of demands
on the learning capacity of the normal in-
dividual and can do the maximum amount
of work; which is to serve as a sort of logi-
cal touchstone to all national languages
and as the standard medium of translation.
It must, ideally, be as superior to any ac-
cepted language as the mathematical
method of expressing quantities and rela-
tions between quantities is to the lumber-
ing verbal form. This is undoubtedly an
ideal which can never be reached, but
ideals are not meant to be reached; they
merely indicate the direction of movement.

logic, richness, and creativeness. We may
begin with simplicity. It is true that Eng-
lish is not as complex in its formal struc-
ture as is German or Latin, but this does
not dispose of the matter. The fact that a
beginner in English has not many para-
digms to learn gives him a feeling of ab-
sence of difficulty, but he soon learns to
his cost that this is only a feeling, diat in
sober fact the very absence of explicit
guide-posts to structure leads him into all
sorts of quandaries.

A few examples will be useful. One of
the glories of English simplicity is the
possibility of using the same word as noun
and verb. We speak, for instance, of "hav-
ing cut the meat" and of "a cut of meat."
We not only "kick a person," but "give
him a kick." One may either "ride horse-
back" or "take a ride." At first blush this
looks like a most engaging rule, but a Htde
examination convinces us that the sup-
posed simplicity of word-building is a
mirage. In the first place, in what sense
may a verb be used as a noun? In the case
of "taking a ride" or "giving a kick" the
noun evidently indicates the act itself. In
the case of "having a cut on the head" or
"eating a cut of meat," it just as clearly
does not indicate the act itself but the re-
sult of the act, and these two examples do
not even illustrate the same kind of result,
for in the former case the cut is conceived
of as the wound that results from cutting,
whereas in the latter case it refers to the
portion of meat which is loosened by the
act of cutting.

Anyone who takes the trouble to exam-
ine these examples carefully will soon see
that behind a superficial appearance of
simplicity there is concealed a perfect hor-
net's nest of bizarre and arbitrary usages.
To those of us who speak English from
our earliest years these difficulties do not
readily appear, but to one who comes to

Four: I he f'rohlcni ofan /nfrrnddondl Auxiluirv l.iini^ua^e

279

Ent^lish from a langii.'tgc which possesses
a totally diflcrcnt structure they arc most
disconcerting.

Again, there is a second difficulty with
the rule, or tendency, which allows us to
use the unmodified verb as a noun. Not
only is the function of the noun obscure
but in a great many cases we cannot use
it at all, or the usage is curiously restricted.
We can "give a person a shove" or "a
push," but we cannot "give him a move"
nor "a drop" (in the sense of causing him
to drop). We can "give one help," but we
"give obedience," not "obey." A complete
examination of all cases in which the verb
functions as a noun would disclose two
exceedingly cheerless facts: that there is a
considerable number of distinct senses in
which the verb may be so employed,
thoucrh no rule can be ^iven as to which
of these possible senses is the proper one
in any particular case or whether only one
or more than one such meaning is pos-
sible; and that in many cases no such
nouns may be formed at all, but that either
nouns of an entirely different formation
must be used or else that they are not pos-
sible at all. We thus have to set up such
rather cranky-looking configurations as

to help: help == to obey: obedience
= to grow: growth
= to drown: drowning,

a set-up which is further complicated by
the fact that such a word as drowning not
only corresponds to such words as help
and growth, but also to such words as
helping and growing.

The precise disentanglement of all !hc:e
relations and the obtaining of anything like
assurance in the use of the words is a task
of no small difficulty. Where, then. Is the
simplicity with which we started.' It is
obviously a phantom. The English-speak-
ing person covers up the difficulty for him-

105

self by speaking vaguely of idioms. The
real point is that behind the vagaries of
idiomatic usage there arc perfectly clear-
cut logical relations which are only weakly
brought out in the overt form of English.
The simplicity of English in its formal
aspect is, therefore, really a pseudo-sim-
plicity or a masked complexity.

Another example of apparent, but only
apparent, simplicity in English is the use
of such vague verbs as "to put" and "io
get." To us the verb put is a very simple
matter, both in form and in use. Actually
it is an amazingly difficult word to learn
to use and no rules can be given either for
its employment or for its avoidance. "To
put at rest" gives us an impression of sim-
plicity because of the overt simplicity of
the structure, but here again the simplicity
is an illusion. "To put at rest" really means
"to cause to rest," and its apparent analogy
to such constructions as "to put it at a great
distance," so far from helping thought,
really hinders it, for the formal analogy
is not paralleled by a conceptual one. "To
put out of danger" is formally analogous
to "to put out of school," but here too the
analogy is utterly misleading, unless, in-
deed, one defines school as a form of
danger.

If we were to define the word put as a
kind of causative operator, we should get
into trouble, for it cannot be safely used as
such in all cases. In such a sentence as "The
ship put to sea," for example, there is no
implied causative relation. If English can-
not give the foreigner clear rules for the
employment of verbs as nouns or for such
apparently simple verbs as put, what ad-
vantage is derived by him from the merely
negative fact that he has not much formal
grammar to learn in these cases? He may
well feel that tlie apparent simplicity of
English is purchased at the price of a be-
wildering obscurity. He may even feel that

280

General Linguistics I

zo6

the mastery of English usage is, in the
long run, much more difficult than the ap-
plication of a fairly large number of rules
for the formation of words, so long as these
rules are unambiguous.

English has no monopoly of this pseudo-
simplicity. French and German illustrate
the misleading character of it just as well.
One e.xample from French will serve our
purpose. There is no doubt that the French
speaker feels that he has in the reflexive
verb a perfectly simple and, on the whole,
unambiguous form of expression. A logi-
cal analysis of reflexive usages in French
shows, however, that this simplicity is an
illusion and that, so far from helping the
foreigner, it is more calculated to bother
him.

In some cases the French reflexive is a
true reflexive; that is, it indicates that the
subject of the sentence is the same as the
object. An example of a reflexive verb of
this sort would be se ttier, "to kill oneself."
To French feeling this sort of verb is doubt-
less identical with the type illustrated by
s'amtiser. Logically, however, one does not
"amuse oneself" in the sense in which one
"kills oneself." The possibility of translat-
ing "to amuse oneself" into "to have a
good time" and the impossibility of trans-
lating "to kill oneself" into "to have a bad
time killing," or something of that sort,
at once shows the weakness of the analogy.
Logically, of course, s'amuser is not a true
reflexive at all, but merely an intransitive
verb of the same general type as "to re-
joice" or "to laugh" or "to play."

Furthermore, the French verb se hattre
gives the Frenchman precisely the same
formal feeling as se tiier and s'amuser. Ac-
tually, it is a reciprocal verb which may be
translated as "to strike one another" and,
therefore, "to fight." Finally, in such a
verb as s'etendre, "to extend" or "to
stretch," the Frenchman distinctly feels the

reflexive force, the stretching of the road,
for instance, being conceived of as a self-
stretching of the road, as though the road
took itself and lengthened itself out. This
type of verb may be called a pseudo-re-
flexive, or a non-agentive, active verb, the
point being that the action, while of a type
that is generally brought about by an out-
side agency, is conceived of as taking place
without definite agency.

In English, verbs of this kind are regu-
larly used without the refle.xive, as in "the
road stretches," "the string breaks," "the
rag tears," "the bag bursts," which are the
non-agentive correspondents of such usages
as "he stretches the rubber band," "he
breaks the string," "he tears the rag," "he
bursts the balloon." It should be clear that
a linguistic usage, such as the French re-
flexive, which throws together four such
logically distinct categories as the true re-
fle.xive, the simple intransitive, the recipro-
cal, and the non-agentive active, purchases
simplicity at a considerable price. For the
Frenchman such usage is convenient
enough and no ambiguity seems to result.
But for the outsider, who comes to French
with a difiPerent alignment of forms in his
mind, the simplicity that is offered is
puzzling and treacherous.

Ill

These examples of the lack of simplicity
in English and French, all appearances to
the contrary, could be multiplied almost
without limit and apply to all national lan-
guages. In fact, one may go so far as to
say that it is precisely the apparent sim-
plicity of structure which is suggested by
the formal simplicity of many languages
which is responsible for much slovenliness
in thought, and even for the creation of
imaginary problems in philosophy. What
has been said of simplicity apphes equally

Four: The Prohlcni of an hurrndnonal Ainiluirv I un^utii^e

281

to regularity and logic, as some of our ex-
amples have already indicated. No impor-
tant national language, at least in the Oc-
cidental world, has complete regularity of
grammatical structure, nor is there a single
logical category which is adequately and
consistently handled in terms of linguistic
symbolism.

A standard international language should
not only be simple, regular, and logical,
but also rich and creative. Richness is a
difficult and subjective concept. It would,
of course, be hopeless to attempt to crowd
into an international language all those
local overtones of meaning which are so
dear to the heart of the nationalist. But
there is a growing fund of common experi-
ence and sentiment which will have to
be expressed in an international language,
and it would be strange if the basic fund of
meanings would not grow in richness
with the interactions of human beings who
make use of the new medium. The sup-
posed inferiority of a constructed language
to a national one on this score is, of course,
no criticism of the idea of a constructed
language. All that it means is that the con-
structed language has not been in long-
continued use. As a matter of fact, a
national language which spreads beyond
its own confines very quickly loses much
of its original richness of content and is in
no better case than a constructed language.
More important is the question of crea-
tivcness. Here there are many illusions. All
languages, even the most primitive, have
very real powers of creating new words
and combinations of words as they are
needed, but the theoretical possibilities of
creation, in most of the national languages
of importance for the International lan-
guage question, are thwarted bv all sorts
of irrelevant factors that would not apply
to a constructed language. Enfjlish. to
name one, has a great many formal rc-

107

sources at its disposal which it seems un-
able to use adequately; for instance, there
is no reason why (he suffix -nesi should not
be used to make up an unlimited number
of words indicating quality, such as small-
ness and opaqueness, yet wc know that
only a limited number of such forms is
possible. One says width, not widcncss;
beauty, not beautifulness. In the same way,
such locutions as "to give a kick" and "to
give a slap" might be supposed to serve as
models for the creation of an unlimited
number of momentaneous verbs, yet the
possibilities of extending this form of usage
are stricdy limited. The truth is that sen-
timent and precedent prevent the national
language, with its accepted tradition, from
doing all it might do, and the logically pos-
sible formations of all kinds which would
be felt as awkward or daring in English,
or even in German, could be accepted as
the merest matters of course in an inter-
national language that was not tied to the
dictates of irrational usage.

We see, then, that no national language
really corresponds in spirit to the analytic
and creative spirit of modern times. Na-
tional languages are all huge systems of
vested interests which sullenly resist criti-
cal inquiry. It may shock the traditionalist
to be told that we are rapidly getting to
the point where our national languages arc
almost more of a hindrance than a help
to clear thinking; yet how true this is is
significantly illustrated bv the necessity
that mathematics and symbolic logic have
been under of developing their own sys-
tems of symbolism.

It is likely that the foundations of a
truly adequate form of international lan-
guage have already been laid in Esperanto
and other proposed international .luxiliary
languages, but it is doubtful if the exacting
ideal that I have sketched is attained bv
anv one of them, or is likclv to be attained

282

General Linguistics I

2.08

for some time to come. It is, therefore,
highly desirable that along with the prac-
tical labor of getting wider recognition of
the international language idea, there go
hand in hand comparative researches
which aim to lay bare the logical struc-
tures that are inadequately symbolized in
our present-day languages, in order that
we may see more clearly than we have yet
been able to see just how much of psycho-
logical insight and logical rigor have been
and can be expressed in linguistic form.

One of the most ambitious and impor-
tant tasks that can be undertaken is the
attempt to work out die relation between
logic and usage in a number of national
and constructed languages, in order that the
eventual problem of adequately symboliz-
ing thought may be seen as the problem it
still is. No doubt it will be impossible, for
a long time to come, to give a definitive
answer to all of the questions that are
raised, but it is something to raise and de-
fine the questions.

I have emphasized the logical advantages
of a constructed international language, but
it is important not to neglect the psycho-
logical ones. The attitude of independence
toward a constructed language which all
national speakers must adopt is really a
great advantage, because it tends to make
man see himself as the master of language
instead of its obedient servant. A common
allegiance to a form of expression that is
identified with no single national unit is
likely to prove one of the most potent sym-
bols of the freedom of the human spirit
that the world has yet known.

A further psychological advantage of a
constructed language has been often re-
ferred to by those who have had experience
with such languages as Esperanto. This is
the removal of fear in the public use of a
language other than one's native tongue.
The use of the wrong gender in French or

any minor violence to English idiom is
construed as a sin of etiquette, and every-
one knows how paralyzing on freedom of
expression is the fear of committing the
slightest breach of etiquette. Who knows
to what extent the discreet utterances of
foreign visitors are really due to their wise
unwillingness to take too many chances
with the vagaries of a foreign language?
Expression in a constructed language has
no such fears as these to reckon with. Er-
rors in Esperanto speech are not sins or
breaches of etiquette; they are merely
trivialities to the extent that they do not
actually misrepresent the meaning of the
speaker, and as such they may be ignored.

A well-constructed international lan-
guage is much more easily learned than a
national language, sharpens one's insight
into the logical structure of expression in
a way that none of these does, and puts
one in possession of a great deal of lexical
material which can be turned to account
in the analysis of both the speaker's Ian-

Four: The f^rohlcni of an Inicruuiioniil .■\ii\iliiit\ I iini^ua^e

283

guagc and of most others that lie is Ukcly
to want to learn. Certain beginnings have
already been made toward the adoption of
international language study as a means to-
ward general language work. Time alcnc
can tell whether this movement is a fruit-
ful one, but it is certainly an aspect of the
international language question that is
worth thinking about, particularly in
America, with its growing impatience of
the largely useless teaching of Latin,
French, German, and Spanish in the high-
schools.

The international language movement
has had, up to the present time, a some-
what cliquish or esoteric air. It now looks
as though it might take on the characteris-
tics of an international Open Forum. The
increasing degree to which linguists,
mathematicians and scientists have been
thinking about the problem is a sign that
promises well for the future. It is a good
thing that the idea of an international lan-
guage is no longer presented in merely

109

idealistic terms, but is more and more tak-
ing on the aspect of a practical or tech-
nological problem and of an exercise in
the cleaning up of the thought process.

The spirit of logical analysis should in
practice blend with the practical pressure
for the adoption of some form of interna-
tional language, but it should not allow
itself to be stampeded by it. It would be
exceedingly unfortunate if an interna-
tional language, whether Esperanto or
English or some form of simplified Eng-
lish, were looked upon as thenceforth
sacred and inviolate. No solution of the
international language problem should be
looked upon as more than a beginning to-
ward the gradual evolution, in the light
of experience and at the hand of all civi-
lized humanity, of an international lan-
guage which is as rich as anv now known
to us, is far more creative in its possibilities,
and is in its structure infinitely simpler,
more regular, and more logical than any
one of them.

Editorial Note
The American Mercury 22 ( U)32-33), 202-209.

ACTES DU DEUXifeME

CONGRES INTERNATIONAL

DE LINGUISTES

GENEVE 25 - 29 AOUT 1931

12. E. SAPIR, Chicago.

The case for a constructed international langdage.
(Resum^.)

It is ver}' important not to confuse the function of an auxiliary international
language with that of a national language normally acquired in childhood. The
latter serves as the complete symbol of the emotionally integrated and locally
oriented personality. The former has the psychologically less far-reaching but,
in the modern world, exceedingly useful function of providing the individual
with a fit symbol of solidarity with the international world on those planes of
interest which have a true international significance.

There is rapidly growing a real international community which is largely
debarred from recognizing itself for what it potentially is by the absence of a fit
symbol of expression. This community is based on transnational functions of an
economic, technological, scientific and ideologicalnature.lt is more or less parallel
to the ecclesiastical and scholarly European medieval community that used Latin

Four: The Proh/i'ni of Hn hilcrmifionul Auxiluirv I (ini^ua^e 285

as a medium of expression far more successfully than it could luvc anyothcrlanguagc.

The multiplication of national languages today means a tremendous waste in
commercial transactions, adds enormously to the difficulties of travel, is respon-
sible for the extreme cumbcrsomencss of international political negotiations, and
has brought about something like an impasse in the scientific world. The usefulness
of exchange professors is more questionable than courtesy allows us to admit.
Radio and talking film are ready to bind the peoples of the world together with
a common speech that transcends the limited usefulness of national languages,
but the present status of human civilization will not allow them to do so.

Tiie educational problem of teaching a variety of separate techniques for the
expression of essentially the same meanings is becoming increasingly serious in
the modern world. An unnccessar}' burden is being laid on smaller European
nationalities and Oriental peoples in the development of world civilization. The
feeling is growing that there is too much to learn that is basically significant for
a grasp of the modern world for so much energy to be spent on acquiring irra-
tionally var)'ing symbolisms, none of which can be mastered satisfactorily.

To meet such conditions in the modern world, a highl)' efficient and max-
imally simple international language needs to be developed.

A soundly constructed auxiliary language has great advantages. It combines an
international function with the elimination of conflicting national claims. It
capitalizes for common purposes the stock of words and grammatical techniques
which lie scattered about in the more impoaant of the national languages of
Europe. It has intellectual value as a help to logical thinking, and as a spur to an
analysis which transcends the largely unconscious implications of particular
national languages. It encourages in the individual a creative and experimental
attitude in the handling of linguistic material.

The term " artificial " does not do psychological nor historical justice to such
construaed languages as those in current use. They are artificial in no profounder
sense than that in which the technique of an opera singer is " artificial " as
compared with the more unconscious technique of a folk-singer.

The current psychological argument as to the supposedly nonvital cliaracter of
a constructed international language is hardly more than a figure of speech. In
learning such a language, for example Esperanto or Interlingua, one builds a new
set of habits on the basis of the old linguistic ones. This is precisely what one
does when one learns another language as an adult. It is quite a mistake to suppose
that an English speaking person's command of French or German is psycholog-
cally in the least equivalent to a Frenchman's or a German's command of his
native language. All that is managed, in the great majority of cases, is a fairly
adequate control of the external features of the foreign language. This incomplete
control has, however, the immense advantage of putting the native speaker and
the foreigner on a footing of approximate mutual understanding, which is
sufficient for the purpose desired.

A constructed international language should be looked at as a system of
communication suited to certain difficult situations arising in the sophisticated
modem world — a system of signs consciously modified and regularized yet
psychologically based on the more unconscious folk-systems of communication,
precisely as mathematical, scientific and technological symbolisms are both inter-
national in scope and based on differing and less systematized folk usages.

286 General Linguistics I

Tlie fear of the splitting up of an international language into mutually
unintelii^'ible dialects is not corroborated by actual experience. The constructed
lanc'uaqcs in use are so simple phonetically that even with considerable latitude
of Tndfvidual pronunciation no ambiguities of moment are likely to arise.
National languages are far more ramified dialectically than a constructed interna-
tional language would be.

One should carefully refrain from injecting into a discussion of the interna-
tional Unguage question any of those romantic concepts in regard to language as
an " organism " which have already done so much harm in the study of
linguistic processes. It is particularly we linguists who stand in danger of making
a fetish of the materials of our study. The romanticism of the past should never
bind the hand or daunt the will of the futureo

Editorial Note

Actes dii deuxieme Congres international de linguistes, Geneve 25 - 29 aout 1931.
Paris: Librairie d'Amerique et d'Orient Adrien Maisonneuve, 1933, pp. 86-88.

The following errors in the originally published version have been corrected

directly into the text printed here (page references are to the original):

p. 86, title: The case for constructed (correct: The case for a constructed)

p. 88, 1. 10: particulary (correct: particularly)

An editorial intervention concerning punctuation has been indicated with <>.

p. 87, 1. I, read: other language (with space between the two words).

ANNEX:

The Stalemcnl of Ihc Inlcrnalional Auxiliary

Language Association

made at the second International C ontcrcncc ol

Linguists (Geneva, 1931)

In the same scclioii as the one to which Edward Sapir conlrihuicd his paper
"The Case for [a] Constructed International Language" a statement was presen-
ted by the I ALA. The statement' is reproduced here because of the information
it contains on the context in which Sapir. Jespersen. and Colhnson wrote on theor-
etical and practical aspects of international auxiliary languages, and because of
its relevance for Sapir's involvement in the project of a "universal conceptual
grammar/* which he was to carry out through his study of formal and semantic
structures corresponding to notions such as "totality," "ending-point" and "grad-
ing" [see section V], and also through innovative research, sponsored by the
American Council of Learned Societies, on English grammar/

' The statement was published as §14 wilhiii the scciu'ii ik\.iiLt.l lo the prohkiu i)( an internalumal au\iliar>'
language (pp. SK-W in the /tc7<'.v).

- The progress reports on this researeh are ineluJeil iii volume 1 1 ol I hr C nllct iid Wnrk.s of Edward Sapir.

238 General Linguistics I

Statement by the International Auxiliary Language Association.

'l"he International Auxiliary Language Association in the United States Incor-
porated (a private organization familiarly known as lALA), carries on, as one of
its main activities, research along three lines : educational, sociological and

linguistic.

The aim of the linguistic research is to furnish material which will aid in the
development of an international language suited to the functions which it should

perform.

At the invitation of lALA and convened by Prof. O. Jespersen, a Meeting of
Linguistic Research was held in Geneva March-April, 1930.

For the first time, eminent protagonists of diverse constructed idioms and
distinguished philologists from European and American Universities met together
not to discuss claims of superiority for any particular auxiliary language but for
the purpose of finding ways in which to collaborate toward a common goal. The
philologists elaborated a comprehensive plan for linguistic research, based on
the suggestions sent by Professors K. Asakawa of Yale University and R. H. Fife
of Columbia University. All participants agreed thas it is reasonable to hope that
the carrying out of the plan might be a potent factor in bringing about ultimate
accord in respect to a definitive form of international language.

The research is planned in three circles :

I. Foundations of Language^ logical and psychological, an approach to the

international language problem. A more philosophical study, according to a
scheme drawn up by Professors Sapir and Collinson.

2. Comparative Studies of four national languages (English, French, German,
Russian) and four international languages. An objective examination of the
structure of the selected languasres, bo:h with regard to details and to the
languages as wholes.

3. Preparation for Synthesis. A comprehensive survey and criticism of the
results of the first two cycles of research with a view to finding data for a syn-
thetic scheme of a definitive language for international use.

The projected research of the first and second circles includes studies in
language structure and vocabulary. The former are to be pursued first and are to
serve a material for further labors, namely the working out of a generalised or
universal conceptual grammar, and of outlines of structure, both of which might
be used as a basis for general language study and as a norm for the structure of an
international language.

The research has been begun and is proceeding under the direction of Pro-
fessor E. Sapir. The work will proceed to completion if sufficient funds are
secured.

lALA has no intention of developing a new language. It believes that after its
research is finished, an independent body of experts should be entrusted with the
task of recommending the form of the ultimate international language. It desires
only to do its share in preparing for such a possible body material which will be
relevant and worthy. It believes that in the evolution of languages unconscious and
conscious processes have gone and must continue to go hand in hand, and that
we are living in an age when creative consciousness can take hold more and more
in the welding for beneficient purposes of the diverse symbolisms produced by
both those processes.

Section Five

Studies in Universal Conceptual Grammar

(1930,1932,1944)

Intrculuctorv Note:
Sapir's Studies in Universal ('oneeplual ( IraninKn

The ihrcc papers included in this sccticm offer an illiisiiaiion ol ihe working
out of a generalised or universal conceptual grammar" referred to in the
"Statement by the liUeriialional .Auxiliary Language Association" [see section
IV, Annex]. They exemplify Sapir's general project of studying, in a global and
integrated manner, the fundamental problems of language structure(s). Irom the
linguistic point of view the three papers constitute a major contribution to the
foundations of linguistic typology (this holds especially for the Ifulinj^-I'oint
Relation paper), while at the same time lhe\ pro\ide meticulous analyses of is-
sues in the grammar and semantics of English. Irom the philosophical pt)int of view
the three papers illustrate in great detail how grammar categorizes experience;
they thus lend empirical content and support to Sapir's general statements on the
relationship between grammar and experience, as well as to the relativitv thesis
formulated by Sapir and Whorl.

The place of the three papers within the general project can be defined as fol-
lows, llie paper on Touiliiy and the paper on (jradini^ fall within the treatment of
"Quantity," ' section 7 of the nucleus "Foundations of Language'. While (irudini^
belongs within "Notions applied to quantification,"' itself a subdivision of the
"General introduction to the notion of quantity." Totality represents one t)f the
types of quanlilicalion proper.' Ilie Endiii};-l\>int Relation paper belongs on the
one hand within the nucleus "Foundations of Language," more particularly under
section 4 "Fundamental relational notions and their linguistic expression," and
under section 8 "Space," and on the other hand within the nucleus (or separate
project) "Comparative studies in selected national and international languages"
(and there it belongs in the seclii)n "lormal elements").'

' The treatment of Quantity includes: (I ) (ieneral intnHluctinn \o the ndtinn of quantity: (2) ("lassifjcalion of
quantifiers and quantificates: (3) lypes of quantificatinn; (4) Negation \\\ quantitative expressions; (.S) transfer
of quantitative concepts. See the "Prefatory Note" to /oitililv.

- See the "Prefatory Note" to louiliiv. In the Ikiitorial Note to the b'.tulin\i-P<mu Rclaiion. this nucleus is iden-
tified as a full-scale project titled "loundations of l.anguage, Logical and Psychological, an .Approach to the

International Language Prohlein."

'ITie "Notions" include: (i) Affirmation and negation: (ii) Identity and difference: (in > Indication; (iv ) (iraJing;
(v) Limiting: excluding and gauging: (vi) Composite wholes: aggregation and distribution, (vii) Ratio and pnv
portion; (viii) Normation; see the "Prefatory Note" to loiniilv.

•' The "Types of quantification" include: (a) Singularity and plurality: (b) Number cardinals and fractions; (c)
Totality: (d) Unity; (e) Duality; (f) {Quantification by parlials; (g) Indefinite quantification.

^ Sec the "Editorial Note" to i.mlini^l'unit Rcluiion.

292 General Linguistics I

The three papers are rich in empirical and theoretical content; their impor-
tance and abiding value are discussed by Professor John Lyons, the author of
foundational works in theoretical linguistics and semantics.' Each of the three
papers wcnild merit a monograph-sized study, based on further empirical work,
each with a different focus, so as to bring out the specific merits of each of the
three. For the present reedition in The Collected Works of Edward Sapir it may
suffice to point briefly to these merits.

ITie paper on Totality offers interesting insights on our apprehension of enti-
ties in the world (as individual objects, as sets or as classes), and makes useful sug-
gestions for the study of whole/part relationships, for the analysis of the linguistic
expression of definiteness and indefiniteness, and for the analysis of universality
and generality in languages. From this, linguists, but also psychologists and logi-
cians^ can derive valuable insights. It will not escape the attention of present-day
readers that much of the analyses contained in Totality can be rephrased in a logi-
cally-based model, using arguments, functions and (first and second level) opera-
tors.

The paper on Grading is of interest to linguists, philosophers, psychologists
and scientists in general, in view of the fact that it reveals essential properties
of any descriptive (meta)language or terminology, and basic characteristics of
(implicit or explicit) judging and measuring.*^ Of fundamental importance are
Sapir's remarks on existents and occurrents, on the difference between a polar
(good/bad) vs. scalar {cold/cool/lukewarm/warm/hot) qualification/quantifica-
tion, on the possibility of a static vs. dynamic perspective, coupled with a specific
directionality. In this paper Sapir reflects on basic mechanisms of our linguistic
thinking and of our (linguistic) dealing with the world. In spite of being the most
theoretical of the three papers the study on Grading contains interesting materi-
al for the descriptive linguist'" (e.g., on the linguistic correlates of explicit and
implicit grading, or on the morphological expression of comparison, as well as the
expression of negation combined with qualification, etc.).

*> See, e.g.. Introduction to Theoretical Linguistics (London/New York, 1968); Semantics (London/New York
1977. 2 vols); Language and Linguistics (London/New York, 1981); Linguistic Semantics: An Introduction
(I-ondon/New York, 1995).

See. e.g., the distinction between subsumption and subordination (and the Hnguistic correlates of this distinc-
tion).

^ Sapir speaks of "perceptions of 'envelopment'."

^ Tills probably has to do with the fact thai the paper was written with the intent "to explore the sadly neglec-
ted field of the congruities and non-congruities of logical and psychological meaning with linguistic form."

Pragmatiaans will also be interested in Sapir's observations on the "kinaesthetic feeling" of graded terms and
on ' subjective grading values."

Five: Sfudii's in i 'nixcrsal C\inccptiiiil (intniniur 293

Tlie contemporary reader of these three papers — still largely iiinored hy the
linguistic community — cannot but appreciate their melhi)dological value and the
scrupulously gathered documentary material; on the other hand, one wonders
why these three papers were conceived and written without any relerenee to
related work or related approaches of European scholars.'-

Pierre Swiggers

" One can think here. e.g.. of Hans Rcichcnbach's model proposed in lAvnu-nis of Symhohc Logic (New York.
1947); see also, in the same vein. Norbert Hornstein. "Towards a Ilieory of lensc". l.inKiiislic Inquiry S ( 1977).
521-557. Reichenbach's (basically ternary) model underlies much of current work in the study of lensc svMenu.

'- Useful insights tin the linguistic expression in French or in Russian of the conceptual rclatu>ns studied by
Sapir and Swadesh can be found in Charles Ballys books Precis de snlisiique ((iencve. 19<)5). Iraiie de sivhs-
tiqiie fratn;ai.\c (Hcidclberg/Paris, UX)9, 2 vols.) and l.ininiisiiqttc i^cncriilc ft liniiuiMiqur (rain,ai-sr (Pans.
1932). in Ferdinand Brunot. La pcnsec cl la Iiim;i4r (Pans. 1922). and I.ucien lesnurr I'mi,- rr.imm.tir,- ru\<,r
(Paris, 19.M).

Introduction to Sapir^s texts ^Totality," ^^Grading," and
''The Expression of the Ending-Point Relation"

Sapir is not generally thought of as a semanticist. When he is cited in this
connection in the Hterature, it tends to be either for his famous assertion of "the
psychological reality" of the word as a meaningful unit of analysis even in pre-
viously unwritten languages such as Nootka' or for the support that he gave
(in several publications but most notably in "The Status of Linguistics as a
Science". 1929) to what subsequently became known as the "Sapir - Whorf
hypothesis."

Sapir's authoritative assertion of the fact that the word is not simply the
product of literacy and scribal practice, as some Hnguists had maintained,
undoubtedly played an important and perhaps decisive part in the resolution
of this particular controversy. But it cannot be said to have influenced the devel-
opment of 20th-century linguistic semantics to any significant degree: after all,
it was generally assumed by traditionalists that the word, rather than the mor-
pheme or the sentence, was the primary unit of semantic (and grammatical ana-
lysis), and those who challenged this view had defensible reasons for doing so.
As to the so-called Sapir - Whorf hypothesis, this has certainly been of very con-
siderable historical importance and, having gone out of fashion (if that is the
right expression) in the 1960s, it is once again on the agenda.' It is now generally
recognized that Sapir's view is far from being that of an out-and-out linguistic
relativist or determinist.

Sapir published very httle on semantics as such; or rather, to make the point
more precisely, he published very Httle that he himself referred to as semantics.
There is no chapter entitled "Semantics" in his influential (but deliberately non-
technical) book Language; and, as far as I know, there are no sections that iden-
tify semantics as a distinct branch of linguistics in any of his other works, except
for the three works devoted explicitly to semantics that are included in the
present volume. It must be remembered, however, that in the 1920s and 1930s
"semantics" had a more restricted sense than it does in present-day linguistics: it
usually referred to what is nowadays called lexical semantics. Moreover, at that
time synchronic lexical semantics as an accepted branch of linguistics was still in
its infancy. In the sense in which we now understand the term "semantics," a good

' E. Sapir. Language (New York, 1921). p. 34.

^See now (amongst an increasingly large number of books and articles) Penny Lee, The Whorf Theory
Complex: a Critical Reconslruciion (Amsterdam/Philadelphia, 1996), and John A. Lucy. Language Diversity and
Thought: a Reformulation of the Linguistic Relativity Hypothesis and Grammatical Categories and Cognition: a
Case Study of the Linguistic Relativity Hypothesis (both Cambridge. 1992).

Five: Studies in Universal C onct'ptiiul Grnnimar 295

deal 1)1 Sapirs published WDik. hoih ihcorotical and descriptive, was on seman-
tics, i.e. on the ineaiiing of graniiiialical eateeoiies and conslruclionv Indeed this
is the case lor Chapter .^ ol his book I.nni^iKii^e.

1 he three works on scFiianlics reprinted here — "Totality." '" Ihe Expression of
the Ending-Point Relation in Lnglish, Irench and (ierman" (co-authored with
Morris Swadcsh) and '(Irading" — provide us with the clearest first-hand state-
ment of Sapir's general views on semantics that we are ever likelv to have. lUit
they need to be conte\tuali/ed and. to some degree, interpreted.

Sapir was a theorist, rather than a theoretician: that is to sav. he did not seek to
formalize or make precise and explicit the theory of the structure of language to
which he subscribed. Most of his theoretical principles, moreover, have to be
inferred from his descriptive practice and from his often allusive, en passant,
explanatory comments upon it. Not only was Sapir not a theoretician. As a theo-
rist, in semantics and more generally, he was. in terms of Isaiah Berlin's historv-
of-ideas metaphor, a fox rather than a hedgehog: the fox. it will be recalled,
knows many things, whereas the hedgehog knows only one.' UnUke many struc-
turalists, he was not possessed of a single synoptic principle —the importance of
contrast, binarism. markedness or whatever — to which he subordinated all else,
systematically and tenaciously, in either his theoretical or his descriptive writings.
As far as semantics — the study of meaning — is concerned, he could not but see,
(Ml the basis of his own experience, the inadequacy of the various kinds of reduc-
tionism that were current, at the time, not only in linguistics, but also in philos-
ophy and psychology. And he was perhaps temperamentally disinclined to commit
himself to a single unified and simplifying view about either the grammatical or
the semantic structure of language. As I have said above, he was not an out-and-
out relativist: he had his own \ iew on what we would now call universal grannnar.
But he knew too much about different languages and cultures, and about their
diversity of patterning (to use one of his favourite terms) for him to commit him-
self prematurely to the simplifying general statements that, it must be admitted,
are often a precondition of theoretical advance. For these and other reasons, in
the historiography of linguistics. Sapir is universally acknowledged as a great
scholar and an inspiring teacher and as a consummate descripli\ ist capable of
brilliant intuitive flashes of insight, but not as a great theorist, still less as a great
theoretician. This generalization holds true perhaps for mainstream linguistics as
a whole; it certainly holds for semantics.

Actually, in saying that the three works that are the subject of the present com-
mentary — "lotalitv."' "Ilie Expression of the I nding-Point Kelatu)n'" and
"Grading" — were seen by Sapir himself as making a coniributu>n to semantic
theory I am j^erhaps going bevond the e\idenee. .All three had their origin in

^ Wticn I wrote ttiis article, I was not consciously aware ttiat lldward Sapirs son I'hilip had cited Ihe s.ime ana-
logy. I am grateful to Pierre Swiggers. the editor of the present volume, for drawing mv attention i.> tlnv |i i^ .if
course remarkably apt as far as I£dward Sapir is concerned.

296 General Linguistics I

Sapir's involvement, together with W.E. ColHnson, in the International Auxiliary
Language Association (lALA)/

it is perhaps no accident that the only one of these three works that appears to
have attracted any attention Irom linguists (most of whom either have had no
interest in the International Language movement or have been positively hos-
tile to it) is "Grading," which, when it was published (posthumously) in 1944, bore
the subtitle "A study in semantics" and, in its final version at least, was written up
as such and made no prefatory mention of the large-scale project sponsored by
lALA of which originally it was, presumably, just as much a constitutive part in
Sapir's mind as were "Totality" and "The Expression of the Ending-Point
Relation." The fact that the work reported in "Grading" had been completed as
part of the same lALA-sponsored project some considerable time ("many
years") before it was written up and submitted for publication is mentioned in
the brief historical note that is appended to it, but there is no reference to this
fact in the actual text of the article. Moreover, the same note (written in the first
person) assures us, as does the subtitle, that "Grading," in "essentially its present
form," was indeed seen by Sapir as a contribution to general semantic theory,
independently of any practical application that it might have. His purpose, we are
told, was to encourage others "to explore the sadly neglected field of the con-
gruities and non-congruities of logical and psychological meaning with linguistic
form." Although we know from other sources that the article had been left by
Sapir in "a relatively unfinished state,"' there is no reason to doubt that this for-
mulation of Sapir's purpose is reliable. The lALA Statement makes it clear,
anyway, that the work in which Sapir and Collinson took the lead was intended
to be of use, not only "as a norm for the structure of an international language,"
but also "as a basis for general language study." And Collinson's 1937 monograph
on "Indication," which appeared in the same series as "Totality" and "Ending-
Point Relation," is commonly cited in the literature as a contribution to seman-
tics. ("Indication" is the term used by both Collinson and Sapir for what is now
called deixis.)

It is perhaps idle to speculate further on the question whether Sapir himself
saw "Totality" and "The Expression of the Ending-Point Relation," as well as
"Grading" as studies in semantic theory. In retrospect, they can certainly be read
as such. Even if we did not have the appended historical note in "Grading" to this
effect (and Mandelbaum's helpful editorial comments)' it would be evident to
anyone reading them from this point of view that all three are terminologically
and conceptually consistent. In "Grading" the descriptive analysis of data is illus-
trative and takes second place to the theoretical points that are being made,
whereas in the other two works the opposite is the case.

* See also the introduction to section IV here.

-"' Cf. Zellig Harris, in his revie'
D.G. Mandelbaum; Berkeley, 1

^ Selected Writings, o.c. pp. 5-6

r^^' w '''! ^^''"^' '" ^'^ review of Edward Sapir: Selected Writings in Language, Culture and Personality (ed.
D.G. Mandelbaum; Berkeley, 1949) in Language 27 (1951), p. 289 [reprinted here].

Five: Sfudii-s in I Hivcrsul Concrplitd! Cnirnrntir 297

It is impossible, in the present eontext. to cK) lull jiistiee to the wealth ol dclail
that is eontained in eaeh ol the three works. Sullicc it lo sav that all ol them are
marked by .Sapir"s aeknouledged bnllianee ol deseriptive iiismhl. lo which I
referred abo\e. and that much oi the em[inieal data that they conlain is. ui my
judgment, \alid and. to the best ol m\ knowledge, original. Semanlics, and more
particularly, the investigation ol "the congruitics and non-congruilies of logical
and psychological meaning with linguistic form" in the domains of vocabulary
and grammar with which Sapir was concerned, is no longer the "s.Klly neglected
field"" that it was in the l^^2()s and earl\ 193()s. The conceptual (and lerminolo-
gical) framework with which, not only linguists, but also psvchologists and logi-
cians, operate nowadays is very different from what it was then. Ilieoretici/ation
and formalization have made great strides in all three disciplines in recent vears;
and theoretically minded foxes, not to mention the theoreticians and hedgehogs
of semantic theory, express themsebes very differentb from the wav in which
most of them did sixty or seventy years ago. Consequential allowances must the-
refore be made by present-day readers of the three works that are the subject of
the present commentary for what might otherwise strike them as looseness of
expression and imprecision. But it is well worth while their making such allow-
ances. There is still much to be learned from them, not only by theorists and
theoreticians of the subject, but also by descriptivists and those of a more empi-
rical bent of mind.

Tlie first general point that needs to be made by way of commentary has to do
with Sapir's structuralism. He was a structuralist, not in the narrower sense that
this term acquired in the 1950s in what we tend to think of as mainstream
American linguistics, but in the broader sense that it had always had in anthro-
pology and the other social sciences and in European linguistics. And he was just
as much a structuralist in semantics as he is universally recognized to ha\e been
in phonology and morphology. He nowhere makes this absolutely explicit. I^ut
his concern with structure (or to use his own term puitcrnin\^) is evident through-
out. All three works are studies in what we would now call structural semantics.

But Sapirs structural semantics, unlike the more classical. Saussurean or post-
Saussurean. \ersions. is dynamic rather than static. I do not mean by this that he
did not respect the distinction between the synchronic and the diachronic: he cer-
tainly did. Nor do I mean that he thought of a language as a metastable system
in which (to use the Prague Circle slogan) there is diachrony in synchrtmy: he
may well have done so. The dynamism that I am referring to here is psycholo-
gical, not chronological.

L.ike all struetina! linguists, he looked upon languages as relational systems —
as systems in which all units (phonological, morphological and lexical) tierive
their value from the relations in contrast and equi\alence that the\ contract with
one another in the system. And Sapir frequently employed the term "relation"" in
this connection exactly as other slruetinalisis ^o. lUii loi hun they were ni>t the
static relations of a kinetic s\stem in which e\er\lhing was in flux or. better,
under tension. This comes out most elearlv, perhaps, in his definition oi equality
as "a more or less temporary point ol passage or equilibrium between "more

298 (h'lH'nil Lifii^uistics I

than ■ and "less than" or as a point of arrival in a scale in which the term which is
graded is eonstantly increasing or diminishing" ("Grading", p. 105). Once we
have reeoizni/ed this as Sapir's view in the case of relations of equivalence (or
equality), it is easy to see that this is also his view for all the semantic relations
that he discusses, not on\y in 'Grading," but also in the other two works. There
are perhaps parallels in the work of some other scholars in the structuralist tra-
dition. But Sapir's view of logical form — for this is, in effect, what it is — is a very
different view from that of logicians or present-day formal semanticists.

The fact that Sapir's notion of logical form differs from the standard view, in
the way that I have indicated, does not mean that it should be dismissed by pres-
ent-day theorists and theoreticians as unworthy of serious consideration.
"Grading" has already had its influence, directly or indirectly, on modern treat-
ments of comparative constructions and antonymy. For example, it is now wide-
Iv. if not universally accepted by linguists that gradable antonyms, such as "good"
and "bad,"" are always implicitly, if not explicitly, comparative; and various ways
of formalizing this insight have been proposed. But there is much more than this
to be learned about gradable antonymy and comparatives from Sapir's work.

There is perhaps even more to be learned from Sapir about the topics that he
deals with in "Totality" and "The Expression of the Ending-Point Relation,"
including quantification, whole-part relations, the mass-count distinction, loca-
tive and directional constructions and telicity (to use current terminology), which
are acknowledged to be of central importance in linguistic semantics and have
been researched intensively in recent years. It is worth noting, in this connection,
that in one respect at least, as far as structural semantics in America is concerned,
Sapir anticipates later theoretical developments: his treatment of what he calls
totalization is often implicitly, and occasionally explicitly, componential.

Sapir's discussion of the data that he adduces is always subtle and interesting;
and the points that he draws to our attention need to be accounted for, even
though, in accounting for them in a contemporary framework, we may find
ourselves invoking distinctions that were not part of the linguist's stock-in-trade
in his day: between sentences and utterances, between competence and perfor-
mance, between semantics and pragmatics, etc. In some cases, we may conclude
that what Sapir attributes to language itself should be handled in terms of
principles or conventions which govern the use of language: e.g., in terms of what
is presupposed or implicated, rather than of what is semantically encoded in
the language-system. We may even conclude that the dynamism that I have noted
as being so characteristic of Sapir's structuralism is a matter of performance,
rather than competence, and should be handled, not in linguistic semantics as
such, but in pragmatics or psycholinguistics. But this conclusion should not be
drawn too hastily. It is still an open question whether the theoretical distinctions
to which I have just referred (as they are currently drawn by most theoreticians)
are soundly based or not.

I must end this brief commentary on a note of regret. One of Sapir's great
strengths was of course his intimate knowledge of a wide range of typologically
different languages operating in a variety of cultures. He frequently draws upon

Five: Sliidit's in I 'nivcrsal Cotucptiml (ininnnur 299

ihis know IclIuc in other works; and he would most ccitaiiiK have done st» in his
projected work on the Liraninialieal eateeor\ ol as|ieet (which would presuniablv
have been consistent with the conceptual Iraiiiework used in "Ilie Impression of
the Ending-I\)int Relation '). In the three works on semantics reprinted here he
often supports ihe generalizations he makes by referring to "manN languages."
but he does not identify these languages by name or family. "Ilie F:xprcssion of
the Ending-PiMnt Relation/" unlike "( Irading" and "■'rotalitv," is. of course, expli-
citly comparati\'e with respect to f-nglish. Irench and Cierman; and there are
interesting points ol dillerenee among these three languages. Bui ihey arc far
from being as interesting from a typological point ol view as the structural differ-
ences to which Sapir. famously, drew the attention of the scholarly world in some
of his other publications. Because it is Sapir who is making the generali/ations
one can perhaps take them on trust. But some o\' them are no doubt checkable
now on the basis o\' work done from a typological point of \iew (much of it bv
Sapir's students) in the years that have passed since the works reprinted in the
present volume were written. It would be good to have them checked for parti-
cular languages and reformulated in the light of more recent advances in gram-
matical and semantic theory. Tlie fact that Sapir's three papers on semantics,
lexical and grammatical, are now being reprinted in his Collecicd Works should
facilitate this task.

Sir John l-^ONS

Totality (1930)

PREFATORY NOTE

Tho present paper on Totality is but the first instalment of a general
work on language entitled Foundations of language, logical and psychologi-
cal, an approach to the international language problem, by William E.
Collinson, Mrs. Alice V. Morris, and Edward Sapir, edited by Mrs.
Alice V. Morris. This work is sponsored by the International Auxiliary
Language Association. In the preUminary work that has already been
done by the three collaborators it soon became evident that many
questions of a theoretical and analytical nature would have to be taken
up that are of perhaps minor interest for the solution of purely practical
problems. The work is intended as an aid toward such solution and
for the eventual use of all those who are interested in fundamental
problems of language structure, whether they wish to make practical
appUcations of the insights secured to the international language
problem or not. This and other papers that may follow from time to
time are printed in this place because of the general linguistic interest
which it is hoped they possess.

The nucleus of Foundations of Language consists of :

1. Introduction

2. Units of communication: the sentence

3. Fundamental types of referents: a natural basis for parts of

speech

4. Fundamental relational notions and their linguistic expression

5. Notions of order

6. Indication

7. Quantity

8. Space

9. Time

10. Existents and their linguistic expression: the noun

11. Occurrents and their linguistic expression: the verb

12. Modes of existence and occurrence : the adjective and the adverb

13. Notional inventory of experience: its expression in nuclear

words

14. Extension of the nuclear vocabulary : word-building

15. Expressiveness in language

Five: Sludics in t'nivrrsul ( Otucpimil (irummar 301

The present paper on Totality is merely a section of that part of
Foundations of Language which deals with Quantity. Its phice in the
total scheme is indicated by the followin^!; table of contJ'nts of the
eventual work on Quantity:

1. General Introduction to notion of quantity

(a) Notions appUed to quantification

i. Affirmation and nep;ation

ii. Identity and difference
iii. Indication
iv. Grading

V. Limiting: excluding and gauging
vi. Composite wholes: aggregation and distribution
vii. Ratio and proportion
viii. Normation

(b) Quantifiables

calculating

(d) Quantificates

2. Classification of quantifiers and quantificates

3. Types of quantification

a. Singularity and plurality

b. Number: cardinals and fractions

c. Totality

d. Unity

e. DuaUty

f. Quantification by partials

g. Indefinite quantification

4. Negation in quantitative expressions

5. Transfer of quantitative concepts

It is a great pleasure to record my indebtedness to Mrs. Morris for
her unflagging zeal in the prosecution of the work, of which Totality is a
harbinger, and for the extraordinarily acute criticism that she hiis
exercised at its every stage. There is hardly a paragraph in the present
paper which has not profited from her keen int-crest. To Professor
Collinson too I am indebted for numerous and valual)le suggestions.

302 General Linguistics I

CONTENTS

Glossary of technical terms 6

I. Whole-part relation and types of totahzed existents and of

totality 7

II. Existents 'individuahzed' and 'indefinitely massed' and their

totalization 11

III. Singularized totaUzers 12

IV. Definite and indefinite totalizers 15

V. Direct and calculated totahzers 16

VI. The all of universal statements 17

VII. Evaluated totaUzers 19

VIII. Modified totahzers 20

IX. Negated totalizers (with table) 21

X. Speciahzed totalizers 23

XI. Quantificates involving totahzation 24

Table of totahzers 28

Five: Siudics in I nnr/Mil C Onccptiial (irummar 303

GLOSSARY OF Ti:CHNICAL TKRMS

1. Quantifiable: anything concerning which quantity may he predicated.

It aiLswers to the question 'So and so much (many) what*'" or 'To
such and such a degree what?'

2. Quantifier: a tenii which expresses any quantitative judgment. It

measures or counts or grades a quantifiable. It answers to the
question 'How much or How many or To what degree so and so?'
A term which cannot answer one of these questions is not a quan-
tifier.

3. Quantifix^ate: a quantified quantifiable; a linguistic expression which

indicates that a quantifiable has been quantified. It answers to
the question 'How much (many) of what?' or 'To what degree
what?' It may serve to quantify further and then l)ecomes a
quantifier called a quantificate-quantifier (specialized quantifier).

Note (A. V. M.): Take the sentence 'Four men are coming'. To the
question 'Four what?' the answer gives the quantifiable 'men'. To
'How many men?' the answer gives the quantifier 'Four'. To 'How
many of what?' the answer gives the quantificate 'Four men'.

In 'That rose is very red', the question 'very what?' eHcits the
quantifiable 'red'; 'To what degree red?' elicits the quantifier 'very';
and 'To what degree what?' elicits the quantificate 'very red'.

'A cupful' as such is a quantificate; it means 'what is contained by a
full cup'. 'A cupful of tea' is a quantificate of a luKher order in
which 'tea' is the quantifiable and 'a cupful of the quantifier (quanti-
ficate-quantifier, specialized quantifier).

4. Direct quantifier: a quantifier which is directly apprehended, without

the necessity of arriving at the intended quantity- by going through
one or more mathematical operations, e.g. /oi/r, sicarjn.

5. Calculated quantifier: a quantifier which is apprehended mediately,

with the necessity of arriving at the intended quantity by going
through one or more mathematical operations, e.g. the sum of,
the whole computed acreage of.

6. Totalizer: a quantifier whose function it is to emphasize the fact

that in the given context the quantifiable is not to be thought of as
capable of increase, e.g. all, the whole flock.

304 General Linguistics I

1. WHOLE-PART RELATION AND TYPES OF TOTALIZED
EXISTENTS AND OF TOTALITY

The notion of totality may be considered as instinctive in the sense
that it so easily and necessarily arises from experience that it is difficult,
and for most individuals probably impossible, to reduce it to simpler
terms. It seems possible, however, to derive it psychologically from
two sorts of experience: (1) the feeling of rest or of inabihty to proceed
after a count, formal or informal, has been made of a set or series or
aggregation of objects; (2) the feeling of inability or unwilUngness to
break up an object into smaller objects. These feelings, which may be
schematically referred to as the 'all' and the 'whole' feefing respectively,
are correlative to each other. They arise naturally from experience
with objects.

The mind views objects as if they functioned in two ways, tending, on
the one hand, to keep their distance from each other — e.g., a table as dis-
tinct from a chair; on the other hand, to cohere in functional units — e.g.,
a table and a chair as necessary parts of a set of furniture and as jointly
excluding such other objects as people. Furthermore, such objects as
tables and chairs may frequently be viewed as falling apart into separate
segments, objects of a secondary or functionally lower order, which
may, actually or in imagination, be reassembled into the 'whole' table
and the 'whole' chair. We may count the segments of a table until
we have 'all' of them needed for the reconstruction of the table, just as
we can count the pieces of furniture needed to make up 'all' the members
of a set of furniture.

But there is an important difference between the segment as related
to the table and the table as related to the set of furniture. The
segments have little or no meaning as such. They are merely constit-
uents or functionally meaningless fragments, which the mind at once
reassembles into a continuous structure. On the other hand, the table
is a significant entity in itself and can be made to cohere with the chair
only in a mental sense because of the unitary 'meaning' given to the
concept of a set of furniture. The more value we attach to this remoter
unity of the set, the greater becomes its psychological resemblance to
the more immediately given unity of the table itself, so that we may

Five: Sdulii's in rnncrsul Ci*t\n'pimil (Iratrinuir 305

finally complain of 'a broken set', with its concotnitant feeling of
vexation. The more vivid the feelinj; of unity of the .s<'t, the more
applicable to it is the term 'the whole set'; the less vivid the frelinp of
unity, the more easily we resign ourselves to saying *all the pieces of the
set'. The more vivid the feeling of unity of the table, tix- more ap-
plicable to it are such terms as 'the whole table' or 'the entire table';
the less vivid the feeling of unity, the less unnatural it l)eromes to
speak of it as 'the aggregate of such segments as go to make up the
table' or 'all the table-segments'.

We may consider these feelings of 'all' and 'whole' as abstracted from
our apprehension of the whole-part relation in existents. If we think
of the whole-part relation, as psychologically we must, as involving
operations or kinaesthetic experiences in keeping, disintegrating, and
aggregating existents, then, if we wish to classify existents in terms of
the whole-part relation, we may describe them in terms of those o [Mira-
tions and we may distinguish them broadly as non-totalizable and
totalizable. The former kind may be called the 'kept' existent, that is,
the object preserved as such, thought of as such, neither analyzed nor
aggregated. Such an existent can not be totahzed, only individualized,
e.g. Hhe table', Hhe cheese'. The totahzable existent is one which is
thought of as susceptible of various kinds of aggregation, either direct or
based on some previous operation of disintegration. When we proceed
to apply to such existents the various operations of totaUzing, we are
driven to analyze them into six types.

1. The existent thought of as divisible into parts but as 'resisting'
such division. Such an existent is given as totahzed from a px)tential
aggregate, e.g. Hhe whole table', Hhe whole quantity oj water'. This type
may be named 'whole existent'.

2. The existent thought of as an aggregate of parts derivable from a
normally undivided existent. Such an aggregate may be said to 'st^ek'
totahzation in the form of type 1, e.g. 'aW the parts of the table', con-
sidered as psychologically 'self-driven' toward 'the whole tal)le'. 'All of
the tal)le' more or less adequately expresses this tyix' of t^)talized
existent. 'All of the table' is to 'the whole table' as 3/3 is to 1. This
type may be named 'summated existent'.

3. compounded of 1 and 2. The existent thought of as di\isible
into parts but apprehended as persistently resisting deformation into
a summated existent (type 2), e.g. 'The irholv of the table is of oak',
Hhe whole o/the water is boiling'. Here the mind seems to linger on the
wholeness of the existent ratlicr than on the existcMit its(>lf. This type

306 General Linguistics I

may be named 'persistently whole existent' or 'reassertedly whole

existent'.

4. The aggregate of existents, each of which is considered as having
functional reality. Such an aggregate does not 'seek' totahzation but
may be said to 'be driven' to it, e.g. 'all the tables (or cheeses)', con-
sidered as psychologically 'driven' toward 'the whole set of tables (or
cheeses)'. The 'all' of such an aggregate consists either of a set of
terms, say an accidental number series or an accidentally isolated
number of existents, that do not cohere except in so far as we make them
cohere by mental segregation from other existents — e.g. 'three'; 'jive
tables'; 'five cheeses'; 'all the tables (or cheeses)' — (totality arrived at
by mere enumeration) ; or else of a set of terms, say 'a set o/ tables' in a
given room, or 'a collection of waters' in an exhibit, which can be thought
of as having function apart from the mere fact of aggregation (totality
arrived at by some kind of formal assemblage). This type may be
named 'aggregate' or 'simple aggregate'.

5. The aggregate thought of as divisible into members or parts and
as 'not resisting' such division. Such a collectivity has been totaUzed
from an actual aggregate, e.g. 'the whole set of tables'. This type may
be named 'whole aggregate'.

6. compounded of 4 and 5. The aggregate apprehended as threat-
ening, as it were, to fall apart into a simple aggregate whose totality is
'air, and such threatened deformation being resisted but resisted in
vain, e.g. 'all of the set of tables', or, more briefly, 'all of the tables (or
cheeses)', which is really a compound totahzed existent in which the
totalizing feehng properly apphcable to a summated existent (type 2)
is applied to a whole aggregate (type 5). We naturally say 'all the
tables' but, preferably, 'all of the tables in this set'. This type may be
named 'relapsed collection' or 'reasserted aggregate'.

It is probable that of the types of part-whole existents the ones that
are nearest to intuitive or primary observation are the 'kept' non-
totalizable existent, the whole existent (type 1) and the simple aggregate
(type 4). The whole aggregate (type 5) follows readily on the heels of 4
by the apphcation to it of the integral feeling characteristic of 1 ; in
other words, aggregates may be looked upon, psychologically, as second-
ary objects created by the transfer of the feeling of individuality to an
aggregate. To put it in slightly different terms, 'the whole set of
tables' is a metaphor, based on 'the whole table', which reinterprets
'all the tables'. Type 2, a summated existent, is perhaps the most
sophisticated of all part-whole existents, and results from the application

Five: Siitdii's in I tiivcrsul C Onrcptiial (inininuir 307

of tlio aggregational fooling characteristic of a simple aggrepate (ty|)e 4)
to a whole existent (tyix* 1). This transfer nicchunisiM may Ik- exprrssed
by saying that 'all the parts of the table' or 'all of the table' is a meta-
phor, based on 'all the tables', which reinterprets 'the whole table'.
Naturally, the experiences which underlie these transfers are, on the one
hand, the coalescence of countable units into functionally close-knit
sets, e.g. of the tables and chairs and certain other ol)jects of a well-
planned room into a definite set of furniture; and, on the other hand,
the frequent psychological l)reaking up of an object into countable
fragments, the object itself being retrospectively analyzed into the
ordered sum of these destined fragments.

Taking the six types of totalized existents, we get the following scheme
of terms of totaUty:

Neutral term : totality.

1. Totality of an existent which resists division, e.g. 'The whole
table is well made', 'the whole /quantity o//water is infected'.
(Totality of a whole existent.)

2. Totality of parts of a normally undivided existent, e.g. 'All parts
of the (or all of the) table are (or is) well made', 'all of the water
is infected'. (Totality of a summated existent.)

3. Totality of an existent which persistently resists division, e.g.
'The whole of the table is of oak', 'the whole of the water is in-
fected'. (TotaUty of a persistently whole existent, or of a
reassertedly whole existent.)

4. Totality of existents in a collection, e.g. 'All the tables have
been brought in', 'all the /kinds o// waters are bottled'. (Total-
ity of an aggregate, or of a simple aggregate.)

5. Totality of an undivided collection or individualizeil aggregate*,
e.g. 'The whole set of tables forms a long line'. (Totality of a
whole aggregate.)

6. Totality of an aggregate which has vainly resisted thR\atened
deformation, e.g. 'All of the set of tables are of oak', 'nil of the
tables (in this set) come from France'. (Totality of a relapsed
collection, or of a reasserted aggregate.)

308 General Linguistics I

II. EXISTENTS 'INDIVIDUALIZED' AND 'INDEFINITELY

MASSED'

So far we have been considering whole-part existents without re-
ference to whether or not they have an assignable structure. Existents
may be conceived of either as having structure, e.g. the table, the land
belonging to X, the cheese formed into a definite object, or as not having
structure, e.g. the wood as material, the land extending indefinitely,
the cheese thought of as food rather than as shaped. The former type
may be called 'individuaUzed', the latter 'indefinitely massed'. In
many languages these two types of objects tend to be differently
totaUzed.

In EngUsh, totality of an individualized object tends to be expressed
as in Hhe whole table', Hhe whole land was annexed'; and totality of an
indefinitely massed object as in 'all the milk has turned sour', 'all the
land was inundated'. Observe that such terms as 'the whole of the land'
(type 3) and 'all of the land' (type 2) apply only to individualized objects.
The all of indefinitely massed objects is particularly suited to such
abstract entities as 'talk', e.g. 'all the talk is unnecessary'. 'Th£
whole talk' necessarily refers to a limited discourse (type 1), 'all of
the talk' to a potentially divided discourse with sections or time-
measured parts (type 2), 'during the whole of the talk' to a discourse
conceived of as a cumulative unit (type 3). Existents which are logi-
cally aggregates, such as 'piled up wood', may metaphorically be con-
ceived of as indefinitely massed objects, hence 'all the wood', rather
than 'all the pieces of wood' collectivized into 'the whole pile of wood'.^

^ A note of warning. It is not claimed for a moment that the ordinary EngHsh
uses of 'the whole', 'all of, 'the whole of, and 'all' necessarily correspond to our
exacting distinctions, merely that they tend to do so. In actual practice there is
considerable confusion.

Five: Sfmlii-s in Universal Cttmcptual (inutiniur ?>W

III. SINGULARIZKD TOTALIZERS

There are still other distinctions in the exprcsvsion of totality that
must be recognized. One of the most important of thcsf is the difference
between the itemizing or singularizing totalizers and the ordinary
aggregating ones. The former, of which 'each' and 'every' are the type,
do not directly express totahty but definitely imply it in a reference
which is individual in form. In other words, 'every a' singles out a
particular a only to emphasize the point that all the other a's of the set
differ in no relevant respect from it. 'Each' is used preferably for the
members of an aggregate of two or, in any event, an aggregate which is
numerically small. There is considerable confusion in usage, however,
just as there is between the corresponding reciprocals 'each other' and
'one another'. 'Each' differs further from 'every' in indi\iduahzing
more clearly at the expense of the notion of totality, so that at times
this notion becomes quite pale, whereas it is always strongly imphed
with 'every'. 'Each' is therefore particularly appropriate in all con-
texts in which the 'every' of an aggregate is looked upon as selected or
otherwise specifically determined; hence 'each member of a series A as
compared with the corresponding member of a series B', where 'every'
would have a blurring, aggregating effect. On the other hand, 'not
every instance', not 'not each instance', because with a negation there
can be no true individuaHzing. Note that the greater individualizing
force of 'each' and its restricted reference to a small or easily com-
prehended aggregate are really two facets of a single psychological
fact ; for the smaller the aggregate, the more individuahzed its meml>ers
tend to become. Perhaps the most accurate distinction that can l)e
made between 'each' and 'every' is this, that 'each' tends to mean 'all,
taken one at a time' (in other words, is the meaning of 'all, taken n at a
time', when n = 1), whereas 'every' is rather an 'nil, accumulating by
increments of one'.

'Each' and 'every' apply primarily to aggregates of type 4, but, just
as the notion of 'all' may blend with that of 'the whole (set)' into 'all of
(type 6), so they too blend with it into notions expressed by such terms
as 'each o/ the men', 'each one o/ the men', 'every one of the men'. It is

310

General Linguistics I

natural to say 'every one of the men in that regiment' but not 'every one of
the men in America', for which 'every man in America' or 'every Ameri-
can' would have to be substituted, inasmuch as the totality of men in
America forms not a significant assemblage but only a statistical
aggregate.

It is interesting to observe that the relative magnitude of 'each' and
'every' is reflected in their positions in the compound cumulative
itemizing totaUzers 'each and every man', 'each and every one of the men'.
This feehng may be symbolized as :

The converse feeling is approximately expressed by 'all and sundry':

Another cumulative totalizer of the 'each and every' type is 'one and
all', in which the itemizing is so rapid that it is swallowed up, as it were,
in the totalizing notion:

Five: Sfudirs in I'nivcrsdl Conceptual Cninmuir

311

The converse of this seems to be expresseti by the use of 'to a inan" in
such locutioDS as 'They were annihilated to a man' :

It is possible also to express the cumulative itemizing feeUng without
explicit reference to the notion of totality, as in 'to the last man' in such
locutions as 'They were annihilated to the last man' :

the last (man)

the last (man) <— to

(all)
(all)

Inasmuch as the logical 'all' is not actually felt here as either a total
arrived at or a total given and then itemized, the feeling of direction in
'to the last man' is ambiguous. It may be interpreted as proceeding
either from the single instance to the totahty imphed by the completion
of the count or from the previsaged totality back to the single instance.

312 General Linguistics I

IV. DEFINITE AND INDEFINITE TOTALIZERS

A more difficult distinction to make is that of definite and indefinite
totalizer. In a sense all totalizers are definite, for in a given aggregate
'all' cannot be ambiguous or indefinite in the sense in which 'some' is
indefinite. Nevertheless there is a real difference as to definiteness
between 'all the people in the room' and 'all the cardinal points',
i.e. 'all four of the cardinal points'. In the former case the totahty is
predicated of a number which need not be known, in the latter of a
known but unexpressed, i.e. an implied, number. 'AH' in 'all the
people' is an indefinite totaUzer; 'all' in 'all the cardinal points' is an
imphcitly definite totahzer; 'all four' in 'all four of the cardinal points'
or 'the four' in 'the four of them' is an expUcitly definite totaHzer.
Many languages possess special terms for certain explicitly definite
totahzers, particularly for an aggregate of two, e.g. Enghsh 'both the
men', 'both of the men', 'both parts of it', where 'both' is the equivalent
of 'all two', cf. French 'tous les deux'. Such terms as 'the regiment'
(type 5) or 'the pint' (type 1) in 'he drank the pint of milk' or 'the deck
of cards' (type 6) may be looked upon as implicitly definite totahzers
of specialized types.

Five: Smdii's in L nivcrsul ConccfUiinl Clnntimiir 313

V. DIRECT AND CALCULATED TOTALIZERS

So far we have been dealing with direct totaHzery, which, like all
quantifiers (i.e., terms indicating quantity), may be one-term (}uantifier8
(e.g. allj the whole of), quanto-quantifiers (e.g. hath = all the two, all the
three parts of), or compound quantifiers (each and every one of).

There are, naturally, also calculated totaUzers, of which the additive
type is particularly common, e.g. the sum (as used in mathematics'),
the sum total, and, of a somewhat more complicated order, terms of the
type net total, in which there is an implication of going through more than
one mathematical operation.

* Dififerent, of course, from the sum in such locutions as 'the sum of money which
I give you', which is really an indefinite quantifier, equivalent to the amount of.

314 General Linguistics I

VI. THE ALL OF UNIVERSAL STATEMENTS

Until now we have dealt only with the concept of totaUty in 'enumera-
tive' (totalized particular) statements (e.g. 'All the people in this room
are wealthy') or in singular statements (e.g. 'The whole cheese is spoiled').
The question arises of whether or not to look upon the all of universal
statements (e.g. 'All men are mortal') as a true totahzer. There is no
question here of a true totaUty, explicitly or implicitly definite or even
indefinite (as in 'all the men in this room', where no count has been
made), but of a class. Any example of the class, namely a man or this
or that particular man, is not, strictly -speaking, a definite part, fraction,
of the class (which might then be said to be diminished by one if the
example is singled out and dismissed), but merely a particularized repre-
sentative of the class itself. Hence 'All men are mortal' says no more
than 'Men are mortal' or 'All the men who can be thought of are mortal'
or 'The men who can be thought of are mortal' or 'Every man is mortal'
or 'Any man is mortal' or 'Any m£n are mortal' or 'A man is mortal'.
All these locutions are merely periphrases for the more logically ex-
pressed universal statement 'Man is mortal'.

All and every in these examples are, then, not totalizers in the strictly
logical sense but class-indicators. Inasmuch as the notion of class
arises, in experience, from the accumulation of particular instances, the
illusory feeling is produced of a prolonged count looking eventually to a
closed total (all). In other words, all in 'All men are mortal' arises
psychologically from the unconscious solution of the illogical proportion :

This man (of those who are here) : all the men (who are here), a sum =
this man (of all possible men) : x, an unknown sum.

But X is not a sum, but a class of indifferent membership as to number.
The pseudo-totalizing nature of all in universal statements comes out
clearly in English when we try to substitute all of (type 6, totality of a
reasserted aggregate) for all, e.g. 'all of men are mortal', which is impos-
sible, or 'all of the men are mortal', which at once takes on the meaning
of a totahzed particular statement, the reference being to a particular
aggregate of men. This is because all of, with discrete existents,

Fix c: Studies in i luvcrsul Conccpiuul Gnuninar 315

expresses totality of a relapsed collection or of a reasserted apprepate,
and the class 'man' is not such a collection or apKr('pat<\ It is, further-
more, significant that in spt^akinp; of continua or discretes thought of as
continuous, whether true (e.g. 'cheese') or metaphorical (e.g. *l)eauty*,
'man'), in universal statements one cannot easily use the totaUzer
appropriate to continua (e.g. 'the whole cheese', 'the whole beauty of her
face', 'the whole of mankind', thought of metaphorically as a collection),
but generally has recourse to the pseudo-totalizing all used for universal
statements in speaking of discretes. Hence 'All cheese is a food',
'all beauty is perishable', 'all man (or mankind) is mortal' are formally
analogous to 'all nieii are mortal'. Such universal concepts of continua
may also be expressed in still other analogous ways, e.g. 'cheese is a food',
'all the cheese that can be thought of is a food'. If whole, the true
totalizer of such existents as 'cheese', cannot be used to universalize
them, it follows that, by analogy, the universalizing all is not even
hnguistically,let alone logically, identical with the true totalizer af/, but is
merely a transferred use of the latter.

316 General Linguistics I

VII. EVALUATED TOTALIZERS

We have dealt only with those totalizers which are abstract and non-
evaluated, that is, pure, denoting quantity and nothing else. But
abstract totaUzers, like other quantifiers, may also be evaluated, that is,
they may include a reference to some notion of an evaluating category,
as in complete, full, enough.

Such terms as full are used so frequently that they seem to have the
value of pure totalizers, but actually they are totalized terms of quality.
Thus, full is a capacity-totahzed term of quality, applying, strictly
speaking, only to volumes in space.

Some of the evaluated totalizers are :

full (totalizer applying primarily to space-bounded existents, with
feeUng-tone of satisfaction at cessation of operation of fiUing;
frequently transferred to other types of existents).

enough, suficient (totalizer of norm, with feeling-tone of satisfaction
at fulfilment of given requirement).

complete (structuralized totalizer, with feeUng-tone of satisfaction
because of attainment of quantity aspect of structure).

entire (totalizer of negated absence, with feeling-tone of satisfaction
because of inability to find any part missing).

intact (totaUzer of negated interference).

perfect (totaUzer of negated blemish, or of negated absence, where
absence is felt as blemish; cf. value quantificates, pp. 25-7).

It is worth mentioning that such terms as all and whole are not in-
frequently used in a secondary, evaluated sense rather than in their
primary, abstract sense. In such a sentence as 'It's all of a mile', 'all of a
mile' really means 'more than enough to be called a mile'; if we rephrase
the sentence as 'It's a good mile', we get a clearer linguistic expression of
the evaluative feehng-tone of all of a. All of a and a good are, in effect,
over-graded totaHzers of norm of explicitly definite type (see section IV).
In such sentences as 'Whole lands were depopulated', 'Whole cheeses
were devoured', the whole is far from being a pure totaUzer; it has the
feeUng of negated absence that belongs to entire. Even more complex
in its logical analysis is such a term as utter in the sentence 'She is an
utter failure', which is a totaUzer of negated absence with strong negative
affect (schematically: deplorably entire).

Five: Sdulics in [ 'ni\ crsal Concvptiml Grunimur 317

VIII. MODIFIED TOTALIZERS

Pure and evaluated totalizers, like other quantifiers, may he modified,
but certain types of modified quantifiers are naturally absent from the
full set of totalizers. Thus, there can hardly be a praded totalizer
parallel to such terms as few, many, more, and most. If we ^'dy fuller or
the most complete, it is not because we naturally grade concepts of
totality but merely because such evaluated totahzers as full and com-
plete may be taken in a derived or fiparative sense which allows of
grading. Thus, fuller imphes a full which really means abundant, a
concept that belongs to the sphere of much or many, not to that of all.
Again, there can hardly be a selective totalizer parallel to such terms as
some or any unless wg consider all of in 'all of the table' (type 2) and
'all of the tables' (type 6) as such selectives, inasmuch as they are the
totahzing limits of such series as 'half the table', 'three fourths of the
table' and 'half of the tables', 'three fourths of the tables' resjDectively.

Among the more important of the modified totahzers are: the dis-
tributive totahzers, corresponding to such terms as one by one, two each,
e.g. distributively all, which may be rephrased in such terms as all of one
{set) by all of another {set), by totals, in whole sets; the selective distributive
totahzers, e.g. all of each; the hmitative totahzers, e.g. quite all, abso-
lutely all, 'just the whole of it must be utihzed, not merely a part'; and
exclusives, e.g. 'you can buy only the whole set, not just one or two
pieces'.

318 General Lini^uistics I

IX. NEGATED TOTALIZERS

Negated totalizers, as. a rule, have the force of partials. In other
words, the idea of totality is negatived or Ufted, as it were, leaving the
mind free to roam over the whole range of quantity l>ing on the minus
side of totahty. In English, however, these partials are not true
affirmatives, but merely positives in the corresponding negative state-
ment, i.e. 'Not everybody came' does not mean 'Some came', which is
implied, but 'Some did not come'. Logically, the negated totaUzer
should include the totaUzed negative, i.e. opposite or contrary, as a
possibiUty, but ordinarily this interpretation is excluded and the
totalized negative (contrary) is expressed by negating the corresponding
unitizer or non-specifying selective. The table of examples on the next
page will make this clear.

The negated totalized negative logically gives some type of partial,
e.g. not none = some, 'it wasn't nothing I got' = 'I got something', but
such usages are not common in standard EngUsh. They are char-
acteristic of Latin, however, e.g. nonnulU 'not none\ i.e. 'som£, quite a
few'; non nunquam 'not never', i.e. 'sometimes'. On the other hand,
when the negation is not of the totalized negative but of the statement
as a whole, the net meaning is that of the contrary, i.e. a totalized
affirmative; e.g. 'there was none but was present' = 'there was none
who was not present' = 'every one was present', 'all were present'.
Such totalizers of double negation may be called corrective totahzers,
for they directly oppose a quantitative negation merely in order em-
phatically to affirm the notion of totahty at the other extreme of the
quantitative gamut. The corrective totahzer is sometimes directly
expressed in EngUsh by all, as in 'it's all wheat', which has a doubly
negative force, i.e. 'it's nothing but wheat', 'it consists of nothing which
is not wheat'.

Five: Studies in l'fii\(rs(il (Ona'ptual Cnimnuir

POSITIVE
TOTALIZER

NEGATED
TOTALIZER

= PARTIAL

negated uni-
tizer; ne-
gated NON-
SPECIKYIN(J
SELECTIVE

totalized

negative

((•ontrart)

all the men

all of it

the whole
table

everybody,
every one

a/ways
enough

not all the
men

some of the
men

no man, none
of the men,
nobody

not all of it some of it

nut one man,
not one of the
men ; not any
man, not a
man, not any
of the men

not one bit of none of it
it; not any of it

not the whole
table

not every-
body, not
every one

some of the
table, a part
of the table

somebody,
some one,
some, some
(people), cer-
tain ones

not one part of
the table; not
any part of the
table

ru)t one, not a
one; not any-
body

no part of the
table, none of
the table

none, nobody

both of them not both of one of the two not either (of neither of

them

not a/ways

not enough

of them

sometimes

to some extent

the two
them)

of them

7wt ever; not at
any time

never, at no
time

not to any^ ei- quite insuffi-
tent, not any ciently
(coll.)

* When any is accented. When any is not definitely stressed, 'not to any
extent' means 'not to any considerable extent,' i.e. 'not enough to satisfy
requirements.'

320 General Linguistics I

X. SPECIALIZED TOTALIZERS

Specialized totalizers are those in which the fundamental quantitative
notion is so limited as to apply only to a particular class of existents.
Thus, the whole swarm of, the complete herd of, every one of the flock of
are speciahzed direct quanto-quantifiers of animal collectivities ; a/^
three bushels of, a whole pint of are specialized direct quanto-quantifiers
of volume or capacity. As with general (abstract) totahzers, so with
specialized ones — they may be direct or calculated, and each of these
may be non-evaluated or evaluated, in turn simple or modified. Ex-
amples of such are: direct specialized totalizers, (a) non-evaluated
simple : all the acres of; (b) non-evaluated modified : only the total acreage
of] (c) evaluated simple: the complete herd of; (d) evaluated modified:
quite a full bushel of, an ample sufficiency of tonnage of = a more than
sufficient weight of (in terms of tons); calculated speciaUzed totahzers,
(a) non-evaluated simple : the whole computed acreage of; (b) non-evalu-
ated modified: absolutely all the remaining battalions of; (c)" evaluated
simple: the complete toll* of (dead); (d) evaluated modified: a merely
suffix:ient quota^ of (immigrant laborers).

•• Toll = counted (or computed) totality of (lost) human beings.
^ Quota = allocated totality of human beings — totality of human beings as-
signed to some part of an implied whole.

Five: Studies in I 'fiivcrsul ( O/ufpiiml (itiiniDuir 321

XI. QUANTIFICATES INVOLVING TOTALIZATION

Specialized totalizers are a species of quantificates (quantified qiian-
tifiahles). They consist of totalized quantificates (quantificates in
which the quantifying element is the notion of totality) applied (or
re-apphed) for quantifying purposes. They may be called quant ificate-
totalizers. In them the notion of totality is used to quantify a (juantita-
tively defined class of existents which is abstracted from the existcnta
themselves (e.g. flock, bushel, acre), and the resultant term is used to
quantify a member of the underlying class (e.g. sheep, potatoes, land).
Like all other quantifiers, quantificate-totalizers answer the question
'How much (many)?' or 'To what degree?'.

We have now to take a glance at the different kinds of totalized
quantificates in which the notion of totaUty determines a non-quantita-
tive type of experience. The total number of possible terms of this
sort is naturally immense and we need only list the types, with an
example or two of each.

TOTALIZED QUANTIFICATES (not primarily totalizers)

A. Totalized quantificates which determine existents or occurrents

L Totalized selection quantificates (totalized selecters), e.g.:
all these, both of those, the whole of a certain

2. Totalized order quantificates (totalized orderers), e.g.: 'at every
point (in the system)', throughout, the last (= the all-th)

2a. Totahzed space quantificates (totalized spacers), e.g. : every-
where (corresponds to both all and whole), at all places (cor-
responds only to all), throughout (may be used of space-
portions), all over

2b. Totalized time quantificates (totalized timers), e.g. : always
(corresponds to both all and 2vhole), at all tirnes (corresponds
only to all), throughout (may be used of time-portions), ever
(e.g. 'I am ever desirous of; not to bo confused with ever
= at any time, e.g. 'if ever I am desirous of)

3. Totalized condition quantificates (totalized conditioners), e.g.:

322 General Linguistics I

'under all conditions* (corresponds to all), 'ivhether (he comes)
or not' (corresponds to both)

4. Totalized purpose quantificates (totalized purposers), e.g.:
for all purposes, for the whole purpose, entirely in order to

5. Totalized cause quantificates (totalized causers), e.g.: for all
reasons, each and every cause, the whole cause, completely deter-
miyied (as to cause)

6. Totalized requirement (norm) quantificates (totalized normers,
e.g.: corresponding to all requirements, complete (in certain
contexts, e.g. 'a complete solution of the difficulty')

7. Totalized manner (quality) quantificates (totalized qualifiers),
e.g.: in every way, of all kinds, 'the whole range (of types)', both
sorts, entirely of a (given) kind, completely dark

8. Totalized value quantificates (totalized valuers), e.g.: perfect

B. Totalized quantificates which are existents or occurrents

9. Totalized existent quantificates (totalized existentials), e.g.:
the whole world, the universe, all, everything, both hemispheres,
'every atom (of his body)'

10. Totalized occurrent quantificates (totalized occurrents), e.g.:
to get completely dark, to finish (related to both all and whole,
e.g. 'to finish the count' and 'to finish the task'), to pervade
(related to whole and all of, but not to all of discretes), to consuine,
to annihilate, to reach up to, 'to embrace (everything)' (related
to cumulative totalizers like each and every, one and all)

It is difficult sometimes to classify actual terms involving quantity
as a determinant, because there are many blended types. We have
grouped manner and quality together, as it is only when specificity is
actuahzed in an existent (e.g. 'a wholly good man') or in an event (e.g.
'to dance wholly well') that the difference between quaUty and manner
arises. Our (in) every way, for instance, is superordinated to 'in every
manner (of happening)' and '(existent) of every sort', somewhat as order
is superordinated to space and time. If desired, the group of totaUzed
qualifiers may be subdivided into the two blend-groups of totahzed
existent-qualifiers and totahzed occurrent-quaUfiers. It is worth
noting that what seem, from the linguistic viewpoint, to be quantifiers
pure and simple are often really disguised quanto-qualifiers, e.g.
'I Uke everyhodf, i.e. 'I like every kind of person', in answer to 'What sort
of people do you Uke?'; 'all land in this city is expensive', i.e. 'all the
varieties of land (even such as is of inferior grade)'. Observe that this

Five: Studies in I 'uivcrsul ( Omcpiitul (irnnittiur 323

disguised totalized qualifier tends to take the form of the pseudo-
totalizer of universal statements — naturally enough, when we consider
the close relationship between the concepts of (juality and class.

Totalized quantificates may be direct or calculated, and any of them
may be modified. For the sake of schematic convenience we may
list examples of totalized calculated quantificates:

TOTALIZED CALCULATED QUANTIFICATES

1. Selection: the sum of these, 100% of a certain

2. Order: at every computed point (in the system), the last (arrived

at by successive subtraction)
2a. Space: the whole area, every accruing allotment (of land)
2b. Time : a whole month, semester, cycle

3. Condition: all odds (in its favor), under enough conditions (to

secure his rights) (evaluated)

4. Purpose : for all calcidated purposes, every calculated goal

5. Cause : for all calculated reasons, altogether determinedly because

6. Norm: corresponding by calculation to all requirements, complete

specifications

7. Quality: in every calcidated way, entirely of a specified ki7id, quite

black

8. Value: calculatedly perfect, perfect on every score (count)

9. Existent: all the (calculated) neurones (calculated quanto-existents

are generally reached through calculated quality)
10. Occurrent: to finish according to schedule (calculated quanto-
occurrents are generally reached through calculated manner)
We need not go through the list of modified totalized quantificates,
as they can be easily formed by determining our quantifiable categories
by means of modified totahzers, e.g. just perfect (limit approximative-
totalized value quantificate).

One may legitimately ask what difference of principle there is between
certain evaluated quantifiers (e.g. enough, complete, entire) and certain
classes of quantificates, say of norm or value, which may be used to
quantify (e.g. perfect, full, complete). Obviously there is no hard and
fast line between these classes of terms. Everything depends on
whether we conceive of the emphasis as a quantitative one, the cjuanti-
tative notion being somehow determined l)y non-quantitative deter-
minants, or conceive of the emphasis as other than quantitative, the
notion of quantity merely coming in as determinant. In entire wr feel
that the emphasis is essentially a totalizing one, with an overtone of

324 General Linguistics I

integration: 'nothing jnissing'. In perfect the concept of totality is
clearly subordinate to that of value, ideal. In other words, while
entire and perfect are within haiUng distance, entire faces such terms
as all, whole, total, while perfect faces rather supreme, excellent. Com-
plete seems to look both to entire and perfect, according to context. In
such a sentence as 'This is a complete set of furniture' the emphasis is
clearly on totality, with an overtone of either the normative or the
integrative, hence it is a normative totalizer. But in 'the complete
angler' the emphasis is not on the enumeration of qualities required for
good fishing but rather on the successful pursuit of the sport by virtue
of all and sundry quaUties required, hence it is a totaUzed normer.
To paraphrase, 'the set of furniture is complete', but 'the angler is a
finished sportsman'. One possesses a normal totality, the other a total
normality.

The touchstone to the classification of any such term in a given context
is the question to which it can appropriately respond. If it responds to

'How much (many) (of) ?' or 'To what degree ?'

it is a quantifier. If it does not, it is some other kind of determinant,
which may, of course, include the notion of quantity.

Finally, we may point out that, properly speaking, such speciafized
totaUzers as a whole bushel of go back to more or less hypothetical quanti-
ficates, of the class a whole bushel (as quantitatively determined portion
of space, not yet as measure of capacity). Theoretically, every quanti-
ficate may become a specialized quantifier.

Thus, everywhere, a space quantificate, becomes a quantifier in such
locutions as 'everywhere in France one drinks wine', which is Unguisti-
cally, rather than conceptually, distinct from 'in the whole of France one
drinks wine'. Similarly, the whole year is clearly a time quantificate
(specifically, a totalized timer), but in 'a whole year 0/ study' it is just as
clearly a specialized totalizer, for it has become a measure of a time-
measurable existent. 'Under enough conditions (to secure his rights)', a
norm-evaluated totahzed conditioned quantificate, may be turned into
'a sufficiently conditioned number of rights', a norm-evaluated totafized
conditioned quantifier. The former term answers the question 'Under
how many conditions?' or, better, 'Conditioned to what extent?' The
latter answers the question 'How many rights?' Hence a sufficiently
conditioned number of, though a very remarkable kind of quantifier, is
just as truly one as^ye or all.

Five: Stuiinw in i Hivcrsiil (Omvptital Gramnuir

325

TABLE OF TOTALIZERS

general
(abstract)

direct

(A. V. M.)

non-eval- [simple all, whole, each, every, both; none

uated s modified absolutely the whole of, diatribu-

(pure) [ lively all; none at all

[simple full, complete, enouRh (sufficient),

evaluated { entire, intact, perfect

[modified quite enough, most complete

calcu-
lated

non-eval-
uated

evaluated

simple the sum of, sum total of, net

total of
^modified just the sum of

simple a calculatedly suflBcient amount
of, netting the entire

modified just a calculatedly sufficient
amount of, netting just enough

specialized
(concrete)

direct

calcu-
lated

non-eval- fsimple swarm, the whole acreage of

uated \modified only the total acreage of

I , , , fsimple the complete herd of

■ evaluated < f,. , i ^ • r * t

\^ [modified an ample sufficiency of tonnage ol

/ . fsimple the whole computed acreage of

s modified absolutely all the remaining bat-

uated * r " f
tahons of

evaluated

simple the complete toll of (dead)

modified a merely sufficient quota of
Note: Examples are merely illustrative.

Fidiloriai Note

Published as no. fi ol language Monographs (Linguistic Soeiet\ i>f America),
Raltiinore. \\'a\erl\ Press. September PX"^().

The Expression of the Ending-Point Relation in
Enghsh, French, and German (1932)

EDITORIAL NOTE

The present paper is the beginning of a detailed study of the variety
of ways of expressing logical relations in language. It is part of a series
of studies sponsored by the International Auxiliary Language Associa-
tion in the United States, Inc. The chief projects which have been
begun are the 'Foundations of Language, Logical and Psychological,
an Approach to the International Language Problem' and 'Compara-
tive Studies in Selected National and International Languages'.^ The
'Ending-Point' study furnishes material which should be useful for the
proposed section on 'Fundamental Relational Notions and their Lin-
guistic Expressions' of 'Foundations of Language' and for the proposed
section on 'Formal Elements' in 'Comparative Studies'.

The outstanding features of the technique are :

1. A 'testing-frame' which may be used for the analysis of any form of
expression, or implication, of a relation or combination of relations,
in any language,

2. The definition of a relation by means of a 'rendering' whose essence
is the naming of the relation of the second term to the first term.
The rendering can be translated approximately literally and used
as a test rendering in any language, e.g., the rendering of the ending-
point relation as 'whose ending-point is'.

3. The use of distinctive and similar type (1) in the sample sentence,
for the locution which expresses or implies the relation treated, and
(2) in the testing-frame, for the words or symbols which give the
meaning of such locution.

The above and other features are explained more fully at the begin-
ning of Part II.

Here, even at the risk of repetition (see p. 30), the 'testing-frame'
device for facilitating the analysis of examples (see p. 36) may be
briefly described as follows. At the left of the page are entered sample
sentences. To the right of the samples there are three 'testing' columns.

* See leaflet: 'Linguistic Research sponsored by the International Auxiliary
Language Association', obtainable on request from the Association, Box 118, 525
West 120th Street, New York.

live: Siuclic.s III Lnivcr.sal ( nnirpiiinl (irnninun 327

The first one is for tlie subject (tlie lof^ifiil first term') of the relation,
that which is oriented, called A. The third coluriin is for the correlate
(the logical second term^) of the relation, that to which A is related,
called B. In these two columns are entered the terms of the relation
in substantival form stripped of idiom and often highly K^neralized.
The middle column is for the relation, symbolized by the entry r. An
interpretation or 'renderinp;' of the relation, given ahead of the testing-
frame, shows how the symbol r is to be read.

During the research which led to the ending-point paper a constant
want was felt for a term to designate a locution or otlier linguistic
device whose sole or principal function is to express a relation or com-
bination of relations — -a word to cover the functions served by preposi-
tions, prepositional phrases and affixes, conjunctions, and case forms.
The term 'relater' was adopted and has proved useful and convenient.

Grateful acknowledgment is made to Professors Algernon Coleman
(University of Chicago) and Henri F. MuUer (Columbia University) for
help with different parts of the French material; to Professor Peter
Hagboldt (University of Chicago), Professor Hermann J. Weigand
(Yale University), Dr. Ludwig Kast and Mr. Reimar von Schaafhausen
for help with different parts of the German material, and to Mi.ss
Frances Faegre for help in preparing the tables in the Appendix. To
other friends and fellow-workers also we are indebted for help along the
roadside, and we take this opportunity to express to them our appre-
ciation.

There remains much to explore in the field of the ubiquitous ending-
point relation. Its expressions in the realm of time have scarcely been
touched upon in the present study, nor have investigations been made
regarding degrees of kinship or identity of 'transitive' and 'dative'
relations with the ending-poing relation, nor concerning the hitter's
kinship to goal, purpose, result, etc.

The hope is expressed that this monograph may prove useful in future
studies of the expression of locative and other relations. If such
studies could be uninterruptedly pursued, in English, French and Ger-
man, by a staff of two or three research assistants guided by skilled
direction, it is believed that within a couple of years there might result
definitions and clas.sifications of relations and combinations of relations,
expressed or masked by ordinary linguistic forms, which are more

* Note that 'first term' and 'second term' have nothing to do with word-order
but refer exlusivcly to the logical priority of the subject which is being oriented
over that by reference to which it.s orientation takes phice.

328 General Linguistics I

fundamental and complete than anything which yet exists. Based
upon the uncovered, classified data, it should eventually be possible
(among other things) to plot out a map of the common relational con-
cepts which pervade language, accompanied by their prepositional and
conjunctional symbols in the languages of the study, — a map to serve as
a guide in translation and interpretation, as a new approach to certain
parts of 'general language study', and as a norm for the simplification
and clarification of the prepositional and conjunctional features of an
international language.

A. V. M.

l-ivc: Studies in Universal C nnccpiunl ( irn/nrnur 32^

CONTENTS
Part I: Discussion

PAGE

Definitions 11

Abbreviations and Conventions 11

The Ending-Point Relation 12

English : General Remarks 14

Locutions which denote the ending-point relation expHcitly or

implicitly 15

French : General Remarks 21

Locutions which denote the ending-point relation expHcitly or

impHcitly 22

German : General Remarks 24

Locutions which denote the ending-point relation explicitly or

implicitly 26

Part II: Sample Sentences^

Explanatory Remarks 30

Sentences with Analyses in Testing-Frame 35

Abbreviations and Conventions 35

Samples of the Expression of the Ending-Point Relation in English. 36

(1) Predicative e-p expression 36

(2) Usual e-p relaters, occasionally with implication of addi-

tional concepts 36

(3) E-p relaters with emphasis on distance traversed 37

(4) E-p relaters in combination with locative relaters 37

(5) Relational locative expressions which may imply e-p

relation 38

(5. a) Those which require an explicit point of reference. . . 38
(5.b) Those which may be used with or without a point of

reference, without change of form 38

' Section numbers correspond to those in Table I of the Appendix ('Compara-
tive Table of Contents of Parts I and IF). Not all classes of ending-point expres-
sions aj)pear in each of the three languages of the study. Hence the omission of
different numbers from the contents of each language.

330 General Linguistics I

(5.c) Those which have different forms when used with or

without a point of reference 41

(5.d) Those which ordinarily do not take an expUcit point

of reference 42

(7) Locutions which consist of or include some expression of

direction 42

(8) Locutions which express position irrespective of context 43

(10) Space-locative pronouns: interrogative, indicative, and

relative 44

(lO.a) Those which essentially include the e-p idea 44

(lO.b) Those which may imply e-p relation 44

(11) Places, no place, home, abroad 45

(12) Combinations of e-p relation with first term of relation and

often also with other concepts (verbs, nouns, adjectives) . . 45

Samples of the Expression of the Ending-Point Relation in French 48

(1) Predicative e-p expression 48

(2) Usual e-p relaters, occasionally with implication of addi-

tional concepts 48

(5) Relational locative expressions which may imply e-p relation 49
(5.a) Those which require an explicit point of reference. ... 49
(5.c) Those which have different forms when used with or

without a point of reference 49

(6) Prefixes which sometimes express e-p relation 51

(8) Locutions which express position irrespective of context ... 53

(10. b) Interrogative, indicative and relative pronouns which may

imply e-p relation 53

(12) Combinations of e-p relation with first term of relation and

often also with other concepts (verbs, nouns, adjectives) . . 53
Samples of the Expression of the Ending-Point Relation in German . 54

(1) Predicative e-p expression 54

(2) Usual e-p relaters, occasionally with implication of addi-

tional concepts 54

(3) Bis 55

(5) Relational locative expressions which may imply e-p relation 55

(5.a-l) Those which with the accusative express both e-p
and a locative relation, and with the dative
express only a locative relation 55

(5.a-2) Same forms as 5.a-l, suffixed to 'da- and 'wo- .... 56

(5.a-3) Those which are construed with the genitive, dative,

or accusative 57

(5.d) Those which never take an explicit point of reference. . 58

Five: Siiidics in Universal C Onccpuiul (inininiur

(7) Locutions which include some expression of direction. . 58

(8) Locutions wliich express position irrespective of context 59

(9) Indicative pronouns which inchide the e-p idea: her and hin

alone, or plus -an, -bei, -cin {-in), -vor, -zu, -zwischn, or

plus a stem wliich expresses a motion (KJ

(10. a) Interrogative, indicative, and relative pronouns which in-
clude the e-p idea: her and hin suffixed to or following
hier; hin after wo, da, dort, uhcrall, irgcndwo, etc til

(11) Heim 01

(12) Combinations of e-p relation with first term of relation and

often also with other concepts (verbs, nouns, adjectives) . . 62

Part III: Notes

Space-locative Pronouns 64

Notes on the English Sample Sentences 70

Appendix

Table I: Comparative Table of Contents of Parts I and II 84

Table II: Summary of Classes of Restrictive Locutions which
denote the Ending-Point Relation, explicitly or implicitly, in the
English, French and German Sample Sentences 88

332 General Linguistics J

PART I: DISCUSSION

Definitions

The ending-point relation is the relation that exists between a
movement and the point at which it ends. It may be extended to
include the relation that exists between a stationary existent or an
extent of space or time and that one of its extremities which is
conceived as its ending-point (ending-part conceived as ending-
point), as in 'England stretches from Kent to land's end', or
between such existent or extent and an outside point which is
conceived as its ending-point (boundary-point conceived as ending-
point), as in 'England stretches from sea to sea'.

A point, as used technically in this study, is a covering word for any
existent, point, spot, place, or area in space, or any occurrent, point,
or period in time.

A POINT OF REFERENCE, as used in this study, is a convenient term for
any existent by reference to which the ending-point is located,
e.g., 'door' in 'He put it outside the door'.

A LOCUTION is a meaningful affix, root, word, or group of words forming
a unit.

A RELATER is a locution whose sole or principal function is to express a
relation or combination of relations. Relaters include prepositions,
prepositional phrases and affixes, conjunctions, and case elements.

Abbreviations and Conventions

e-p 'ending-point'
sec 'section'
....... 'so and so', 'such and such'

When in the text a rendering is given of the first term of the relation
(i.e., the subject) plus the relation plus the second term, the renderings
of the terms of the relation are usually printed in capital letters, and the
rendering of the relation in italics, e.g., "The verb 'to enter' without
exphcit second term expresses a movement whose ending-point is a
point in something."

live: Slutlic.s m I ni\cf\itl ( i>n(i-]Uti,il ( ii.inmuir }y}i})

The Endincj-I'oint Hklation

The encIinp;-point relation exists between two terms when one of thcin
is given as the ending-point of the other. Tims, in 'Our tour ended at
Boston', 'Boston' is stated to be the ending-point of 'our tour.' In
*We went to Chicago', 'Chicago' is the ending-point of 'our going.'

To end at and to both serve to express the ending-point rehition, but
there is an obvious difference between them. Tlie one asserts or
predicates tlie rehition, wliile the other expresses it as quahfying or
restricting one of the terms of a predication. 'Our tour ended at Boston'
asserts an ending-point relation as existing between 'our tour' and
'Boston.' 'We went to Chicago' asserts a going, and the fact that the
ending-point of the going is Chicago is incidental to the main assertion.
It will be convenient to call such locutions as to end at predicative and
such locutions as to restrictive.

A proposition involving the predicative ending-point relation may be
reduced to the formula: A has as ending-point B. When the relation is
restrictive the relationship may be formulated: A, luhose ending-point is
B, or A, haning as ending-point B. In all cases, we shall call the first
term of the relation A, the second term B.

Linguistically, the relation may be expressed by an isolable locution,
as is the case with 'to end at' and 'to'; or the relation may be combined
with a substantial notion in a single locution. Such a case is the word
thither, which means having as ending-point that place: 'his journey
thither' = 'his journey <o that place.'

Again, the relation may be not specifically expresseil but implied by
the context. This is illustrated by 'He stumbled and fell\ Here we
presume, unless specifically told otherwise, that the ending-point of the
fall was the ground. In such and similar cases an implication both of
ending-point and of ending-point relation is forced by the nature of
things, physical or social. But this occasional phenomenon is only of
passing interest in a general discussion of the expression of a relation.

Another type of implication is of first importance, namely, the case of a
given word or class of words imj)lying a relation in a given type «)f con-
text. The word 'home' always implies the ending-point relation when
used with words expressing a motion, e.g., 'The journe\' home', 'He went
home\ 'Take me home' , etc.

Several classes of locative expressions when used with words denoting
a movement imply the ending-point relation in contexts where the
second term responds to the question 'Whither goes the movement '.''

334 General Linguistics I

Three classes of such words deserve special mention. One comprises
locative prepositions, e.g., 'When it rained, they moved under the
canopy'. Another class consists of elliptic locative pronouns, e.g.,
'Show them in' . In still another class are found interrogative, indica-
tive and relative locative pronouns, e.g., ^ Where do you think of going?'
'It takes half an hour to walk there', 'I'm going where the wild thyme
grows'. (See note on 'Space-locative pronouns' at beginning of Part
III.)

The study of the expression of the ending-point relation in English,
French, and German necessarily includes a treatment of combinations
of this relation with various locative relations, for this type of combina-
tion is quite common in those languages. Thus, English into combines
the ending-point relation with the locative 'in' relation. Under in 'A
mouse ran under the bed' combines the ending-point relation with the
locative relation expressed by 'under'. The logical content of such
expressions as into, and under in certain contexts, includes a substantial
element which is the second term of the ending-point relation and the
first term of the locative relation. This intermediary term, if expressed,
would be something like 'a point' or 'a place'. It is sometimes actually
expressed, as in 'They rowed to a point north of the pier'. Into
(explicit locative in plus explicit ending-point to) may be paraphrased
'whose ending-point is a point in'; under (implied ending-point to plus
explicit locative under), 'whose ending-point is a point under'.

All types of impHcation of ending-point relation of which examples
have been given, as well as words of the 'thither' type in which the
ending-point idea is essentially included, are combinations of the rela-
tion with the second term, whether explicit, as in 'He went there (to
THAT place)', or implicit, as in 'They went under the balcony (to a
point under the balcony)'. The latter example is typical of one of the
ordinary linguistic methods of locating an ending-point, namely, impH-
cation of the actual ending-point by explicit reference to something to
which it is spatially related. In some combinations with the second
terra both the ending-point and the point of reference are impUed. This
is true of elliptic locative pronouns, e.g., in 'He went below', 'below'
means to an implied point below the implied point of reference
(the point where he was standing)'.

There exist also combinations which include the first term and often
other concepts as well. Such combinations are found chiefly in predica-
tive expressions and in locutions derived from such. To reach is 'a
movement whose ending-point is' (first term plus e-p relation); to

I ivc: Siiiilit's in Univt-rsiil Ctmccpinul iiruninutr 335

embark denotes 'a movement whose ending-point is a point on some-
thing (or SOME ship)' (first term plus implied e-p relation plus
IMPLIED SECOND TERM plus implied locative 'on' plus implied point
OF reference); a landing expresses 'an implied movement whose end-
ing-point is A point on the land' (implied first term plus implied
e-p relation plus implied second term plus explicit point of refer-
ence). Occasionally, the first term is implied in restrictive expres-
sions (in cases of ellipsis of the verb), e.g., To bed, you rascals', where to
includes the idea of a movement.

A summary of the restrictive classes of locutions illustrated in the
English, French and German sample sentences is given in the Appendix,
Table II.

Little attention is paid in this study to the predicative expression of
the ending-point, for a full understanding of such locutions would in-
volve us in complicated questions of the expression of predication in
general and of the 'reification' of relations into entities; e.g., end, as
noun, as a 'reified' ending-point relation. As examples of the bewilder-
ing variet}^ of possible renderings in actual speech of the predicated
ending-point relation we may note the following German ways of express-
ing the journey ended at Berlin: 'Die Reise endete in Berlin', 'Die
Reise war in Berlin beendet', 'Die Reise wurde in Berlin beendet', 'Die
Reise fand ihr Ende in Berlin', 'Die Reise fand ihren Abschluss in
Berlin', 'Die Reise kam zu Ende in Berlin', 'Die Reise wurde beendigt in
Berlin', 'Die Reise fand ihre Beendigung in Berlin', 'Die Reise kam zu
ihrem Abschluss in Berlin', 'Die Reise wurde in Berlin zu Ende ge-
bracht', 'Die Reise wurde in Berlin zu Ende gefiihrt', 'In Berlin war
das Ende der Reise', 'In Berhn war die Reise abgeschlossen', 'Das
Ende der Reise war Berhn'. There is, of course, a host of other possi-
bilities.

English

general remarks

The principal device for expressing tlie ending-point relation in
English is the relater to. Most locative relaters can in certain contexts
imply the ending-point relation. In such cases there is no specific
ending-point relater, e.g., 'He went inside the enclosure'. The same
holds for such locutions as here ('He came here'), upsldirs ('He went
upstairs'), etc. Other devices will be mentioned in their t urn.

33t) General Linguistics I

LOCUTIONS WHICH DENOTE THE ENDING-POINT RELATION EX-
PLICITLY OR IMPLICITLY

(1) EOOP: The common predicative ending-point expressions in

English are to end at, to terminate at, to finish at.
There are many others, but in the case of locutions of this type it is
impossible to assemble all the words of the class. We have had to be
content here, and in other groups, with assembling a few representa-
tive locutions.

(2) E002-015: The usual ending-point relater is to, as used in 'He

rode to Chester', 'the trip to Rome', 'cut to the bone'.
Unto is archaic, but expresses virtually the same thing as to; e.g., 'My
throat is cut unto the bone'. Unto may be used not only as a preposi-
tion but also as a suffix to here and to there; e.g., 'Let us hasten there-
unto'. Occasionally to implies a locative relation, as in 'She put the
baby to bed', i.e. Ho a point in bed'. Sometimes, through ellipsis of the
verb, to may imply the first term of the ending-point relation, as in 'He
asked him out to a round of golf', where to equals to go to, though it is
probably more accurate to say that the implied to go to is expressed
'straddle' fashion by asked out and to.

(3) E016-017: The ending-point relation, with emphasis on the dis-

tance traversed, is expressed in English as in the fol-
lowing sentences: 'He went with her as /ar as the gate', 'Joan ran all
the way to the post office', etc.

(4) E018-027: The two relaters, into and onto, combine the ending-

point relation explicitly with a locative relation.
'He ran into the house' asserts a running whose ending-point is some
POINT inside the HOUSE. 'The plaster fell onto the floor' asserts a
FALLING whose ending-point is some point on the floor. Into may
be used in archaic style as a suffix to there (thereinto).

(5) E028-096: Practically all locative expressions in English may

imply the ending-point relation when used with
words denotmg a movement. Thus, 'He went under the balcony'
asserts a going whose ending-point is a point under the balcony.
'He ran behind his mother' asserts a running which has as ending-paint
A point behind his mother. To counteract this implication of the
ending-point relation when one desires to use the locative relaters in

1 Numbers refer to those of sample sentences in Part H.

Ii\ c SiuiJic.s in Universal ( omcpiuiil (iiiinmuir 337

their primary meaning, one often has to use some expression hke alutu)
or about, as in 'He went along under the balcony' or 'He ran c^out beliind
his mother'. Sometimes, liowever, the use of the locative rchiters for
the en(Unp;-point rehition is ambiguous or the interpretation is depend-
ent on the context, e.g., 'He dropped the bucket in tlie well', 'Something
splashed in the well'.

(5.a) E028-O37: A number of these locative expressions recjuire an

explicit point of reference in the second term:
at ('Up and at them.')
beside ('He brought up his chair beside her.')
east of, west of, north of, south of, etc, ('Afterwards we went east of the

Alleghenies', i.e., to a point or region cast of the Alleghenies.)
upon, which in addition to its prepositional use may in archaic style be

suffixed to there and to here ('He threw it tipon the table.' 'He

placed it thereupon.')
{a)round ('He went around the corner' = 'He went to a point around

the corner.' 'It fell around his neck' = 'It fell to a position aroujui

his neck (at his neck)'. (See Note on Elll, page 81.)
against ('He threw the stone against a tree.')
with, used as preposition and also as suffix in hereunth, therewith.

(5.b) E038-077: A number of other locatives may be used with or
without an explicit point of reference. When un-
expressed, the point of reference is supplied by the context. Many
such locatives function both with and without an explicit point of refer-
ence without change of form. We can say 'He went in the house' or
'He went in.' Relaters of this type are :

aboard, about, above, across, along, alongside, athwart, before,

behind, below, between, betwixt, beyond, in, inside, within, outside,

without, on, over, under.

In, on, and under are capable of suffiixation to here and to there. In

and over, and sometimes some of the others, are used to express the

ending-point relation as prefixes, thus, 'the zVirush of the waters' =

THE RUSHING whose ending-paint is a point in something; 'She oirrshot

her port in the night' expresses a movement having as ending-point a

point beyond the port. The Latin prefix in- in insert is not to be

confused with the Kngli.sh prefix in- in inrush. Externally the Latin

in- is diiTerent in that it never takes the accent while tlie luigli.'^h in- is

always accented. As to meaning, the in- of Latin derivatives is much

more closely joined in meaning to the rest of the word, so that its funda-

338 General Linguistics I

mental meaning is more or less obscured in the combination. We have,
therefore, not attempted to isolate the in- prefix of Latin derivatives.

(5.c) E078-093 : Some locative expressions have different forms when
used with and without an explicit point of reference.
The following add an of when the point of reference is expressed (e.g.,
'He threw them on top of the trunk', 'He threw them on top') :

on top (of), in front (of), short (of);

to the east, west, north, south (of).
One must not make the mistake of always analyzing expressions like
to the east into the ending-point relater to plus a substantial the east,
for we must bear in mind that to the east may be used without any
implication of ending-point. The sentence 'He went to the east over
the holidays' evidently expresses a movement whose ending-point
is THE EAST, i.e., the eastern part of the country. In the sentence
'They sailed off to the east', however, 'to the east' is no longer analyzable
into ending-point relater 'to' plus substantial 'the east', but must be
taken as a unit; the analysis is rather a sail whose ending-point is a

PLACE TO THE EAST OF THE STARTING-POINT. 'To the Cast' aS it OCCUrs

in this second sentence functions as a directional expression equivalent
to 'in an easterly direction'. In such a sentence as 'Haiti lies to the
west of Santo Domingo' to the west of is equivalent to in a westerly
direction from.

The following locative expressions, many of them involving a locative
relation qualified by some judgment of distance, have different forms
without the use of 'of, when used with and without an explicit point of
reference.

Without point of reference With point of reference

He came close close to us

He came near near (to) us

Pull them up up to us

Don't go far (away) far (away) from home

Often he would go apart apart from us

He ran away away from us

He stepped aside aside from us

He went back back to them

(5.d) E094-096: Another group of locative expressions do not ordi-
narily take an explicit point of reference:
upfront ('Take the gentleman upfront to a better seat.')
uppermost ('The spike of my ax turned uppermost.')

live: Siuclu's III inivciMil C o/uf/niinl (nunittutr 339

together ('Tlie coinini^ toijether oi ;i crowd takes but ;i few ininutea.')

(7) IC097-111: The idea of direction is often coiiihini-d with the
ending-point idea.

Some directives are used with an ending-point relatcr: 'I'm going
down to Tucson', 'When are you going up to Wisconsin'*'

There are a few directives which in certain contexts imply the ending-
point relation with the point of reference unexpressed and drawn from
the context, e.g., 'He set the jug down and went out.'

It is possible to analyze in ways other than shown in Part II directives
which are associated with the ending-point relation. For example :

a movement

a a point of

direction reference

down-

(...)

Tucson

ward

from

down-

(...)

(r)

(the busi-

ward

ness sec-

froju

tion of)
town

down-

(...)

(r)

(...)

I am going down to a move-
Tucson, ment

I am going down a move-
town, ment

He waited for her a move-
to come down. ment ward

from

Down with your {an action down- (. . .) (r) (. . .)

sails! impelling ward

a move- from
ment)

up, down, out, over, about (in 'to turn about').

(8) El 12-115: Certain locutions express position irrespective of
context, like upright, upside down, back to back, on
one^s feet, to stand, to lie, to sit, and may express also the ending-point
relation in an appropriate context, e.g., 'turning the bucket upside
down.' We include here those which are not predicative. Samples of
those which are predicative (such as to stand, to sit, to Iw) will be found
in other sections.

340 General Linguistics I

(10) E116-137: There are some interrogative, indicative, and relative

pronouns of space-location which essentially include
or which may imply the ending-point idea.

(lO.a) E116-121: The explicit combinations are whither? (interrogative),
hither, thither, whither (relative), somewhither, any-
whither, whithersoever, nowhither, which may be paraphrased having as
ending-point what place?, this place, that place, the place to
WHICH, etc. It may be noted that all of these locutions are archaic.

(lO.b) E122-137: A larger number of such space locative pronouns do
not essentially include the ending-point idea but
may imply it when used with a word denoting a movement, e.g., 'Our
journey here', 'He went there'. Such are where? (interrog.), here, there,
thereaway s, yonder, where (relative), anywhere, elsewhere, everywhere,
nowhere, somewhere, somewhere else, wherever, wheresoever. The differ-
ence between this group of words and the hither group corresponds to
that existing between locative expressions like at, aboard, etc. (see
above, sections 5 to 5.d) and combinations like into, onto (see above,
section 4). In the one group are expressions which may imply the
ending-point relation, in the other are those in which that relation is
explicit.

(11) E138-142: A few stray words which may imply the ending-point

relation when used with a verb of motion constitute
a group by themselves, e.g., places, no place, some place, any place, home,
abroad, and no doubt others.

(12) E143-19o: A considerable number of locutions in English com-

bine the ending-point relation with the first term and
often also with other concepts at the same time. To come upon (e.g.,
'I cam£ upon an old friend in the market place.') is such a one. Most of
these refer to the ending phase of a movement and to the movement
itself only by implication. Thus, 'They made port' cannot be para-
phrased 'They performed a making which had as ending-point port',
for that would be nonsense. We are rather to see in to make an expres-
sion implying a movement which acts as the first term of the relation.
This applies, perhaps, to all the locutions listed below:
to come up with ('He came up with McPherson on the third lap'). This

locution implies that the point of reference of the second term is in

motion.
to catch up to {with)

Five: Studies in Univi-rsul iOnn-piuul (inininiur 341

to (jdin ('W'c liopc to gain Boston by iii^htfuU')

to hit ('Waukegan was the next town we hit')

to make ('They made port')

to join ('Try to join me before noon')

to rejoin

to reach ('We still hoped to reach our destination')

to draw up to ('A carriage drew up to the gate')

to surmount [10143-154]

/() lai/, to set, and to stand {'Stand it in the corner') are like the above

except that they require locative relaters or locutions involving a

loc;itivo relation as part of the second term. To put usually has this

same requirement; it is also used with or without the ending-point

relater to in different contexts. [E155-158]

Some locutions which combine the ending-point relation and the first
term sometimes absorb the second term as well. We say 'Ho entered
the house', asserting a movement whose ending-point is a point in the
HOUSE. If we do not express a point of reference (as in 'He entered'),
the sentence is still meaningful, for to enter comes in this case to mean
A movement whose ending-point is a point in something. To enter
may be looked upon as a combination of first term plus the ending-point
relation plus *a point in' (when used with an explicit point of reference)
or as first term plus the ending-point relation plus indefinite second term
(when used without an explicit point of reference). [EloO-lGl]

To admit ('He was admitted by the butler') means 'to cause or permit
to enter', and 'to enter', we have seen, expresses a movement whose
ending-point is a point inside something. [E1G2-163]

Other locutions of this type are:
to approach, to near {= to have as ending-point a point near so.mething

or points successively nearer something)
to board {= to have as ending-point a point aboard something)
to mount {= to have as ending-point a point on top of somktiiing)
to stop belongs here but takes locative relaters when occurring with a

complement (*He stopped', but 'He stopped at the gate'.) [E1G4 1701

A word like to emhark asserts a movement which has as ending-point
A point or place aboard something. Thus it combines the ending-
point relation with both the first and the second terms of the relation.

Other examples (not all perfectly clear cases) are: to incise, to imhihc,
to import, to immerse, to insert, to introduce, to land, to {a)light, to tahlc,
to arrive, to turn up, to deliver. To converge, corresponding with the
restrictive locative together, expresses movements hatnng as ending-point

342 General Lingiiisdcs I

A POINT OCCUPIED BY ALL. To relum denotes a movement whose
ending-point is the place formerly occupied. [E 17 1-189]

The reservation is to be made that most of these locutions (as well as
to come, below) often are used in conjunction with ending-point relaters
like to, into and others. [E159-189, 195]

To down and the slang expression to out, although in form non-sub-
stantial, are complete expressions (chiefly by implication) of both terms
of the ending-point relation and of the relation. [E 190-194]

To come, in the absence of an explicit second term, is likewise a complete
expression and includes the further concept of 'direction towards'.
Often, if not always, it asserts a (hitherward) movement whose
ending point is an indefinite point; thus, 'He came an hour ago, but
would not wait'. [E195]

French
general remarks

In French the relaters which most commonly express the ending-point
relation are a and jusque. But d, like dans, y. Id, may also function as a
locative: 'II vient a Paris', but also 'II demeure a Paris'. Context
determines whether a is to be interpreted as expressing the ending-point
relation or a locative relation. The general rule is that with a first term
denoting a linear movement a is usually to be interpreted as ending-
point: 'Tout chemin mene a Rome'. On the other hand, with a first
term denoting a non-Hnear movement or a stationary notion the relater
is usually interpreted as simply locative. Thus, 'lis dansent au centre
de la salle' would be interpreted as 'They dance in the center of the
room'.

To say 'They dance to the center of the room' one employs the word
jusque ('lis dansent jusqu'au centre de la salle'.) This word jusque
serves ordinarily to emphasize the distance traversed and corresponds to
such English expressions as all the way to and as far as (also German
bis), but is quite often used with French locative locutions to obviate
ambiguity as to the ending-point relation : 'lis marchent jusque devant
le palais'. One uses jusque practically always when the first term is a
non-moving entity, as 'Le chemin va jusqu'd Toulouse', 'La foret
s'etend jusqu'd la mer'.

The expressions of action and the direction or ending-point of action
are frequently crowded into the French verb, while in English and Ger-
man the tendency is rather to use a verb with a broad meaning and a

i'ivc: Suidu's ill inivcrsul C nmcpiiuil (inininiar 343

preposition or adverb to complete tlie idea of direction or ondinR-point.
Thus, Fr, ^accoster le navire' = Kng. 'to come alongside the ship', Ger.
*sich langseit dem Dampfer legeri' ; Fr. descendre = \iT\%. to come down,
Ger. herunterkommen .

LOCUTIONS WHICH DENOTE THE ENDING-POINT RELATION EX-
PLICITLY OH IMPLICITLY

(1) FOOl: French has the predicative ending-point expressions

finir a and terminer a.

(2) F002-010: For d and jusque, tlie usual ending-point relaters, see

above under General Remarks.

(5) FOll-041: In French, as in English, locative expressions, when

used with words of motion, may imply the ending-
point relation. Some require an explicit point of reference and others
may be used with or without one,

(5. a) FOl 1-018: Relaters requiring an explicit point of reference:

hors de, chez, dans, en, sous, sur, contre, parmi, cntre.

(5,c) F019-041: Relaters not requiring an explicit point of reference*
but capable of taking one; they modify their form by
the addition of dewhen so used ('II le met au c6ti, but 'au c6tt dw chemin').
{au, en) dega, {au, en) deld, (au, en) dehors,
{au, en) dedans, {au, en) dessous, {au, en) dessus,
autour, {au, en) devant, derricre, {au, en) haut,
d la tete, au cdte, au pied, alentour.
Of the above, devant, dessous, dessus, dedans, contre may be suffixed
to Id- and to ci-; e.g,, 'II est alld ld-dedans\

(6) F042-066: There are a number of prefixes of somewhat general

meaning which sometimes express the ending-poini
relation,^ Most of them are of variable form because they assimilate
phonetically to the following element.

* The reader will remember that to arrive, to approach, to import ami others
were treated in the discussion of English locutions as units. These EnRlish wor«ls
arc etymologically the same as French arrivcr, appruchcr, importer, but j)ro.sent
day English speech is not aware of the components as is French. The rea-son is
obvious: the Frenchman feels that arriver consists of d + rive -f- verb-ending
-er, approchcr of d -|- proche + -er, importer of en + porter. In English the parta
of the words are meaningless and so the word is taken as a whulc and not ,ia a
combination of separate parts (except in etymology).

344 General Linguistics I

ad-, in attahler, arriver, approcher, admettre.

en-, in emballer, enchatner.

in-, in inserer, importer.

inter-, in interposer, entremeler.

outre-, in outrepasser.

trans-, in transvaser, transporter.
The relation of the prefix to the stem to which it is attached varies.
This will become evident if we roughly paraphrase some of the forms.

approcher: mettre quelque chose proche a.

attahler: mettre a table.

emballer: mettre dans une balle.

importer: porter dans.

transvaser: placer dans une autre vase.

transporter: porter dans un autre endroit.
That is, in attabler, emballer, transvaser, the stem represents the point
of reference of the second term of the relation. In importer and trans-
porter, the stem represents the first term of the relation. In approcher,
the stem represents a locative relation to the point of reference of the
second term. Approcher would be paraphrased: to bring about a
MOVEMENT whosc ending-point is a point near . . . Words composed
with prefixes which imply the ending-point relation, if analyzed as a
whole, may be classed as combinations of the relation with the first
term (see below, section 12).

(8) F067-068 : French has a number of special expressions of position
which may imply the ending-point relation, as in
'II se met debouV. 'Allons, debout, il est d^j^ grand jour' = 'Come,
get up {into an upright position), it is broad daylight.' The close logical
relation between such terms as upright and to stand is indicated by such
facts as that French can hardly express the idea of standing except by
predicating adverbial terms for uprightness: etre debout = to stand.

(lO.b) F069-079: In French all interrogative, indicative and relative
pronouns of space location may imply the ending-
point relation without change of form: 'II est Id' and 'Je vais Id,'
'J'y etais et 'yy retournerai.' Locutions of this type are:

Ouf (interrog.), ici (ci-), Id, y,

Id-bas, ceans, leans (little used nowadays),

ou, (relative), quelque part, nulle part, partout, autre part,

ailleurs, and others.
Note that Id and ci- (for ici) can be used with suJExed devant, dessous,
dessus, dedans, contre (e.g., 'II est entre Id-dedans').

Five: Studies in Universal ( oncepituil ( iinninuit 345

(12) F080-090: French has a considembk' riurnlj<'r of IcK'utions which
combine the ending-point reUition with the first term
and often also with other concepts, e.g., rejoirulre, alteindre, and se
rapprocher de. The last two involve, however, the prefix (id- discu.'vsed
above in section 6. All the samples with preii.xes in Part II (see page
51, section 6) combine the ending-point relation with the first term
(as.neger, atteindre); many express or im{)ly in addition a locative idea
{approcher, interpoler, assaut); and some include also a point of reference
and thus are combinations of the ending-point relation with both the
first and second terms {attabler, encadrer, transvaser) .

Entrer and parvenir combine the ending-point relation with both the
first and second terms. They may be paraphrased a movement wliosc
ending-point is a point in something and a movement whose ending-point
is some point respectively. To introduce an explicit point of reference
for the second term, one must employ a locative relater or a locution
involving a locative relation. Thus, 'II entre\ 'II entre dans la salle',
*I1 y entre\ 'II est parvenu', 'II est parvenu a la ville', 'II y est parvenu .
Venir might also be considered a case of the combination of the ending-
point relation with both the first and second terms.

S'arreter is a locution like English to stop, to arrive, which, referring
to the ending phase of a movement, implies the movement which is the
first term of the relation (cf. E170, G115, G116).

German
general remarks

German makes a careful distinction between locative relations and the
ending-point relation. In addition to a distinctive ending-point relater,
zu, it has a device whereby the ending-point relation is unambiguously
brought out when used with nine of the principal locative relaters. We
refer to the rule that an, auf, hinter, in, neben, iiber, unter, vor, and
zivischen govern the accusative case when their meaning includes the
ending-point idea, whereas they govern the dative when their meaning
is purely locative. Further, the distinction between 'Whither?' and
'Where?' is carried out fully with interrogative, indicative and relative
pronouns of space location (see sections 9, lO.aV The differentiation
between expressions which respond to 'Where'.*' and 'Whither?' is,
however, not complete. It cannot be made in cases like 'Er setzt sich
gegeniiber dem Alten', as opposed to 'Er sitzt gcgcnidtcr dem Alten.'

The accusative case after an, auf, etc., is considered to denote expli-
citly the ending-point relation because another case (the dative) is used

.^4(1 General Linguistics I

after the same prepositions when there is no question of ending-point.
For this reason, in the analyses of ending-point phrases composed with
an, aiif, etc., the ending-point relater is shown without parentheses (see
Part II, Ger., sec. 5.a-l). Contrariwise, a case form which is not distinc-
tive for the ending-point relation is not considered to be an expUcit
denotation of that relation, and in the analyses of ending-point phrases
composed with prepositions (other than zu and nach) which govern but
one case, irrespective of the ending-point idea, the ending-point relater
is shown within parentheses (see Part II, Ger., sec. 5.a-3).

Of special interest in German is the use of her and hin as expressions
of the ending-point relation. The meanings of her and hin depend
upon context. It is beyond the scope of this discussion to touch upon
all their meanings, but mention should be made of such as bear upon the
ending-point relation. Never forgetting that whatever is said below is
in reference to her and hin only in the kinds of contexts illustrated, their
meanings as ending-point expressions may be summarized as follows:

1. 'whose e-p is': her and hin as e-p relaters, (a) her when suffixed to
'hier', and hin when used in connection with 'wo' ('irgendwo', 'anderswo',
etc.), 'hier', 'da', 'dort', e.g., 'Hier/ier kommen nur Reiche', 'Bitte
sagen Sie mir wohin dieser Zug fahrt', 'Wo kamm der Waidmann hinf^
(see Ger., sec. 10. a); (b) in combinations which express a direction when
the specifically directive idea resides entirely in the other element of the
combination, as in 'HerAVS mit eurem Flederwisch!' where her merely
performs the function of an e-p relater between the (implied) movement
of the sword and the movement's ending-point expressed in terms of
direction by aus: 'a point outward from . . . (the scabbard)'.

2. 'whose e-p, in direction of (away from) speaker, is': her and hin
as e-p directives: (a) as prefixes in combinations in which both elements
express a direction, her or hin expressing a direction toward or away from
speaker and the other element expressing a specific direction away from
the starting-point, e.g., 'Die Menge quoll /ler aus (hinAVsY: whose e-p,
in direction of {away from) speaker, is a point outward from . . . (the
starting-point); (b) her and hin as prefixes in combinations in which
they are the only expression of direction, e.g., 'Da kommst du schon her-
voRgehiipft' whose e-p, in direction of speaker, is a point before . . .
(the starting point). (See Ger., sec. 7.)

3. 'whose e-p is this (that) place' : her and hin as e-p indicative pro-
nouns, when used alone or in connection with an, bei, ein (in), zu, etc.,
or with a verb-stem which expresses a motion, e.g., 'vSie kam her (to this
placeY, 'die //erfahrt, Hiniahrt (the journey to this, that, place)', 'Sie

/•'i\c: Sluilic.s m L niwrsul ( t>nn'i>iuul (iiiininuii Ml

kam herzu, hinzu {to this, tluit place, object, person}'. In the lasi two
examples 'zu' merely repeatB and emphasizes the e-p idea alrciuiy essen-
tially embodied in her or kin alongside of the substantial idea of place
or the like. (See Ger., sec. 9.)

II cr and hin alone and in the kinds of combinations mentioned above
are never used as mere locatives. For this reason their meanings (in
such contexts as illustrated) are analysed as essentially including the e-p
idea and not merely as implying it,

LOCUTIONS WHICH DENOTE THE ENDING-POINT RELATION EX-
PLICITLY OR IMPLICITLY

(1) GOOl: Predication of the ending-point relation has various

forms in German, e.g., 'Die Reise findet in der Stadt
Berlin ihren Abschluss.' 'In Berlin wird die Reise zu Ende gebracht.'
'Die Reise kmnmt in Berlin zu Ende.' (See page 14).

(2) G002-008: Zu is the usual ending-point relater (e.g., 'Er geht

zuT Schule'). Nach sometimes has the same func-
tion, particularly with place names (*Er will nach Heidelberg', 'Er
fiihrt nach Amerika'). Both of these relators may be used as preposi-
tions or suffixed to da and ico. However, wozu and dazu generally
mean 'for what reason?' and 'for that reason' respectively and rarely
have their literal meaning. Her, when suffixed to /iter-, and hin, when
suffixed to or following wo, da, dort and the like, also serve as ending-
point relators, as already mentioned. (See above. General Remarks,
and below, section 10. a.)

(3) G009-012: Bis sometimes functions as a kind of ending-point

relater. In meaning it is much the same as French
jusque (or Latin usque = all the loaij to, as far as), but the German word
can be used independently, e.g., 'Ich fahre bis Diisseldorf mit'. It is
not an ending-point relater in the same sense as zu is one, but it does
involve the ending-point relation.

Bis used in conjunction with another relater may serve merely to
obviate a possible ambiguity, e.g., 'Der Adler flog bis iiber den Hof
means 'The eagle flew to a point over\ where 'Der Adler flog libcr den
Hof would mean 'The eagle flew over (and past) the yard'.

(5) G013-063: Locative expressions which may imply ending-point
relation.

(5.a-l) G013-032: An, auf, hinter, in, ncben, iiber, unkr, vor and zwischen
take the dative in the simple locative meaning, the

348 General Linguistics I

accusative to indicate the ending-point relation. *Ich gehe an deine
Tur' (ace.) is thus clearly distinguished from 'Ich stehe an deiner Tiir'
(dat.) — so much so that it is common in German to leave the first term
of the relation unexpressed, e.g., 'Ich will unter den Baum (ace), denn
es ist hier zu heiss!'

(5.9.-2) G033-049: These relaters may be suffixed to wo- and to da-
{woraujf, darauf), in which case the distinguishing
mark between the locative and the ending-point plus locative use is
lost, for da- and wo- do not show case: 'Er stellt sich dazwischen', *Er
steht dazwischen'.

(5.a-3) G050-060: There are quite a number of locative relaters, con-
strued with the genitive, dative, or accusative case,
which show no distinction between the purely locative and the ending-
point relation plus locative uses: 'Er setzt sich links des Baumes',
'Er sitzt links des Baumes'. These locatives are not used as frequently
as their number might suggest. They are here Usted alphabetically
and not according to importance :

ahseitSy ahwdris, aufwdrts, ausserhalb, heiderseits, diesseits,
gegenwdrts, halbwegs, herseits, herwdrts, hinseits, hinterhalb,
hinterriicks, hinterwdrts, hinwdrts, hiiben und drilben, inmitten,
innerhalb, inwdrts, jenseits, Idngs, links, niederwdrts, nordlich,
nordwdrts, oberhalb, oberwdrts, osten, ostlich, ostwdrts, rechts,
rings, ringsum, rittlings, riickwdrts, seitab, seitwdrts, sudlich,
siidwdrts, unfern, unterhalb, unterwdris, unweit, vorwdrts, west-
lich, westwdrts, generally governing the genitive.
bei, benebst, entgegen, gegeniiber, ldngs{i), mit, ndchsty nebst, gen-
erally governing the dative.

durch, entlang, gegen, um, wider, governing the accusative.
Of these, bei, mit, durch, gegen, um, wider, and occasionally some of
the others, can be suflBxed to wo and da.

(5.d) GO6I7O63: Beiseite and zusammen (= at or to a place occupied
by all) never take an expUcit point of reference.
They may express either the ending-point relation plus the second term
or a locative relation plus the second term. *Sie stand beiseite mich
erwartend,' 'Sie nahm mich beiseite und fliisterte mir etwas ins Ohr.'
'Zwsamwenbitten' = 'to invite together, i.e., to a place occupied by all.'

(6) Relaters of the an, auf, hinter group, zu, wider, gegen,

um, and some others are used as prefixes, but as such

do not often retain their hteral meaning. In-, for example, in the

I-'ivc: Studies in I 'ni\ crsiil ( Out cpimil (inunnuir ^4W

form cin- forms the compound cingchcn, wliich iiu'hils not 'to (jo in'
but 'to shrink'; unlergehen means *to succumb'; Widersland meanB
'opposition'. On the other hand, there arc compounds Uke zugehen
'to approach', nuftischc7i 'to set upon the table', umlegcn 'to surround',
which retain the hteral meaning of the rehiters. (icgtnuUrr and hvi ju<
jirefixes tend to retain their hteral meaning ('Gepenu/jerstellung' = 'plac-
ing opposite', '6ezspannen' = 'to hitch next to something'). Samples of
this type of composition are given in Part II, section 12, where they are
analj'zed as a whole as being representative of combinations of the
ending-point relation with the first term.

(7) G064-076: Her and hin plus -ah, -ciuf, -aus, -liber, -unter; her-

plus -vor, form directives which es.sentially include
the ending-point relation (cf. Eng. iip, down, out): 'Warte ein bi.schen,
ich komme bald hinaus'. (See above. General Remarks.) Ilerum is
used both as a locative non-directive ('Die Kinder stehen um den Tisch
herum') and as an e-p directive ('Er kehrte die Miinze heruni). In the
latter case it can be rendered ^whose e-p is a reverse position (at ...)'.
Zuruck, also, is a directive which may imply the ending-point relation;
e.g., 'Er ist schon zuruck\

(8) G077-079: German, Uke EngUsh and French, can combine in one

locution the special expression of a position and an
implication of the ending-point relation: 'Er stellte sich anf den Kopf.'

(9) G080-094: German has words which express an indicated sub-

stantial in combination with the ending-point idea,
namely, in certain contexts, her and hin alone or plus -an, -bei, -ein (-in),
-vor, -zu, -zwischen or plus a stem which expresses a motion. These indi-
cative ending-point pronouns (akin to English hither and thither, and to
ending-point here and there) may refer (1) to a location: 'Wie weit ist's
hin? (How far is it there? to that place?),' 'die //erfahrt (a trip to this
place)' ; (2) to an object: 'Er sah einen Feigenbaum an dem Wege, und
ging hinzu (He saw a fig tree near the road, and went to it)'; or (3) to a
person: 'Samed zittert herzu, und umarmt ihn (Samed tremblingly
goes to the man [implied by context] and embraces him)'. '!•> geht
hinzwischen' = 'He goes to a point between the (indicated) ^ and -v,. '!■>
geht hinzu' = ^He goes to the (indicated) ^ (supplied by context)'.

(lO.a) G095-108: Her suffixed to hier, and hin .suffixed to or following

luo, irgendwo, nirgend{s)wo, etc., hicr, da, dart, uhrrall,

form interrogative, indicative and relative pronouns of space-location in

350 General Linguistics I

combination with the ending-point relation. In these combinations it is
her or hin which expresses merely the ending-point relation (contrast
section 9 above) and wo, hier, dort, etc., which constitute the explicit
second term. 'Komme hierher,' *Gehe dahin, dorthin, uberallhin, nir-
gendswohin.'

It is interesting to note that her, when in combination with hier, like
hin serves as an e-p relater, but when in combination with da, dort, wo,
etc., in sharp contrast to hin, serves as a starting-point relater. And
yet 'Gehe daherein' expresses a movement whose ending-point is a point
in there.

Such an expression as dahinein is placed in this section, hin being
interpreted as being merely an e-p relater, whereas hinein belongs to
section 9, where, in the absence of any such pronoun as da, hin is con-
sidered to be an e-p pronoun.

(11) G109-111: Heim, when used with a word which denotes or

implies movement, may imply the ending-point
relation, e.g., ^Heimweg', 'Schon ist der i/emgang und lieb das Heimsein.'

(12) 0112-138: A number of locutions in German combine the ending-

point relation with an implied first term. Many of
these also include an implied, and some an expressed, point of reference.
Many of them include also a locative relation. 'Er ndherte sich' asserts
A MOVEMENT which hos OS ending-point points successively nearer
SOMETHING. 'Dauu bUeb er stehen^ asserts an implied movement
whose ending-point is some point. Auftischen = a moving whose ending-
point is a point on the table.

/•/It'. Siuilii's in Universal Conccpiuul (iruninuir 351

PART II: SAMPLI-: SENTENCES
Explanatory Remarks

Sentences, each containing some expression of the encling-point
relation, accompanied by an analysis of such expression, will serve to
illustrate, clarify, and substantiate the remarks made in the discussion
in Part I. The samples for each language may be related to the discus-
sion by means of the marginal numbers, which correspond with the
numbers occurring in the discussion.

Samples are entered in column S, and a schematic analysis of each
sample or the pertinent part thereof is given in columns A, r, H. Column
A shows the first term of the relation. Column r gives the relation,
i.e., the ending-point relation, which is to be read 'has as ending-point'
(-r) or 'whose ending-point is' (r)^ Column B gives the second term
of the relation. Italics are used in the sample (column S) to show the
linguistic element which expresses the relation, either term of the rela-
tion, part of such a term, or any combination of these and in columns A,
r, B to show the corresponding elements of meaning.

Examples :

S A r B

The tour ended at Boston.

a tour

-r- Boston

He rode to Chester.

a riding

r' Chester

The journey hither.

a journey

r this place

A trip into the mountains.

a trip

rap* in the
mountains

If more than one linguistic element connotes the ending-point rela-
tion, only one of them is treated at once, and itahcs in colunms A, r, B
show only the meaning of the element being treated ; indeed, convenience
may require that the extra element be ignored. Thus, in the sentence
'The maid admitted him into the front parlor', the ending-pt)int relation

* For the difference between these two renderings see pp. 12-14.

* Read: 'has as ending-point'.

'Read: 'whose ending-point is' ('whose ending-points are') or 'having as
ending-point'. In the analyzed samples the former reading is assumed.

* *p' is to be read 'point' or any convenient synonym; often 'place'.

352

General Linguistics I

is expressed not only in 'into', but also in 'admitted'; for 'to admit'
means 'to cause or permit to enter', and 'to enter' expresses a movement
whose ending-point is a point inside something. When we treat the
word 'admitted', the analysis is:

S A r

5. The maid admitted him an induced r

into the front parlor movement

When we treat 'into', the analysis becomes:

S A r

6. The maid admitted him an induced r

into the front parlor movement

a p inside the
front parlor

Notions implied but unexpressed are included in the entries of the
analyses in columns A, r, B, being either placed in parentheses or
represented by some special means. An impUed 'point' which is the
logical second term of the ending-point relation and first term of some
other relation is represented by the abbreviation p (see above, examples
4, 5, 6, and below, examples 8, 9, 10). An implied ending-point or
point of reference is symbolized by three dots (small dots corresponding
to ordinary type, . . . , and fat dots to italics, . • . ) which can be read
'so and so' (see example 8). If the implication is part of the linguistic
element being treated, the entry is also italicized. Thus:

S
7. It is two miles to Dover.

8. Go behind and look for it.

9. He went in there.

10. She put the baby to bed.

(an extent of
space measur-
ing) two miles^

a movement

a moving

(r)

Dover

a p behind • . .
a pin that place
a p (in) bed

Pronouns are entered in column B in the nominative when they are
without context in that column (examples 11 and 12). They are
entered in their ordinary grammatical form when they are part of a
phrase all of which is in column B (example 13). Thus:

^ The ending-point relation is used not only for movement but also in cases
where there is no movement and consequently no ending at the point in question.
This is true in measures of space, where we make the transfer, from expressions of
limited movement, of seeing one extremity of the extent as the starting-point, the
other as the ending-point, of a hypothetical movement. A somewhat similar
transfer takes place with notions of looking, pointing, facing, directing. (See
Definitions, p. 11.)

l-ivc: Studies in Liiuvcrsul ( onu-fHiinl (ininunitr 353

(an aggressive

(r)

thoy

movement)

spccificalhj di-

ir)

rected actions

impelling

movements

a movement

(r)

a p close Id rn<

11. Up and at them.

12. Als hiitton mehr donn

hundcrt Festen ihm Die
Schliissel ihrer Tore
dargesandt. /Broxter-
mann, in Grimm

13. Come close by me, and tell

me what is the matter.

It is to be noted that the word place is used as a general expression
of the substantial element in interrogative, indicative, and relative
locative pronouns such as where?, here, ivherever (examples 3, 9).

It is further to be noted that *a movement' is used in Column A as a
general term usually preferred to the various specific types of intransitive
movements.

The formula 'a moving' is often used for one type of transitive verb
which appears in the sample sentences, namely, a verb which expresses a
causative transitive motion initiated by one subject (the agent) and
implies a caused intransitive motion ('a movement') by another subject,
the two motions being simultaneous, pursuing the same path and having
the same ending-point, or ending-points so nearly the same that they
can be described in broad outline by identical language. This kind of
verb may conveniently be referred to as a verb of two motions with one
path and one ending-point. To illustrate, *to lay' in the sentence
'Sarah laid her pencil on the desk' expresses a transitive motion by
Sarah and implies a simultaneous parallel intransitive motion by the
pencil. The latter is caused by the former, and of both can be said:
'whose ending-point is a point over the desk'. Verbs of this class
include (in most contexts): put, lay ('Lay it down'), set ('Set it on the
table'), stand ('Stand him in the corner'), pull, draw ('She drew the
curtain'), drag, hand, carry, bring, lead, paste. Every durative causative
verb based on an active intransitive belongs to the category described.

Another formula is used for a different type of transitive verb, namely,
the verb which expresses or implies two motions with different paths
and ending-points, one of the motions in some way causing the other.
The most general formula appropriate for sucli a verb is 'an notion
causing a movement'. Instead of 'causing', a somewhat more sjH^cial
word may be used, such as 'impelling' or 'compeHing', where the
performance of the intransitive movement is not voluntary . or 'imiucing'
where such performance is voluntary.

:?54 General Linguistics I

Examples :

S A r B

14. He dropped the bucket an action impel- (r) a p in the well

in the well. ling a move-

ment

15. Down with your sails. {an actionimpel- (r) a p downward

ling a move- from . • .

ment)

16. He threw me a ball. an action (a (r) I

moving) impel-
a movement

17. He shooed the cat out of an action com- (r) a p outside the

the house. pelting a move- house

ment

18. Advertised by two Sym- an action {an (r) a p before . . .

erons, whom he sent order) compel-
before. ling a move-

ment

19. The butler admitted the an action indue- r a p in ,,,, in.

visitor into the parlor. ing a move- the parlor

ment.

20. We import large quanti- actions {con- r pp in one's

ties of coffee every year. tracts or orders) country

inducing mov-
ings

The caused movement may accompany or follow the causative
action, or may in part accompany and in part follow, as illustrated in
the above examples. Thus :

Analysis of transitive verb. Time relation of caused movement

to causative action.

14. An action impelUng a move- succeeding.

ment.

15. An action impeUing a move- simultaneous.

ment.

16. An action (a moving) impel- simultaneous and succeeding.

ling a movement.

17. An action compelUng a move- simultaneous and succeeding.

ment.

Five: Sriidics in I 'nivcr\al ( Onccpimil ( iruninuir 355

18. An action (an order) compel- succeedinR.

ling a movement.

19. An action inducing a move- simultaneous and succeeding.

ment.

20. Actions (contracts or orders) simultaneous and succeeding, if

inducing movings. causative action be considered

as durative, e.g., a standing
contract between importer and
exporter;

succeeding, if causative action be
considered as non-durative, e.g.,
an order once given by importer
and accepted by exporter.

With a verb of two motions with different paths the ending-jjoint
of the initial action (whether physical or volitive) is not ordinarily
expressed but only that of the caused movement. For e.xample: 'Sarah
poured water into the jug' pictures an action (by Sarah) impelling a
MOVEMENT (by the water) whose ending-point is points in the jit..
The unexpressed ending-point of the causative action is a point at which
the vessel of supply, from which Sarah wishes to cause water to flow,
reaches a position which impels the contained water to pour forth. In
throwing a ball (No. 16), the impelling action of the hand ends when the
ball leaves it. But the impelled movement of the ball continues, ami it
is that, and not the action of the hand, whose ending-point is T. In
scaring away a cat (No. 17), the man may not have touched the cat,
but may have compelled her by chasing and shouting to run to a point
outside of the house while he remained within. In the case of an
admitting (No. 19), the butler may hav^e remained in the hall and
merely opened the parlor door and bowed to the visitor to enter, in which
case the ending-point of the butler's action was a point in the hall,
whereas that of the visitor's was a point in the parlor.

It now remains to indicate the sources for material used in this study.
In general, we have sought our samples in the standard dictionaries of
the languages, especially in dictionaries which quote known authors anil
general usage, A few samples are taken from other book sources and
a number are purely ad hoc. The source, if any, is given after each
sample. The following abbreviations are employed:

356

General Linguistics I

NED A New English Dictionary on Historical Principles; Oxford,

Clarendon, 1884-1908.
Littr6 E. Littre, Dictionnaire de la Langue Fran^aise; Paris,

Hachette, 1885-1889.
Grimm Jacob and Wilhelm Grimm, Deutsches Worterbuch; Leipzig,

S. Hirzel, 1854-
S ANDERS Sanders, Worterbuch der Deutschen Sprache; Leipzig, Otto

Wigand, 1876.
Muret-Sanders Muret-Sanders, Encyclopaedic English-German
and German-English Dictionary; Berlin, Langenscheidtsche
Verlags-Buchhandlung, 1891-1901.
CuRME Curme, A Grammar of the German Language; New York,

The Macmillan Company, 1922.

Sentences with Analyses in Testing-Frame

Some of the rubrics at the heads of sections, in order to be complete,
should be supplemented by the words ^and which in the given context imply
{do imply, do express) the ending-point relation'. To avoid cumber some-
ness such a phrase has been omitted. The reader is asked to supply it where
needed.

abbreviations and conventions

e-p 'ending-point',
fig. 'figurative',
a movement 'an intransitive movement',
a moving 'a transitive movement'.

n 'Note'. A raised n indicates that there is a note in

Part III.
p 'point', 'place', or any convenient synonym, p repre-
sents an implied 'point' which is the logical second
term of the relation being treated. (See Definitions,
page 11.)
pp 'points', 'places', 'successive points, places'.
P 'position'.
PP 'positions', 'successive positions',
-r 'has as ending-point'. Used only in predicative expres-
sions of the e-p relation (section 1 of each language).
r 'whose ending-point is' ('whose ending-points are') or
'having as ending-point'.

I-'i\c: Siuilics in I'nivctsiil ( (>/uc[>ituil (iidniniar

357

'so and so', 'such and such'. Term to be suppHcd from
context. Three dots represent tin implied second term
of the e-p relation or an implied '\)umi of reference' (sec
Definitions, page 11).

Large dots rei)resent italicized dots, with same inter-
pretation as small dots. Large dots form part of the
interpretation of the locution which is italicized in the
sample.

represents an expressed second term or point of reference.

indicate an omission.

Samples of the Expression of the Ending-Point Relation in

English

Sample Sentence

First term of
relation

Re-
la-
tion

Second term uf
relation

(1) Predicative e-p expression.

EOOl The journey ended at Chi- a journey
cago.

-r Chicago

additional concepts,
r Chester

(2) Usual e-p relaters, occasionally vnth implication of

E002 He rode to Chester. a movement

E003 She threw^' crumbs to^ the a moving impelling r the birds

birds. a movement

EOOl To bed^, you rascals! (a movement)

£005 He pointed to a clump of a line of indication

trees. /Blackw. Mag., in NED

E006 He went to his death'*. a movement

r {a p in) bed
r a clump of trees

E007 He asked him out to^ a (a movement)
round of golf.

E008 She has been to confession'*, (a movement)

E009 Which is the road to Joliet?

EOlO It is eleven miles to Witney.
/NED

EOl 1 Protestant to the backbone^
/Eraser's Mag., in NED (fig.)

E012 You should know what it is
to wear'* iron <o your bone! /An-
derson, Gods of the Lightning, p.
35

a road

(an extent

space)
(an extent)

a wearing (thought
of aa having ex-
tent)

r (the p at which) his

death (occurred)
r (the p at which

they will play) a

round of golf
r (the p at which

one makes one's)

confession
r Joliet
r Witney

r the backbone

r your bone

358

General Linguistics I

E013 Then speed we Hermes

the Flitter, to go Unto the isle
Ogygia. /Morris, in NED

E013a Majestic men who looked
unto the skies. /Aird, in NED

E014 My throat is cut^ unto the
bone. /Wordsw., in NED

E014a The hope thus to press thee

f/n^o my fond bosom. /R.Allan,

in NED
E015 When God had brought me

theTeunto^. /Myers, in NED

under 'thereunto'

a movement

lines of sight

a cutting (or: an
extent of cut-
ting)

a pressing

a moving

(3) E-p relaters, with emphasis on distance traversed,

E016 If you walk as far as Mt. a movement

Hermon, you'll be tired out.
E017 The new state road is laid all an extent

the way to Pocantico.

(4) E-p relaters in combination with locative relaters.

E018 Come into the garden, Maud.

/Tennyson, in NED
E019 At dusk he harries the

Abazai — at dawn he is into^

Bonair. /Kipling, in NED
E020 These smaller off-drains

should be flushed into the main

street drain. /Kendall Merc,

in NED
E021 The child walked out into""

the street.
E022 He examined into every fis-
sure in the crags. /J. F. Cooper,

in NED
E023 He dropped the bucket into"^

the well.
E024 The poor comedian runs^

into^ a tree.
E025 And the black blood flowed

thereinto. /Morris, in NED

under 'thereinto'

a movement
(a movement)

an action impel-
ling a movement

a movement
a searching

an action impel-
ling a movement
a movement

a flowing

r the isle Ogygia

r the skies

r the bone

r my fond bosom

r that place

r Mt. Hermon

r Pocantico

r a pin the garden

r a p in Bonair

r a p in the main
street drain

r a p in the street
r pp in every fissure

r a p in the well
rap against a tree
r pp in that place

'Into' is here used figuratively.

live: Smdu's in Lnncr.sul ( Oncfptnnl (imnmiur

359

E026 A bit of ceiling dropped onlo a movement

the floor.
E027 It rolled onto the sidewalk, a rolling

r a p on the floor

r a p on the 8i<lf'»Hlk
(5) Relational locative expressions which may imply ending-point relation.
(5. a) Those which require an explicit point of reference.
E028 Up and at^ them. (an aggressive (r) pp in contact with

E029 I looked at"^ him.

movement) them

a directed line of (r) he
sight

E030 A large loligo had thrown a movement

itself high and dry upon the

beach. /H. Miller, in NED
E031 We went round the corner, a movement

/Dickens.
E032 My mother put her arms a moving

around my neck. /J. Wilson, in

NED
E033 The ship was dashed against a movement

the pier-head. /NED

(r) a p on the beach

(r) a p around the
corner

(r) a P around my
neck (at my
neck)

(r) a P directed to-
wards the pier-
head (at the
pier-head)

(/•) a p in contact with
whom?

(r) a p on him

E034 Guess whom I ran* against^ a movement

in London the other day? /NED
E035 Her gray eyes absolutely a flaming

flamed upon him. /Mrs. Oli-

phant, in NED
E036 I was rudely thrown upon^ a moving impel- (r) a P with remain-

my back. ling a move- ing portion of

ment body over back

(on ground)
E037 The collocation of such a moving (r) the p occupied by

names as those of Aeglamour such others

and Earine with such others as

March and Maudlin. /Swin-
burne, in NED

(5.b) Those which may be used with or without a point of reference, without
change of form.

E038 Don't think I want to get (a movement)
aboard your ship, /G. Mac-
donald, in NED

(r) (J p aboard your
shii>

* 'To run against' is used figuratively. Cf. 'into' in M021. above.

360

General Linguistics I

E039 The Syracusans got aboard,
and rowed along-shore. /Grote,
in NED

E040 All aboard]

E041 Scatter seeds about. /NED

E042 He looked about him' for
some means or way To keep this
unexpected holiday. /Longf.,
in NED

E043 In another corner a wooden
stair leading above.
/Dickens, in NED

E044 Lay it across^ the entrance.

E045 I jumped right on to the ice,
and how I got across^ I don't
know. /Mrs. Stowe, in NED

E046 He planted roses along the
wall.

E047 The naked hulk alongside^
came. /Coleridge, in NED

E048 The telescope is swayed

by the earth's rotation athwart the
rich regions of the galaxy. /R.
A. Proctor. Expanse of Heaven,
Funk and Wagnalls New Stand-
ard Dictionary

E049 When many meats are set
before me. /Hooker, in NED

E050 As ill an action as any that
comes before the Magistrate.
/Steele, in NED

E051 Advertised by two Symer-
ons, whom he sent before.^
/Johnson, in NED

E052 The Benjamites looked be-
hind them. /Bible, in NED

E053 Go behind and look for it.
/NED

a movement

(a movement)
movings impelling

movements
lines of sight

a stair

a moving

(a movement)

movings

a movement

a moving (of lines
of sight)

a moving
a movement

an action compell-
ing a movement

a line of sight

a movement

(r) a p aboard , .

(r) a p aboard . * .
(r) divers pp around

. . . (you)
(r) divers pp around

him

(r) a p above . . .

(r) a P across the
entrance line (at
the entrance)

(r) a p on the other
side of , » *

(r) pp along the wall

(r) a P lengthwise of

... (at the side

of,..)

(?•) PP across the rich

regions of the

galaxy

(r) a p before me

(r) a p before the mag-
istrate

(r) a p before . . .

(r) a p behind them
(r) a p behind . . .

' Metaphorical; mental operations involving a purpose are expressed in terms
of directed sight, a type of movement with ending-point.

fhi': Sluilics in I nivcrsal Coficcpliiul (iruninmr

>u I

E054 It being the turn of our a movement

watch to go below. /R. Dana,

in NED
E055 This new Jehu Instructs amoving

the beast To take the bit

between his teeth and fly. /Dry-
den, in NED
E056 If Mrs. B. had not thrown a movement

herself betwixt us. /Dickens, in

NED
E057 Lofty souls who look beyond lines of sight

the tomb. /Beattie, in NED
E058 Never able to pass a step a movement

beyond the self-drawn circle.

/Spalding, in NED

E059 [The snake] retires And

in some secret Cranny slowly
glides. /Dryden, in NED

E060 The most judicious mode of
putting a kicker in harness. /M.
J. Higgins, in NED

E060a In here, please!

E061 Show them in.

E062 /nbound vessels told of pas-
sing these valiant American
schooners. /Outing, in NED

E063 Norway's inarming melan-
choly sea. /F. W. H. Myers, in
NED

E064 Running down the middle of
the triangular plate is the cen-
tral string of tissue, the rachis,
and at its end the incurrent
blood-vessel. /Stud. Biol. Lab.
Johns Hopkins, in NED

E065 I have enclosed herewith a
copy of the letter about which
you inquired.

E066 I went inside and waited.

E067 Admission within the fold.

E068 He put it just outside^ the
door.

a movement

an action compell-
ing a movement

(a movement)

an action impelling

a movement
a movement

an extension

a flow

a moving

a movement

(an action inducing

a movement)
a moving

(r) a p below . . .

(r) a p betwern bin
teeth

(r) a p between us

(r) a p beyoTui the
tomb

(r) a p beyond (the cir-
cumference of)
the self-drawn
circle

(r) a p in some secret
cranny

(r) a p in harness

(r) a p in this place
(r) a p in * . ,

(r) a p in . . .

(r) a p in . . ,
(r) a p in . . .

(r) a p in . » . (this
envcloy>o)

(r) a p in . . .
(r) a p in the fold

(r) a p outside (be-
yond) the door

362

General Linguistics I

E069 The men and women were a movement

ordered to come outside. /R.

Boldrewood, in NED
E070 Pinchas betook himself

unceremoniously vnthout. /Zang-

will, in NED
E071 Just lay it on the table.
E072 When up reached the elder

his hands thereo?i to lay. /Mor-
ris, in NED under 'thereon'
E073 Polly put the kettle on.
E074 Let us draw a veil over this

dismal spectacle. /Temple Bar

Mag., in NED
E075 Throw it under the bench.

a movement

a movmg
a moving

a moving
a moving

E076 Rude Boreas, who likes to
let daylight under the focussing
cloth. /Photogr. Ann., in NED

E077 He went under for the third
time.

a moving impelling

a movement
(a movement)

a movement

(r) a p outside . . .

(r) a p outside . .

(r) a p on the table
(r) a p on that place

(r) a p on , . ,
(r) a p over this spec-
tacle

(r) a p under the
bench

(r) a p under the fo-
cussing cloth

(r) a p under . . . (the
surface of the
water)

(5.c) Those which have different forms when used
reference.

E078 She put a chair on the table a climbing

and climbed on top.
E079 The arrow fell short of its a falling

mark by no more than a foot,
E080 They sailed off to the south^. a sailing

E081 Turn to the east^.

a turning

EX)81a Then Coyote went off west a movement

of^ the mountains.
E082 We went out of^ the room a movement

and waited there for poor Tommy.
E082a Out] Out with you! (a movement)

with or without a point of

(r) a p on top of * * »

(r) a p before its mark

(r) a p at the south of
. . . (their starting-
point)
(r) a P facing the east
of, . .(your pres-
ent location) (at
. . .your present
location)
(r) a p west of the

mountains
(r) a p outside of the

room
(r) a p outside of (this
place)

/he: Simlics in L ninrsul C onu'/niuil (irumniur

E083 Come close by me, and tell a movement

me what is the matter. /NED

under 'by'
E084 Now they come nearer, a movement

/Mrs. Radcliffe, in NED
E085 He lost no time in speaking a movement

as soon as she came near him.

/Geo. Eliot, in NED
E086 Pull up^ your chairs. a moving

E087 He walked up io^ the parson, a movement

E088 Get as far from here as a movement

possible.
E089 Often he would go apart a movement

seeking solitude.
E090 She folded her work, and a moving

laid it aiyay^'. /Longf., in NED
E091 One gentleman drew another a movement

aside to speak in an under-tone

about Scotch bullocks. /Geo.

Eliot, in NED
E092 To evade and slip aside /rom a movement

difficulty. /Burke, in NED
E093 I came from Arizona and I'm no movement

not going back.

(r) a p close to nie

(r) a p ncartir to . . ,
(r) a p near him

(r) a p near . , ,
(speaker)

(r) a p nearer the par-
son

(r) a p as far as possi-
ble frojn here

(r) a p away from . . .
(other people)

(r) another (but not
too distant) place

(r) a p to one side of

(r) the p formerly occu-
pied by me

(6.d) Those which ordinarily do not take an explicit point of reference.

E094 Take the gentleman tip a moving
front^ to a better seat.

E095 In making this effort the a movement
spike of my axe turned upper-
most. /Tyndall, in NED

E096 The coming together of a movements
crowd takes but a few minutes.

(r) a p in the front
part of this place

(r) a p above all (of
remaining por-
tion of a.xe)

(r) a p occupied by
them all

(7) Locutions which consist of or include some expression of direction
E097 I'm going down^ to Tucson, a movement

E098 I'm going up to Wisconsin, a movement

E099 He sat in his car and waited a movement

for her to come down.
ElOO I think that this case must a movement

go down for a new trial. /Ld.

Watson, in NED

r Tucson, a p down-
ward from . . .
r Wisconsin, a p up-
ward from . . .
(r) a p downward from

(r) a p downxrard from

364

General Linguistics I

E104 He is now in Newcastle, but a movement
is coming down south next week.
/NED

E105 Take this downstairs for me. a moving

El 06 I'm going upstairs.

a movement

E107 He seldom goes out in this a movement

weather. /NED
E108 Why should they put me a moving

out?
E109 Turn him over^ on his face, a moving

/NED

EllO Saying that she checked and a movement
sharply turned about^ to hide her
face. /Tennyson, in NED

ElU Sit up^ and pay attention, a movement

a p in town down-
ward from . . •

ElOl Down^ with your sails, (an action impel- (r) a p downward from

Motteux, in /NED ling a movement)

E102 Today when I go down town, a movement (r)

I shall subscribe for the 'New

York Observer' for you. /Gray,

in NED under 'down'.
E103 'Taking her down south'? a moving (r)

said the man. /Mrs. Stowe, in

NED

a p in the southern
part of the coun-
try downward
from . . .

(r) o p in the southern
part of the coun-
try downward
from . . .

(r) a p lower than the
stairs downward
from . . .

(r) a p above the stairs
upward from . . .

(r) a p outward from

• • *

(r) a p outward from

• • •

(r) a horizontal P fac-
ing the opposite
direction from
that of previous
P (at . . .)

(r) a vertical P facing
the opposite di-
rection from
that of previous
P (at . . .)

(r) an upright P (at

(8) Locutions which express position irrespective of context.

E112 A tall figure reared itself up- a movement (r) an

right at her approach. /L. B.

Walford, in NED
E113 The films are thick enough to movings (r)

place in racks to wash, or to

stand upright to dry. /Photogr.

Ann., in NED
E114 He put them upside down on movings (r) vertically

the tables. PP (at

upright P (at
, . . , a horizon-
tal surface)
upright PP (at . . .)

reversed
..)

Five: Sliidii's in Universal C <>m rpiiiiil ( irdninuir

ViS

El 15 To start this game the play- movements
era place themselves back to back
in two lines down the center of
the room.

(r) bark-to-baek I'l'
(at . . .)

(10) Space-locative pronouns: interrogative, indicative, and relative.
(10. a) Those which essentially include the e-p idea.

El 16 Whither away? (a movement)

E117 Come hither, my boy, and let a movement

me see you.
El 18 The road thither leaves the a road

main road at right angles.

/Jenkinson, in NED
El 19 And thither he went, taking a movement

his young wife Annabelle.
E119a I have hereunto set my a moving

hand and seal.
El 19b The feast where[un]to we a movement

hasten.
El 19c Go whither he went. a movement

E120 Go whither^ you will. a movement

E121 Wandering they knew not a movement
whither. /Dickens, in NED

(10. b) Those which may imply e-p relation.

E122 Where do you think you're a movement

going?
E123 We moved here last May. a movement
E124 Give it herc^. (a moving)

E125 It takes a half hour to walk a movement

there.
E126 I'm going over there tomor- a movement

row.
E127 What's taking ye i/icreaways/ amoving

/Buchan, in NED

E128 I'm going up yonder the first a movement

day my back's feeling better.

E129 She followed on to the place a movement (by

where he had gone. him)

E130 Let's go where it's quiet. a movement

r what placet
r this place

T that place

r that place

rap {on) this object

r which (feast)

r the place to which

he went
r whatever place (is

desired)
r what place

(r) what place?

(r) this place
(r) thi^ place
(r) that place

(r) that place

(r) that place selected
out of many jws-
sible OMS

(r) yon place

(r) which (place)

(r) a place in which
it is quiet

366

General Lini^uistics I

E131 O! take them anywhere, but a moving

leave me in peace.
El 32 We have nothing here for a movement

you, go elsewhere.
E133 Our mail system reaches a movement

everywhere.
E134 You will go nowhere^, I tell a possible move- (r) no place

you. ment, any move-

ment
E135 They must have gone some- a movement

where.
E136 There's no room there, put it a moving

somewhere else.
E137 Wherever^ I go, I find the movements

same sadness.

(r) any place
(r) another place
(r) every place

(r) some place

(r) some other place

(r) any place which
(is reached)

(11) Places, no place, home, abroad,

E138 You look like you're going a movement
places^.

E139 All dressed up and no place to a possible move-
go^, ment

E140 It's late enough for us to be a movement
going home.

E141 Her final argument is that a movement
she will go home to mother.

E142 He is going abroad for his movements
vacation.

(r) several places

(r) no place

(r) home

(r) pp in foreign lands

(12) Combinations of e-p relation with first term of relation and often also
other concepts (verbs, nouns, adjectives).

E143 I came upon^ a flower. a movement

E144 We came across^ an old de- a movement

serted cottage.
E145 The Russians came up a movement

with his rear. /Ann. Reg., in

NED under 'up'
E146 They will never catch up to a movement

that horse. /F. Hume, in NED

under 'up'
E147 K we can only gain Portland (a movement)

before dark.
E148 We will hope to /it7 Pendleton a movement

before nightfall.
E149 In spite of the heavy sea they a movement

managed to make land without

mishap.

(r) a flower

(r) an old deserted

cottage
(r) pp at which his

rear is

(r) any p at which that
horse is

(r) Portland

(r) Pendleton

(r) the land

Five: Sluilii's in i'ni\i-i\nl Ctnucptuiil (irnrnnuir

367

E150 Would you care to join us a movement

a^ain later?
E151 The squad succeeded in re- a movement

joining its brigade that day.

E152 Only by a ruse did we man- a movement

age to reach the inside of the

palace.
E153 A carriage with six mules o movement

drew up to the guard-house.

/Southey, in NED under 'up'
E154 The captain's party is the a climbing

one that has thus far succeeded

in surmounting the peak.
E155 She had put^ the baby to^ an action compel-

bed and now sat reading a book. ling a movement

E156 He put^ the onion to"^ his a moving

nose.
E157 He put his pack down at the a moving

gate.
E158 Stand^ it in the corner. a moving

E159 The judge enters from the

rear.
E160 The entrance of the police

was received with whispering and

sidelong glances.
E161 I entered the house, but saw

no one.
E162 We too often get tlie notion

that the butler's sole function is

to open doors and admit visitors.
E163 Knock twice and you will

be admitted.
E164 Approaching the shack, I

came upon an old cistern.
E165 We neared the shore without

intending, however, to land.
E166 We hoarded the train and

started looking for our berths.
E167 He mounted and rode off.

a movement
a movement

a movement

an action inducing
a movement

an induced tnove-

ment
a movement

a movement

E168 They wmn< their soap-boxes a movement
and hold forth to the motley
crowd, whose sole occupation
seems to be to listen to these
harangues.

r the p occupied by Ufl
all

r a customarily oc-
cupied p with it.q
brigade

r the inside oi the
palace

r the guard hou.sc

r a p on top of the
peak

rap (in) bed

rap (be/ore) his

nose
r a pat the gate

r a vertical P at a [)

in the corner
r a p in . » .

rap in . . .

r a p in , . . the
house

rap in . . .

r a p in . , ,

rap near . . . the

shack
rap iiKir . . . the

shore
rap aboard . . .

the train
rap on ... (a

horse)
r a p on . . . their

8oap-boxe8

368

General Linguistics I

E169 They mounted the gun as a moving

quickly as possible.
E170 She will stop at the hotel (a movement)

north of here.
E171 It's time to embark.
E172 Two stones with incised

crosses. /Lubbock, in NED

a movement
extent of cuttings

E173 He was not your ordinary

drunkard, who will imbibe just

anything.
E174 We import large quantities

of coffee every year.
E175 The price of our imported

cotton is relatively high.
E176 When the metal reaches the

proper temperature, it is im-
mersed in cold water.
E177 It is like inserting a monkey

wrench into the mechanism.
E178 Introduce a few drops of the

acid into the nitrate.
E179 He landed poorly and was

off-stride for the next hurdle.
E180 The landing of the pilgrims.
E181 He landed on the champion

with his right.
E182 Cats always light on their

feet.

E183 Mr. Thurston of the finance
committee wishes to table the re-
solution.

E184 The more our arrival is un-
heralded, the better will be the
results.

E185 We arrived in Paris.

E186 Osgood didn't turn up after
all.

E187 To cut and deliver the ma-
terials at the spot. /G. White, in
NED under 'at'

E188 These streamers seem to con-
verge at a point beyond the ze-
nith. /Sc. Monthly, in NED

a drinking

actions inducing
movings

an action compel-
ling a movement

a movement

a movement
a moving

a movement

a moving

a movement

a movement
a movement

a moving

movements from vari-
ous directions

r a p on , , »

r . . ., a p at the
hotel

r apon » , * (a boat)

r pp in . . . stones,
(which have the
form of) crosses

rap inside oneself

r pp in one's country

r a p in liquid, in
cold water

r a p in ..», in the

mechanism
r a p in ..», in the

nitrate
r a p on the ground

r a p on the land

r a p on ,,,, on
the champion

r a p on t , , (the
ground) at which
a P with remain-
ing portion of
body above feet

r a p on the table
(fig.)

r . . .

r . . ., a p in Paris
r a p at , , ,

r . . ., a p at the
spot

rap occupied by
them all

Five: Sludii'.s in Lhiixfrsul CO/u I'ptunl (iramituir

;(>w

E189 He returned late last night, a movement

E190 The masked wrestler having
downed all the professional ath-
letes. /Gunter, in NED

E191 If caught in a hard sudden
squall, <ioum helm at once. /Bed-
ford, in NED

E192 She could not afford two out-
ings in the year. /Illustr. Lond.
News, in NED

E193 With that heaps and he oufs.
/Doyle, in NED

E194 TheWesternermay out pistol
and shoot you if you annoy him.
/M. O'Rell, in NED

E195 Someone is always coming to
pay us a visit.

an action compel-
ling movements

a moving

trips

a movement

a moving

a movement

r the p formrrly occu-
pied by him
(r) pp downward from
. . . («<•//)

(r) a p downward from

r pp outward Jrom
, . . (urban home)

rap outward from

rap outward from
. . . (pocket)

(r) (this place)

Samples of the Expression of the Ending-point Relation in

French

Sample Sentence

First term of
relation

r
Re-
la-
tion

H
Second term of
relation

(1) Predicative e-p expression.

FOOl Son voyage s'est termini a a voyage
Adrianople.

-r Adrianople

(2) Usual e-p relaters, occasionally with implication of additional concepts.

(a 7novemc7it) r the thief

(a movement) r the station

(a 7novemcnt) r port

a flight r foreign lands

F002 ^u voleur!

F003 A lagare!

FOOl lis ont touchc'- au port.

F005 Je mdditais ma fuite aux
terres <?trang^res. /Racine, in
Littr6

F006 Enfin je viens d vous. /Ra-
cine, in Littr£

F007 Je respecte Idamd jusqu'en
son 6poux mfime. /Volt., in
LiTTRfi under 'jusque'

F008 II est mont6 jusque sur les
toits.

a movement

(an extent)

a movement

r vou

r her spouse

rap on the roofs

370

General Linguistics 1

F009 Jusqu'oii? (a movement) r what place?

FOlO Apportez-les jusqu'ici, s'il a moving r this place

V0U8 plait.

(5) Relational locative expressions which may imply e-p relation.

(5. a) Those which require an explicit point of reference.

Foil C'esttoujoursuncherplaisir a movement

que de retourner chez soi apr^s un

long voyage.
F012 II saute dans la tranch6e a jump

avant tous.
F013 II va en Italie. a movement

F014 Le comte est all^ en pro- a movement

vince.
F015 Je me jette sous un arbre a movement

pour lire.
F016 La mouette descend lente- a movement

ment sur I'eau.
F017 Place-toi contre ce logis. (a movement)

/Mol., in LiTTRfi
F018 La revolution de la lune in- a movement

terpose ce satellite entre le soleil

et la terre. /Littre under 'in-

terposer'

(r) a p at one's own
place

(r) a p in the trench

(r) a p in Italy

(r) a p in the country

(r) a p under a tree

(r) a p on the water

(r) a p opposite this

dwelling
(r) a p between the sun

and the earth

(5.c) Those which have different forms when used with or without a point of
reference.

F019 Qu'on passe deux fois Au a movement

degd du rivage bleme. /Malh.,

in Littr6 under 'dega'
F020 Peuples qui erraient degd et a wandering

deR sur des chariots. /Boss., in

Littr6 under 'dega'
F021 Qu'appelle-t-on franchir les a movement

bornes de toute pudeur et passer

au deld de toute impudence ?

/Pasc.jinLiTTRf; under 'dela'
F022 Porter deld les mers ses a moving

hautes destinees. /Corn., in

Littr6
F023 Le comte Raimond mettait a moving

en dehors toute son dme. /Stael,

in LiTTRf; under 'dehors'
F024 II est all6 au dehors de la a movement

ville.

(r) a p this side of the
shore

(r) a p this side of * . .

(r) a p beyond all
impudence

(r) ap beyond the seas
(r) a p outside . . •

(r) a p outside of the
city

/■ivc: Studies in Universal Co/ucfiiuil (munnuu

371

F02.5 Ce hntiment va mettre de-
hors. /Littr6
F025a Dehorsl
F026 Je n'entre pas \a.-dedans.

/Pascal, in Littr6 under '\k'
F027 L'eau ne nous venait qu'au

dessous dn genou. /Littu6

under 'dessous'
F028 Mettez ce paquet \k-dessous.

/LiTTRfi under 'la'
F029 II tficha de me mettre des-
sous. /F6n., in Littr6
F030 Elle avait place son tombeau

au dessiis d'une des portes les plus

remarquables de la ville. /Rol-

lin, in Littr£ under 'dessus'
F031 Mettez ce livre \k-dessus.

/Littr6 under 'la'
F031a lA-dessus, s'il vous plait.
F032 Ce qui est sous la table met-

tez-le dessus. /Littr6
F033 II mit les chaises aulour de

la table.
F034 La foule se portait au-devant

d\i prince. /Littr^ under 'de-

vant'
F035 Les services d'Hipal en ce

meme moment Lui reviennent

devant la vue. /La Font., in

Littr6
F036 Allez devant.
F037 La troupe s'arrete derriere

la foret.
F038 II s'est retird derriere un re-

tranchement. /d'Ablancourt, in

Littr6
F039 Monsieur s'assidra au haul de

la table.
F040 Aien^our du muet toutes huit

accoururent. /La Font., in

LlTTKfi

F041 Elle rcpandait alenlour du
1)1(5 pour les poules.

a moving

(a tnovirnrnt)
no movement

a movement

a moving
a moving
a placing

a moving

(a moving)
a moving

a moving

a movement

a movement
a movement

a movement

a movement
a running

movings impelling
movements

(r) a p outside of , . .

(r) a J) outside of . . .
(r) a p in that place

(r) a p below the knee

(r) a p btlinc that

place
(r) a p below , . .

(r) a p above one of the
doors

(r) a p above that

place
(r) a p on that place
(r) a p above . . .(table)

(r) pp about the table

(r) a p before the
prince

(r) a p before his sight

(r) a p before . . .
(r) a p behind the

forest
(r) a p behind an in-

trenchment

(r) a p at the head of

the table
(r) pp about the dumb

man

(r) pp about • . . (self)

372

General Linguistics I

(6) Prefixes which sometimes express e-p relation.

F042 Le mar6chal s'etant ache- a movement

mine pour aller aTreves /Sdv.,

in Littr6 under 'acheminer'
F043 On conclut Qu'il ne fallait a movement

s'aHablerdavantage. /La Font.,

in Littr6 under 'attabler'
F044 Le bruit en arriva aux oreil- a movement

les de la police. /Littre under

'arriver'
F045 Les arrivants etaient nomb- a movement

F046 II lui ass6na un coup de ba-
ton sur la tete. /Littre under
'ass^ner'

F047 Les glorieux assauts de plus
de cent murailles. /Corn., in
Littr6 under 'assaut'

F048 Je n'assiege pas la porte des
grands. /Boss., in Littre under
'assi^ger'

F049 Get enfant brise tout ce qu'il
atteint. /Littre under 'attein-
dre'

F050 Approchez-vous du feu.

F051 II s'embarqua sur un paque-
bot. /Littr6 under 'embarquer'

F052 Je fais encadrer nos dessins.
/J. J. Rouss., in Littre under
'encadrer'

F053 Le harnais ^clatant qu'il
avait endosse. /Tristan in Lit-
TRfi under 'endosser'

F054 La vieillesse chagrine inces-
samment amasse, Garde non pas
pour soi les tr^sors qu'elle en-
tasse. /Boil., in Littr6 under
'entasser'

F055 Et, pour gagner, emballent
et d^chargent toutes sortes de
marchandises prohib^es et dMen-
dues. /Arret du Gonseil d'Etat,
in LiTTR^ under 'emballer'

an action impel-
ling a movement

a movement

a movement
a movement

movings
a moving
a moving

a moving

r the road

r a p at the table

r . . . , the ears of
the police

r a p on his head

r pp at more than a
hundred walls

r the door

r everything

rap near the fire
(r) a p ni . . . , on a

merchant ship
(r) pp in frames

(r) ap on his back

(r) pp in a pile

(r) pp in a package

Five: Sludii'.s in Univfr.\nl ( Dnccpiiial (ifittn/nur

373

F056 Quand j'avais empoch6 mon
livre, je ne songeais plus d rien.
/J. J. Rouss., in LiTTu6 under
'empocher'

F057 Si on vendnit le feu et I'eau,
il devrait ^tre permis de les im-
porter et de les exporter d'un
bout de la France a I'autre.
/Volt., in Littr6 under 'importer'

F758 On n'mcise point I'arhre du
cote de la terre, mais du cote de
I'eau, pour qu'il tombe sur le
courant. /Chateaub., in Littr6
under 'inciser'

F059 L'eau s'lnfiltre dans le bois
le plus dur. /Littr^ under 'in-
filtrer'

F060 II faudrait avoir un moyen
d'cvaluei la quantite d'eau que
les plantes Vmbibent de cette
maniere. /Bonnet, in Littr6
under 'imbiber'

F061 Le gouvernement qu'on es-
saya d'mplanter en ce pays.
/LiTTR^ under 'implanter'

F062 La r6volution de la lune
interpose ce satellite entre le
soleil et ia terre. /Littr6 under
interposer'

F063 Entremeler des fleurs rouges
a des fleurs blanches. /Littr6
under 'entremeler'

F064 II est aver6 aujourd'hui que
les cinq ou six lignes qu'on attri-
bue k Josephe sur J6sus, ont 6t6
Vn<erpol6es par une fraude tros-
maladroite. /Volt., in Littr6
under 'interpoler'

F065 II faut ne /ransvaser les vins
que lorsqu'ils sont bien faits.
/Genlis, in Littr^

F066 II a fransporte leurs bagages
d'une gare k I'autre.

a moving

actions inducing
movings

a cutting

a movement

a moving

a movement

a moving

actions impelling
movements

a moving

(r) a p in my pocket

(r) pp in . . .

(r) a p in the tree

(r) a p in . . , the
toughest wood

(r) pp in (themselves)

(r) a p in ... this
country

(r) a p between . . .

and . . . the sun
and the earth

(r) pp among . . .
white flowers

(r) a p between . . .
and . . .

(r) pp in other vessels

(r) pp in another sta-
tion

374

General Linguistics I

(8) Locutions which express position irrespective of context

F067 Je me mets debout. a movement

F068 Elle a mis son bonnet sens a moving

devant derri^re. /Littr6 under

'devant'

(r) an erect P (at . . .)
(r) a reversed P (at

(10. b) Interrogative, indicative, and relative pronouns which may imply e-p
relation.

F069 Ou vont toutes ces personnes?

F070 Val6re, ici. /Destouches,
in LiTTRfi

F071 Vous savez quel sujet con-
duit ici leurs pas. /Rac, in
Littr6 (fig.)

F072 D'ici Id, on compte deux
lieues. /Littr^

F073 C'est la que je veux aller.

/LlTTRfi

F074 Nous n'avons pas assez d'ar-
gent pour y voyager.

F074a L'eau limpide s'y deverse.

F074b J 'y ai appos6 ma signature.

F075 Nous allons Id-bas de temps
en temps.

F076 Si son clerc vient ceans, fais-
lui go<iter mon vin. /Rac, in
Littr6

F077 Je r^vite partout, partout il
me poursuit. /Rac, in Littr^

F078 Quelque part et quelque loin
que Ton ait p(5n6tr6 depuis la per-
fection de I'art de la navigation,
I'homme a trouv6 partout des
hommes. /Buff., in Littr^ under
'part'

F079 Je vais oit le vent me m^ne.

a movement
(a movement)

a moving

(an extent of space)

a movement

a movement
a moving
a movement

a movement

a movement
movements

(r) what place?
(r) this place

(r) this place

(r) that place

(r) pp (in) that object
(r) a p (on) that object
(r) yonder place

(r) a p in this place

(r) every place

(r) pp in any place
which (has been
reached)

a movement

(r) any place to which
the wind has or
will have taken
me

(12) Combinations of e-p relation with first term of relation and often also with
other concepts (verbs, nouns, adjectives).

F080 O mdnes de mon pere! a movement r a formerly occupied

N'ayant pu vous venger, je vous p near you

i rai rejoindre . /Corn ., in Littr:^

Five: Studies in I 'niversul i'oiieeptual (Inutumir

375

F081 Ces deux rivit'res confluent movements

iiu-dessus de Paris. /Littr6

under 'confluer'
F082 i&n/rons dans le salon. /Lit- a movement

F083 II est temps de rentrer dans a movement
la salle.

F084 II monle a cheval.'

a movement

F085 II ne put jamais parvenir au a movement

haut delamontagne. /Littr6
F086 II y a longtemps qu'il est a movement

venu.
F087 Toutes les marchandises se movings

trouvent enfin transportees.
F088 Les papiers sont emballis a moving

dans une boite.
F0S9 Le rapprochement est com- movements

plet.

F090 C'est un bon atterrissage. a movement

r (I place occupied by
both below Pariu

rap in . . . , in the
salon

rap formerly occu-
pied, in the room

r a p on . . . , on a

horse
r . . . , a p at the top

of the mountain
(r) (this place)

r other places

(r) pp in ... , in a
box

r pp [formerly] occu-
pied by them
in common (fig.)
r a p on the ground

Samples of the Expression of the Ending-Point Relation in

German

S
Sample Sentence

First term of
relation

r
Re-
la-
tion

Second term of
relation

(1) Predicative e-p expression.

GOOl Die Reise kommt in Berlin a movement
zu Ende.

-r Berlin

(2) Usual e-p relaters, occasionally with implication of additional concepts.

G002 Er geht zut Schule. /Grimm a movement r the school

G003 Niihert sich zu ihm. /Loh- a movement r he

enstein, in Gkimm
GOOl Man fahrt Steine zur Stadt. a moving r tlie city

/CURME, p. 372
G005 Ich will Sie zu ihm fiihren. a moving r he

/CURME, p. 372

' Monter d cheval can mean 'to ride horseback' or *to get up on a horse'.

376

General Linguistics I

G006 Er hat die Feder zu den a moving

ubrigen gelegt. /Curme, p. 372
GOO? Wie komme ich nach der a movement

Friedrichstrasse? /Curme, p.

369
G008 Er geht nach der Stadt zu a movement

seinem Bruder. /Curme, p. 372

mit bis Wien. a movement

rap near the others
r Friedrichstrasse

r the city
r Vienna

rap above

r (a p) in
room

rap below

Sara's

(3) Bis.

G009 Ich reise

/Grimm
GOlO Der Herzog und Staff sind a movement

bis herauf gegangen. /Gothe, in

Grimm under 'herauf
GOU Sie wiirde mich bis^ in dem a movement

Zimmer der Sara suchen. /Les-

sing, in Grimm
G012 Er habe sie noch totfcishin- a moving

unter ins Leuker Bad gebracht.

/Gothe, in Grimm under 'bis'

(5) Relational locative expressions which may imply e-p relation.

(5.a-l) Those which with the accusative express both e-p and a locative relation,
and xvith the dative express only a locative relation. (See beginning of page 25.)

G013 Wir Ziehen den Kahn ans a moving

Ufer heran. /Grimm under

'heran'
GOU Er trat an das Bett. a movement

/Curme, p. 372
G015 Er setzte sich an meine a movement

Seite. /Curme, p. 380
G016 Das Wasser reichte bis an an extending

die Knie. /Curme, p. 380
G017 Er klettert auf den Baum. a climbing

/Curme, p. 381
G018 Sie fahren aufs Land, a movement

/Curme, p. 381
G019 Er setzte sich au/ den Stuhl. a movement

/Grimm

r the shore

r the bed

r a p by my side

r the knees

r a p on the tree

r a p in the country

r a sitting P at o p
on the chair

^ Bis functions here as the sole means of indicating the ending- point relation,
for 'in' used with the dative case has the locative function. Contrast the next
sample where 'in' with the accusative repeats the ending-point relation.

Five: Studies in Lhuvcrsul COticcptiuil (inirnnuir

311

G020 Sie wircl am Ende doch er-

raten, dass sie hinter die Sache

gekommen sind. /Gellert, in

Grimm
G021 Sie sollen sich hinter die

Tapeten verstecken. /Schiller,

in Grimm
G022 Die Abendrote, die grade in

sein Gesicht fiel. /J. Paul, in

Grimm
G023 Er ging in den Garten.

/CURME, p. 384
G024 Wasser in den Krug schiit-

ten. /Grimm
G025 Er hat sein Haus neben das

meinige gebaut. /Curme, p. 384
G026 Er setzte sich neben mich.

/Curme, p. 384
G027 Der Adler erhebt sich iiber

die Wolken. /Curme, p. 385
G028 Wir setzten uns untcr den

Baum. /Curme, p. 386
G029 Er spannt die Pferde vor den

Wagen. /Curme, p. 386

G030 Er wirft seine Perlen vor die

Saue. /Curme, p. 386
G031 Ich blickte zum Baume

empor, eine Frucht fiel herab vor

meine Fiisse. /Grimm under

'herab'
G032 Sie setzte sich zmschen mich

und ihren Bruder. /Curme, p.

386

a movement

an action impel-
ling a movement
a moving

a movement

an action compel-
ling movements

a moving impel-
ling a movement
a movement

a movement

(5.a-2) Same forms as 6.a-l, suffixed to 'da-' and 'wo-' .

a moving

G033 Er klebt einen Zettel daran.

/Grimm under 'daran'
G034 Er giesst VVein daran.

/Grimm under 'daran'
G035 Wir legen unsre Klauen

drauf. /Gothe, in Grimm under

'darauf
G036 Es war nicht mehr auf dem

Tische, worau/ichesgelegthatte.

an action impel-
ling a movement
a moving

a moving

rap behind the
affair (fig.)

rap behind the
carpets

r a p on his face

r a pin the garden

r a p in the jug

rap near my house

r a sitting P at a p

near me
rap over the clouds

r a sitting P at a p
under the tree

r side by side PP
at a p before the
wagon

rap before the sows

rap before my feet

rap between me and
her brother

(r) a p on it
(r) a p on it
(r) a p on it

(r) a p on which (the
table)

378

General Linguistics I

(r) a p behind it

G037 In dem Zimmer stand eine a movement

spanische Wand, sobald er jem-

and kommen horte, machte er

sich da.hinter. /Grimm under

'dahinter'
G037a Da.hinter, bitte! (a movement)

G038 Wohinter konnte sich der a movement

Knabe versteckt haben?
G039 Die Sonne steige nie aus a movement

roter Flut Und sinke nie darez'n.

/Kleist, in Grimm under 'darein'
G040 Dies ist das Wasser, worin a movement

der Ring gefallen ist.
G041 Er stellte sich unmittelbar a movement

daneben.
G042 Da zersprang die Maschine, a moving

woneben er die Lampe gestellt

hatte.
G043 Das Dienstmadchen legte ein a moving

weisses Tuch daTiiber.
G044 . . . und haben da,Tunter die a moving

Kranken getragen. /Schuppius,

in Grimm under 'darunter'
G045 'Er ging unter die Briicke/ (a movement)

sagte ich. 'Worunter?' fragte er.
G046 Sie zog einen Vorhang davor. a moving

/Grimm under 'davor'
G047 Bringen Sie mich zu dem a movement

Gasthaus, wovor die Post halt.
G048 Die Dazwischenkxinit des a movement

Kaisers und ein Bannstral aus

Rom bewaffnete gegen ihn seine

Landstande und sein Capitel.

/Schiller, in Grimm
G049 Die Baume, vfozwischen der a movement

Apfel gefallen war, fingen an zu

streiten.

(5.a-3) Those which are construed with the genitive, dative, or accusative.

With the genitive.

GOoO Immer wider bringst du die a moving
kotigen Schuhe innerhalb meiner
Kiiche.

(r) o p behind that

place
(r) a p behind what?

(r) a p in it

(r) a p in which (the

water)
(r) o p near it

(r) a p near which (the
machine)

(r) a p over it

(r) a p under it

(r) a p under what?
(r) a p before it

(r) a p before which

(the inn)
(r) a p between them

(the different

factors of the

situation)

(r) a p between which
(the trees)

(r) a p in my kitchen

I'ivc: Snulic.s in Lmvcrsul i Diucpimil (inunmur

379

G051 Sie kamen diesseits des Ge- a movement

birges und besiedelten ein grosses

Gebiet.
G052 Es ist gefuhllos von Ihnen, an action compel-

dass Sie die arme Katze ausser- ling a movement

halb des Hauses bei dieser Kalte

jagen.

With the dative.

G053 Kehren Sie, bitte, bei mir a movement

ein.
G054 Er nahm ihm grad gegeniiber a movement

Platz. /CuRME, p. 367
G055 Man stelle sie mir gegeniiber. a moving

/Schiller, in Grimm
G056 Er stellte Karl an eine Seite a moving

der Ttir Heinrich direkt gegen-
iiber.
G057 Ndchst ihr stellt' er sich sel- a moving

ber den schongebildeten Sessel.

/Voss, in Grimm

With the accusative.

G058 Er lenkt seine Schritte gegen a directed turning

Westen. /Curme, p. 369
G059 Die Giiste setzten sich utn a movement

den Tisch.
G060 Sie lehnte den Besen gegen a moving

(wider) die Wand.

(r) a p Ihia aide of the
mountain range

(r) a p outside the
house

(r) a p at my place

(r)

a P facing him (at

. . .)
a P facing me (at

. . .)
a P facing Henry

(at . . .)

(r) a vertical P at a p
next to her

(r) pp facing the west

(at . . .)
(j) sitting PP at pp

about the table
(r) a leaning P at a p

against the wall

(5.d) Those which never take an explicit point of reference

G061 indem er ihn beiseite a moving

nahm. /Gothe, in Grimm
G062 Sie nahmen den Vorleser a moving

beiseit und beschworen ihn.

/Klingers, in Grimm
G063 Sie kommen jede VVoche a movement

zusammen.

(7) Locutions ichich include some expression of direction.
G064 //era6 vom Pferde ! /Grimm (a movement)

G065 Als er nun Aeraftgelangt, ihr a movement
unter den hohen Biiumen am
landlichen Tische gegeniiber sass
/Gothe, in Grimm

(r) a p to one side of
(;•) a p to one side of

(r) a p occupied by
them all

a p downward from

. . . the horse
a p downward from

380

General Linguistics I

G066 Sie waren kaum heraufge- a movement

kommen, als der Hauptmann

schon da war.
G067 Kinder die aus der Schule a movement

/lerausbrausen. /J. Paul, in

Grimm
G068 Du kannst die Freude bald a moving

erleben, Das Kesselchen heraus-

zuheben. /Gothe, in Grimm
G069 Heraus mit eurem Fleder- (a moving)

wisch! /Gothe, in Grimm
G070 Die Menge quoll vor das a movement

Tor hinaus. /Grimm

G071 Ich werde mich iiber die a movement

Menschen hinausreissen. /J.

Paul in Grimm
G072 Als der Mond iiber den (a movement)

Apennin heruber war. /J. Paul,

in Grimm
G073 Der Alte brachte aus einem a moving

Kasten allerlei Raritaten hervor.

/Grimm
G074 Da kommst du schon hervor- a movement

gehiipft. /Gothe, in Grimm

rap upward from

rap outward from
. . . the school-
house

rap outward from

rap outward from
. . . , before the
gate

rap outward from

rap hitherward from

and past . . . the

Apennines
rap hitherward from

and before . • .

a box
rap hitherward from

and before . . .

G075 Er kehrte den Stein herum. a turning

/Grimm

G076 Hermann kommt aus der a reverse

Schule zuriick. movement

(r) a reverse P (at

. . .)
(r) the original p for-
merly occupied
by him.

(8) Locutions which express position irrespective of context

G077 Unser Boot legte sich lang- a movement
seit dem englischen Dampfer.
/Gerstacker, in Curme, p. 362

G078 Die Infanterie entwickelte
sich ri tilings der Strasse.
/Moltke, in Curme, p. 363

G079 Setzt sich der Aff rittlings
aufs Holz und spaltets mit der
Axt. /Lehmann, in Grimm

movements

a movement

(r) a P lengthwise of
the English
steamship (at
the side of the
steamship)

(r) PP straddling (fig.)
the street (at
. . . the street)

(r) a P straddling the
wood (at . . .
the wood)

J'ixc: Studies in L Hivcrsal iOnccptital (irnnittuir

381

(9) Indicative pronouns which include the e-p idea: hek and iii.n alone, or plut
-AN, -BEi, -EiN (-in), -vok, -zu, -zwischen, Or plus a stem which expresses a motion.

G080 Komm hcrf

CJOSl Auf der Hinfahrt trafen wir

keine, auf der //crfahrt recht

angenehmeGesellschaft. /Grimm
G082 Dass ich die bohmschen

Spielleut /icrbestellte. /Korner,

in Grimm
G083 Herbci, ihr Manner, gute

Leute, helft! Gewalt! Gewalt!

Sie f uhren ihn gef angen ! /Schil-
ler, in Grimm
G084 Wenn sich Baylens Schat-

tengestalt durch Beschworungen

/icr6eizaubern Hesse. /Abbt, in

Grimm
G085 Die Sonne veranlasst den

Herbeizxig der nordlichen Luft.

/Kant, in Grimm
G086 Herein] herein Gesellen

alle! /Schiller, in Grimm
G087 Als Eduard erwachte, schien

ihm der Tag ahnungsvoU herein-

zublicken.
G088 Samed zittert herzu, und

umarmt ihn. /Klopstock, in

Grimm
G089 Sie tratnaher/ierZM. /Gothe

in Sanders
G090 Wie weit ists kin? /Gothe,

in Grimm
G091 Dem wachst das Herze

gleich, wenn je ein Waldgeschrei

Von Hunden wird erweckt, da

macht er sich hinbei. /Morhof,

in GniMM
G092 DasGrubenlicht immer blei-

cher /iineiTiflimmernd in die ein-

same Nacht. /H. Heine, in

Grimm
G093 Sie begeben sich in das

Haus hinein. /Gothe, in Guimm
G094 Er sah einen Feigenbaum

an dem Wege, und ging hiniu.

/Matth., in Grimm

a movement
a trip

an action compel-
ling movements

(a movement)

this place
this place

this place

an action [an

in-

the {indicated) X (a

cantation] com-

place near them)

pelling a move-

ment

a movement

the (indicated) X

(a movement)

a p in this place

a line of sight

a p in the {indi-
cated) A' (his
room)

a movement

the {indicated)
X (he)

a movement

a p nearer to the
{indicated) A'

an extent

that placet

(a movement)

that place

lines of light

a movement

a p in that place,
in the night

a p in that place,

in the house
it, a fig-tree

382

General Linguistics I

(lO.a) Interrogative, indicative, and relative pronouns which include the e-p idea:
HER and HiN suffixed to or following hier; hin after wo, da, dort, uberall,

IRGENDWO, etc.

(a movement) r what place?

a movement r what place?

G095 iro/immiissenSiesofrlih?
G096 Wo kam der Waidmann hin,

mit dem ich sprach? /Schiller,

in Grimm under 'hin'
G097 Hierher kommen nur Reiche.
G098 Hierhin musst du sehen,

nicht dahin. /Grimm
G099 Die das Abenteuer ihrer Da-

hinknnft zu erfahren hochst be-

gierige Thusnelde. /Lohenst.,

in Grimm
GlOO Ich weiss nicht, wie ich je

dahin kommen werde.
GlOl Ich will esda/izneznstecken.

/Grimm
G102 Man sagt, er sei schon dort-

hin gegangen.
G103 Et woUte uberallhin konnen.

/Woldemar, in Grimm
G104 Konnten Sie mir bitte sagen,

wohin dieser Zug fahrt?
G105 Schicken Sie es wohin Sie

wollen, ich habe nichts dagegen.
G105a Wir gehen wohin er gegangen

ist.
G106 Ich werde irgendwohin

gehen, nur nicht zu ihm.
G107 Ich weiss nirgendswohin.

/Muret-Sanders
G108 Anderswohin abweichend

verkiindigen. /Voss, in Grimm

a movement
a line of sight

a movement

(movements)

a movement

an action compel-
ling a movement
a movement

a movement

(a possible,
movement)
(a movement)

any

(11) Noun which may imply e-p relation: heim

G109 Der Hirt treibt die Herde

heim. /Grimm
GllO Die Fiirsten in Schlesien auf

die //eimfiihrung einzuladen.

/Schweinichen, in Grimm

this place
this place

r that place

r a p in that place

r that place

r every place

r what place?

r any place which
you choose

r the place to which
he has gone

r any place whatso-
ever

r no place^

r another place

an action compel- (r) home

ling a movement
a movement (r) home

2 A possible movement to no place = any movement negated to any place.
Cf. note to E134.

Five: Smdifs in Liuvcrsul C iniccpimil (inininuir

3S3

GUI Es waren ihm aber ein paar

Knotten in die Schuhe gefal-

len, die driickten ihn auf deni
Ifcimweg. /Grimm

(12) Combinations of e-p relation
other concepts (verbs, nouns, adjectiv

G112 Diese verwilderte Ortschaft

liisst sich nicht leicht erreichen.
G113 Konnten wir nur den See

erreichen.
G114 Die Kleinen gcrietcn in einen

Krossen Saal, wo alles von Zuck-

erwerk gemacht war.
Gilo Es ist schon Zeit halt zu

mac hen.
G116 Der .... Alte macht an

jedem Port, auf jeder Insel Ilalte.

/Wieland, in Sanders
G117 Hast du gehort dass der

Onkel gestern gekommen ist.'
G118 Der Berg ist ersteigbar.

G119 Crusoe erstieg den Berg und

sah umher.
G120 Aufsteigenl
G121 Da wurde eine grosse Gans

aiifgelragen.
G122 Er setzte den Hut auf.
G123 Als ich zum erstenmal vor

dem Publicum als Schriftsteller

auftrat. /Gokingk, in Grimm
G124 Ich kann dir keine Leckerei-

en auftischen. /Grimm
G125 Ich hatte des friih Angekom-
inenen schon vergessen.
G126 Ich werde ihm gleich meine

Ankunft melden.
G127 Es ist ihm Gift in Wein

beigebracht worden. /Grimm
G128 Es miissen noch zwei Pferde

beigcspannt werden. /Grimm
G129 Er ist bei der Quelle steficn

geblicbcn.

a road

(r) home

with first term of relation and often also with
es).

a movement
a movement
a movement

a movement

movements

a movement

a possible climbing

a climbing

a climbing
a carrying

a inoving
a movement

a moving

a movement

a moving

an action compel-
ling movements
a movement

r thi.s wild country

r the lake

r a pin 0. large room

r . . . pp at every
\iovi and every
island
(r) [this place)

r (the top of) the

hill
r (the top of) the

hill
(r) pp on , . »
(r) apon . . .

(r) a p on . . .
(r) a p before . . .
the public

(r) a p on the table

(r) a p before . . .

him
(r) (aide by side PP)

at pp besidr . . .
r a standing P (at a

p) before the

384

General Linguistics I

G130 Das Speer blieb in seiner
linken Seite sleeken

G131 Er stelUe sich vor den Grab-
stein seines Vaters und weinte.
G132 Er stelUe sich auf den Kopf .

G133 Setzen Sie sich (nieder) wo es

Ihnen am bequemsten ist.
G134 Der schlafrige Knabe legte

sich miide unter den Baum.
G135 Sie sah mich wunderfreund-

lich an Und bot den Mund mir dar

zum Kuss. /Holty, in Grimm
G136 Und reichte Dem Gon-

ner eine Bittschrift dar. /Gel-

lert, in Grimm
G137 Als hatten mehr denn

hundert Festen ihm Die Schliis-

sel ihrer Tore dargesandt.

/Broxtermann, in Grimm
G138 Das dahingleitende Schiff

verschwindet in der Feme.

(a movement)

a movement
a movement

a movement

a specifically directed
moving

specifically directed
actions compel-
ling movements

a movement

r a protruding P (at
a p) in his left
side

r a vertical P before
the gravestone

r a vertical P with
remaining por-
tion of body
above head (on

r a sitting P (on

r a horizontal P
under the tree
(r)I

r the patron

(r) he

r that place

Five: Studii's in L'nivcrsnl Conceptual Cratnnuir 3K5

PART III: NOTES
Space-Locative Pronouns

Space-locative pronouns (or, in other words, pronouns of space-loca-
tion) play so conspicuous a role in any study of the expressions of the
ending-point relation that it seems not out of order to devote a few pages
to observations on some of their characteristics.

Of space locative pronouns one may say in general that they have the
meaning of nouns with preposition and the syntax of pronouns. Their
logical content includes both substantial and relational elements. For
example, thither ma}"" be rendered ^ which has as ending-point that place' ;
it expresses the ending-point relation plus the substantial 'that place'.

Space-locative pronouns occur in great variety in the languages of
this study. The various forms are based on distinctions in the compo-
nent elements of meaning with regard to mode, method of reference
(whether immediate or mediate), notions of indication, etc.

In the interrogative mode, the different forms are based only on the
relation with which the substantial is combined. This may be illus-
trated from English, in which where?, whitherf and whence? represent
combinations of a substantial term in the interrogative mode with the
locative (or ending-point), the ending-point, and the starting-point
relations respectively. Where? may be rendered whose location is (or
whose ending-point is) what place?; whither?, whose ending-point is
WHAT place?; whence?, whose starting-point is what place? German
has a set of pronouns (worunter?, woriiber?, etc.) which express the in-
terrogative substantial plus an indirect locative relation and which may
be rendered: whose location is (whose ending-point is) a point under
what place?, a point over what place?, etc.

A pronoun of the type of those last named, on the basis of the fact that
its referent is denoted through the medium of its relation to something
else (a point of reference), may be called a pronoun of mediate reference
or a mediate pronoun. On the basis of the fact that its logical content
contains two substantial terms it may be called a two-point pronoun.
Where?, there, whither?, and all other pronouns of immediate reference
logically involving only one substantial term would accordingly be one-
point pronouns. Below, in e.g. 'He's below' (at a point below . . . ), and
other mediate expressions like it would be two-point pronouns.

386 General Linguistics I

In the affirmative mode immediate (one-point) pronouns may be
distinguished as indicative and relative.

'Indicative' pronouns may be so called because their meaning includes
some type of indication^, e.g.,

definite: at, to, from the place there thither thence

indefinite: « << " a " somewhere somewhither

hypothetical: " " " any " anywhere any whither

Demonstrative pronouns, which are special cases of indicative pro-
nouns of definite reference, generally occur in pairs: here, there; id, la.
But German and some dialects of English have sets of three: hier, da,
dort; here, there, yonder. There are languages with still more compli-
cated indicative systems. The different forms in each set are based on
differences in remoteness from the speaker, and languages vary in what
they consider near enough to be called here or remote enough to be called
there or yonder. An adequate treatment of these distinctions is outside
the scope of this paper. We shall recognize three degrees of remoteness,
of which the third does not often occur in the languages in this study:

this place, point
that place, point
yon place, point

Everywhere is a case of the combination of an indicative locative
pronoun with the notion of totahty.

Elsewhere, ailleurs, and anderswo{hin) express the contradictory
indicative notion 'at (to) a place other than . . .'. The contradictory
locative is expressed with nouns by the use of other: at the other, another,
some other, any other place.

A few words may now be said concerning affirmative indicative loca-
tive pronouns in each language of this study, separately.

ENGLISH

EngHsh has three sets of such pronouns.

1, here, yonder, there, somewhere, anywhere, everywhere, elsewhere,
somewhere else, anywhere else, everywhere else.

All the above forms can be used with ending-point meaning (i.e.,
'whose ending-point is the place, some place, any place,' etc.).

^ For the term 'indicative', as applied to locutions whose meaning includes
some type of indication, we are indebted to Prof. W. E. Collinson. In his mono-
graph on 'Indication', now in preparation, he defines indication as the use of a
word or gesture whose function is to direct attention towards or away from an
item or items.

Five: Sii((lit's in Vnivcrsul Com cpiunl (innnituir

387

2. In addition, another set essentially includes the ending-point idea.
Formerly the only ending-point pronouns, they arc now Ix'ginning to be
archaic:

hither, thither, somewhither, anywhither, everywhither, elsewhither.
Someu'hither else, anywhither else have long been obsolete.

3. Starting-point pronouns are hence, thence, but the more usual
expressions are from here, from there.

FRENCH

French has the following affirmative indicative locative and ending-
point pronouns (there is no distinction between simple locative and
ending-point ^locative) :

at, to this place

at, to that place

at, to the (indicated) place

at, to some place, any place, a place

at, to any place

at, to all places, every place

at, to a place other than • . •

tct
la

quelque part

n'importe oil, ou que ce soit

partout

autre part, ailleurs

Ld-bas is often equivalent to English yonder.

GERMAN

Ending-point forms in standard German are definitely distinct from
the merely locative forms, being derived from the latter by the suffixation
of -hin. This ending-point suffix -hin alternates in usage with the
ending-point word hin. The former is more common in formal \sTitten
German, the latter in colloquial. Locative forms (i.e. without [-]hin)
for ending-point are dialectic.

Locative Locative Pronoun Logical Rendering

Pronoun which includes

Ending-Point

Relation

at, to this place

at, to the {mentioned) place

at, to that place

at, to yon place
at, to any place

at, to some place
at, to any places

at, to every place
at, to a place other than . . .
at, to any place other than . . .

hier

hierher

dahin

dort

dorthin

irgendwo

irgendwohin

vberall

uherallhin

anderswo

anderswohin

sonstwo

sonstwohin

388 General Linguistics I

The indicative ending-point form dahin that should correspond with the
middle indicative locative da seems to have lost its force as an inde-
pendent pronoun. Dorthin is therefore the independent ending-point
form of both da and dort, as well as the referential form for dort.

An indicative pronoun can be regarded as independent or referential.
It is 'independent' if it represents something not previously mentioned
or not at the moment clearly in mind but whose recollection is being
sought, e.g., 'Look at that' (with a gesture of pointing). A pronoun is
'referential' if it refers to something previously mentioned, e.g., 'Do you
remember the tree where we built a fort as children? Let us go there
now.'

No formal differentiation between independent and referential functions
of indicatives is made in the languages of this study, and such a word as
there may mean either 'at (to) the place at which I am pointing' or 'at (to)
the place referred to\ Take for instance this situation: A man is sitting
in his study reading a book, his back turned toward the door. A knock
is heard. The man, raising his eyes from his book, but without turning
around, calls, 'Who's there?' Is 'there' a referential or an independent
form? Does it mean 'at the place that has just come into my con-
sciousness' or merely 'at that place' (implying a substitute behavior for
pointing at, a mental pointing at, as it were)? Does 'Who's there?'
correspond to 'Who is it?' or to 'Who's that?' Both explanations are
possible. In oral speech, the stress and intonation would reveal the
type (independent there is stressed, referential there is relatively un-
stressed), but these criteria are absent in written forms.

In contrast to immediate indicative pronouns, immediate relative
pronouns may now be considered. Relative pronouns are basically, of
course, a syntactic specialization, but they have certain characteristics
requiring special treatment. They may be broadly classed as simple
and compound.

Simple relative pronouns are referential. They refer to an explicit
noun and may imply the ending-point relation, e.g., 'Let us Hve in
Paris where art and gayety join hands', 'In Paris where we're going
tomorrow'. Such pronouns have a grammatical relation to only one
verb.

Compound relative pronouns are used without any explicit noun.
They might well be called 'double pronouns', for they are best con-
sidered blends of two pronouns, one independent, the other referential.
For example, in 'They live where art and gayety join hands', the 'double'
locative pronoun where can be analyzed into the independent locution

I-'ivc: Sluilic.s in Liiivcrsal i Omcpimtl (itiininiur >'^*^>

'in a place' (= the pronoun 'somewhere') plus the referentiiil pronoun
'in which'. The included referential idea may be looked upon as merely
referential (as in the analysis just given) or as indicative referential.
In the latter case where may be taken as a blend of 'a place' and 'there'.
For example, 'let's go where it's quiet' may be interpreted as 'Ixit's go
to a place' — 'It is (may be) quiet there.' Of course, roughly the same
idea may be expressed in English by 'Let's go to a quiet place', but the
relative clause cannot be reduced to a simple modifier in cases like 'I^et's
go where he went.' A 'double pronoun' has a grammatical relation to
two verbs. The implied independent idea bears a definite syntactic
relation to one verb, and the referential idea bears the same or another
syntactic relation to the other verb, as may be seen in the following
analysis:

where the fairies dwell

to a place

independent
element re-
lated to 'go'

at which

referential
element re-
lated to
'dweir

This twofold relationship is apt to obscure the nature of the 'double
pronoun' in respect to its being a one-point or a two-point pronoun,
because any linguistic rendering of its logical content cannot escape
including two different symbols of a substantial, e.g., place and which.
This does not mean, however, that two different substantials are sym-
bohzed. A compound relative pronoun is as truly a one-point pronoun
as is a simple relative, i.e., its logical content includes only one sub-
stantial; in other words, behind the linguistic place and which there is
only one objective 'point'. The difTerence between the 'points' under-
lying the simple and the compound relative is a difference not in number
but in kind. The former might be likened to a one-facet 'point', the
latter to a two-facet 'point'. The distinguishing characteristics are
those already mentioned, namely, the linguistic expression of the one-
facet point has a grammatical relation to one verb, that of the two-facet
point to two verbs.

The following one-point relative space pronouns are found in our
three languages:

390

General Linguistics I

English

French

German

Locative

where

wherever

wheresoever^

anywhere (that)

ou que

wo nur, wo auch

Ending-point

where

wohin

plus locative

wherever
wheresoever^
whither^
whithersoever^

anywhere (that)

ou que

wohin nur,
wohin auch

d'ou

woher

Starting-point jrom where
plus locative whence"^

Most of the foregoing forms are used also for expressing interrogation.
English anywhere is used also as an indicative without any relative
function, e.g., 'He can go anywhere'. Contrast 'He can go anywhere
he likes' or 'He can go anywhere that he likes or 'He can go wherever
(archaic: whithersoever) he likes'.

So far, in the affirmative mode, we have been considering immediate
or one-point pronouns. There are also mediate or two-point pronouns,
and these may be classed as indicative, elliptic and relative.

The indicatives comprise two-point forms of definite reference, like
thereunder, therewith, darunter, dariiber, etc.

Then there is a type of locution based on the ellipsis of the point
of reference after certain locative relaters, e.g., below ('He's below',
'He went below'), above, under, etc.; these may be called elliptic locative
pronouns. This type of pronoun abounds both in EngUsh and in
French, with the same form for the simple locative and for the locative
which includes the ending-point idea, e.g., 'He is in now, safe and
sound', 'Show the others in\ In German there are very few instances
of this type of locative pronouns, and they are apt to be hmited to the
locative relation without implying ending-point, e.g., oben, unten.
The corresponding ending-point forms are herauf, hinauf, herunter,
hinunter. As an instance of an elliptic locative pronoun with ending-
point implication may be cited zusammen in such a context as 'Sie
kommen jede Woche fiir ein Kaffeeklatsch zusammen (Every week they
come together for a good gossip)', i.e., 'to a place occupied by them all'.

* Archaic.

l-'ive: Studies in Universal ( onccpituil ( irnniniar 3^1

Two-point or mediate relative pronouns are found in lOnKli-sli and
German. In both languages there are forms which serve l>oth jls a
simple locative and as a locative plus the ending-point idea: wherc-
against, wherein, whereon, ivhcrelhrouyh, whereunder, whereupon; wodurch,
wohinter, woneben, worauf, woriiber, ivorunter, wovor, wozwischen. Tlierc
are also exclusive ending-point forms: whereinio; wohinab, wohinaus,
ivorein. The starting-point forms are: wherefrom, wherehence, where-
out; woher, woraus. In German the given forms are constantly used,
in ]^nglish rarely. English prefers in which, on which, etc. Thus,
although like German it has mediate relative pronouns, its common
usage is like French, which is devoid of such and has no choice but to
to use a collocation of formally separate linguistic symbols, e.g., dans
lequel (laquelle, lesquelles), sur lequel, etc.

In the negative mode, the negative, nowhere, is closely related to
anywhere, but does not constitute a separate type of indication. It
may be looked upon as a substitute device for negating the entire
sentence. This is evident when we remember that 'nowhere' becomes
'anywhere' when the verb is made negative. That is: 'It is done
nowhere' = 'It is not done anywhere.' Nowhere, nirgend{s)wo, nir-
gend{s)wohin, nulle part = not + cit or to any place. The contradictory
notion is expressed by nowhere else, nulle autre part, nirgend{s)wo
anders.

Notes on the English Sample Sentences'

E003: She threw crumbs to the birds. To throw is an interesting word,
parallel to to send, e.g., 'He sent a messenger to the king'. To send
means to order or cause to go, with the emphasis on the causing rather
than on the going. But the e-p used with to send is not the e-p of the
sending, but that of the locomotion which ensues on the action of
sending. To send to so-and-so parallels to order to so-and-so (e.g., 'He
ordered the troops to the capital'), but in the latter case we feel that
there is an ellipsis of to go, whereas we do not in the former.

There is, however, an important difference in e-p connotation between
the to of 'He sent a messenger to the king' and 'She threw crumbs to the
birds.' In the former case, there is no special implied relation Ix^tween
'messenger' and 'king' other than the e-p relation which obtains properly
between 'sent' and 'king' and only mediately between 'messenger'and
'king.' In the latter case, however, there is a very special implied

' The numbers which are at the head of each note are those which the sample
sentences bear in Part II.

392 General Linguistics I

relation between 'crumbs' and 'birds', namely the dative or 'giving'
relation, expressed by to in 'She gave crumbs to the birds' and hy for in
'She threw crumbs for the birds.' In other words, the sample sentence
E003 is really a blend between two logically distinct sentences:
'She threw crumbs to (or at) the birds' (where neither to nor at adequately
renders the simple e-p relation) and 'She threw crumbs for (or to) the
birds' (where for clearly, and to less unambiguously, renders the dative
relation). It is a curious fact that with such a verb as throw one cannot
express the psychologically unmodified e-p relation, but must, wittingly
or unwittingly, include a connotation of 'friendly' or 'hostile' attitude
('with intent to give to x' or 'with intent to hit x'), hence 'throw to^
or 'throw at.' If we insist on the expression of an unmodified e-p
relation in E003, we have to use the clumsy locution 'She threw crumbs
as far as the birds.'

The correctness of the above analysis, which seems needlessly com-
pHcated at first sight, is borne out by two facts: first, that to send itself
is capable of being used in two shghtly different senses — with implication
of e-p relation alone, as in 'He sent a messenger to the king,' and with
implication of dative relation, as in 'He sent tribute to the king' (where
'tribute' is not 'caused to go', but is 'caused to be gone with [by a
messenger] /or [the king]'); second, that in sentences with dative imph-
cation to may be omitted, hence 'He sent the king tribute' and 'She
threw the birds what crumbs she had,' but not 'He sent the king a
messenger,' unless it is intended to imply that the messenger was not
sent to deliver a message to him but was rather sent for him, that the
king might use him as his own messenger (cf. 'He sent the king a cook').

E004: To bed, you rascals. Note ellipsis of the verb of motion. 'To
bed' for 'go to bed' is not common in English, but sometimes occurs.
'Go to bed' is not the same as 'go to the bed' or even 'go to a bed,' for
'go to bed' implies the conventional sequel 'for the purpose of sleeping'.
We might well imagine a child all tucked in to whom its mother says,
'Now close your eyes. Tommy, and go to bed.'

Another matter of interest in 'to bed' is that in this phrase the ending-
point relater (to) impUes a locative relation (in), the reverse of the much
more usual imphcation of the ending-point relation by a locative relater.
Another instance: 'He put the onion to his nose', where to can be an-
alysed 'whose ending-point is a point before'. Cf. Note E155-156.

The logical expansion of the formula 'whose ending-point is' to 'whose
ending-point is a p locatively related to', of which 'to bed' in

I-'ivc: Sliiilii'.s in L'tii\cr\ul C Oihcpnnil (itiininutr 393

E004 ( = 'whose e-p is a p in bkd') and 'to his nose' ( = 'whose e-p is a p
BEFORE HIS nose') are rather clear special cxaniplos, is possible in a
great many other locutions in which a casual examination reveals only a
simple e-p relation. 'On to Rome!' would ordinarily he analyzed as
'(a movement) whose e-p is rome', as when one thinks of Rome as the
goal of attainment, the climax of a series of ventures. But if one plans
to stop at many places in Italy, 'doing' Venice and Milan and Genoa and
Rome in conscientious detail, such a phrase as 'On to Rome!' might
rather have the connotation of '(a move.ment) whose e-p is pp in ro.me'.
In this case it becomes impossible to decide definitely between an analy-
sis of 'to' as 'whose e-p is pp in' or that of 'Rome' as 'pp in home', for
the overtones of contextual implications can never be quite unambigu-
ously allocated to the overt terms of a linguistic expression.

E006: He went to his death is an interesting type of transfer. It is a
psychological He went to . . . blended with a logical He went and died,
the second event being looked upon as the e-p of the first. The vivid
feeling of fatality is due to the reinterprctation of a time sequence as an
e-p relation, which is possible here because the first verb is a verb of
motion with implied literal e-p (say 'the battlefield'). Schematically:

A r B

1. He went to (the battlefield),

and 2. He died at that place,

contract to 3. He went to his death.

1 is the syntactic model for 3, but 3 really = 1 -f 2. Hence, his
death is a condensation of 2, which swamps out the implied e-p in 1.
We can symbolize this type of metaphor as :

Ai ri IBi] H- Ao [t, B2] > Ai ri B3

A schematically similar example is He rose to the toast = He rose in
order to give the toast, where the e-p element to coalesces with tlu> purpo-
sive to by virtue of a pun-like condensation. Apparently every sub-
sequent act can, by metaphor, become the e-p of a verb of motion,
e.g.. He escaped and became a bandit > He escaped into banditry.

E007: He asked him out to a round of golf. The expression 'to ask
out to' is in a class with 'to invite out to', 'to take out to', 'to have out
to' and others as applied to social functions and activities, e.g., a dance,
a bridge-party, supper, the opera, etc.

E008: She has been to confession. Expres.^^ions of this sort (cf. 'I
have been to the opera,' 'I have been to town,' 'They've been to Europe')

394 General Linguistics I

can only be used in the past, more particularly the perfect tenses. 'I
am to the opera' and *I shall be to the opera' are impossible and 'I was
to the opera' is a httle strange unless quahfied by a definite time refer-
ence, e.g., 'I was to the opera yesterday.' The reason seems to be that
the locution 'to he to x' requires the type of sentence that rather
definitely sets off the achieved e-p relation from the logically subsequent
locative relation. In other words, the he to construction may be looked
upon as a condensation of a go to and of a he at construction.

£011: Protestant to the backbone. Note the double transfer or meta-
phor. First, intensity of quality is rendered, implicitly, in terms of
measurable extent. Secondly, this metaphorical extent is conceived
in terms of a movement with ending-point, as if to say 'as far as (or all
the way to) the backbone'. Thus to in the given context may be con-
sidered an e-p relater with emphasis on distance traversed. Cf. Notes
E012, E014, E087.

£012: You should know what it is to wear iron to your hone. 'To wear
iron to your bone' is a very subtle expression. The underlying form is
that of, e.g., 'to wear wool next to the body' but there is much more than
this packed into the 'to' of 'to your bone.' This is due to the fact that, by
a somewhat violent ellipsis of both words and meaning, the type 'to wear
next to X ' is assimilated to the more expressive type of, e.g., 'it went to
his very vitals,' with a frankly e-p term. 'To wear', as such, can have
no e-p relation but the cruel pressure of the thing worn is conceived of as
moving to an e-p, and it is this overtone which swamps out, with e-p
'to', the more static 'next to' of the normal idiom,

£014: My throat is cut unto the hone. This sentence is understandable
in two slightly different senses, hence double rendering in column A.
(1) 'My throat is cut, having been cut unto the bone', rendered 'a
cutting'. (2) 'My throat is with this quahty, that it is with a cut unto
the bone', rendered 'an extent of cutting'.

£015: When God had brought me thereunto. Thereunto illustrates an
old property of English, formerly quite important but which now remains
only in vestiges. There and here once combined freely with most
prepositions (just as German da-, dort-, and hier-). All compounds of
this type that still exist are much restricted in usage or are purely
archaic.

Five: Studies in Universal C Oncepiunl (iiumnuir 395

E019: At dusk he harries the Abazai—at dawti he is into Bonair. This
use of into without specific mention of any movement is of course poeti-
cal. Into carries an impHcation of motion for the reason that it ia
definitely an ending-point relater.

E021: The child walked out into the street. 'Into the street' and the
corresponding 'in the street' are difificult idioms. 'Street' seems to
refer to the space between the two sides rather than to the paved
surface, and we speak of 'in the given area'. But the analogy between
'street' and 'area' is not complete, for we cannot say 'inside the street'
in the same way that we say 'inside the given area'. On the other hand,
we sometimes speak of 'on the street', especially in an indefinite sense,
as in 'That child is on the street again' (also 'Keep children off the
streets'). Again, a building, actually at one side of the street, is said
to be 'on so-and-so street' and one lives 'on such and such a street'.

£023: He dropped the bucket into the well. The context here would
imply the ending-point relation even if the unambiguous relater 'into'
had not been used. That is, one could say without ambiguity, 'He
dropped the bucket in the well'. It may be pointed out that certain
expressions of movement, like 'to go', 'to fall', 'to carrj^', 'to throw',
'to drop', often suggest the ending-point relation, and the locative rela-
ters are readily used with them to imply the ending-point relation toge-
ther with the locative relation. Other ideas, like 'to run,' 'to ride',
'to skip', 'to march', do not suggest an ending-point relation, and one
has to make use of some expression like to a point under, over, behind,
near, etc., in order to express the ending-point relation with a locative
relation, e.g., 'They are marching to a point near Brussels'.

E024: 'The poor comedian runs into a tree.' Into in this usage may be
explained as a means of expressing lack of intention. It is as though one
were trying to run to the inside of something when one is suddenly
stopped short by the outside of that something. The running up
against is therefore unexpected, unintended, sudden. Runs into a tree,
in other words, may be paraphrased as runs as though it had been possible
to run (hterally) into a tree. Cf. Notes E034, E143, E144.

E028: Up and at them.
E029: / looked at him.

The relater at in these and countless other examples ('I threw it at
him' 'I was talking at him, not to him') is peculiarly different from

396 General Linguistics I

the simple locative at of 'I'm staying at the house', 'Meet me at the
fountain'. It is an e-p relater like to and unto but with a distinct impli-
cation of goal or more or less aggressive purpose. It might almost be
considered a blend of the e-p relater to and an underlying purposive
relater which we may phrase as 'with intent to strike x', 'aimed to-
ward x'. It is curious that the two uses of at here defined are so
clearly separated from each other by the simple e-p relater to. The
psychological, as distinct from the Unguistic, distribution of the three
relaters might be thus schematized :

Simple locative relater ati
E-p relater to

Aggressive e-p relater at2

Such a configuration suggests strongly that ati and at2 are two distinct
'words'.

E034: Guess whom I ran against in London the other day. This is
British English. Americans would probably use to run across. Both
usages are figurative and seem to imply unexpectedness.

E036: I was rudely thrown upon my back. 'Upon one's back' is an
expression of Position of a particular type and, like all expressions of
Position, implies Direction. It may be paraphrased as 'with face
directed away from the ground and with back on it'. In the sentence the
impHed e-p of the movement is the ground, but the expressed e-p is that
of position away from the ground. Such expressions as 'on (upon) one's
back, face, feet, hands, side' are an interesting group of positional e-p
locutions in which the position which characterizes the implied e-p is
somewhat metaphorically expressed by considering the individual as
resting on, or weighing down upon, a part of himself, namely, that part of
himself which is actually on the ground or other surface. There is here,
in on, a sort of blend of 'body on the back (face, etc.)' and 'back (face,
etc.) on . . . ' The positional e-p relation of the type illustrated in
E036 may be schematized as:

A r B

I e-P : A POSITION WITH REMAINING PORTION

A movemi;nt whose e-p is ■{ of body above x

[e-p: A POINT ON . . .

where x = 'face, back, side, etc' and . . . = 'surface, ground, bed, etc/
Cf. Notes E109,E111.

Five: Studies in I 'niversul ( Oneeptiml (irnninutr 397

E044: Lay it across the entrance.

E045: / jumped right on to the ice, and how I got across I don't know.

Note that these two sentences illustrate two different meanings of
across. In the second sentence across connotes two points or Hnes or areas
which arc separated by an area. In the first sentence across connotes
two areas which arc separated by a bounding Unc (in this case 'the
entrance' conceived as a Hne, 'the threshold'). Schematically:

'A > B'

E045 .

. across to .

E044 A h^p\ B

The psj'chological intuition which leads to the use of the same word
(across) for two such distinct relational concepts is obviously that of
movement through the abstract geometrical configuration of two like
elements (dis)connected by a third unHke element: e.g., two points
(dis)connected by a line drawn from one to the other; two areas (dis)-
connected by a line held in common; two land-masses (dis)connected
by water; two intelligences (dis)connected by non-inteUigent spaces,
hence 'I got it across to him.'

E047: The naked hulk alongside came. The meaning of alongside is
dependent upon the context. In such instances as the given sample it
expresses 'a position lengthwise of something, in other words 'side by
side with'. But it can mean merely 'near' or 'by the side of, as in
'With his dog alongside of him, he read by the hour.' Here, as likely
as not, the dog may be facing the side of the master. Contrast German
langseit, which is etyniologically the same as alongside, but which ex-
presses, irrespective of context, 'a position lengthwise of something, as in
sample G077, 'Unser Boot legte sich langseit dem englischen Dampfer.'
French characteristically expresses 'to go alongside' by means of a verb,
e.g., 'Notre navire a accoste le vapeur anglais'.

It may be noted that in our three languages the 'alongside' idea seems
to refer both to the side of the object referred to in the movement,
denoting its position when it reaches the ending-point of the movement,
and also to the side of that which is the ending-point and by reference to
which its location is indicated. 'Alongside' (Ger., 'langseit'; Fr., the
'c6t6' idea in 'accoster') is a brief way of saying 'in a side-long (length-
wise) position at points by the side of. 'Alongside' is predicated of one
object in relation to another. If both objects were spoken of, they would
be described as 'side by side'.

398 General Linguistics I

E051: Advertised by two Symerons, whom he sent before. Before may-
be either a temporal or a spatial relater. In this sentence it is probably-
intended to be spatial.

E068: He put it just outside the door. Outside may be used in three
ways:

He put it outside the house.

He put it outside the walls of the house.

He put it outside the door, the window.
That is, outside may be used in connection with a container, its bounding
walls, or any opening in the walls.

E080: They sailed off to the south.

E081 : Turn to the east.

E081a : Then Coyote went off west of the mountains.

Of these samples, EOS la might have been equally well classed under
section 5. a (see Part I, English), for such locutions as west o/and south of
are felt rather as units than strictly as 'of forms, with expressed points
of reference, of west and south. The reason for this will become clear
in a moment.

In E081 starting-point and ending-point are identical as to location
but are oppositely characterized as to position. One turns from . . . ,
characterized positionally as 'facing the west (or other cardinal point but
east) of your location' to the same • • • , characterized as 'facing the
east of your location.' We have here a distinctive type of positional
e-p (or of e-P) relation, which we may term the 'reflexive ending-point
relation'. Other samples are 'He turned around' and G075: 'Er kehrte
den Stein herum.'

Cardinal point relations are very troublesome in English. This is
because the fundamental terms are used in at least four measurably dis-
tinct relational senses: 1, locative, e.g., 'He's working (down) easV, i.e.,
'at a p reached by moving in an eastward direction from . . . '; 2,
ending-point, e.g., 'He went (down) east', i.e., 'he made a movement
whose ending-point is a point east of his starting-point'; 3, direc-
tional, e.g., 'He walked east', i.e., 'eastward, in an easterly direction'; 4,
positional, including reflexive ending-point, e.g., 'He faced east', i.e.,
'He was in the position of having his face to the east', and 'He turned
east', i.e., 'He turned so as to be in the position of having his face to the
east'. From these relations are abstracted, further, a noun east, the
east, which can only be used in the sense of 'that part of a whole which is
characterized by being in an easterly direction from the rest', e.g., 'He

/•ivc: Stiiilics in Universal C oncc/niuil (iiiuninai

;oo

lives in the cast of Chicago', whereas, if cad were a true locative to \)Ofpn
with, we should say 'He lives east in Chicago' (Hke 'He lives there in
Chicago', 'He lives up town in Chicago'); if it were :i true non-partitive
noun of location, hke region (in some uses), place, we should say 'He
lives in the east in Chicago' (like 'He hves in that place in Chicago').
Clearly, the cardinal point terms are only hnguistically primary, not
psychologically and logically so. They symbolize fundamentally direc-
tional concepts, hence such terms as eastward and to the east uf, linguis-
tically secondary though they be, are logically prior to east itself.

When used with points of reference, the simple formula of adding 'of
cannot be safely applied in all cases, hence 'east of gets to be felt as
another kind of unit than 'out of or 'close to' in their relation to 'out'
and 'close'. The following table will help to make this clear:

Without point of reference With point of reference

(to walk) east; eastward; to the east of x; in an

toward the east; in an easterly (eastward) direc-

easterly (eastward) direc- tion from x

tion
(to go) east; to the east; out east of x

(down, back) east
(to face) east; in an easterly

(eastward) direction

Direction

Ending-
Point
Position

east of x; in
(eastward)
from . . .

an easterly
direction

Reflexive
ending-
point

Location

(to turn) east;* to the east

(to stay) east; out (down,
back) east; in the east

east of X

£082: We went out of the room and waited there for poor Tommy.
Out of is generally a starting-point expression, but sometimes, as in this
case, it may be interpreted as an ending-point relater meaning to a
point at no great distance from.

E086: Pull up your chairs.

E087: He walked up to the parson.

The up of these sentences has the function of adding to the ba,sic e-p
notion that of increasing nearness, e.g., 'He walked progressively nearer
to the parson'. This complex notion is not dissimilar to that of e-p

* *To turn east' may also mean 'to change one's direction eastward while
moving'. Cf. terms for Direction.

4(X) General Linguistics I

with reference to extent covered, as in German bis and French jusque
(jusqu'd) but is not identical with it. Periphrases like as far as and
all the way to must be used in that case. The spatial up to is not a true
unit but is to be analyzed as up + to. In other words, 'He walked up to
x' is properly 'He walked progressively nearer to x + (he walked)
to x'. But when up to is applied to time, it becomes the equivalent
of till, until and bis, jusqu'd.

E090: She folded her work and laid it away. Away generally means
'to any place other than the starting-point or the point of reference.'
In speaking of 'putting' things away there is an added connotation of
putting them in the appropriate places for storing or keeping.

E094: Take the gentleman up front to a better seat. Without context
and without speech intonation, this sentence is ambiguous. It might
mean either 'Take the gentleman who is up front to a better seat' or
'Take the gentleman to a place up front, to a better seat.'

E097: I'm going down to Tucson. This use of down illustrates a whole
series of curious adverbial usages in English involving out, up, down,
over, and possibly others, in other than their literal meanings. Without
attempting to explain the details of the usage, we cite a few examples :

over there

I'm going out to San Francisco.

down to London town

out west, up north, down south, down east (note also back east)

down on the farm

out on the open sea

wpstate, downstsite

uptown, downtown

up (down) to the corner

over to the Joneses

pacing up and down

The use of such expressions presents a complicated problem. It
varies markedly in different parts of the English-speaking world, so
that rules of usage would have to be restricted as to locality. The
following remarks on the use of up and down may prove suggestive of
the considerations involved. Elevation, when clearly evident, as when
marked by mountain slopes or the direction of rivers, is certainly a
factor, e.g., down to New Orleans, up to Duluth. The more important
or larger place is likely to be honored with the up, e.g., up to Chicago,
down to Springfield. The older part of town is down, the newer up.

rive: Studies in Lnivrrscil Conn'fUuiil (irunifnur

401

Map directions often seem to be the determininp; fuctor, north Ix'inK up,
south being down, e.g., up to Seattle, down to San Francisco.

Perhaps the most interesting thing about these expressions, to which
right may be added, is that they seem to map out a space-configuration
involving psychological distance. Right is nearest the speaker or point
of interest, over a little farther away, up and down are moderately
distant, and out is farthest away. Note:

1
It's right there
where you are.
(Speaker thinks of
'you' as point of
departure.)

It's over there.
(E.g., at the house,
not distant, you
and I being here in
the field.)

Take this road and
you're right in New
York.

I live in New
Haven, but when
I want to do any
important shop-
ping I just run
over to New York.

It's up there at the
house. (House is
psychologically
more distant than
in 2.) It's down
there by the mill.
(More of a walk
suggested than in
'It's over there by
the mill'.)

Even Philadelphi-
ans frequently run
up to New York.
It isn't so much of
a trip from Albany
down to New York.

It's out there some-
where. (Suggests
considerable labor
to get at.)

It isn't often a
Californian man-
ages to take a trip
out to New York.

Up town and down town are expressions in every-day use by English-
speaking people in many localities. There is no French equivalent.
Except in hill towns, such as Carcassonne and Quebec, where there is a
ville haute and ville basse, the usual expression corresponding to our down-
town is en ville, which refers to the business, amusement, shopping or
other section in contradistinction to chez soi. When a Parisian returns
from en ville he has no such 'blanket' expression as our up toum. He
must express with more precision where he is going: 'Je vais remonter
chez moiVJevais rent rer (I am returning home)', 'Jerentre;\ Vaugirard',
'Je remonte a Monmartre'. He uses remonter for going to the residential
district, even though he does not go up any more than we do when we go
up town. A Frenchman residing in New York adopts the English turn
of thought in saying ' Je vais dans le haul dc la ville, dans Ic has dc la ville'.

The nearest French equivalent to out is found in the en of 'Je m'en

vais '. In 'Je m'en vais '1 Paris' the mental set is that of a person

starting out from his present location on a trip whose ending-point is Pans.

402 General Linguistics I

ElOl: Down with your sails. The English 'Down with your sails!'
would be expressed in French by 'Abaissez les voiles!' the verb being used
to denote a direction. The French expression 'd bas x' has a feeling-
tone of scorn or hatred when addressed to or used for apostrophizing
persons: 'A bas les rois! (Down with kings!)' It is a proper command
for animals, however, and contains only the implication of sternness.
'A bas les pattes! (Get your paws downl)'. The German 'Heruntef
damit!' has the same feeling-tone as the English 'Down with it!' but is
like the French in that the forms herunter and a bas cannot be used
purely locatively.

E109: Turn him over on his face.

£110: She turned about to hide her face.

Over in 'to turn over' and about in 'to turn about' may both be trans-
lated 'to a position facing in the opposite direction from that of the
previous position', but over refers to turning on a horizontal axis,
about to turning on a vertical axis.

In E109 the e-p is characterized by two positional expressions — a
dynamic one, over, which defines the changing, and resultingly changed,
position, and a static one, on his face, which defines, in the manner
explained in note E036, the changed position. Such phrases as 'over on
his face', therefore, are conceptually parallel, for position, to such e-p
and locative terms as 'into', for location.

£111: Sit up and pay attention. When the location (ending-point)
is thought of as united with the particular 'stasis' which results from a
characteristic movement into a characteristic locatively determined
being, we have 'position'. Position is not a type of location but is a
more complex type of entity, involving or implying location, frequently
ending-point of preceding movement. In actual usage position and
ending-point get blended. Sometimes the ending-point impHcation
is clearer, sometimes the position ('directional result') is clearer. In
'He sat down on the chair' (cf. G019), ending-point is clearly expressed;
in 'He sat down', position is clearer and involves ending-point only by
further imphcation within position-impUcation. The result, involving
both position and ending-point, of the movement is 'a sitting position
on . . .'. In 'He stood up', 'up' is position-result, not strictly ending-
point, but further implies ending-point 'on . . . (a horizontal surface)'.
Perhaps the most logical formula for expressing (in column B) both posi-
tion and ending-point would be '. . . (or a point on . . .), at which a

position', but this is a bit clumsy, and so the simpler formula has been

Five: Siiidirs in I nivcrsul Conceptual (ininutuir 403

adopted 'a position at (on) . . .'. In the ca.so of 'He sat down',

the renderings according to the two fonnulae would b)c: '(a point on
. . ., at which) a sitting position', *a sitting position (on . . .)'•

E120: Go whither you urill. The relative pronoun may be considered a
blend of two pronouns, one going with each clause. Thus, 'Go whither
3'ou will' may be taken as a blend of 'You want to go somewhere' and

'Go there'.

E124: Give it here. This colloquial expression seems to blend the
notions of, say, 'Give it to me' and 'Bring it here'.

E134: You will go nowhere, I tell you. The negative pronoun is
really only a substitute device for negating the sentence as a whole.
'You will go nowhere' means 'You will not go anywhere'. The sug-
gested analysis of the sample follows English linguistic usage rather
closely. We can redistribute the negative and possibility terms and
then the analysis becomes :

a possible movement not (r) a possible place
any movement any place

E137: Wherever I go I find the same sadness. Wherever, the relative
pronoun corresponding to anywhere, is used in this sample as a locative
pronoun dependent on 'find' and as an ending-point pronoun dependent
on 'go'. The sample may be reworded: 'I find the same sadness in
any (= all the) places to which I go.' (Cf. Appendix, Table II, sec-
tion G.)

E138: You look like you're going places.

E139: All dressed up and no place to go.

'To go places' is colloquial and restricted to this sort of application,
where it is the opposite of 'to go no place'. Places acts as a plural to
some place. The first sample would be expressed in more refined u^age
by 'You appear to be going somewhere'.

E143: / came upon a flower. The expression to come upon has some
interesting implications. Bundled into this simple quasi-metaphorical
expression are the ideas (1) that the e-p was not purposed, for if that
particular e-p were intended, we should use the verb to go: 'He went to a
flower'; (2) that the total movement did not stop there, for to indicate
that the e-p is final, we should use to: 'He came to a flower'. We should
hardly say 'He came upon the well he had been making for.' On the
basis of distinction as to whether or not the e-p is final ami whether or
not it is purposed, we get the following four types:

404 General Linguistics I

purposive, conclusive : He went to John.

non-purposive, conclusive: {In his journeying) he came to John.

purposive, non-conclusive : He dropped in on John.

non-purposive, non-conclusive: He came upon John.

On the analogy of to come upon, evidently, appear synonymous expres-
sions that do not even have a verb of motion, e.g., to chance upon, to
happen upon.

£144: We came across an old deserted cottage. 'To come across' is the
modern expression corresponding to the more archaic 'to come upon'.
See note E143.

£155: She had put the baby to bed and now sat reading a book.

£156: He put the onion to his nose.

'To put' is one of those verbs which refer to the ending phase of a
moving, and to the movement largely by implication; it might be para-
phrased 'to finish moving something somewhere'. Nevertheless it
often requires an ending-point relater to complete its meaning. Cf.
Note E004.

£158: Stand it in the corner. Part of the meaning of 'to stand'
defines the position of the object after it has reached the ending-point
of its movement. Many other verbs have an implication of position,
e.g., 'to erect', 'to lay', 'to seat'.

I'ivc: Studies in Universal C Dnci'piuiil ( iruninuir

405

APPENDIX

TABLE I

COMPARATIVE TABLE OF CONTENTS OF PARTS I

AND II

[E = English. F = French. G =

German.]

Page No. Sample Number

Pagp No.

in Part I

in Part II

1 (EFG) Predicative expressions 15

EOOl

of the ending-point relation. 22

FOOl

14, 26

GOOl

2 (EFG) Usual e-p relaters (occa-

sionally with implication of

additional concepts).

to, unto. 15

E002-015

d, jusque. 22

F002-010

zu, nach. 26

G002-008

(EG) Ending-point relaters
which always or usually in-
clude emphasis on distance
traversed.

as far as, all the way to.

bis.

4 (E) Ending-point relater in 15

combination with locative
relaters, into, onto.

5 (EFG) Relational locative ex- 15

pressions which may imply 22
ending-point relation. 26

5.a (EF) Those which require 16
an explicit point of reference 22
in the second term. {Prepo-
sitions.)

E016-017

G009-012

E018-O27

E028-O37
FOll-018

38
49

4(Ki

General Linguistics I

TABLE I— continued

\E = English. F = French.

Page No.
in Part I

5.a-l (G) Those which with 26
the accusative express both
the ending-point and a loca-
tive relation, and with the
dative express only a loca-
tive relation. (Preposi-
tions.)

5.a-2 (G) Same forms as 27
5.a-l suffixed to 'da-' and
'wo-'. (The resulting com-
binations constitute inter-
rogative, indicative, or rel-
ative pronouns of space
location. If treated as a
whole, they would be
classed under section 10. b.)

5.a-3 (G) Those which are 27
construed with the geni-
tive, dative, or accusative.
(Prepositions.)

5.b (E) Those which may be 16
used 1, with or 2, without a
point of reference, without
change of form. (1, Preposi-
tions; 2, elliptic locative pro-
nouns.)

5.0 (EF) Those which have 17
different forms when used 1, 22
with or 2, without a point of
reference. (1, Prepositions
and prepositional phrases; 2,
elliptic locative pronouns.)

G = German.]
Sample Number

G013-032

Page No.
in Part II

G033-049

G050-060

E038-O77

E078-093
F019-041

41
49

Five: Studies in l't]i\crsiil Conccptunl (Initnnnir

4()7

TABLE I— continued

\E = English. F = French. G

PaRC No.
in Part I

5.d (EG) Those which do not 17

ordinarily or ever take an 27

explicit point of reference.

(Elliptic locative pronouns.)

6 (F) Prefixes which sometimes 22

express the ending-point re- 27
lation. (The given com-
binations of prefix plus stem
if treated as a whole would
be classed under section 12.)

7 (EG) Locutions which consist

of or include some expression
of direction and which may
imply ending-point relation.

about, down, out, over, up, 18
alone or as first member of
a combination.

her- and hin- plus -ab, -auf, 28
-aus, •4iber, -unter; her-
plus -vor, -um; zuriick.

8 (EFG) Locutions which express 18

position irrespective of con- 23
text and which may imply 28
ending-point relation.

9 (G) Indicative space-locative
pronouns which include the
ending-point idea: composed
of her or hin (as indicative
ending-point locative pro-
noun) alone, or plus -an, -bei,
-ein (-in), -vor, -zu, -zwischen,
or plus a stem which expres-
ses a motion.

— (j'erman.l
Sample Number

i:091 00 G
GUOl 003

F042-066

E097-111

G064-076

Pagf No.
in Part II

E112-115

F067-068

G077-079

G080-094

408

General Linguistics I

TABLE I— concluded

[E = English. F = French. G = German.]

Page No.
in Part I

10 (EFG) Space-locative pro- 19

nouns: interrogative, indica-
tive and relative.

10. a (EG) Those which essen-
tially include the ending-point
idea.

whither, hither, thither. 19

her or hin (as ending-point 28
relater) suffixed to or follow-
ing a locative pronoun:
hier -\- her or hin; wo, da,
dort, uberall, irgendwo, etc.,
-f- hin.

lO.b (EF) Those which may 19
imply ending-point relation. 23
(For G, cf. 5.a-2.)

11 (EG) Stray words which may

imply ending-point relation.

'places, no place, home, 19
abroad.

heim. 29

12 (EFG) Combinations of the 19

ending-point relation with 24
the first term of the relation 29
and often also with other
concepts. (Verbs, nouns, ad-
jectives.) (Cf. 6.)

Sample Number

E116-121
G095-108

E122-137
F069-O79

Page No.
in Part II

44
61

44
53

E138-142

G109-111

E143-195

F080-090

G112-138

[p. 88 blank in the original]

Five Studies iti I'tiivcrsal ( Dnrrptuiil (ininirtuir

4()y

TABLK ir

SUMMARY OF CLASSES OF UKSTRICTIVE LOCITIONS

WHICH DENOTE THE ENDINC-POhXT RELATION,

EXPLICITLY OR IMPLICITLY, IN Till-:

ENGLISH, FRENCH AND GERMAN

SAMPLE SENTENCES

By Alice V. Morris

Index

(A) Expressions of e-p relation un-
accompanied by psycho-
logical emphasis, by other
relations, or by pronominal
concepts.

cl 01 He rode to Chester.
cl 02 On to Rome!

(B) Expressions of e-p relation
with psychological empha-
sis on a particular feature
of the total situation:

(B.l) With emphasis on dis-
tance traversed.

(B.2) With emphasis on aggres-
sive purpose.

(B.3) With emphasis on unex-
pectedness.

cl03

He went as far as the

WALL.

cl04

He threw a stone at the

CALF.

cl05

At it\

cl06

He ran across mr. smith

to-day.

cl 07.a The bucket fell into tlie

WELL.

cl 07. b He went to bed.

cl 07. c He put the kicker in

HAUNKSS.

410

General Linguistics I

cl 08.a

cl 08.b

cl09

cllO

At dusk he harries the
Abazai, at dawn he is

into BONAIR.

To BED, you rascals!
They were shown in.
Out with you !

(D) Expressions or implications cl ll.a He is going owf ^o tucson.

of e-p relation in combina-
tion with a from-ward di-
rective relation.

cl ll.b He is going down town.

cl 12. a Er ging hinunter.

cl 12.b He went down to greet

her.
cl 13. a Herunier damit!
cl 13.b Down with your sails!

(E) Implications of e-p relation
in expressions of position.

cl 14 He put them upside down

on the TABLE,
cl 15 Sit up and pay attention!
cl 16 Our ship came alongside

the English steamer.
cl 17 The naked hulk alongside

came.

(F) Expressions or implications
of e-p relation in interroga-
tive and indicative pro-
nouns of space-location.

cl 18. a Thither he went.

cl 18.b He went there at noon.

cl 19. a Whither away?

cl 19. b There f you fool!

cl 20. a Let them which are in

Judaea flee and let

not them that are in the
countries enter there-
into.

cl 20.b I have hereunto set my
hand and seal.

cl 20.C He strode forward and
sat thereon.

cl 21. a Herein, gefaUigst!

cl 21. b In here?

live: Sliulics in I luvcrsul ( Omrpiiuil (iranmuir

(G) Expressions or implications
of e-p relation in relative
pronouns of space-loca-
tion.

cl 22. a The feast whereto he has-
tened was proviiietl by
the king.

cl 22.b She followed on to the
place where he had g<jne.

cl 23 The tome wiis still on the
table whereon he had
laid it.

cl 24 He wandered he knew
not whither.

cl 25. a He went whither she had
gone.

cl 25. b He went where she had
gone.

cl 26 He went where it is quiet.

cl 27 Wherever he paddled he
found plenty of bass.

(H) Implications of e-p relation
in stray words.

cl 28 He went home.

Explanatory Remarks

Table II is based on a survey of the samples of restrictive^ ending-
point locutions in Part II, but differs in details of classification and pres-
entation. In Part II samples are often grouped primarily according
to linguistic form and there is considerable variety in the method of
grouping in the different languages. Samples are given ahead of
analyses and without any special limitation as to number. Table II
is an attempt to summarize and classify the types of restrictive ending-
point locutions primarily according to the nature and number of logical
concepts which they denote, explicitly or implicitly. For each type a
formula is given first, followed by the minimum amount of illustration
needed to show whether and how each type is expressed in each of the
three languages of the study.

In Table II the ending-point locutions are grouped in eight main
divisions (A-H). Each division is composed of from one to six clas.'^es,
twenty-eight in all. Classes are subdivided when it is necessary to show,
for the same conceptual combinations, types of locutions marked by a

•See Purt I, page 12.

4 1 2 General Linguistics I

difference in the kinds of elements which they express or imply. For
example, in 'Whereto do you hasten?' and 'Are you going somewhereV ,
whereto and somewhere are conceptually of the same class, namely, a
combination of the ending-point relation with a questioned or indicated
place, object or person. But the class is subdivided linguistically, be-
cause in whereto both the relational and substantial ideas are expressed
(cl 18. a), whereas in somewhere an indicated place is alone expressed and
the ending-point relation is only implied (cl 18. b).

Another and less fundamental difference in the linguistic symbohza-
tion of like conceptual combinations is merely a matter of outward
appearance, as in thither, whereto? and jusqu^ou? In all three of these
locutions the concepts of ending-point relation and an indicated or
questioned place are both expressed, but in thither by an inseparable
linguistic blend, in whereto by a collocation of two separable elements
with the ending-point element last, and in jusqu'ou by a like collocation
but with the ending-point element first. This kind of difference is
noted in the annotations.

With combinations which are expressed by blends wherein the differ-
ent concepts are linguistically inseparable, any attempt to show
schematically the correspondence between linguistic form and concep-
tual elements is confronted by a delicate problem, namely: should a
relation which is included in the meaning of the 'blend' be considered
to be expressed or implied and therefore pictured as free or as enclosed
within parentheses? In connection with locative pronouns this ques-
tion frequently occurs concerning ending-point and locative relations.
In answer, the following procedure has been adopted : for words which
always include the ending-point idea and can never be used merely
locatively, the ending-point element is left without parentheses both
in the test formula and in the English rendering, e.g., Eng. whither,
hither, thither, Ger. her, hin, with meaning given as whose e-p is what,
this, that place; and for words which can be used merely locatively, loca-
tive relations are left without parentheses and ending-point elements
are placed within them, e.g., 'Let's go where it's quiet' with meaning
of where given as (whose e-p is) a place in which.

In reading the table it will be a help to remember that every numbered
class differs from every other in the nature of its conceptual content as
represented by itaHcized symbols in columns A, r, B, and that all sub-
divisions (a, b, c,) of the same numbered class are identical in the nature
of their conceptual content and therefore have identical itaUcized sym-
bols, but differ in linguistic expression and therefore differ in the paren-
thesizing of their symbols.

live: SlH(lic\ III I ni\crs(il ( otufftiiail ( iraitunar 413

Nature of Contents of Columns

First column: A, First term of relation hot worn A and H,

Second column: e-p r or loc r, The endin^i-point relation (ci 01-26,
28) or a locative relation (cl 27) hetween A and H.

Third column: B, A generalization of the linguistic expression of
second term of relation between A and B.

In classes other than those of relative pronouns column B is sub-
divided into three flutings:

(1) ending-point, the ending-point of A and often also the logical first
term of a relation to another substantial either expressed or implied
(B POINT OF reference);

(2) r, locative, directive or e-p relation between B ending-point and
B POINT of reference;

(3) POINT OF REFERENCE, any existeut by reference to which the
ending-point is located.

In classes of relative pronouns (division G) column B is not sub-
divided. It is provided with descriptive headings for different groups
of entries.

Fourth column: S, Sample sentence in English (E), French (F),
and German (G). The sentences are selected from such only as depict
a motion. The formulae are equally applicable to locutions found in
sentences which depict a stationary existent or an extent of space or
time conceived of as having an ending-point.

Fifth column : English rendering of instance of A, e-p r, B. Liberty
has been exercised in rendering an implied first term (A) on the basis
of a supposed context. Other contexts would supply as many other
renderings.

Sixth column: References, Annotations. References apply to Parts
I, II and III, abbreviated T, 'IF, and 'III', respectively. The
letters E, F, G stand for English, French and German. Section num-
bers in Parts I and II are enclosed within parentheses; sample sentences
are referred to by their number without parentheses.

Opposite the headings of lettered divisions are given references
which have a bearing on the entire division. The sections thus referred
to for Part II include all or the bulk of examples of any class within the
division.

414

General Linguistics I

Illustrations in Part II, for each class other than in divisions E, F and
G, are first to be sought in the sections referred to opposite the division
heading or class formula. Then may be sought the given isolated
samples. These include all relevant samples not contained in the
referred-to sections.

For classes in divisions E, F and G, references to Part II cover all
relevant samples both within and apart from sections cited.

Occasional annotations call attention to items of interest.

Abbreviations and Conventions

SMALL ]

CAPITAL

LETTERS

e-p

dir

Gnl Rmks

ind'd

italics

loc
M

obj
P

pers

qst'd

are often used to indicate the ending-point or the point of
reference.

'class'

'ending-point'

'directive'

'General Remarks'

'indicated'

are used in the sample sentence to mark the locution being
illustrated, and in the formula (columns A, e-p r, B)
and the English rendering to mark the corresponding
meaning.

'locative'

'a motion'. M represents any kind of motion, including
'movement', 'moving', 'action impelling or inducing
a movement', etc. (See Part II, Explanatory Remarks,
page 32 ff.)

'object'

'point', 'place', 'points', 'places', 'successive points, places'
or any convenient synonym

'position', 'positions', 'successive positions'

'person'

'place'

'questioned'

'relation'

'so and so', 'such and such'. Term to be supplied from
context. Three large dots form part of the interpreta-
tion of the locution which is italicised in the sample.
In Table II they represent an implied point of reference.

Five: Sdidic.s in LUivcrsal Coiucpiiml (iriinmuir 415

'^ represents a subordiiuite verb or proposition. The aym-
bol -^ may be read in various ways. In the analysis of
'Go whither he went ' (*M e-p r place to which ~'), ^
may be read in the specific words of the text as 'he
went'. In the analysis of the subordinate clause 'where-
to we hasten' (''^ e-p r which'), '^ may be read as a
reified verb, either 'a hastening', or, more generalized,
'a movement'.

( ) Parentheses in a formula or an English rendering encloi^e
an implication, such as may be found, for example, in
'To bed, you rascals!' where to masks an implied move-
ment and an implied point in bed, and the implications
are placed w^ithin parentheses, thus: '{M) e-p r (p) (lac r)
POINT OF reference'; '(Go) to (a place) (in) bed!*

[ ] Square brackets which enclose entire sentences in the
columns 'Sample sentence' and 'English rendering' in-
dicate that such bracketed illustrations belong to a class
similar to but not in every way identical with the class
given in columns A, e-p r, B. Square brackets are also
used for minor services in order to avoid confusion which
might result from a more liberal use of curved paren-
theses instead of reserving the latter ahnost exclusively
to indicate conceptual imphcations.

416

General Linguistics I

TABLE II
Note that the table runs across two pages. Each lettered division and each numbered
class must be read across both pages.

First
term
of re-
lation

e-p r

End-
ing-
point
rela-
tion

Second term of relation

S
Sample sentence

ENDING-
POINT

POINT OP
REFERENCE

(A) Expressions of e-p relation unaccompanied by psychological emphasis, by other
relations, or by pronominal concepts.

cl 01 M e-p r ending-

point

El He rode to Chester.

E2 So Joshua sent mes-
sengers, and they
ran unto the tent.

Fl J'ai voyag^ a rome.

F2 Apportez-les jusqu'-
ici, s'il vous plait.

Gl Er geht zut schule.

G2 Wo gehen Sie kin?

03 Ich gehe nach der

STADT.

cl 02 (M) e-p r ending-

point

E On to ROME !

F Ala. gare!

Gl Zut stadt!

G2 Nach Berlin!

Five: Sfmlirs in I'nivcrsal C'onccpluiil (inininuir

All

TABLK II

English rendering of instance
of A, c-p r, B

References
Annotations

(A)

cl 01 El A ridinp; whose e-p is Chester.
E2 A running whose e-p is the

TENT.

I & II, Gnl Rmks; EFG (2), F
(6), G (10. a).

Ill, E003, OOG, 007, 008.

Fl A trip whose e-p is rome.

F2 A commanded bringing whose

e-p is THIS PLACE.

Gl A going whose e-p is school.
G2 A going ivhose e-p is what

PLACE?

G3 A going whose e-p is the city.

cl 02 E (.4 commanded movement) whose

e-p is ROME.

F (A commanded movement) whose
e-p is the station.

Gl (.4 commanded movement) whose
e-p is the city.

G2 (An announced movement) whose
e-p is BERLIN, [= (All
aboard. The train is about
to move on) to Berlin.]

II, F003; III, E004.

418

General Linguistics I

TABLE II— Continued

e-p r

First

End-

term

ing-

of re-

point

lation

rela-

tion

Second term of relation

EN'DING-
POINT

r POINT OF

REFERENCE

Sample sentence

(B) Expressions of e-p relation with psychological emphasis on a particular feature
of the total situation:

(B. 1) With emphasis on distance traversed.

cl 03 M e-p r ending-

With POINT

empha-
sis on
distance
traversed

El As far as [up to] here
the tourists may
come, but no fur-
ther.

F2
Gl

I went all the way to
the WALL.

Jusqu'ici on est per-
mis de venir, mais
pas plus loin.

Je suis alle jusqu^au

MUR.

Bis HiER/ier darf man
kommen, aber nicht
welter.

Ich ging his an die

MAUER.

Five: Sfndics in I'nivcrsul C Onci'pimil Gninmuir 4U)

TABLE II— CorUinued

English rendering of instance RcforenceB

of A, e-p r, B Annutatiuntf

(B)

(B. 1) I, E(3), F Gnl Rmks, G(3); II,

EG(3).

cl03 El A movement whose e-p, with II, F007-10; III, KOll-12, 015,
emphasis on distance tra- 086-7.
versed, is here.

As far as denotes, irrespective of
context, the relation of e-p with
emphasis on distance traversed.
The meaning of ^^p to depends on
context. Cf. E087.
E2 A movement whose e-p, with
emphasis on distance tra-
versed, is the WALL.
Fl Same as El.

F2 Same as E2.

01 Same as El.

02 Same as E2.

420

General Linguistics I

TABLE II— Continued

e-p r

First

End-

term

mg-

of re-

point

lation

rela-

tion

Second term of relation

S
Sample sentence

ENDING-
POINT

POINT OF
REFERENCE

(B. 2) With emphasis on aggressive purpose.

cl04

cl05

(M)

e-p-r ENDING-

vrith POINT

empha-
sis on
aggres-
sive pur-
pose

e-p r ENDING-
toith POINT

empha-
sis on
aggres-
sive pur-
pose

(B. 3) With emphasis on unexpectedness.

cl 06 M e-p r ending-

mith point

empha-
sis on
unexpect-
edness

E He picked up a stone
and threw it at the

CALF.

Fl II a jet^ une pierre au

VEAU.

F2 On a tir^ sur lui.

G Er warf den Stein
nach dem kalb.

E AfiT!

Fl Au voleur!

F2 [Allez-y!]

G Draw/los!

E Guess WHOM I came
across, upon, [to
run across, against,
into; to turn up.]

Fl Donner dans le pan-
neau.

F2 Je tombai sur d'AUTRES
mati^res.

G Auf eine sache stos-
sen.

100

Five: Studies in [ nivcrsiil C Omcpiiml (Inirnnmr 42

TABLE 11— Continued

English rendcrinR of instance RcfcrencTB

of A, e-p r, B Annotaliona

(B. 2) III, E028-9.

cl 04 E A throwing whose aggressively II, E029, 035.
intended e-p is the calf.

Fl Ditto.

F2 A shooting ichose aggressively

intended e-p is he.
G Same as E.

cl 05 E (.1 movement) lohose aggressively II, E028, F002.

intended e-p is it.
Fl (A movement) whose aggressively

intended e-p is the robber.
F2 [A movement (whose aggressively

intended e-p is) it.]
G Same as E.

(B.3)

cl 06 E A movement whose unexpected II, E024, 034, 143-4, ISO; III,
e-p is WHO. E024, 034, 143-4.

Fl A movement ivhose unexpected

e-p is a SNARE.
F2 A movement whose unexpected

e-p is OTHER THINGS, [fig,]

G A movement whose unexpected
e-p is SOMETHING [Hteral or
fig.].

101

i'M

General linguistics I

TABLE 11— Continued

A e-p r

First End-
term ing-
of re- point
lation rela-
tion

Second term of relation

ENDING- r POINT OP

POINT REFERENCE

Sample sentence

cl 07. a M e-p r (p) loc r point E He dropped the buck-

OF REF- et into the well.

ERENCE

cl 07. b

e-p r (p)

(loc r) POINT

OF REF-
ERENCE

F II est monte jusgue
sur les ToiTS.

Gl Er klettert auf den
BAUM (ace).

G2 Er suchte den Dieb
bis in dem gewolbe
des Schlosses.

E He went to bed.

F Se mettre au lit.
G Er ging zu bett!

cl 07. M (e-p r) (p) loc r point E The most judicious

OF REF- mode of putting a

erence kicker in harness.

La mouette descend

sur Z'eau.
Er nahm ihm gege-

nuber Platz.

102

I'ivc: Sfudics in I nivcrsul ( Omcpiunl ( ininmuir 423

TABLE U— Continued

English rondoririK of instance References

of A, e-p r, B Aonutatiotia

EF(5).

cl07. a E An action impelling a move- I & II,E(4), G(5.a-1); II, F007-
ment whose e-p is (a point) in 8, G114, 131, 134; III, K021, 023.
the WELL.

In English, combinations of e-p
and locative relators vary in form
and meaning. Into is used in cer-
tain idiomatic expressions to point
to an unexpected e-p (II & III,
E024; Table II (B. 3)). The
combination of to and on is found
as onto (II, E026-7) and on to
(II, E045).
F An ascent whose e-p is (points)

on the ROOFS.
Gl A climb whose e-p is (a point)

on the TREE.
G2 A searching whose e-p is (a
point) in the vaults of the
castle.

cl 07. b E A movement whose e-p is (a II, E155-6, G006; III, E155-6.
point) (in) bed.
F Ditto.
G Ditto.

cl 07. c E An action compelling a move- 1, EF (5); I & II, E (5. a, b, c),
ment (whose e-p is) (a point) F (5. a, c), G(5.a-3); II, E145, 158;
wnthin the harness. G124; III, E03G, 044, 068, 080-la,

082, lo8.
F A descent (whose e-p is) (a

point) on the water.
G A movement (whose e-p is) (a
point) opposite him.

103

424

General Linguistics I

A e-p r

First End-
term ing-
of re- point
lation rela-
tion

TABLE II— Continued

Second term of relation

Sample sentence

ENDING-
POINT

cl 08. a (M) e-p r (p)

POINT OF
REFERENCE

lOC r POINT

OF REF-
ERENCE

E At dusk he harries the
Abazai — at dawn he

is into BONAIR.

F [Au crepuscule il dd-
pouille les Abazai —
a I'aube il est dans

BONAIR.]

G [Morgen friih sind wir

in BONAIR.]

cl08. b {M) e-pr (p)

{loc r) POINT E To BED, you rascals !

OF REF-
ERENCE F Au lit!
G Zu bett!

cl 09 M (e-p r) (p)

loc r (...)

E They were shown in.

F II nous a pris d part.

cUO

(M) (e-pr) (p)

loc r (...)

G Er nahm ihn beiseite.
E Out! Out with you!

F Dehors! Dedans! En
haut! En bas!

G [Of. cl 13.a, G.]

104

Ihc: Siudtfs in ['nivrrsul i Omcptitiil (Inimnuir 425

TABLE U— Continued

English rendering; of instance Itprcrenres

of A, e-j) r, B Aiinutattuoa

I cl 08. a E (A movement) whose e-p is (a II, E019; III, K019.

point) in bonair.

F [ (A movement) (whose e-p is)
(a point) in bonair.]

G [Ditto.]

cl OS. b E (A commanded movement) whose II, E004; III, E004.
e-p is (a point) (in) bed,
F Ditto.
G Ditto.

cl 09 E A guiding (ivhose e-p is) (a I & II, E(5. b, c, d), F(5. c),
point) within (the place occu- G(5.d); II, G120-3, 12.5-8; III,
pied by the others). Elliptic locative pronouns, page G9,

F A moving (whose e-p is) (a E045, 047, 051, 080-la, 090, 094.
point) at one side of (the pres-
ent location).

G Ditto.

cl 10 E (A commanded movement) (whose II, E082a, F025a.

e-p is) (a point) outside (this

place). Cf . cl 13. a-b in which the direc-

F (A commanded movement) (whose tion of a movement is stressed. In

e-p is) (a point) outside, in, cl 10 the e-p is stressed.

above, below (this, that place).
G —

105

426

General Linguistics I

TABLE 11— Continued

e-p r

First

End-

term

ing-

of re-

point

lation

rela-

tion

B
Second term of relation

ENDING-
POINT

POINT OF
REFERENCE

Sample sentence

(D) Expressions or implications of e-p relation in combination with a from-ward
directive relation.

cl 11. a M e-p r ending-
point

from- (...) E I am going out to Tuc-

ward SON.
dir r

F [Je m'ew vais d paris.]

G [Ich gehe zur stadt
hinaus.]

cl 11. b M (e-pr) ending- from- (...) E I am going down

POINT ward town, up town.

dir r

Fl [Je vais en ville.]

F2 [Nous allons dans le
has, dans le haul
de la ville.]

F3 [Nous allons d la ville
basse, haute.\

F4 [Nous allons remonter.

G Der Zug nahert sich
NiederB AYERTSi.

106

Five: Stuilii'.s in L'nivi-rsdl iOnccpimil (irtittinuir

All

TABLE n— Continued

English renderinR of instance
of A, e-p r, B

Hcfereneei!

Annutiitions

(D)

cl 11. a E

cl 11. b E

A going whose e-p is tucson
outward from {my present
location).

[A movement outward from this
place whose e-p is paris.]

[A going, whose e-p is the city,
whose e-p is (a point) outward
from {my present location).]

A going {whose e-p is) (the
PART of) town downward
from {the residential district),
upward from {the business
district).

[A going {whose e-p is) {a point)
in (the business part of)

TOWN.]

[A going {whose e-p is) {a point)
in (the part of) town down-
ward from {the residential dis-
trict), upward from {the 6im-
ness district). ]

[A going {whose e-p is) (the
part of) town downward
from {the upper part) , upward
from {the lower part). ]

[A going {whose e-p is) (the
place) h.abitually (occu-
pied) upward from {the busi-
ness section of toion).]

A coming nearer {whose e-p is)
(the part of) BAVARIA down-
ward from {the upper part).

I, E(7), F Gnl Rmks, G Gnl
Rmks, (7); II, KG(7); III,
E080-la, 097.

III. E097.

Note repetition of e-p idea by
zu and hin.

Ill, E082, 097.

107

428

General Linguistics I

TABLE 11— Continued

First
term
of re-
lation

e-p r

End-
ing-
point
rela-
tion

Second term of relation

S
Sample sentence

ENDINQ-
rOINT

POINT OF
REFERENCE

cl 12. a

e-p r

(P)

from-
ward
dir r

(...)

E
F
G

[Cf. cl 12.b, E.]
[Cf. cl 12.b, F.]
Sie ging hinunter und
griisste ihn.

cl 12. b

(e-p r)

ip)

from-
ward
dir r

(...)

El
E2

He waited for her to
come down.

To go east.

cl 13. a (M) e-p r (p)

cl 13. b (M) (e-p r) (p)

from- (, . .)

ward

dir r

from- (, . .)

ward

dir r

F [II a attendu qu'elle
descende.]

G [Cf. cl 12.a, G.]

E [Cf. cl 13.b, E.]
F A bas les pattes!

G Herunter damit !

E Down with your sails!

F [A 6ame2 les voiles!]

G [Cf. cl IS.a, G.]

108

Five: Smdii's in Ihiivcrsal ( Omcpiuiil (irnmnutr

424

TABLE U— Continued

English rendering of instance
of A, c-p r, H

RoforcnoM
Atuiututiuntt

cl 12. a E
F
G

A going whose e-p is (a point)
dommoard from {her present
location).

cl 12. b El A coming (ivhose e-p is) (a
point) downward from (her
present location).

E2 A going (ivhose e-p is) (a point)
eastward from (the starting-
point).

F [A movemeni (whose e-p is) (a
point) downward from (her
present location).]

G —

II, E157.

cl 13. a E
F

(A commanded movement) ivhose
e-p is (a point) downward from
(the present location).

Ditto.

Ill, ElOl.

cl 13. b E (A commanded movement) (whose
e-p is) (a point) downward
from (the present location).

F [A commanded movement whose
e-p is (a point) downward from
(the present location).]

G —

II, E190-4; III, ElOl.

109

430

General Linguistics I

TABLE 11— Continued

First
term
of re-
lation

e-p r

End-
inp-
point
rela-
tion

Second term of relation

ENDINQ-
POINT

POINT OF
REFERENCE

Sample sentence

(E) Implications of e-p relation in expressions of position.

cll4

M (e-p r) P (at p) too r

POINT
OF REF-
ERENCE

E He put them upside
down on the table.

F II les a mis sur la

TABLE sens dessus
dessous.
G Setzt sich der Aff ritt-
lings aufs holz.

110

I'ivc: Stticlic.s in Lnncrsal C(>mcf)[mil (irmtinmr

431

TABLE \l— Continued

English rendering of instance
of A, e-p r, B

Rcforrncps

Aiiiiiitntiunji

(E)

I ct II, EFG(8); III, E080-

la.

In the samples in Part II and Table II are found three chief ways
of denoting position: (i) by a restrictive expression, e.g., upright,
alongside; (ii) by a predicative and a restrictive expression, e.g.,
sit up; (iii) by a predicative expression, e.g., stand, sit, lie.

References given below for Part II are grouped in accordance
with the aforesaid ways of denoting position.

cl 14 E A moving (whose e-p is) a verti-
cally reversed position (at
points) on the table.
F Ditto.

G A movement (whose e-p is) a
straddling position (at a point)
on the WOOD.

II, (i) E114-5; (ii) GOoG, 079;
(iii) E158, G019, 026, 028-9, 057,
059-60, 129-31, 134.

Class 14 is one in which position
is denoted irrespective of its loca-
tion, with its location expressed.
Expression of the location of an e-p
position may be (a) by means of a
locative relater plus point of refer-
ence, or (b) by means of some
other kind of locative phrase, (a)
is illustrated in the samples here
given and referred to. An in-
stance of (b) is 'He threw himself
down on his back where the pine-
needles were thickest' which ex-
presses 'a movement (whose e-p is)
a horizontal P with remaining por-
tion of body over back at a place on
which the pine-needles were thick-
est'. (Of. G133.)

In sentences of the tyjx^s of
classes 14 and 15 position is de-
noted in any of the three ways
noted at the head of this diviiiion.

Ill

432

General Linguistics I

TABLE U— Continued

e-p r

First

End-

term

ing-

of re-

point

lation

rela-

tion

Second term of relation

ENDING-
POINT

POINT OF
REFERENCE

Sample sentence

cl 15 M (e-p r) P (at p) (loc r) (...) E Sit up and pay atten-

tion.

F Elle a mis son bonnet
sens devant derriere.

G Er stellte das Kind
auf die Fiisse.

cll6

(e-p r) P (at p) loc r

POINT E Our ship came along-
OF REF- side the English

ERENCE STEAMER.

F [Notre navire a accoste
le VAPEUR anglais.]

G Unser Boot legte sich
langseit dem Eng-
lischen dampfer.

112

Five: Studies in Universal Concepiiml (inininuir

433

TABLE n—Coidinued

English rendering of instance
of A, e-p r, B

Ilcfercncea
Annotntiomi

cl 15 E A commanded movement (whose
e-p is) an upright position
(at a point) (un) (your
chair).

F A moving (whose e-p is) a hind-
side-foremost position (at
points) (on) (the head).

G A moving (whose e-p is) a verti-
cal position iHth remaining
portion of body above feet (at a
point) (on) (the ground).

cl 16 E A coming (whose e-p is) a length-
wise position (at points) by the
side of the steamer.

F [.1 movement whose e-p is a length-
wise position (at points) by the
side of the steamer.]

G Same as E.

II, (i) E0:J6, 081, 109 10, 182,
F0G7-8, G07o; (ii) El 11-3, G132;
III, K0:JG, 109 10, 111.

Class 15 is one in which positioo
is denoted irrespective of its loca-
tion, with its location unexprcsaed.

II, (i) E032-3, 048, G054, 078;
(ii) E044, G055-6, 077.

Class 16 is one in which position
is denoted by reference to its loca-
tion, with its location expressed by
means of the same locution which
portrays said position, e.g., 'My
mother put her arms around my
neck.' Here around denotes the
position of the encircling arms by
reference to their location 'around
MY neck'. The sentence expresses
'a motion (whose e-p is) a position
around neck (at points) around
neck'.

In sentences of the typos of
classes 16 and 17 position is de-
noted by a restrictive expression
or by both a restrictive and a pre-
dicative expression, but not by
merely a predicative expression.

113

434

General Linguistics I

TABLE II— Continued

First
term
of re-
lation

e-p r

End-
ing-
point
rela-
tion

Second term of relation

ENDING-
POINT

r POINT OF

REFERENCE

Sample sentence

cll7

(e-p r) P (at p) loc r (...) E

The naked hulk along-
side came.

F [Cf. cl 16, F.]
G Ein Boot kam

setts.

Idng-

(F) Expressions or implications of e-p relation in interrogative and indicative pro-
nouns of space-location.

cl 18. a M

e-p r qst'd 1 , .

ind'd ]'*^

^ [pers

El Thither he went.

E2 Whereto do you has-
ten?
F Jusgu'oii allez-vous?

Gl Sie kam hierher, hier-
hin, her, herzu; da-
hin, dorthin, hin,
hinzu.

G2 Auf der Hinfahrt tra-
fen wir keine, auf
der Herfahrt recht
angenahme Gesell-
schaft.

114

/•Ivc: Sdidii'.s in I'niM-rsdl (Omcptual (inininntr

435

TABLE n— Continued

English rendering of instance
of A, c-p r, B

Rcfcroncea
Annotations

cl 17 E

F
G

A movement (whose e-p
position lengthwise of(,,
points) hj the side of (. ,

Same as E.

is) a 11, (i) i:u47; III, i:047.

.) (at Class 17 is one in which position
.)• is donotctl by reference to its im-
plied location, with its location
implied by means of the same locu-
tion which portrays said position.
'The naked hulk alongside came'
expresses 'a movement {whose e-p
is) a position lengthwise of (implied
location) (at points) alongside (im-
plied location)'.

(F) I, E(10), G, Gnl. Kmks.; I^fcll,

E(10. a, b), F(lO.b), G(5.a-2, 9,
10. a); III, Interrogative pronouns,
page 64, Indicative pronouns, pages
65ff, 69.

cl 18. a El A going whose e-p is that place, II, E015, 117-9, FOlO, 0080-2,
object, person. 084-5, 088-91, 094, 096-100, 102-

3, 138; III, E015.
E2 A questioned going whose e-p is

what place.
F A questioned going ivhose e-p is jiisqu'ouf; jusqii'ici; jusque Id =
what place. 'to' + 'what[this,that]-pIace(obj,

Gl A coming ivhose e-p is this, that pers]'

place, object, person. whereto? wohinf; hereto, hicrher,

hierhin; thereto, dahin, dorthin =
G2 A journey whose e-p is that, this 'what[this, that]-place [obj, pers]-
place. + '-to'

ichithcrf; hither, her; thither, hin =
'to-what[this, that]-place[obj,
pers]'

herzu; hinzu = 'to-this[tha(l-placc
[obj, pers]-' + '-to'

115

436

General Linguistics I

First
term
of re-
lation

TABLE 11— Continued

e-p r

End-
ing-
point
rela-
tion

Second term of relation

ENDING-
POINT

POINT OF
REFERENCE

Sample sentence

cl 18. b M (e-p r) gst'd ) (pi

ind'd) \ohj

E Are you going some-
where'^
Fl Partout il me poursuit.

F2 Rendez-vous-1/.
G Hier stieg er aus.

cl 19. a {M) e-p r

qst'd
ind'd

pers

cl 19. b (M) (e-p r) qsVd \ [ pl

ind'd J \ ohj

E Whither away?

F JusquHd et pas plus

loin,
G Er will hierher, hier-

hin, her, herzu; da-

hin, dorthin, hin,

hinzu.

E T/iere, you fool!
F Valere, ici\
G Hier? frage ich.

cl 20. a M e-pr (p) locr qst'd) (pi

ind'd) \obj

E The black blood flowed

thereinto.
F [Cf. cl 20. c, F.]
G Stelle es dorthinein,

hinein.

116

/■'ivc: Snulics in Universal ('otucpiiKil (iriunniar 437

TABLl'J ll—CoiUinucd

English rendering of instance References

of A, e-p r, B Annotationa

cl 18. b E A questioned movement {whose II, E122-8, WW 0, F009, 071-4,
e-p is) some place. 075-7; III, E124, 134,

Fl Movements {ivhose c-p is) all
places.

F2 A going (whose e-p is) that place.

G A getting off (whose e-p is) this
place.

cl 19. a E (A going) whose e-p is what place, II, F009.

object, person?
F (A permitted going) whose e-p is

this place.
G An intended (going) whose e-p is

this, that place, object, person.

el 19. b E (A commanded movement) (whose II, F070.

e-p is) that place.
F (.4 commanded movement) (whose

e-p is) this place.
G (A questioned moving) (whose

e-p is) this place.

cl 20. a E A flowing whose e-p is (points) II, E025, G087, 101.
in that object.
F — thereinto = 'that place-' +

G A commanded movement whose '-in-' + '-to'
e-p is (a point) in that place.

dorthinein = 'that place-' 4-
'-to-' 4- '-in'

hinein = 'to-that-place-' 4- '-in'
(See I, G(lO.a), last paragraph.)

117

438

General Linguistics I

TABLE Il—Contimied

A e-p r

First End-
term ing-
of re- point
latioQ rela-
tion

Second term of relation

ENDING-
POINT

POINT OF
REFERENCE

S
Sample sentence

cl 20. b M e-p r (p)

cl 20. c

(e-pr) (p)

Qoc r) qst'd \ (pi E I have hereunto set my
ind'dj \obj hand and seal.

F [Cf. cl 20. c, F.]

G [Cf. cl 20. a, G.]

he r qst'd \ fpl E Gazing in exaltation
ind'dj \obj at the throne, he

strode forward and
sat thereon.
Fl Mettez le Id-dessus.

F2 L'eau limpide s'y di-
verse.

F3 J'y ai appos^ ma sig-
nature.

G Er giesst Wein daran.

cl21.a (M) e-pr (p) locr qst'd\ (pi E [Cf. cl 21. b, E.]

ind'dj \obj

F [Cf. cl 21. b, F.]

G Herein!

cl21.b (M) (e-pr) (p)

loc r qsVd 1 (pi E In here?
ind'dj [obj

F Ld-dedans, s'il vous

plait.
G Dahinter, bitte !

118

Five: Smdii's in Universal Com cpuml ( irainnuir -^y)

TABLE ll—Coniinued

English rendering of instance References

of A, e-p r, B Annotations

cl 20. b E A moving whose e-p is (a point) II, EI19a.
(on) this object.
F —
G —

cl20. c E A movement (whose e-p is) a II, E072, F02G, 02S, 031, f;033-
sitting position (at) (a point) 5, 037,038-9, 041, 043-4, 04G, 048.
071 it.

Fl A comnianded movement (whose

e-p is) (a point) on it.
F2 A movement (ivhose e-p is) II, F074a, b.

(points) in that object, it.
F3 A moving (whose e-p is) (a

point) on it.
G A moving (ivhose e-p is) (a

point) on it.

cl 21. a E — II, G086.

F —

G (A requested movement) whose
e-p is (a point) in this place.

cl 21. b E (.4 questioned movement) (whose II, F031a, G037a, 045.
e-p is) (a point) in this place.
F (.4 commanded movement) (whose
e-p is) (a point) in that object.
G (A commanded movement) (whose
e-p is) (a point) behind that
object.

119

440

General Linguistics I

First
term
of re-
latioD

e-p r

End-
ing-
point
rela-
tion

TABLE U— Continued

ENDING-POINT OF MOTION

EXPRESSED BY SUBORDINATE

VERB'^

Sample sentence

(G) Expressions or implications of e-p relaiion in relative pronouns of space-location.

cl22. a

e-p r

which

E The feast where[un]to
we hasten.

F Le banquet auquel
nous nous rendons
en toute hate.

G Das Fest zu dem, zu
welchem, nach dem,
nach welchem, wir
eilen.

cl 22. b

(e-p r)

which

cl23

(e-p r)

(p) on which
(p) in which
(p) near which
etc.

E She followed on to the

place where he had

gone.
F Elle le suivit jusqu'£t

I'endroit ou il 6tait

all6.
G [Cf. cl. 25.a, Gl, G2.]

E The tome is still on the
table where[up]on I
had laid it.

F [Le hvre est encore
sur la table, sur
laquelle je I'avais
pos^.]

G Das Buch ist noch
auf dem Tische,
worauf ich es gelegt
hatte.

120

Five: Studies in Universal Conccptmil (irmnmur
TABLE ll-Conlinucd

English rendering of instance
of A, e-p r, B

Rcferenceji
Annotationa

(G)

Kt II, i:(10.H, b), F(lO.b), G(5.
a-2, 10. a); III, Relative pronouns,
page 67.

cl 22. a E A hastening whose e-p is the II, E119b.

referred-to occasion. ichere[un]to = 'which-' + '-to'

F Ditto. auquel = 'to-' + '-which'

zu dem, etc. = 'to' + 'which'

G Ditto.

cl 22. b E A going (whose e-p is) the re-
ferred-to place.

F Ditto.

II, E129

G —

cl 23 E A laying (whose e-p is) (a point)
on the referred-to object.

F [Ditto.]

G Ditto.

II, G036, 040, 042, 047, 049.

where[up]on = 'which-' + '-(to
a point) on'

sur lequel, laquelle, ksqiicUes, =
'(to a point)on' + 'the-which'

worauf = 'which-' + '-(to a
point)on'

121

442

General Linguistics I

TABLE U— Continued

e-p r

First

End-

ENDING-POINT (symboHzed by

Sample sentence

term

ing-

'place') OF MOTION EXPRESSED

of re-

point

BY PRINCIPAL VERB M, AND DI-

lation

rela-

RECT OBJECT OR ENDING-POINT

tion

OR LOCATION (symbolized by

'which') OF OCCURRENCE EX-
PRESSED BY SUBORDINATE VERB

(symbolized by ~)

cl24

M e-p r place which

E Wandering they knew

not whither.
F [II est parvenu d un

endroit qu^il ne con-

naissait pas.]
G Sie wanderten wohin

sie wollten.

cl 25. a M

e-p r place to
which

E Go whither he went.

cl 25. b M

(e-p r) place (to)
which ~'

F [Of. cl 24, F.]

Gl Wir gehen loohin er

gegangen ist.
G2 Wir gehen dahin wo

er gegangen ist.

E We shall go where he

went.
F Je vais ou le vent me

mene.

cl 26 M (e-p r) place at

which

G Wir reisen ihm nach
wo er gegangen ist.
[Colloquial. Of. cl
25. a, G.]

E Let's go where it's
quiet.

F Nous irons ou chant-
ent les hirondelles.

G Wir gehen wo es still
ist.

122

Five: Smdlcs in I 'nivcrsal ( Omcputal ( iraninuir

443

TABLi: il-Coniinucd

English rendering of instance
of A, e-p r, LJ

Rcfercncea
AnnutationB

cl 24 E A wandering whose e-p is places
which they know not.
F [A movement whose e-p is a
place which he did not know.]

G A wandering whose e-p is places
which they desired.

II, E120-1, GlOo; III, E120.
u-hilhcr = 'to-a-i)hice-which'
a un endrait que = 'to' + 'a

place' + 'which'
wohin = 'a-place- which-' -f

'-to'

cl 25. a E A commanded movement M"/jose II, E119c, GlOoa.

e-p is the place to which he whither = 'to-a-place-to-whicli'
went.

F — wohin = 'a-place-(to-) which- -f-
Gl Ditto. '-to'

dahin wo = t he-place-' -f
G2 Ditto. '-to' + '(to-)which'

cl 25. b E A movement (whose e-p is) the II, F079.

place (to) which he went. ivhere = ('to-)the-placc-(to-)
F A going (whose e-p is) any place which'

(to) which the wind has or ou = Ditto.

will have taken me. wo = Ditto.
G Same as E.

cl 26 E A requested movement (whose II, E130.

e-p is) any place at xchich

it is quiet. where = '(to-)a-placc-at-which'

A going (whose e-p is) the place oil = Ditto.

at which nightingales sing. wo = Ditto.

A going (whose c-pis) any place

at ichich it is quiet.

123

444

General Linguistics I

OCCURRENCE

TABLE II— Concluded

r
loc r

PLACE WHERE

Sample sentence

cl 27 OCCUR- loc r any place (to)
RENCE which

E Wherever I go, I find
the same sadness.

F Quelque part que Ton
ait p^netr6, I'hom-
me a trouv^ par-
tout des hommes.

G [Wohin ich auch im-
mer gehe, finde ich
Kummer.]

(H) Implications of e-p relation in stray words.

First
term
of re-
lation

e-p r

End-
ing-
point
rela-
tion

ENDING-
POINT

POINT OP
REFERENCE

cl 28 M {e-p r) ending-

point

E He went home.

F [II est alle chez lvi.]

G Der Hirt treibt
Herde heim.

die

124

Five: Studies in Universal Concepdnil (Inimnuir 445

TABLE n -Concluded

English rendering of instance Refercneefl

of A, loc r, H, or of A, e-p r, IJ AnnotationB

cl 27 E The finding of sadness in any II, El 37, F07.S; III, 137.
place (to) which I go.

Wherever = 'in-place-(to-) which-'
+ -any'
F The finding of men in any place

(to) which one has penetrated. Quelque part que = '(in) any' -|-

'place' + '(to-) which'

G [The finding of trouble in any [Wohin auch = 'in-place-which-*
place to which I go.] -f '-to' + 'any']

(H) I&II, EG(ll).

cl 28 E A going (whose e-p is) home. [Cf. II, FOIL] III, E138-9.

F [A going (whose e-p is) (a point)

AT ONE'S OWN PLACE.]

G A compelling of a movement
(whose e-p is) home.

125

446 General Linguistics I

Editorial Note

By Edward Sapir and Morris Swadesh; edited by Alice V. Morris. Published as no.
10 of Language Monographs (Linguistic Society of America), Baltimore, Waverly
Press. March V)32.

Hic following errors in the originally published version have been corrected

directly into the text printed here (page references are to the original):

p. 21, 1. 34-35: Le foret etend (correct: La foret s'etend)

p. 80, 1. 37: (ie le ville (correct: de la ville)

p. 98. under cl 03, Fl: permi (correct: permis)

p. 120, under cl 22. b, F: le suivi (correct: le suivit)

Further corrections to be made (page references are to the original):

p. 22, 1. 19: ail cote; correct into: a cote [2 x]

p. 22, 1. 20: dega; correct into: degd

p. 23, 1. 13: une autre vase; correct into: un autre vase

p. 49. 1. 29: 'dega': correct into: 'dega'

p. 50, 1. 14-15: one of the doors; correct into: one of the gates

p. 51, 1. 14-15: ba-ton; correct into: ba-ton

p. 64, 1. 7: space locative; correct into: space-locative

p. 70, 1. 11-12: but to to use; correct into: but to use

p. 80, 1. 34: Monmartre; correct into: Montmartre

p. 118 under class 20c, Fl: Mettez le; correct into: Mettez-le

The spelling of words as quoted from lexicographical sources has not been
modernized.

Five: Sditliis in I nivcisul C Onu'plitul (iiiininuir 447

GRADING, A STUDY IN SEMANTICS
EDWARD SAI'IIl
1. THE PSYCHOLOGY OF GRADING

The first thing to realize about pradinp; as a psyohological process is that it
precedes measurement and counting. Judgments of the ty|)e "A is larger than
B" or "This can contains less milk than that" are made long Ix^fore it is possible
to say, e.g., "A is twice as large as B" or "A has a volume of 25 cubic feet, B a
volume of 20 cubic feet, therefore A is larger than B by 5 cubic feet," or "This
can contains a quart of milk, that one 3 quarts of milk, therefore the former has
less milk in it." In other words, judgments of quantity in terms of units of
measure or in terms of number always presuppose, explicitly or implicitly, pre-
liminary judgments of grading. The term Jour means something only when it is
known to refer to a number which is "less than" certain others, say five, six,
seven, arranged in an ordered series of relative mores and ksses, and "more than"
certain others, say one, two, three, arranged in an ordered series of relative mores
and lesses. Similarly, a foot as a unit of linear measure has no meaning what-
ever unless it is known to be more than some other stretch, say an inch, and less
than a third stretch, say a yard.

Judgments of "more than" and "less than" may be said to be based on F)er-
ceptions of "envelopment." If A can be "enveloped by" B, contained by it, so
placed in contact with B, either actually or by the imagination, as to seem to be
held within its compass instead of extending beyond it, it is judged to be "less
than" B, while B is judged to be "more than" A. With only two existents of
the same class, A and B, the judgments "A is less than B" and "B is more than
A" can be translated into the form "A is small" and "B is large." In the ca.se of
the two cans of milk, we may say "There is little milk in this can" and "There is
much milk in that can." Again, if there are three men in one room and seven in
another, we may either say "The first room ha.?, fewer men in it than the second"
and "The second room has more men in it than the first" or, if we prefer, "The
first room has few men in it" and "The second room has many men in it."'
Such contrasts as small and large, little and much, few and many, give us a decep-
tive feeling of absolute values within the field of (juantity comparable to such
qualitative differences as red and green within the field of color perception. This
feeling is an illusion, however, which is largely due to the linguistic fact that the
grading which is implicit in these terms is not formally indicated, wherea^s it is
made explicit in such judgments as "There were fewer people there than here"
or "He has more milk than I." In other words, many, to take but one example,
embodies no class of judgments clustering about a given (luantity norm which is
applicable to every type of experience, in the sense in which red or green is appli-
cable to every experience in which color can have a place, l)ut is, prop<^rly sj^-ak-
ing, a purely relative term which loses all significance when (lcpri\«'d of it.s conno-

*"Few" and "many" in a relative sense, of course. More of this anon.

448 General Linguistics I

tation of "more than" and "less than." Many merely means any number,
definite or indefinite, which is more than some other number taken as point of
departure. This point of departure obviously varies enormously according to
context. For one observing the stars on a clear night thirty may be but "few,"
for a proof-reader correcting mistakes on a page of galley the same number may
be not only "many" but "very many." Five pounds of meat may be embar-
rassingly "much" for a family of two but less than "little" from the standpoint
of one ordering provisions for a regiment.

2. DEGREES OF EXPLICITNESS IN GRADING

We may bring these remarks to a focus by saying that all quantifiables
(terms that may be quantified) and all quantificates (terms to which notions of
quantity have been applied) involve the concept of grading in four degrees of
explicitness.

(1) Every quantifiable, whether existent (say house) or occurrent (say run)
or quality of existent (say red) or quality of occurrent (say gracefully), is intrin-
sically gradable. No two houses are exactly identical in size nor are they identi-
cal in any other feature that can be predicated of them. Any two houses selected
at random offer the contrast of "more" and "less" on hundreds of features which
are constitutive of the concept "house." Thus, house A is higher but house B
is roomier, while existent C is so much smaller than either A or B that it is "less
of a house" than they and may be put in the class "toy" or at best "shack."
Similarly, the concept of "running," involving, as it does, experience of many
distinct acts of running which differ on numerous points of "more" and "less,"
such as speed, excitement of runner, length of time, and degree of resemblance to
walking, is as gradable as that of "house." Different examples of "red" similarly
exhibit "mores" and "lesses" with respect to intensity, size of surface or volume
characterized as red, and degree of conformity to some accepted standard of
redness. And "gracefully" is quite unthinkable except as implying a whole
gamut of activities which may be arranged in a graded series on the score of
gracefulness. Every quantifiable, then, not yet explicitly quantified, is gradable.
Such terms may be called implicitly gradable but ungraded.

(2) As soon as a quantifiable has been quantified, the resulting quantificate
necessarily takes its place in an infinite set of graded quantificates. Thus, three
houses and the whole house belong to infinite sets in which they are respectively
"less than" four houses, five houses, six houses, • • • , and "more than" }ialf of the
house, a third of the house, a fourth of the house, • • • . Such terms may be called
implicitly graded by quantification. The process of grading is here of interest
only insofar as quantification is impossible without it.

(3) Instead of directly quantifying a quantifiable in terms of count or measure,
e.g. one hundred men or a gill of milk, one may content oneself with an indirect
quantification by means of quantifiers which are thought of as occupying posi-
tions in a sliding scale of values of "more" and "less," e.g. many men or a little
milk. Such terms may be called quantified by implicit grading. Here the grading
is of essential interest but is assumed as accomplished rather than stated as

Fivt': SiKilifs in I nivcrsul ( Omcpuml ( iriinirndr 449

taking place. Such terms as many arc psycholoKJcally miflway Ix'twM^n U»rmH
like more than and hundred. First, a set A is iM-rceived jus capable of envelopment
by another set B, which latter is then declared to Ik' "more than" A. Next, B
is declared to be "many," the reference to sets of tyi>e Ai , Aj , Aj , • • • , all of which
are "less than" B, being purely implicit. Finally, the "many" of B is di.scovered
to consist of a definite number of terms, say "one hundred," at which point grad-
ing as such has ceased to be of interest. In the realm of quantity "one hundre<l"
is a gradable but ungraded absolute in approximately the same sen.se in which in
the realm of existent^ "hou.se" is a gradable but ungraded ab.solute.

(4) Instead, finally, of quantifying by means of terms which grade only by
implication we may grade explicitly and say, e.g., ^'More men are in this room
than in that." Such a statement emphasizes the fact of grading itself, the (juan-
tifying judgment (e.g. "Mamj men are in this room but few in that" or ''Few
men are in this room but even fewer in that") being left implicit. Such terms as
"more men" may be called explicitly graded and implicitly quantified.

The following scheme conveniently summarizes the grading gamut:

1. Implicitly gradable but ungraded: house; houses

2. Implicitly graded by quantification: half of the house; a house SO ft. mde;

ten houses

3. Quantified by implicit grading: much of the house; a large house; many houses

4. Explicitly graded and implicitly quantified: mx)re of the house {than); a

larger house; mx)re houses (than)

3. GRADING FROM DIFFERENT POINTS OF VIEW

Only the last two types of terms are of further interest to us here. We shall
briefly refer to the quantifying elements of terms of class 3 as implicitly graded
quantifiers, to explicitly grading terms as graders (more than, less than), and
to the implicitly quantifying elements of terms of class 4 as explicitly grading
quantifiers. It is very important to realize that psychologically all comparatives
are primary' in relation to their corresponding absolutes ("positives"). Just as
more men precedes both some men and many men, so better precedes both good
and very good, nearer (= at a less distance from) precedes both at some distance
from and near (= at a srnall distance from). Linguistic u.sage tends to start
from the graded concept, e.g. good (= better than indifferent), bail (= worse than
indifferent), large (= larger than of average size), small (= smaller than of average
size), much (= more than a fair amount) , few (= less than a fair number), for the
ob\ious reason that in experience it is the strikingly high-graded orlow-gradcil
concept that has significance, while the generalized concept Nshich includes all
the members of a graded .series is arrived at by a gradual process of striking the
balance between the.se graded terms. The purely logical, the psychological,
and the linguistic orders of primacy, therefore, do not necessarily correspond.
Thus, the set near, nearer, far, farther, and at a normal distanrr from and the .sot
good, better, bad, worse, of average quality, show the following orders of complica-
tion from these three points of view:

A. LoGic.\L Grading:

450 General Linguistics I

Type I. Graded with reference to norm:

(1) Norm: at a normal distance from; of average quality

(2) Lower-graded: at a less than normal distance from = nearer or less far (from)
(explicitly graded), near or not far {from) (implicitly graded) ; of less than average
quality = worse or less good (explicitly graded), bad or not good (implicitly graded)

(3) Upper-graded: at a more than normal distance from = farther or less near
(explicitly graded), far or not near (implicitly graded) ; of more than average qual-
ity = better or less bad (explicitly graded), good or not bad (implicitly graded)

Type II. Graded with reference to terms of comparison:

(1) Lower-graded: at a less distance than = relatively nearer or relatively less far
(explicitly graded), relatively near or relatively not far (implicitly graded); of less
quality than = relatively worse or relatively less good (explicitly graded), relatively
bad or relatively not good (implicitly graded)

(2) Upper-graded: at a greater distance than = relatively farther or relatively
less near (explicitly graded), relatively far or relatively not near (implicitly graded) ;
of greater quality than = relatively better or relatively less bad (explicitly graded),
relatively good or relatively not bad (implicitly graded).

Note on A (Logical Grading). In type I, "graded with reference to norm,"
any "nearer" or "near" is nearer than any "farther" or "far," any "worse" or
"bad" is worse than any "better" or "good"; correlatively, any "farther" or
"far" is farther than any "nearer" or "near," any "better" or "good" is better
than any "worse" or "bad." But in type II "graded with reference to terms of
comparison," "nearer" and "near" do not need to be near but may actually, i.e.
according to some norm, be far, "worse" and "bad" do not need to be bad but
may actually be good; correlatively, "farther" and "far" do not need to be far
but may actually be near, "better" and "good" do not need to be good but may
actually be bad. Hence specific "nears" and "bads" may factually be respec-
tively farther and better than specific "fars" and "goods."

A warning: These are logical terms, not terms of actual usage, which exhibit
great confusion. In certain cases usage preferentially follows type I, e.g. "more
brilliant" and "brilhant" connote, as a rule, some degree of noteworthy ability,
"more brilliant" being rarely equivalent merely to "not so stupid"; "good"
follows type I, but "better" follows type II, being equivalent to "relatively
better, not so bad," e.g. "My pen is better than yours, but I confess that both are
bad" (on the other hand, "A is more brilliant than B, but both are stupid" is
meaningless except as irony, which always implies a psychological transfer);
"near" tends to follow type I, "nearer" follows type II, but "near" may fre-
quently be used like a type II term, e.g. "From the point of view of America,
France is on the near side of Europe," i.e. "nearer than most of Europe, though
actually far." Interestingly enough, the correlatives of these terms do not
exactly correspond. "Stupid" and "less stupid" follow type I, "less stupid"
being never equivalent to "more brilliant" (except, again, ironically); "less
brilliant" is still "brilliant" as a rule, just as "less stupid" is still "stupid."
"Bad" and "less bad," differing in this respect from "good" and "better," both
follow type I; "less bad" is still "bad" but "better" (with reference to another

Five: Sliidirs in I'nivcrsal iOmcpluiil (irninnntr 451

term) may be even worse. (The "more" of inverse terms, e.g. "more stupid"
and "worse," has a negative direction, as we shall sw later.) "Far" tenrls to
follow type I, "farther" follows (ypo II, but "far" may frecjuently 1m' u.sod likf a
type II term, e.g. "He is sitting at the /ar end of the table," i.e. "at the end that
is farther, though actually near." Needless to say, a logical analysi.s must
proceed regardless of linguistic usage. On the whole, u.sage tends to juKsign
comparative terms to type II of grading, positive terms to type I of grading,
though this tendency never hardens into a definite nile. The linguistic types
will be tabulated under C below.

According to strict logic, we should start from, say, good = of average quality
(type I) or of a certain quality (type II) and grade all other cjualities as follows:

Type I: better, less good (explicit), corresponding to ordinary better, worse;
good indeed, indeed not good (irtiplicit), corresponding to ordinary good, bad.

Type II: relatively better, relatively less good (explicit); rekUively good indeed,
relatively indeed not good (implicit).

How embarrassing logically such linguistic couplets as good'.bad, far '.near,
much'.little really are comes out in asking a question. "How good is it?" "How
far was he?" and "How much have you?" really mean "Of what quality is it?"
"At what distance was he?" and "What quantity have you?" and may be an-
swered, with a superficial character of paradox, by "Very bad," "Quite near,"
and "Almost nothing" respectively.

B. Psychological Grading (a is graded with reference to b, which is either
some other term comparable to a or stands for some norm) :
Type I. Open-gamut grading: a, b, c, • • • , n.

(1) Explicit: a is less than b = bis more than a: a is nearer than b = b is farther;
than a, a is worse thdn b = b is better than a. Similarly for a:c; • • • ; a:n; b:c.
• • •; b:n; • • •; c:n; • • •

(2) Implicit: a is little = b is much: a is near = b is far, a is bad = b is good.
Similarly for other cases.

Type II. Conjunct closed-gamut grading: a, b, c, • • •, n [ ] o, p, q, • • •, t ((vg
series of colors graded from a, vivid green, to /, vivid yellow).

(1) Explicit: a is less green than b = b is greener than a; ■ • ■ (judgments of more
or less green or yellow] o is less yellow than p = p is yellower than o; • • • In the
brackets [ ] we have indeterminate field of marginal greens and marginal yellows,
in which ai'.ai is greener than 6i = 6i is yellower than ai = Oi is less yellow than bi .
In other words, at some point, n, crest of green is reached and more green as
grader gives way to more yellow, with establishment in transition zone, [ ]. (if
secondary more green always coming before less green.

(2) Implicit: a, b, c, • • • , n are shades of green; [judgments of green or yellow];
0, p, q, • • • , t are shades of yellow. In the brackets [ ] we havo yellowish greens
and greenish yellows.

Type III. Open-gamut grading (/) interpreted in terms of conjunct closed-gamut
grading {II): "a, b, c, • • •, n" interpreted, bv analog\' of (II), as "a, b, c, •• -,
glJh, i, j, •••,n[]."

452 General Linguistics I

(1 ) Explicit: a is less than b = bis more than a: a is less far than b = bis farther
than a, a is less good than b = b is better than a; ■ ■ • []his less near than i = i is
nearer than h, h is less bad than i = i is worse than h; - • • . In transition zone [ ]
we have psychologically indeterminate field of marginal fars (goods) and marginal
nears (bads), in which aiibi is interpreted as bi is less far (good) than ai = ai
is farther (better) than bi = 6i is nearer (worse) than ai = Oi is less near (bad)
than bi. In other words, at some point, g, crest of far (good) is reached and
farther (better) as grader gives way to nearer (worse), with establishment in transi-
tion zone, [ ], of secondary /ar//ier (better) always coming before less far (less good).
Type III, however, differs from type II in that it has a second psychologically
indeterminate field of marginal nears (bads) and marginal fars (goods), in which
hi:ii is interpreted as ii is less near (bad) than hi = hi is nearer (worse than ii =
ii is farther (better) than hi = hi is less far (good) than t'l . In other words, at some
point, n, crest of near (bad) is reached and nearer (worse) as grader gives way to
farther (better), with establishment in second transition zone, [ ], of secondary
nearer (worse) always coming before less near (less bad). Obviously, our second
[ ] brings us back to a, b, c, • • • , g. Type III of psychological grading (far-
near, good-bad) is circular in configuration, as we shall see more clearly later on,
while type II (violet-blue-green-yellow-orange-red) is successively semicircular.
Type II may be called conjunct semicircular closed-gamut grading or conjunct
closed-gamut grading with open ends; type III, conjunct circular closed-gamut
grading or conjunct closed-gamut grading with meeting ends.

(2) Implicit: a, b, c, • ■ -, g are far (good) in varying degree; h, i, j, ■ • •, n are
near (bad) in varying degree. In first transition zone [ ] we have psychological
blends of type not near (bad), not really near (bad), in second transition zone [ ],
psychological blends of type not far (good), not really far (good).

Type IV. Disjunct closed-gamut grading: a, b, c, • • •, g [e.g. neither blue nor
yellow] o, p, q, • • • , t.

(1) Explicit: a is less blv£ than b = bis blu£r than a; • • • [zone of indifference in
which neither blu£, nor yellow strictly applies] o is less yellow than p = pis yellower
thano; ■ • • . There is no psychological interest in zone of indifference, [ ], which is
only gradually spanned with increasing experience and demand for continuity.
When zone of indifference [ ] is recognized as h, i, j, • • • , n, it may: (a), take on
distinctive character, e.g. green, in which case type IV becomes identical with II,
for with establishment of continuity certain blues now become greenish blues,
bluish greens are created, and certain yellows now become greenish yellows; or
(b), be characterized negatively, in which case we cannot do better than say
h is neither blue nor yellow, neither h nor i is blue or yellow, but h is more nearly
blue than i and i is more nearly yellow than h, j is more nearly blue than yellow (is
bluer than it is yellow), k is mx)re nearly yellow than blue (is yellower than it is blue).
In other words, for grading are substituted other techniques, which have grading
implications, e.g. intermediate placement (between blue and yellow), goal-gauging
(nearly blv£), graded goal-gauging (more nearly blu£, nearer yellow than), negation
of alternatives (neither blue nor yellow), compromise (blu£-yellow^) .

*To be understood as theoretical tag for green.

Five: Sliuliis in Universal Conceptual Craniniar 453

(2) Implicit: a,b,c, ■ ■ ■ , g are shades of blue; o, p, q, ■ ■■ , t are shades of yellow.
For zone of indifTerence [ ] see (1).

Type V. Open-gamut grading (I) interpreted in terms of disjunei closed-gamut
grading (IV): "a, b, c, • • •, n" interpreted, by analogy of (IV), em "a, b, c, • • •,
e[]j, k,l, •••,n."

(1) Explicit: a is less hot, old, brilliant, good than b = b is holier, (even) older,
more brilliant, (even) better than a; ■ ■ ■ [zone of indifTerence in which neither hot
nor cold, neither old nor young, neither brilliant nor stupid, neither good nor bad
strictly appHes] J is less cold, young, stupid, bad than b = b is colder, younger, more
stupid, worse than a; ■ • . When zone of indifTerence [ ] is gradually recognized
as f, g, • • •, i, it may: (a), take on distinctive character, e.g. temperate, muidle-
aged, of normal intelligence, of average quality, such terms establishing filling-in
norms rather than mores and lesses of primary fields (e.g. more than middle-aged
rather than more middle-aged, of more than normal intelligence rather than more
normal, which would generally be understood as an ellipsis for more nearly nor-
mal), in which case type V becomes identical \vith type I, f, g, • • • , i being inter-
calated between j, k, 1, • • •, n and reversed field e, d, c, • • •, a; or (b), be char-
acterized negatively, e.g. f if neither hot nor cold, neither f nor g is old or young but
f is more nearly old than g, h is more nearly stupid than brilliant. In other words,
for grading are substituted other techniques, which have grading implication,
e.g. intermediate placement {betwixt old and young), goal-gauging {nearly good),
graded goal-gauging {nearer cold than hot = implicitly graded cool), negation of
alternatives {neither good nor bad), compromise {good or bad, deperuiing on one's
standard).

(2) Implicit: a, b, c, • ■ ■ , e are hot, old, brilliant, good in varying degree; j, k, I,
' • • , n are cold, young, stupid, bad in varying degree. For zone of indifference
[]see (1).

Note on B (Psychological Grading). It must be carefully borne in mind that
these five psychological types of grading, which naturally do not preclude the
possibility of still other, and more complex, grading configurations, are by no
means mutually exclusive types. The same objective elements of experience,
e.g. good'.bad, may be graded according to more than one type. Thus, when we
say "A is better than B," though A and B are both bad, we are obviously treating
better as an incremental grader in an open .series in which the movement is assumed
to be toward the relatively good and away from the relatively bad. "A is better
than B" therefore illustrates type I, open-gamut grading, which is the prototype
of all logical grading. On the other hand, when we .^ay "A is irorse than B,
which in turn '\s fairly good," we do not mean to imply that A too is jxTliaps not
too far from good, rather that A belongs distinctly to the lower end of the gamut,
that good and bad are psychologically distinct <iualities (not, like logically graded
terms of type I, merely a more and a less of a single (luality), but that these
distinct qualities are psychologically contiguous and capable of being fitted into
a single .series with two crests or maxima. All of this means that in this case we
are fitting the concepts of good and bad into a conjunct closed-gamut grading
scheme, and since the natural, or rather logical, type to which good .bad belongs

454 General Linguistics I

100

is type I, we speak of a transfer on the analogy of type II and create a blend type
III. Finally, when we say "A is better than B but both are good, C is of quite a
different order and is actually bad, while D, being neither good nor bad, is of no
interest," we are thinking in terms of a type of grading in which psychologically
distinct qualities are connected, by intercalation, into an open series of the dis-
junct closed-gamut grading type, namely type V.

Type I recognizes no crest, only a norm at best, which, in the logical form of
the grading (A), sinks to an objective or statistical norm — in other words, an
average. Type III recognizes two crests and two areas of blend, but no norm
except at the points where psychology, via neutral judgments, fades away into
logic. Type V recognizes two extreme and opposed crests and a trough of nor-
mality between them. Types I, II, and IV (near: far, green '.yellow, blue'.yellow)
are given us directly through our sensations or perceptions. Type III is prob-
ably the most natural type for psychologically subjective, as contrasted with
objective, judgments; even such simple contrasts as near: far and good:bad
probably present themselves, first of all, as contiguous areas of contrasting qual-
ity, not as points above or below a norm with which they intergrade in an open
series. After considerable experience with socially determined acceptances and
rejections, familiarities and strangenesses, contrasting qualities are felt as of a
relatively absolute nature, so to speak, and good and bad, for instance, even far
and near, have as true a psychological specificity as green and yellow. Hence the
logical norm between them is not felt as a true norm but rather as a blend area
in which qualities grading in opposite directions meet. To the naive, every
person is either good or bad ; if he cannot be easily placed, he is rather part good
and part bad than just humanly normal or neither good nor bad. Type V
represents the most sophisticated type of judgment, for it combines psychological
contrast with the objective continuum of more and less and recognizes the norm
as a true area of primary grading, not as a secondary area produced by blending.

We can easily see now that the confused psychological state of our grading
judgments and terminology, also the unsatisfactory nature of our logical grading
terminology, is due to a number of factors, the chief of which are: 1, the tendency
to conceive of certain points in an evenly graded series as primarily distinct and
opposed to each other instead of directly capable of connection by grading in
terms of more and less (this tendency is, of course, carried over even into the
realm of abstract quantity, and even a mathematically trained person may find it
somewhat paradoxical to call 7 "many" and 100 "few," though the 7 belong to a
context in which 9, say, is the maximum, and 100 to another context in which 500
is the norm) ; 2, the contrary direction of grading in two such contrasted qualities,
the "more" of one being logically, but not quite psychologically, equivalent to
the "less" of the other (e.g. logically better = less bad, worse = less good, but
psychologically this is not quite true; contrast nearer = less far, farther = less
near, where logic and psychology more nearly correspond) ; 3, a preference for the
upper or favored quality, in its relative sense, as grader (e.g. better and heavier
more easily serve as incremental upward graders, of more quality and of more

/Vk'. S[ii<li('\ in Universal Conceptual (Iraniniar 455

101

weight, than do worse and lighter as incremental downward graders; tlii.s haniipi
together with 5); 4, the confhet with psycholoKical grading brought in hy a more
sophisticated attempt to establish an absolute continuity of grading in a logical
sense (problems of interpretation of how, of two contrasted terms, a and b,
"more a" is related to "less b," and of whether the neutral area Ix'twcen a and b
is to be understood as a "both and" area, a "neither nor" area, or logically as
a tie between a and b, which thereupon lose their distinctiveness and one of which,
in consequence, must change its direction of grading so that a complete open-
gamut grading may be established); 5, the different psychological value of a
given grade according to whether it is reached positively, e.g. fairly good from
poor, or negatively, e.g. fairly good from very good (the latter "fairly good" is
almost necessarily an "only fairly good," i.e. a "fairly good" with emotional
coloring of "poor").

C. Linguistic Grading (elaboration of terms) :
Type I. Explicit:

(1) Abstract: more than, less than. These terms are general upward and
downward grading terms and carry no implication as to class of graded terms or
as to presence or absence of norms or crests. Certain other terms, of originally
specialized and normated application, such as greater, larger, and smaller, have
taken on abstract significance (e.g. a greater amount of = more • • • than, a larger
number of = more • • • than, a smxiller number of = less, fewer). More and less
apply to both count and measure. Fewer, as equivalent of less, applies to
counted terms only, e.g. fewer people = less people, but is secondary as exphcit
grader, being based on few, which is implicitly graded. There is no special
count term in English corresponding to more. More and less are old comparatives
in form, but are not really referable to much and little.

(2) Specialized. There are no explicit specialized graders in English which
are not based, generally by use of more and less or suffixing of comparative -er,
on linguistically primary graded terms which imply above or Ix^low a logical
norm. Thus, heavier, based on heavy (= of more than average weight), means
heavy to a greater extent (thari another heavy object) to begin with, and only second-
arily takes on, in its specialized sphere of weight, the purely relative grading
quality of more; similarly with /ess heavy as parallel to less. Such terms as of
more weight or more ivcighted, of less linear extent, of more temperature, less in vol-
ume are not in ordinary use and have to be replaced by comparatives of such
terms as heavy, short, rearm, small, which are not neutral in reference as to graded
area.

Type II. Implicit:

(1) Abstract: much and little for measured terms; many and few for counted
terms. Note that implicitly graded terms can themselves be taken as new points
of departure for grading, e.g. /r.s.s- tfian many, more than a few, many and a few
being respectively arrived at by grading upward and downward from a certain
norm. "How much?'' and "How many?" show how helpless language tends to be

456 General Linguistics I

102

in de\ising neutral, implicitly graded abstract terms; linguistically upper-graded
terms for logically neutral ones are also used in such terms as so and so many,
as much as.

(2) Specialized: A great variety of terms, most of which appear as pairs of
opposites. We may distinguish:

(a) One-term sets (graded as more and less; there is no true contrary) : capacious,
silvery, distant (in its strictly scientific sense of at such and such a distance, near
and /ar being "psychologized" forms of it). Such terms are either of notions of a
relatively ungradable type or are of scientific rather than popular appUcation.
Such terms as "how /arf", "how long?", "2mm. wide," "how warm?", "as heavy
as one tenth of a gram," "old enough to know better" again show how helpless
language tends to be in devising specialized single terms which are logically neu-
tral as to grading.

(b) Two-term sets. Two types are both common: 1, linguistically unrelated
terms indicating opposites, e.g. goodihad, far '.near, high: low, long '.short, full:
empty, heavy '.light, friend:enemy, hard: soft, old:young; 2, linguistically related
terms which are implicitly affirmative and explicitly contrary (formally negative)
terms, e.g. friendly: unfriendly (also type (b) 1, friendly: hostile, inimical), usual:
unu,suxil, normal: abnormal, frequent:infrequent, discreet: indiscreet. These for-
mally negative terms frequently take on as distinctive a meaning as type (a)
contraries and can be as freely graded, "upward" and "downward," e.g. more
and less infrequent are as good usage as rarer and less rare.

Note on (6), Two-term sets. As regards grading relations, two-term sets (con-
trary terms) tend to fall into three types :

I. Symmetrically reversible, e.g.
far, farther near, nearer

not near, less near not far, less far

II. Partly reversible, e.g.

good, better bad, worse

not bad, less bad not good, less good

III. Irreversible, e.g.
brilliant, more brilliant stupid, more stupid

not stupid, less stupid not brilliant, less brilliant

Note that implicitly graded specialized terms can themselves be taken as new
points of departure for grading, e.g. more than good, less than bad = better than bad.

(c) Three-term sets. These are not as common as type (b) (two-term sets)
in ordinary usage but are constantly required for accurate grading. Generally
one takes opposite terms of type (b) and constructs a middle term by qualifying
the upper-graded one, e.g. bad, averagely (or moderately or normally) good, good.
Sometimes a middle term comes in by way of transfer from another field, e.g. bad,

l-i\c: Siiidifs in Universal Conn'piiuil (irnminur 457

103

fair, good. Sf)ecific middle terms, however, tend to gravitate toward one or the
other of the two opposites, e.g. fair, on the whole, leans more to good than bad.
If we further insert poor, again transferred from another field, we get type(d),
four-torm sets: had, poor, fair, good. (The reason why poor, when transferred to
the bad: good scale, does not quite fall in with had is that poor .rich ha« not quite
as great a scale amplitude as bad: good (zero to maximum) but is felt aH corre-
sponding rather to a scale of little to maximum. Zero, lower average, higher
average, much is the implicit measure of having nothing (= destitute, penniless)
having little (= poor), having a moderate amount (= fairly well off), having much
(= rich). Hence poor stresses something, though little and cannot entirely parallel
had, which includes its logical extreme.) On the whole, three-term sets do not
easily maintain themselves because psychology, with its tendency to simple
contrast, contradicts exact knowledge, with its insistence on the norm, the
"neither nor." True three-term sets are probably confined to such colorless
concepts as: inferior, average, superior, in which the middle term cannot well be
graded.

(d) Four-term sets: cold, cool, warm, hot. These are formed from type (b) by
grading each of the opposites into a psychologically lower and higher. The new
terms become psychological opposites (or sub-opposites) of a smaller scale. It
is important to note that the two middle terms do not correspond to the middle
term of type (c) (three-term sets), i.e. warm is psychologically no nearer to cool
than superior is to inferior. In other words, cold-cool contrasts with wartn-hot
precisely as does very bad-bad with good-very good. The problem of connecting
cool and warm has to be solved, psychologically, by blend-grading {coolish;
warmish, lukewarm) or, more objectively, by norming (of ordinary, normal, tem-
perature). As usual, the normed term is quasi-scientific rather than popular
in character.

More complex linguistic sets are of course possible. We may summarize
these analyses of the grading process by saying that logical grading is of the open-
gamut type and may be with or without reference to an objective norm or sta-
tistical average, while psychological grading and linguistic grading tend strongly
to emphasize closed-gamut grading, whether of the conjunct or disjunct typx*,
and have difficulty in combining the notions of grading and norming into that
of a normed field within which grading applies. Furthermore, it is worth noting
that the dilTerence between explicit and implicit grading is of little importance
logically, of considerable importance psychologically (with constant conflict of
the relative and fixed points of view), and of paramount importance linguistically.

4. IMPLICATIONS OF MOVEMENT IN GIUDING

The main operational concepts that we have u.sed in developing our notions
of grading up to this pc^int have been: the successive envelopment ui values by
later ones (giving us a set of "lesses" in an open series); the establishment of a
nf)rm somewhere in such an open series; the placement of values "above" and
"below" this norm; the contrasting of specific gradable values which l)clong to
the same class; the establishment of continuity Ix'tween such contra^^ting values

45S General Linguistics I

104

by means of intercalation; and certain implicit directional notions (upward,'
e.g. good'.helter, bodiless bad; downward,' e.g. good'.less good, bad: worse; contrary,
e.g. good-better: bad-worse).

The directional ideas so far employed have merely implied a consistent in-
crease or decrease in value of the terms which are seriated and graded. Thus,
of a set of terms "a, b, c, • • •, n," in which a is less than any of the terms "b, c,
• • • , n," and b is less than any of the terms "c, • • • , n," and c is less than any of
the terms " • • • n," and no term is more than n, we have established an upward
grading direction, consistently from less to more, but the terms themselves are
not necessarily thought of as having been arrived at either by moving up from a
or do\vn from, say, c. Logically, as mathematically, b increased from a = b
decreased from c. Psychologically, however, and therefore also linguistically,
the expUcit or implicit trend is frequently in a specific direction. It is this
tendency to sUp kinaesthetic implications into speech, with the complicating
effects of favorable affect linked with an upward trend and of unfavorable affect
linked with a downward trend, that so often renders a purely logical analysis of
speech insufficient or even misleading.

We can easily test the kinaesthetic aspect of grading by observing the latent
direction and associated feeling tone of an implicitly graded term like "few."
If some one asks me "How many books have you?" I may answer "A few,"
which is, on the whole, a static term which, though indefinite, takes the place of
any fixed quantity, say 25, deemed small in this particular context. But if I
answer, "I have few books," the questioner is likely to feel that I have said more
than is necessary, for I have not only fixed the quantity, namely "a few," but
implicitly added the comment that I might be expected to have a larger number.
In other words, "few" suggests grading downward from something more, while
"a few" is essentially noncommittal on the score of direction of grading. The
difference here in implicit grading is not one of magnitude, but of direction only.
The psychological relation between "a few" and "few" is very similar to the
psychological relation between "nearly" and "hardly," which belong to the
conceptual sphere of gauging.

Can "a few" be given an upward trend? Not as simply and directly as the
change to "few" gives a downward trend, but there are many contexts in which
the upward trend is unmistakable. If I am told "You haven't any books, have
you?" and answer "Oh yes, I have a few," there is like to be a tonal peculiarity
in the reply (upward melody of end of "few") which suggests upward grading
from zero. Language, in other words, here ekes out the notional and psycho-
logical need for an upward-tending quantitative term as best it can. If I use
"quite," which has normally an upward-tending feeling tone, and say "Quite
a few," the kinaesthetic momentum carries me beyond the static "a few," so
that "quite a few" is well on toward "a considerable number."

The kinaesthetic feeling of certain graded terms can easily be tested by trying

^^^ "Upward" and "downward" are used in the sense of "in the direction of increase" and
"in the direction of decrease" respectively. This purely notional kinaesthesis may be, and
probably generally is, strengthened by a concomitant spatial kinaesthesis.

Five: Siiulii's in LInivcrsal Conccpiuiil (Iruninuir 459

105

to use them with terms whose kinaesthetic latency is of a difTerent nature and
noting the baffled effect they produce due to implied contradictions of movement.
Thus, we can say "barely a few" or "hardly a few" lx.'cause "a few" is conceived
of as a fixed point in the neighborhood of which one can take up a position or
toward which one can move, positively or negatively. But /'nearly few" is
baffling, and even amusing, for there is no fixed "few" to be near t(j. "Hardly
few" is psychologically improper too, for "hardly" suggests a falling short, and
inasmuch as "few" is downwardly oriented, it is hard to see how one can fall
short of it. "Hardly few" has the same fantastic improbability as the concept
of A moving on to a supposedly fixed point B, which it "hardly" expected to
reach, and finding that B was actually moWng toward A's starting point, and
eventually reaching it, without ever passing A. Again, "all but" requires a
psychologically fixed term to complete it, e.g. "all but half," "all but a few."
"All but few" suggests a remainder which is not even a remainder. Again, "all
but quite a few," even if "quite a few" is no more factually than a small propor-
tion of the whole, is psychologically difficult because "quite a few" is no more
static than "few." The "all but" form is implicitly static, hence "all but few"
and "all but quite a few" ring false, involving, as they do, down-tending and
up-tending elements respectively.

5. THE CONCEPT OF EQUALITY

We are now in a position to arrive at a simple psychological conception of
"equal to." "Equal to" may be defined as the quantitative application of the
qualitative "same as," "more than" and "less than" being the two possible
kinds of quantitative "different from." But it seems more satisfactorj-, on the
whole, to define "equal to" in a more negative spirit, as a more or less temporary-
point of passage or equilibrium between "more than" and "less than" or as a
point of arrival in a scale in which the term which is to be graded is constantly
increasing or diminishing. In other words, if we take q as defined to begin with,
we can give meaning to a = g by saying that: (1) a is less than q to begin with,
gradually increases while still less than q, and is later found to be more than (.\,
having i)assod through some point at which it was neither less than nor more
than (j; or (2) a is more than q to begin with, gradually decreases while still more
than q, and is later found to be less than q, ha\ing passed through some point
at which it nas neither more than nor less than (]: or (3) a is less than q to Ix^gin
with, gradually increases while still less than (j, and finally rests at some point
at which it is neither less than nor more than (i; or (4) a is more than q to begin
with, gradually decreases while still more than (], and finally rests at some point
at which it is neither more than nor le.ss than q. The.se four types of eciuality
may be classified as:

(1) While increasing toward and away

T T-. 1- • 1 I • J from

I. Explicitly dynamic <j ^2) While decreasing toward and away

from

460 General Linguistics I

106

. ((1) Having increased toward

. mp y yTi I ^2) Having decreased toward

A fifth type of equality, that of kinaesthetic indifference, is the limiting or neutral
type which alone is recognized in logic:

III, Non-dynamic: Statically "equal to."

So far are these psychological distinctions from being useless that, as a matter of
fact, a httle self-observation will soon convince one that it is hardly possible to
conceive of equality except as a medium state or equilibrated state in an imagined
back and forth of "more than" and "less than." It is safe to say that if we had
no experience of lesses increasing and of mores decreasing, one could have no
tangible conception of how obviously distinct existents, occurrents, and modes
could be said to be "equal to each other" in a given respect.

6. THE CLASSIFICATION OF TYPES OF GRADING JUDGMENT

The classification of "equals" applies, of course, equally well to "mores" and
"lesses," so that we have, psychologically speaking, 15 fundamental judgments
of grading to deal with, of which the 3 logical ones ("more than," "equal to,"
and "less than") are the kinaesthetically neutral judgments. The best way to
understand this enlarged grading scheme is to express it symbolically. Let
a — > g be understood to mean "a is less than q and is increasing toward it,"
a*— qto mean "a is less than q and is decreasing away from it," g' — > a to mean
"a is more than q and is increasing away from it," g <— a to mean "a is more than
q and is decreasing toward it." In other words, "to the left of" means "less
than," "to the right of" means "more than," while an arrow pointing to the right
means "increasing," an arrow pointing to the left means "decreasing." An
arrow pointing dowTiward will mean "having increased," an arrow pointing up-
ward will mean "having decreased," and an arrow superimposed will mean
"equal to, \nth imphcation of actual or prior movement." We then have the
following symboHcally expressed notional scheme of grading judgments which
can be made of two entities of the same class, a and q, of which q is supposed to
be known and fixed. In the symbolism a will be understood as the subject of the

implied proposition.

I. Explicit
d)aiamic

II. ImpUcit
dynamic

Types of Grading Judgment

Increasing

1. ->q 6.^ 11. q-
Decreasing

2. <-q 7."^ 12. q^

Increased I

3- i q 8. q 13. q j

Decreased

4. I q 9. q| 14. q |

Five: Stuilies in Universal ( Oncrpiuiil (iiiutiniur 461

107

III. Nondynamic 5. ja = ql 10. fa = (U 15. fa

^q = a/ lq = a/ \q

These syinl)()ls may Ix; read as follows:

1. "is being less than (i, though increasing" ( =
"still falls short of")

2. "is being less than q, and decreasing" ( =
"falls shorter and shorter of")

3. "is less than q, though increased from still
less" (= is still short of")

•1. "is less than q, and decreased from more"
(= "is even short of")

5. "a is less than q" (= is short of) = "q is more than a"

6. "is equalling q, on its way from less to more"

7. "is equalling q, on its way from more to less"

8. "is equal to q, having increased to it"

9. "is equal to q, ha\'ing decreased to it"

10. "is equal to q"

11. "is being more than q, and increasing"
(= "exceeds more and more")

12. "is being more than q, though decreasing"
(= "still exceeds")

13. "is more than q, and increased from less"
(= "is even in excess of")

14. "is more than q, though decreased from more"
(= "is still in excess of")

15. "a is more than q" (= "is in excess of") =
"q is less than a."

The symbols for nos. 5, 10, and 15 are of course the ordinary mathematical ones,
a < q and q > a being considered equivalent notations. The sign of equaUty,
= , may, if one likes, be looked upon as the neutralized forms of nos. 6 and 7: ^.
In order to give more reality to these theoretically distinct types of grading, it
may be of some service to give simple examples of them. For this purpose we
shall take 5 (miles, pounds, hours) as illustrative of q, thus applying our notions
of grading to the sphere of quantity.

1. "He has run less than five miles" : -* 5 (answer to question: "How far has
he run by now?")

2. "He has less than Jive hours to finish his job" : <— 5 (answer to (juestion:
"How much time can he count on to finish his job?")

3. "Ho ran until he came to a point that was less than Jive miles from his start-
ing point": i 5 (answer to cjuestion: "How far had he got wlien he stopjx'd
running?")

4. "He got weaker and weaker until he could lift less than Jiir pouwis" : ] 5
(answer to question: "How much could he still lift when he had to give up?")

5. "Jersey City is less than five miles from New York": a < 5 (answer to
question: "How far [a, i.e. required distance] is Jersey City from New \ork?")

4f,2 General Linguistics I

108

— >

6. "He has run {as much as) five miles": 5 (answer to question : "How far has he
run by now?")

7. "He has {just, still) five hours to finish his job: 5 (answer to question: "How
much time can he count on to finish his job?")

8. "He ran until he came to a point that was {just, as much as, already*) five

i
miles from his starting point": 5 (answer to question: "How far had he got

when he stopped running?")

9. "He got weaker and weaker until he could lift {ju^t, only, no more than)

five pounds": 5 (answer to question: "How much could he still lift when he had
to give up?")

10. "A is {just) five miles from B": a = 5 (answer to question: "How far [a]
is A from B?")

11. "He has run more than five miles": 5 — > (answer to question: "How far has
he run by now?")

12. "He {still) has more than five hours to finish his job": 5 -^ (answer to ques-
tion: "How much time can he count on to finish his job?")

13. "He ran until he came to a point that was {even) more than five miles from
his starting point": 5i (answer to question: "How far had he got when he
stopped running?")

14. "He got weaker and weaker until he could lift hardly more than five
pounds": 5 t (answer to question: "How much could he still lift when he had to
give up?")

15. "Philadelphia is more than five miles from New York": 5 < a (answer to
question: "How far [a] is Philadelphia from New York?")

7. AFFECT IN GRADING

It will be observed that such terms as as much as, just, still, already, only, no
more than, even, hardly, and others not illustrated in our examples help along,
as best they can, to bring out the latent kinaesthetic element in the logical
concepts "less than," "equal to" and "more than" when these are applied to
experience, but at best they are only a weak prop. Most languages suffer from
the inability to express the explicitly dynamic, implicitly dynamic, and non-
dynamic aspects of grading in an unambiguous manner, though the notional
framework of fifteen grading judgments that we have developed is intuited by all
normal individuals. Such English terms as we have suggested are really un-
acceptable for two reasons: 1, they are transfers from other types of judgment
than dynamic and non-dynamic grading (e.g. "only" is properly an exclusive
limiter; "hardly" and "just" are goal-gauging Hmiters; "still" has time implica-
tion, at least in origin); and 2, they unavoidably color the judgment \Wth their
latent affect of approval or disapproval (e.g. "as much as" smuggles in a note of
satisfaction; "only" and "hardly" tend to voice disappointment).

*More idiomatic in German: schon.

Five: Sludics in IJnivcrsul COtucptinil (iianunnr 463

109

Even the simple graders "more than" and "less than" tend to have a definite
affective c}uality in given contexts. Thus, if a (juantitative goal is to Ix? reached
by increase, say "ten pages of reading," more than necessarily has an approving
ring (e.g. "I have already read more tluin three pages," though it may actually Ix;
less than four), less than a disapproving ring (e.g. "I have only read le.sn than right
pages," though it may actually be more than seven). On the other hand, if the
quantitative goal is to be reached by decrea.se, say "no more reading to do,"
more than has a disapproving ring (e.g. "I have still more than three pages to do,"
though actually less than four remain to be done), less than an approving ring
(e.g. "I have less than eight pages to do," though more than seven pages remain
to be done out of a total of ten). In other words, grading and affect are inter-
twined, or, to put it differently, more than and less than tend to have both an
objective grading value and a subjective grading value dependent on a desired
or undesired increase or decrease. This means that linguistic awkwardnesses
arise when it is desired to combine an objective more than with a subjectively
desired decrease or an objective less than with a subjectively desired increase.
Thus, if the more than three days in "I have more than three days to wait" is to
convey the approving connotation of "only four or five days," we cannot say "I
have only more than three days to wait" (as contrasted with a possible more than
ten days) but must recapture the note of approval by minimizing the implied
excess, hence "I have only a little more than three days to wait." An approved
more than (a slight quantity) in a desiredly decreasing scale, though logically
defensible, goes against the psychological grain of language. Again, it is hard to
say "I have only more than fifty dollars in the bank," ior fifty dollars plus a slight
amount (by implication) is on an upgoing trend, as it were. We have to grade
down from fifty-one dollars, say, and say "I have less than fifty-one dollars m
the bank." To put it differently, if S50.99 is disapproved of, it must lx> graded
downward as less than fifty-one dollars; if $50.01 is approved of, it can be graded
upward as more than fifty dollars. The difficult word hardly frequently reorients
the normally implied affect, hence "I have hardly lyiore than three days to wait"
(approval), "I have hardly more than fifty dollars in the bank" (disapproval).

If we had a subjective grading symbolism that was independent of objective
grading, it would be possible to convey very compactly every po.<;sible tyjx? of
grading judgment — static, implicitly dynamic, and explicitly dynamic grades
independently combined with neutral, approving, not di.sappnn-ing, disapprov-
ing, and not approving affect. How complex, in actual speech, our grading
judgments, or rather intuitions, really are from a psychological standpoint,
however simple they may seem to be from a purely logical or merely linguistic
standpoint, may be exemplified by considering the meanings of such apparently
simple statements as "I have three pages to read," "I have more than three pages
to read" and "I have less than three pages to read." In the first place, it makes a
difference if "three pages" (or "reading matter equal to three pages") is conceived
non-dynamically or dynamically, e.g. "three pages as an a.ssigned task" (grade
10: non-dynamic "equal to") or "more than three pages in a rapidly accumulating
series of M8 pages submitted for approval" (grade 1 1 : explicit dynamic increas-
ing "more than") or "more than three pages still to do in the passage from a total

454 General Linguistics I

110

of ten pages to do to the goal of no pages left to do" (grade 12: explicit dynamic
decreasing "more than") or "less than three pages yet accumulated in a long MS
report which one desires to read" (grade 3: implicit dynamic increased "less
than"). Ordinarily, the affective valuation involved in such statements does
not clearly rise in consciousness because "more than" and "less than" pool the
energies, as it were, of the grading process itself and the approval or disapproval
of increase (growing exhilaration, growing fatigue) or decrease (growing relief,
growing disappointment). We cannot possibly go into all the involvements of
this very difficult field of inquiry, but a general idea of its nature may be had by
considering one case, say the expHcit dynamic decreasing forms of "less than"
(grade 2), "equal to" (grade 7) and "more than" (grade 12).

Our type statements will be "less than three pages (to read)" "(still as much as)
three pages (to read)" and "more than three pages (to read)." These will be

symbolized, in the first instance, by <— 3, 3, and 3 <— respectively. If, in the
statement "I have less than three pages to read," the reading is conceived of as a
task which is to be accomplished, say a certain amount of Latin to be prepared
for translation, the statement will be normally interpreted as implying approval
of decrease (growing relief), the implication being that of "only." Had we
wished to imply disapproval of increase (growing fatigue), we should normally
have put it not at "less than three pages" but at "more than two pages," with an
implication of "still." We could combine the form of approving "less than"
statement with that of disapproving "still" and say "I have less than three pages
to read, to be sure, but there is still some of my assignment to read." In other
words, w'hen the goal, zero, is approved, any form of statement implying decrease
toward that goal involves approval, and the factual disapproval of having still so
much left to do has normally to be rendered by terms implying reversal of judg-
ment, such as to be sure, but, still. Our linguistic awkwardness in expressing
disapproval of a state which is kinaesthetically committed, as it were, to approval,
is on a par with, though less obviously helpless than, such periphrases for the
potential mode as "He will come, he will not come," a naive substitute for "Per-
haps he will come" or "He may come." Let us, for the sake of brevity, reduce
the complete circle of valuation in judgment to the two simple forms of approval
and disapproval, symbolized respectively by ^ and J . Then ^ 3 symbolizes
an explicit dynamic decreasing "less than 3" which is approved of, the "less than
3" of growing relief inadequately rendered in English by "less than 3" or the
rather unidiomatic "already less than 3" or the round-about "only 3, in fact
less." And

symbolizes an explicit dynamic decreasing "less than 3" which is disapproved
of, the "less than 3" of growing fatigue, which cannot easily be rendered in Eng-
lish except by such periphrases as "still some, though less than 3."

Further consideration of the implied "only" and "still" of these statements
shows that they may indicate exactly the opposite affects if w^e assume that the
goal of decrease is not desired but resisted. Thus, if my desire is to read all I

/Vkv .S7/ /<//<• s /// Universal Comcpiunl (iriunniiir 465

111

can get, an approving <— 3 can not imply that I am rcliovcd to find that what I
still have left to read is even less than three pages, but that I am glad to know
that while there are less than three pages left, at least there is 8lUl left more than
nothing. This, then, is an approving "still." Correlatively, the di.'^ip|)roving
"only" of ♦— 3 implies that neither the (juantity on hand nor its proximal*;
extinguishment is approved of. In other words, two distinct affective judgments
arc involved, that of the grade itself and that of the goal of its imi)lied tendency.
How can we distinguish the +^3 of growing fatigue from the *— 3 of growing
disappointment? Obviously we must have some way of indicating the affect
attaching to the factual goal, which gives the whole grading process its signifi-
cance. We shall therefore use a symbol for limit of tendency, | , in which q

stands for any quantity, and express the four affective types of explicit dynamic

decreasing "less than" as follows:
/--^

1. I <— ^3 (both decreasing quantity and zero-limit are approved: "I have
^ only [a little] less than 3 pages [still] to read"

2. I <-=^3 (quantity disapproved, zero-limit approved: "I have slill to read

[only a little] less than 3 pages," "I have hardly less than 3 pages
^ [stili] to read"

K-^3 (quantity approved, zero-limit disapproved: "I still have for read-
ing [but a little] less than 3 pages")
r-^3 (both decreasing quantity and zero-limit are disapproved: "I
have merely less than 3 pages left for reading")
The four affective types of explicit dynamic decreasing "more than" are as

follows:
r

1. I 3'^<— ("I have only [a Httle] more than 3 pages [still] to read," "I have
^ hardly more than 3 pages [still] to read"

2. I 3^<— ("I have still to read more than 3 pages")

3. ] 3 <— ("I still have for reading more than 3 pages")

4. J 3'+— ("I have merely [a little] more than 3 pages left for reading,"

"I have hardly more than 3 pages left for reading")
And the four affective types of explicit dynamic decreasing "etjual to" or 'jis
much as" are as follows:

r S-

1. I 3 ("I have only [no more than] 3 pages [still] to read")

r ♦_

2. I 3 ("I have still to read [no less than, as nuich as] 3 pages")
-^ ^

3. I 3 ("I still have [no less than, as nuich as] 3 pages for reading")

4. I 3 ("1 have merely [:us much as] 3 pages left for reading")

Needless to say, analogous distinctions are to be made for the other grading
cases. Here, as in every other phase of linguistic inquiry, we find that the more

466 General Linguistics I

112

closely we study actual linguistic forms, the more we are driven to realize that
they nev^er express merely static, affectively neutral, concepts and judgments,
but classes of concepts and judgments in which nuclear notions, capable of logical
definition, are colored by unavowed dynamic and affective determinants. These
determinants must be laboriously ferreted out and set in their own configuration
of possible scale or types, so that the nuclear notions themselves may stand out
^vith logical rigor. Certain of these dynamic and affective determinants are
primary or typical, because arising naturally in experience; others are complex,
invoking a blending of features in logically permissible but psychologically
atypical form, as when a logically static concept is blended with a dynamic im-
plication and two opposed affects. So far as the primary, maximally natural,
blends of dynamic tendency and affect with logically static grading concepts are

concerned, we have probably to reckon with the following five types :

1. "More than" of growing exhilaration: q^ |

2. "More than" of growing fatigue: <\-^ I

3. "Less than" of growing relief : j <-^q

4. "Less than" of growing disappointment: | <^q

'Equal to" of balanced satisfaction: < ^ ^>

The neutral, logical, "more than" is probably derived from nos. 1 and 2 by
progressive elimination of upward tendency (stage 1 : q -^ ; stage 2 : q i ; stage
stage 3: q <) and affect; the neutral, logical, "less than" is probably derived from
nos. 3 and 4 by progressive elimination of downward tendency (stage 1: *— q;
stage 2: | q; stage 3: < q) and affect; the neutral, logical, "equal to" is probably

derived from no. 5 by eUmination of balancing (stage 1: a A q; stage 2: a = q)

q •(-^ a
and affect (stage 3 : a = q) . Once the kinaesthesis and affect are rooted out of
the psychology of grading, the human spirit is free to create richer and more
complex meanings by recombining the elements of grading, of direction, of move-
ment, halt, and status, and of immediate and prospective affect, into novel
configurations in which inhere conflicts that have been reconciled.

8. THE SUPERLATIVE

"More than," "less than," and "equal to" are the most general grading terms
and concepts we have. Owing to our habit of thinking of such triplets as good-
better-besl, bad-worse-worst, famous-more famous-most famous, and famous-less
famous-least famous as possessing a logical structure which is analogous to their
linguistic form, we tend to consider the concepts expressed by most and least
as of the same nature as more than and less than. A little reflection shows that

/■'ivr: Studies in Ihiivcrsul C'lmccpntiil ( ininunar 467

113

this feeling is an illusion and that the linguistically suKgestod [iroportion good:
better — better'.best is logically incorrect. If a, b, and c are arranged in a M'rie.s of
relative (jualities, a may he said to be "good," b "Ix'tter than" a, and c "Ix'tter
than" b. But c is just as truly "better than" a as it is "lj«'tter than" b, in fact
more unreservedly or a fortiori so. We cannot say that c is "lx>.st" unless we
know either (a), that a, b, and c are the only memlx-rs of the series that are to \)c
graded, in which case c is "best" not because it is bett^'r than b as well as bettor
than a but because there is no other member of the series which is l)etter than it;
or (b), that the quality possessed by c is equal to that grade which is known not
to be exceeded by any other possible member of the whole class of gradahle mem-
bers. In the former case c may soon cease to be "best" as other memlx'rs (d, e,
f, • • •, n) are added to the series, though it always remains "Ix'tter than" certain
other fixed members of it. In the latter case c remains "best" throughout.
These two meanings of the superlative form are really quite distinct, though they
are often confused linguistically. Type (a), e.g. the most- ■ ■ of them, the leaf<t- ■ •
of them, the farthest of them, the best of them, the nearest of them, the worst of them,
may be called the "conditioned superlative" or "relative superlative." The
other type, (b), e.g. the mast- • ■ possible (= as-- as possible), the lecust- ■ - possi-
ble, the farthest (possible), the best (possible), the nearest (possible), the worst (possi-
ble), may be called the "unconditioned superlative" or "absolute superlative."
Both represent unique grades, though in differently ordered contexts, at the upper
or lower end of a series. If we characterize a class of individuals, say as "good,"
the criterion of membership, good, applies to all; better (or less bad) applies to all
but one of the class, which is thought of as least good (or worst) ; less good (or
worse) applies to all but one of the class, polar to the member excluded from the
sub-class "better," which is thought of as 6^5^ (or least bad); best (or least bad)
applies to only one member, the extreme of the sub-class "better"; and least
good (or worst) appUes to only one member, the extreme of the sub-class "less
good."

Whether the terms worst and least bad properly apply to any of the meml)ers
of the class depends, of course, on whether good and bad are thought of a.s mu-
tually exclusive classes separated by a normative line of division (logical grading:
A, I, with reference to norm) or as relative terms applying to the "more" and
"less" of a single class (logical grading: A, II, with reference to terms of com-
parison). Hence arise certain ambiguities in the ust> of least. Least good may
either mean the least good of good individuals, i.e. the first grade toward "U'st"
beyond the dividing line of neutraUty, as when we say "The least good, if gwKl at
all, will do"; or, more naturally, the least good of good and fxul indiiidiuds. i.e.
the worst, as when we say "77/c least good is indistinguishable from the worst."
Similarly with least bad, except that here it is the normative usage that seems the
more natural. Correlative ambiguities, though le.ss eiusily, may ari.>^» for most.
Paradoxically enough, language so handles hast and 7twst that least goml (of good
ones) and least bad (of bad ones) are often next door to each other, though least
good and least Ixul may in other contexts be \wh\v extremes, while best and xrorst
are typically polar extremes. The set best (of bad ones) and worst (of good ones)

46S

General Linguistics I

114

is not generally thought of as a natural neighborhood. It is only in "open-
gamut grading" (psychological grading: B, 1) that least and most can be defined
as identical concepts arrived at by opposite movement of grading (farthest =
least near, nearest = least far). We may conveniently speak of "open-gamut
superlatives" (of which there are only two possible in the unconditioned type,
namely most and least, e.g. best and worst) and of "closed-gamut superlatives"
(of which there are typically four in the unconditioned type, e.g. best, least good,
least bad, worst; or any higher even number, depending on the nature of the
grading).

It is interesting to note that the superlative form is often used to denote a
high grade, but not necessarily an apical grade, of the graded quality. Thus,
Latin amatissimiis means not only "most beloved, the most beloved" but also
"greatly beloved." Similarly, we say in English, "He had a most pleasing
personality," i.e., not '^the most pleasing personality" among some implied
number of individuals but simply "a very pleasing personality." It is probable
that this logically unreasonable, but psychologically somehow inevitable, usage
is due to a transfer of conditioned superlatives (type a) to the grading gamut in
which unconditioned superlatives (type b) occur as polar points. The following
diagram illustrates the process for unnormed grading:

the worst better worse the best

(type a,
conditioned superlative :)

the best

(type b,
unconditioned superlative :)

very bad very good

In other words, a conditioned superlative, true of some limited range of instances,
becomes, when seen in the wider perspective of all possible instances, not a true
superlative at all but an up-graded or down-graded comparative fixed at some
point psychologically near the unconditioned extremes. This process at the
same time involves a translation of explicit superlative grading into implicitly
quantified grading, a more sophisticated type of grading judgment. Hence, to
reverse the direction of transfer, it seems natural, because psychologically archaic,
to see such judgments as "very bad" or "very good" as conditioned superlatives
in an imaginary series in which all other graded terras fall below. It is as though
one felt that what is merely "very good" in this context or the context of all
values is actually "the best" in some other imagined context.

9. POLAR GRADING

At first sight it seems that the differences between explicit and implicit grading
can not be carried out for the superlative. But there are, as a matter of fact,

ri\c: Siudics in Universal Cotucpiuiil (I riinmiur AM

115

quite a number of implicitly superlative terms which have, however, thin lin-
guistic and psychological peculiarity, that they are not felt afl end jMjintw of a
graded series but as points of polar normality. These outer pointH, though
logically arrived at by the cumulative grading process that gives uh "most"
and "least," are not, psychologically speaking, worked up to via "more than"
but can only be fallen short of via "less than." If, for instance, a Heries

a, b, c, • • • , k, 1, m, n

is graded via increments of "more than" up to n, "the most," and we then accept
this n as a new norm, we note: 1, that there can be no up-graded terms which are
"more than" n; 2, that such terms as c, • • • , 1, m, which could in the first instance
be defined as progressively "more than" such lower terras as b, ■ • •, k, 1, respec-
tively, can now only be defined in an opposite sense as progressively "less than"
the unique term n. We thus arrive at what amounts to a new type of grading,
which we may term "polar grading."

A good example of a transfer from ordinary grading to polar grading is shown
in the following normed scheme:

bad

good — > better -^ better — > the best (possible)

far from still less less (than) j

perfect perfect perfect perfect

1 m

Observe that the "less perfect" of B is really as illogical as "more perfect" would
be. It may be considered an eUipsis for the logical "less than perfect" or "less
nearly perfect" based on a secondary extension of the range of meaning of the
term "perfect." The superlative implication of "perfect," which should make
of it a unique and ungradable term, tends to be lost sight of for the simple reason
that it belongs to the class of es.sentially gradabie terms (e.g. "good"). Such
terms as "less perfect" are psychologically blends of unicjue terms of the type
"perfect" and graded terms of the type "less good." The polar term is stretched
a little, as it were, .so as to take in at least the uppermost (or nethermost) segment
of the gradabie gamut of reality. Observe that at the worst the term which is
farthest in significance from the unique value of t!>e jxtlar t<'rin under which it
comes does not ordinarily relapse into the normal area of the term which im-
plicitly underlies this polar term. Thus, "least perfect" is generally U'tter than
the merely normal "good," e.g. "the least perfect of these poems," which could
hardly he said of a poem that did not belong to a set of pixnus which could he
described, most of them, as "jxTfect." On the other hand, a complication arises
when we fix the polar point not so much objectively as on the basis of a de.'^ired
upper norm, as when we say "even the least perfect of God's creatures." which

4 70 General Linguistics I

116

is a way of saying "even the worst of God's creatures, of whom we would all were
perfect." As a result of such affective interferences, polar terms may be second-
arily graded down (or up) to their polar contraries.

"Perfect" is perhaps the best example of a polar term. "Complete" and "full"
are others of the implicitly up-graded type; "empty" and "barren," of the im-
plicitly down-graded type. Implicit superlatives and polar grading offer many
psychological subtleties, of which we have only touched the more obvious.
Through the habit of using polar terms only to indicate some measure of falling
short of their proper significance they may finally take on a less than polar func-
tion. Thus, "perfect" comes to mean to some people, and to all people in certain
contexts, merely "very good." This paves the way for the secondary grading of
polar terms in a positive direction, e.g. "more perfect" and "most perfect."
Logically such terms might be interpreted to mean "more nearly perfect" and
"most nearly perfect" (conditioned superlative with polar goal); actually, that is
psychologically, they denote rather "better" and "best" in an upper tract of
"good."

This paper was finished, in essentially its present form, many years ago as part
of a larger study carried on, in collaboration with Professor W. Collinson, for the
International Auxiliary Language Association (I ALA). My original purpose
was to carry the analysis of grading considerably further but it seems best to
offer this fragmentary contribution to semantics in the hope that others may be
induced to explore the sadly neglected field of the congruities and non-congruities
of logical and psychological meaning with linguistic form.

My thanks are due the lALA and Mrs. Alice V. Morris for permission to pub-
lish this paper here. I am also indebted to Mrs. Morris for her careful reading
of the manuscript and for a number of critical observations from which I have
profited greatly.

E. S.

Editorial Note

Posthumously published in Philosophy of Science 11:2 (March 1944), 93-116.
[Reprinted in Edward Sapir, Selected Writings in Language, Culture, and
Personality. Edited by David G. Mandelbaum. Berkeley: University of
California Press, 1949, pp. 122-149]

Section Six

Patterns of Language in Relation to History
AND Society (1931-1933-[I947|)

Inlrcxiuction:
Sapir's General Linguistics in ihc 1^^3()s

TTiis section contains articles written hy Sapir in the earl\ 193()s. I1ic last paper.
"The Relation of American Indian Linguistics to Cieneral Lingmstics. " was uni-
ten alter l^-)33. hut did not appear before Sapir's death; it was pubhshed poslhuni-
ously from a manuscript contained in the Boas collection m the l.ibrarv of the
American Philosophical Society.

Although this section contains papers of di\ergent scope and purpose — three
of the papers were written for the Encyclopaedia of i/ic Social Sciences, one was
written for a handbook on methods in social sciences, another item is the synop-
sis of a lecture given by Sapir, and the paper on the psychological reality of
phonemes appeared in a special issue of the Journal dc P.sycholoi^ic. intended to
show the manifold connections between linguistics and psychology— there are a
number of unifying features linking these papers:

(1) There is first the o\erall presence oi the notion t)f ■patterning" in language,
which Sapir spots at various levels: that o{ the insertion of language within
society, that of its ties with the indi\ idual. then also the level of patterning of lan-
guage throughout history, and, basically, the level of language structure itself.

(2) The papers in this section also testify to Sapir's broadening of linguistics as a
social and cultural science; they are thus in perfect harmon\ with Sapir's publi-
cations of the same period on socit>Iogy and on the psychology of the indi\ idual.

(3) As publications in general linguistics the papers also testify to the ci>nlmuil\
of the ideas developed already in Sapir's book lAin^iiage (1921 ): it is striking to
see how the gist of that book, written by Sapir in his mid-thirties, remained intact
throughout Sapir's later writings, albeit enriched with new data and integrated in
a much more comprehensive view on the social and psychological status o{ lan-
guage (see Sapir's writings included in I'lic Collected Works oj i.dward Sapir, \ol.
III).

(4) A final recurrent feature is the emphasis put on methodological aspects o\ hn-
guistic work, and especially the need to check generalizations about language
through a careful, objective analysis of languages belonging to lamilies other
than the one(s) the linguist may be acc|uainted with.

' This special issue t)f the Journal de psyduilo^ic normalc el iuilli(>l(>i;iiiuf was devnted lt> ■■PsychiiK>gic du Ian-
gage": it is divided into six sections: "Tiieorie du langage". "Linguistique generale". "Systi'me materiel du Ian-
gage", "Systeme lormel du langage". "Acquisilion du langage" and "Palhulogie du langage" Ilie list »>( contrib-
utors includes, apart from Sapir. Iirnsl (assirer. Albert Sechehaye. \\ili>ld Doros/ewski. Karl Huhler. Aniomc
Meillet. Joseph \endryes. I'lero Meriggi. Viggo Hrondal. Antoine dregoire. Marcel (tihen, Nikola) Iruhct/k«»y.
Jacques van Ciinneken. Alf Sommerlelt and Otto Jespersen. Sapirs tcvi is iiuiudid in du- section "Svsteme
materiel du langage".

-The first classical statement by Sapir was in his \'-)2> article "Sound Patterns in [.^mguage '. laiigujfic 1 (1925).
37-51 (reprinted here in section III).

474 General Linguistics I

In "The Concept of Phonetic Law," written for an audience of social scientists,'
Sapir discusses a methodological problem. The topic is a classic theme of histor-
ical-comparative grammar and modern linguistics, viz. the notion of phonetic law.
From the very outset. Sapir rejects the older naturalistic conception of phonetic
laws as absolute regularities:^ phonetic laws are to be seen as a posteriori gener-
alizations, the validity of which is limited in time and space. They are "laws of
formulas" (p. 297; here, formulas should be understood as correspondences rela-
ting linguistic forms).' Phonetic laws are set up by inference; they are based
on observed regularities (and on the general assumption of the regularity of
change*'), and express a transitional directionality (p. 298).^

The purpose of Sapir's article'' is to show that the type of phonetic laws posit-
ed by scholars in the field of Indo-European, Semitic or Finno-Ugric languages
is also attested in "primitive" languages, such as the American Indian languages.
The general implication is that the comparative method as used in Indo-
European or Semitic comparative grammar is equally valid for American Indian
languages. Interestingly, the exemplification'^ is first drawn from the work of the
other key figure of American linguistics in the first half of the 20th century,
Leonard Bloomfield. Bloomfield's work on comparative Algonkian"' is presen-
ted here from the point of view of its methodological interest, which lies in the
establishing of proto-forms through recursive (and "crossed") triangulation of
forms from parent languages/dialects. This technique normally leads to positing a
maximal number of proto-forms in order to account for split correspondence sets
(reconstruction favours "backwards" sphtting rather than merging, unless there

■'' The text was written in 1928, and revised by Sapir in 1929.

'' See Gisela Schneider, Zum Begriff des Lautgesetzes in der Sprachwissenschaft seit den Junggrammatikern
(Tubingen. 1973): Terence H. Wilbur ed.. The Lautgesetz-Controversy: A Documentation (1885-1886)
(Amsterdam, 1977) (with further bibhography).

5 See p. 298: "phonetic formulas which tie up related words." Elsewhere in the article (p. 302) the term formula
(■'or tag") is used in a different sense, viz. as a reconstructed form in a pattern. Note that the classical statement
by Antoine Meillet, La methode comparative en linguistique historique (Paris, 1925, esp. chap. VIII), equates the
diachronic laws with "formules generates de changement."

6 See p. 302. This is the general problem of "induction on induction."

' In the absence of external historical evidence, the directionality of a sound change can be stated on the basis
of principles of panchronic phonology: such a principle is mentioned here by Sapir (p. 298: "stopped" conson-
ants, i.e. stops or occlusives, more often become spirants than vice versa).

« Sapir's article and his comparative method are discussed, from the point of view of Athabaskan and Algonkian
comparative linguistics, by Michael Krauss ("Edward Sapir and Athabaskan Linguistics") and I. Goddard
("Sapir's Comparative Method"), in New Perspectives in Language, Culture and Personalitv: Proceedings of the
Edward Sapir Centenary Conference (Ottawa, 1-3 Oct. 1984), edited by William Cowan, Michael K. Foster and
Konrad Koerner (Amsterdam/Philadelphia, 1986), pp. 147-190 and 191-214.

■^ A few years earlier Sapir had discovered a phonetic law in Chinook: see "A Chinookan Phonetic Law",
International Journal of American Linguistics 4 (1926), 105-110 [reprinted in The Collected Works of Edward
Sapir, vol. VI, pp. 275-280].

1" Another major contributor to the comparative grammar of Algonkian languages was Truman Michelson;
Sapir, however, was not a great admirer of Michelson's work.

.SV.v." I.tini^iiiii^c in Rclaiion i<> llisiorv and Society 475

is clear evidence ol anak)giciil processes). ' Hie recoiisirucled pri)li)-U)rnis ihal
Bloomtiekl had postulated were confirmed by the data he later loiind in the
Swampy ("rec dialect ol Manitoba.

Sapir then iii\es a parallel example from his own work m .Athabaskan. \shich
concerns the treatment ol initial ctinsonants. Here also the reconstruction posit-
ed on the basis ot the application ol the prmciplc ol "phonetic law" was lalcr
confirmed by Sapir's fieldwork on the Hupa languaue (pp. 30.^-305).

The demonstration given is. of course, confined to examples of diachronic
methodologv. but Sapir's concluding remarks extend beyond the field of histor-
ical linguistics: apart from cniphasi/ing that language jxitlerns fit \siihin cultural
beha\'iour (p. 306). ' he also points out that what is fundamental in language and
in linguistic description is the '■patlernr not the "materiar" (or "content"). ' I"he
linguist should therefore be interested in how changes affect the pattern of lan-
guage; and Sapir approvingly quotes Bloomfield who wrote that what we call
"sound change" is in fact a statement about phonemic change.

"Conceptual Categories in Primitive Languages" is the synopsis of a lecture
given by Sapir at the meeting of the National Academy of Sciences in New
Ha\en. 16-18 November 1931. This short abstract testifies to Sapir's comprehen-
sive definition of language" as a symbolic organization, serving to express
individual and social experiences in a cultural setting. As far as one can judge
from the abstract, the lecture stressed the methodological value of comparing
languages "of extremely different structures,"'" and the need for relativizing our
conceptions concerning the universality of categories familiar to us.

The specific value of this published abstract lies in its succinct formulation
of what later has been called the "Sapir - Whorf hypothesis" on the relationship

" Sec tabic I, with 5 different clusters for "PrimitiNe C entral Aigonkian" (completed with table \).or table Hi

'- See also Leonard Bloomfield. I Aini;iuii;c (New York. 1^.^.^). pp. 359-.^6t) (in the chapter on "Phonelic chan-
ge") and liloomfields article "Algonciuian". in Harry Hoijer <7 <;/.. Linf^uistic Striulures of Native Amencu (New
York, 1946). pp. S.S 129 (iiflO. willi self-correcting note 10).

'^ See also E. Sapir. "nie Unconscious Patterning of Behavior in St)ciely". in t thel S Dummer ed.. Tin-
Unconscious: A Symposium (New York. 192S), pp. 1 14 142 [reprinted in //;<• Collccud Works of Edward Sapir.
vol. III. pp. l.Sft-l 72]. For a very explicit demonstration, see K.I.. F'ike. I tmi:uii\;c in Rchiion lo a I tufted Theory
of (he Siruclurc of Humun Hchavior (Santa Ana. 19.S4 19W). 3 parts: reedition: Ilie Hague. 1*>67). For an ana-
lysis of Sapir's view of patterning in language and in culture, see Maria Xose lernande/ Ca.sas. Ldward Sapir
en Id hni^iiisiicii aciuiil. Linens de conlinuidud en la hisioriu de la linf-iiixliea (Verba. Ancxo .M) (Santiago dc
Compostela. 2(K)4). pp. 191-20.^ and 2.U-2.^7.

'•* As noted by Sapir. languages may be different in their sounds, but similar or even identical in pattern (p. .VM.
with reference to Hupa and Chipewyan). For a similar remark with application lo the conser>ation of a pattern
in the history of a language (Old High (iernian). see Sapir. l.ani:uai;e (New York. 1^21). pp. 1*M-I9.S.

'■'' See also his article "Language" (19.'^.^) |rcprinlcd in this scclionl

'" See also -Language" {\^)^^) and • ITie Relation of .American Indian Linguistics to ( ieneral Linguistio (1''47)
(also reprinted in this section).

476 General Linguistics I

between language and world view.'^ Here this hypothesis is formulated in terms
of the elaboration, within a language, of categories originally abstracted from
experience, and the subsequent imposition of these elaborated forms upon "our
orientation in the world." As such, each language elaborates its own "system of
coordinates," and Sapir likens the incommensurability of languages to the incom-
mensurability of divergent (geo)metrical systems. The notion of language as
form seems to have been pervasive in the lecture: in the abstract we find terms
such as "formal completeness," "self-contained [...] system," "formal limitations",
and "linguistic form." In his lecture Sapir seems to have depicted language as a
"mathematical system."

This extremely concise text is a very important testimony on Sapir's "relativ-
istic" view of the relationship between experience of the world and symbohc
expression, and his characterization of language as a formal structure and a sym-
bolic device.

The three articles written for the Encyclopaedia of the Social Sciences included
in this volume'* exemplify the continuity of Sapir's thought, when compared with
some classic statements in his Language (1921), but at the same time they also
mark the evolution of Sapir's conception of language as related to culture, society
and personality.

The article "Language," which opens with a remark on the universality of the
phenomenon of language,''^ first defines language as a system of phonetic symbols
for expressing experience.''^ Sapir then presents the systemic properties of all lan-
guages: their phonetic, phonemic and morphemic structure or patterning.^' The
distinction between phonetic elements and phonemes, between distinct sounds
and distinctive sound classes is clearly drawn (pp. 155-156). The levelled pattern-
ing of language leads then to a definition of grammar as a system of formal
economies; as noted by Sapir, all languages have form,^^ precisely because of
this organized economy of patterns. Further on, he compares language structures
to "quasi mathematical patterns,"'- which combine with expressive patterns
(p. 158).

'^ See note 30 in the introduction to section III.

'8 Sapir also contributed articles on "Custom," "Fashion," "Group," "Personality" and "Symbolism" to the
Encyclopaedia of the Social Sciences [these articles are all reprinted in The Collected Works of Edward Sapir,
vol. Ill (section I)].

'^ See E. Sapir, Language (New York, 1921), pp. 21-23.
^^ See Language, o.c. p. 7.

21 The term "morphology" is used (p. 156). On Sapir's use of morphological criteria in typology, see Stephen R.
Anderson, "Sapir's Approach to Typology and Current Issues in Morphology", in Contemporary Morphology,
edited by Wolfgang Dressier, Hans Luschutzky, Oskar Pfeiffer and John Rennison (Berlin/New York, 1990),

277-295.

22 See also Language, o.c, pp. 132-133.

23 As in "Conceptual Categories in Primitive Languages" (1931) [reprinted in this section].

Six: I.dHi^udi^i' in Relation to History iiml Socictv All

Sapir ihcii proceeds lo deline language as a scll-sutlieieiil (p. l."S<). p. 15.S) sym-
bolic system, which interpenetrates with experience (p. 157): the relationship be-
tween speech and action, which was hardly discussed in Lun^iuiin' ( 1921 ), is given
due weight here, a fact which icstilies lo Sapir's interest in the svmbolic nature of
language, an increasingly prominenl theme in his writmgs following his readmg
and discussion of Ogden and Richardss ///<- Mauun^ of Meiminj^.-^Wxc section
on the definition of language thus includes a discussion of the formal character-
istics of language (= its systemic nature) and its psychological characteristics
(symbolic system; relation to experience; language as the carrier of qualified
expressiveness; the possibility of substitution by secondary systems).

Tlie next section deals with the origin of language.' a problem which according
to Sapir remains unsolvable from a linguistic point of view, in his brief survey of
views and theories proposed in the past, Sapir notes that the basic question of
how language emerged has never been answered satisfactorily. He outlines his
own view of the evolution of language as a product of a peculiar (symbolic) tech-
nique.

The third section deals with the functions of language. Here Sapir discusses the
relationship between language and society (language as a force of sociali/alion.
and a factor of cohesion, solidarity and intimacy),^" between language and culture
(pp. 159-160), and between language and personality.

The following sections respect to a large extent the structure of Sapir's 1921
book Lani^iiage: they deal successively with language classification, linguistic
change, and the (supposed) relations between language, race and culture.

The section on language classification outlines the basic differences between a
genetic and a structural classification (pp. 161-163): the two may converge (p.
163), but they are based on very different principles. Genetic classification invol-
ves the "technique'* of comparative grammar; it has to face the problem that one
never can prove that two languages are not (ultimately) related (p. 1^3). Sapir
pays much more attention to structural classification (or typology); while the dis-
cussion here is less elaborate than in his book Laniinai;c,-' the text provides a suc-
cinct view of the parameters of synthesis, technique and expression of (relatio-
nal) concepts."* It is interesting to note that Sapir expresses some doubts about
the operational character of the latter parameter (p. 162).

Linguistic change has either inherent or external causes, but the borderlines
between the two may not always be clear. Inherent change is related to • ilritts"

-* See the reprint of Sapir's review ot llic Mcanini; of Mdininy,. in scdion III i>l this volume.

-^ See the writings reproduced here in section I and section II. and I unauagt' (New ^ork. \^)2\). pp f> 7

^^ One should note Sapir's recognition ol the •phatic function" (in lakohsons lerminology) of language; sec
Sapir's remarks on "small talk" (p. IW)).

-^ See Language, o.c. chapters IV. V and VI.

28 In Language, o.c, the parameter involves the expression of basic, derivational and relational conccpt.s.

478 General Linguistics I

in the language.- "The levels at which linguistic change operates (phonetic, gram-
matical, lexical) are briefly discussed by Sapir; the substance of his discussion is
fairly traditional, and takes up the gist of his treatment in Language, but one will
note two new features:

(a) the recognition of the role of age groups;^"

(b) the importance of bilingual subjects for language variation and change.^'
The last part of this encyclopedia article deals with language, culture, and

race. As in his Language,^^^ Sapir rejects correspondences established between
language forms and cultural forms; the form (grammar) of language is a self-
contained, unconscious system which changes only very slowly. Culture is subject
to rapid changes of fashion; its nearest linguistic match is in the vocabulary, which
reflects cultural changes. As themes of particular interest Sapir singles out taboos
and special languages. As to the possible correlation between language, race, and
culture, the article — published in 1933! — highlights (misplaced) nationalistic
beliefs (or propaganda)," and emphatically denies any correlation between lan-
guage, race and culture.^^ Sapir also notes the phenomenon of language imposi-
tion and of language cult (especially for minority languages). ^^

The article ends with a plea for an international language, a theme absent from
the book Language (1921). Its presence here reflects Sapir's involvement in the
question of an international auxiliary language; ^^ a more elaborate treatment can
be found at the end of the article "Communication," where Sapir emphasizes the
need of one language for intercommunication, while noting at the same time the

-"> In Language, o.c, pp. 157-182, Sapir uses the singular "-drift:" On this term, see Yakov Malkiel, "What did
Edward Sapir mean by "Drift" ?", Romance Philology 30 (1976-77), 622 and "Drift, Slant and Slope:
Background of, and variations upon, a Sapirian ihcmt". Language 57 (1981), 535-570, and see the literature
referred to in note 9 of the introduction to Zellig Harris's review of the Selected Writings.

30 It may be that Sapir was influenced here by the classic statement in Louis Gauchat, "L'unite phonetique dans
le patois d'une commune", in Festschrift Heinrich Morf zur Feier seiner flin fund zwanzigjahrigen Lehrtiitigkeit
von seinen Schidern dargehracht (Halle, 1905), pp. 175-232; cf. Pierre Swiggers, "Louis Gauchat et I'idee de varia-
tion linguistiquc". in Ricarda Liver - Iwar Werlen - Peter Wunderli (Hrsg.), Sprachtheorie und Theorie der
Sprachwissenschaft: Geschichte und Perspektiven. Festschrift fiir Rudolf Engler zum 60. Geburtstag (Tubingen,
1990), pp. 284-298.

3' Here Sapir may have been indebted to the work of the French Indo-Europeanist and general linguist Antoine
Meillet who in the 1910s had recognized the importance of bilingualism for understanding linguistic changes;
see P Swiggers, "La linguistiquc historico-comparative d'Antoine Meillet: theorie et methode", Cahiers
Ferdinand de Saussure 39 (1985). 181-195, "La linguistique historique devant la variation: le cas de Meillet",
Recherches sur le franqais parle 1 (1986), 61-74 and "Le probleme du changement linguistique dans Fceuvre
d'Antome Meillet", Histoire, Epistemologie, Langage 10:2 (1988), 155-166.

32 See Language, o.c, chapter X.

33 See also the article "Dialect" [reprinted in this section].

34 See also "The Relation of American Indian Linguistics to General Linguistics", and note 63 below.

35 It may be that Sapir relied here on the information provided in Antoine Meillet, Les langues dans I'Europe
nouvelle (Pans, 1918) and its revised edition [by A. Meillet and Lucien Tesnierel, Les laneues dans I'Europe
nouvelle (Pans, \928).

3^ See sections IV and V in this volume.

.S7.v; lAini^uiii^c in Rclmion i<> lltsiorv and Society 479

dangers oK di dIisUicIcs Id ihc process ol cxlciulnm the radius, and ol nicrcasnii;
the rapidity of coniniunicalion: ihc irruption into tlic pri\alc sphere, cultural
degeneration, and the desire ol non-uiulerstanding (as evidenced bv cryplo-
graphy).

The article "C'oniinunication," in which Sapir delincs societ\ as a network o\
understandings and a totality ol insiiiulit)ns grouiuied ni acts of communication.
puts forward a definition of language as a referential system, lor this reality-
directed function, languages are self-sufficient everywhere and thus equivalent. ''
Communication, as culturally patterned, involves primary processes — language.
gesture (including "intonations" "'), overt behaviour and "social suggestion " (p.
7cS) — and a number of secondary techniques. While the primary processes arc
universal, the secondary techniques are culturally linked. Sapir ideniilies three
main classes of techniques facilitating communication:

(1) transfers or transposed systems: such systems (e.g., writing or morse code)
maintain an isomorphism with a primary process-system:

(2) symbolisms used in special situations: these symboli/alions (e.g.. railrt)ad
signs, smoke signals) are more restricted in referential scope, and more depen-
dent on contiguity:

(3) physical conditions for extending communication.

With respect to the recent increase of the radius and rapidity of communication.
Sapir formulates some thoughts on the "opening up'" of the world and the dimin-
ished importance of geographical and personal contiguity or contact. In a passage
which reminds one of Heidegger's musings, ' Sapir welcomes the creation of
"new worlds,"" of sociological, cultural or technological texture (p. 80). These new
worlds correlate with new (sub)groups in society (p. 79).

Tlie article "Dialect."" written at a time when dialect studies were fUnirishing.'
hardly goes into typically dialectological issues. Sapir discusses "dialect" from the
point of view of the theoretical linguist and sociologist. From the linguistic point
of view dialects are languages, both historically" and systemically. Integrating the
sociological point of view, one can define dialects as the socialized form ol the
tendency towards variation in languages.'- This is a universal phenomenon, which
has both inherent causes (such as "drift"") and extrinsic ones (such as mixture ol
groups). Sapir mentions the popular conception of "dialect"" as being a de\iation

''^ Sec alsi) lAiiixntiifie, ox:, p. 22 and pp. 2.^3-2.^4.

^'^'Ilic tL-rm is used to rclcr to the uencral iiuidiilation of the voice, not to inlonali«Mi.iI patterns (which tvlong
to language).

^'^ See Martin Heidegger. Holzwfi^i- (franklurt. \')>[)). especially the essay •Die /eit dcs Wellhildcs" (Icxt of a
conference of 19.38).

•"• For a survey of the Romance field, see Piel Desmet Peter I auwers Pierre Swiggcrx •Dialectology.
Philology, and Linguistics in the Romance field. Methodological Developments and Inlcraclions". Bclfiian
Jounuiiof l.iniiuistics 13(1 9W ). 1 77-203.

■" See Antoine Meillet. l.cs ilialeclts inili>-fiiroin-in\ (Pans. I*)()S).

■•- See Language, o.c. pp. l.S<S-lb3.

480 General Linguistics I

or corruption; such a view ignores the fact that standard languages are nothing
but the elaborated form of a dialect that has won out'' (pp. 123-124). Dialects
often have an ancestry which is at least equal to that of the standard language (in
which case one can call them "co-dialects"). Some dialects, however, are later
developments of a standardized language (such ''post-dialects" are the result of
either language exportation or regional diversification).

Surveying the contemporary situation of dialects Sapir notes their persistence
in some areas of Europe (like Italy), and singles out their socially symbolizing
function. Dialectalization (or regionalization) has political, cultural '^ and ideolo-
gical reasons. Sapir's text reflects the then recent rise of "new nationalisms," '' but
he seems to underestimate their importance in comparison with the pressure of
the modern "realistic and pragmatic" mind (p. 125), favouring unification and
normativism. As noted by Sapir, some institutions (education, army, etc.) contrib-
ute to the process of unification.

The article ends with a remark on secondary symbolization, i.e. the sublimated
cult of a dialect as a symbol of the local group (which thus inverts its inferiority
status) and with a note on the (emotional) ties of the individual to his "dialectic
habits of speech" (p. 126).

The paper on the psychological reality of phonemes was published in French
under the title "La realite psychologique des phonemes", but first written in
English by Sapir. A carbon copy,'" corrected by Sapir, of the English source text
has been preserved. The English version published in 1949'' is slightly different
from the corrected carbon copy, and seems to be based on another (uncorrected)
copy, collated with the published French text.'** None of the divergences between
the versions is of major importance for the content.''^

This classic paper"' deals in fact with the psychological reality of morphophon-
emes rather than phonemes: the first two examples concern morphophonemic

*^ For a nice example, see R. Anthony Lodge, French, from Dialect to Standard (London, 1993). On the process
of standardization in Europe, see John E. Joseph, Eloquence and Power: the Rise of Language Standards and
Standard Languages (London/New York, 1987).

'♦^ As an example Sapir mentions the influence of Romanticism (p. 125).

'*^ These were reported upon by A. Meillet and L. Tesniere in the works referred to in note 35.

'** The original copy has not been preserved.

47 Published under the title "The Psychological Reahty of Phonemes" in Edward Sapir: Selected Writings in
Language, Culture, and Personality (ed. D.G. Mandelbaum, Berkeley, 1949), pp. 46-60. This version was reprint-
ed in Valerie Becker Makkai ed.. Phonological Theory: Evolution and Current Practice (New York, 1972), pp.
22-31.

*^ An offprint of the French text with Sapir's corrections has also been preserved. See the "Corrections to the
French version published in 1933."

*'^ Both the published French and English versions are reprinted here: both deserve their place in The Collected
Works of Edward Sapir, especially because the two papers had a different reception history and have thus found
their place within different research traditions.

50 On the adumbration of its contents in the correspondence with Alfred L. Kroeber, see Victor Golla, The Sapir
- Kroeber Correspondence. Letters between Edward Sapir and A.L. Kroeber 1905-1925 (Berkeley, 1984), letter

Six: Language in Relation to Htstorv and Sociely 481

altcrnalii)ns. the loLirlh one invDlvcs an armiiiicnl hascd on morpliDphoncmics,
and the litlh example is a case ot subphoneniic "projeelion;' onK ihc third exam-
ple exemplifies the psychological conscience ol "phonemes." I ioueser. Sapir did
not have at his disposal the notion o\' morphojihoneme (in lael. the term f)lu>n-
ctnc was at that time still relatively new to him), llie basic idea ot the paper is
that native speakers "perceive" the organization (or categorization, or calibra-
tion) of their language not in direct relation to the (material) phonetic data, but
with respect to the (underlying) functionaP' structures (and prt)cesses affecting
them).'nie judgement of native speakers betrays a "phonemic' (morphophon-
emic) intuition, and reveals their grasp of the "dynamic reality' of language. ITic
key concept in this paper is that of functional pattern (a term translated as
"fonnc^ in French), which is the basis for the psychological understanding the
native speaker has of his language.

Sapir discusses five examples drawn from his fieldwork'' and his teaching,
all showing the difference between "objective facts" and "subjective categoriza-
tion":''

( 1 ) an example from Southern Paiute, fully analysed in his grammar of Southern
Paiute.' which involves the treatment of non-initial consonants in postvocalic
contexts with morphophonemic conditioning:'"

140. p. 13y {correspondence of 28th May 1914. Idler by Sapir). letters l.S2and 153. pp. 1.5S-16()(correspt>ndcncc
oi late October 1914. letters by Kroeber). letter 208. pp. 220-222 (correspondence of Sth September 191h. let-
ter by Sapir): see also Sapir's Lan^iiatie (New York. 1921). p. 58 n. 16: "The conception of the ideal phonetic
system, the phonetic pattern, of a language is not as well understood by linguistic students as it should be. In
this respect the unschooled recorder o( language, provided he has a good ear and a genuine instinct for lan-
guage, is often at a great advantage as compared with the minute phonetician, who is apt to be swamped by his
mass of observations. I have already employed my experience in teaching Indians to wnte their own language
for its testing value in another connection. It yields equally valuable evidence here. 1 found that it was difficult
or impossible to teach an Indian to make phonetic distinctions that did not correspond ti> "points in the pattern
of his language." however these differences might strike our objective ear. but that subtle, barely audible, phon-
etic differences, if only they hit the "points in the pattern." were easily and voluntarily expressed in writing. In
watching my Nootka interpreter write his language, I often had the curious feeling that he was transcribing an
ideal flow of phonetic elements which he heard, inadequately from a purely objective standpi>int. as the inten-
tion of the actual rumble of speech."

■"' In his Laniiiiui^e (New York. 1921). Sapir did not use the term phoneme: the distinction snund phinumr is
however implicitly made. Ilie first theoretical discussion, within American structuralist linguistics 4>f the phi>-
neme is William F. Twaddell, On Dcfininfi the Phoneme (lialtimore. 19.^5).

''' As shown by Sapir at the beginning of his paper, these functional structures are "vitally" important

"•^ For a survey of Sapir's field work (languages, dates, locations and informations, sec The Collected Works of

Edward Sapir. \o\. IV, pp. 255-257.

''•* As noted by Sapir. the latter can be influenced by "etymology" (i.e. insight into the history ol the language).

^^ See E. Sapir. The Southern Paiute Lan^iuaf^e (3 parts). 19.^0-1931 |reprinled in Ihe Collected Works of
Edward Sapir. \o\. X]; see part I. pp. 45-70. Ilie inlormant Sapir refers to is loiu Tilloh.ish (a Kaibab l\nule).
who worked during four months in 1910 with Sapir in Philadelphia (in UXW Sapir had worked with the
llncompahgre and Uintah lJtes):see Catherine S. Fowler - Don I). Fowler. "Fdward Sapir. Fony TilU>hash and
Southern Paiute Studies", in New Perspectives in l.an^uai;e. Culture and Personality: Proceedings of the Edward
Sapir Centenary Conference (Ottawa. /-.? Oct. NS4). edited by William (\iwan. Michael K. Foster and Konrad
Koerner (Amsterdam/Philadelphia. 1986), pp. 41-65.

-''*' Sapir draws a parallel with morphophonemic processes in Old Irish (spiranti/ation, na.sali/atiun. ...).

482 General Linguistics I

(2) an example from Sarcee, involving a case of homophony (or merger) of two
stems (one of them to be described as ending in morphophonemic °T\ of which
the dental consonantal element can be omitted in certain contexts; Sapir speaks
here of "consonantal latency");

(3) & (4) are two examples from Nootka, based on Sapir's fieldwork with Alex
Thomas. The first example concerns cases of phonemic discrimination as opposed
to phonetic reality; the second example concerns the (phonemic) homogeneous
treatment of postglottalization (of stops and affricates) and preglottalization (of
nasals and semivowels) by the informant. Sapir provides three reasons for such
a phonemic homogenization, the first and second referring to phonotactic condi-
tions, while the third involves morphophonemic considerations.

(5) the last example concerns a case of "projection" of subphonemic distinctions
from one's mother language to another language (with a different patterning).
Here the "psychological reality" becomes an instance of "illusion" (and inaccur-
ate notation).

In "The Relation of American Indian Linguistics to General Linguistics"'^
Sapir wants to give an "object lesson in linguistic methodology" (p. 3).'*^ The
experience acquired in Americanist work is used as a corrective against general
conclusions often drawn on the basis of data limited to Indo-European and
Semitic.''* At the same time, the paper can be read as a tribute to Boas, who had
warned his young student Sapir against hasty generalizations, and had categor-
ically rejected equations too easily posited between language, race, and culture. In
the present paper Sapir seems inclined to give more weight to the "diffusionist"
view (held by Boas) than he does in his Language^ or in his reconstructivist work
on American Indian languages, but this may be explained by his primary concern
of correcting some preconceived ideas of linguists working exclusively within one
linguistic family (Indo-European or Semitic).

This short paper is structured as follows. First, Sapir shows the methodological
value of American Indian languages for general linguistics, with examples taken
from the domains of phonology and morphology. For phonology, the American
Indian languages are referred to in order to show that the distribution of phon-
etic elements is not necessarily tied up with genetic affiliation; the languages of

-•'^The text must have been written after 1933 (reference is made to Leonard Bioomfield's Language, published
in 1933), but probably not much later.

•■'« Sapir also speaks of an "object lesson" for a "general theory of historical phonetics" (p. 2), of correcting "pos-
sible misconceptions" (p. 3), and ends his paper with a general word of caution: "These are but a few out of
hundreds of examples of what may be learned from American Indian languages of basic linguistic concepts, or
rather of the grammatical treatment of basic concepts. There is hardly a classificatory peculiarity which does not
receive a wealth ot illumination from American Indian languages. It is safe to say that no sound general treat-
ment of language is possible without constant recourse to these materials" (p. 4).

59 Sapir also uses the term "Hamitic-Semitic" (p. 2, p. 4); see also the introduction to section III.
^' See Language (New York, 1921), pp. 219-220.

.S"/a; I.dni^iKi^r in Relation m Uisiorv und Society 483

the North-Wcst Coast testily to the arcal spread of teatures transcending and dis-
rupting genetic ties."' A similar example is given lor morphology: here Sapir uses
examples from liokan and Athabaskan to shou that morphological "re-forma-
tions" can occur owing to contact, but that there are also languages which arc
extremely conser\ati\e and resistent to foreign intrusions.

Alter the section on the "corrective methodological value." Sapir prt)ceeds
to demonstrate the importance of the intrinsic analysis of American Indian lan-
guages, which display a great variety of types.' Here again, the paper turns into
a lesson of methodology: given the high degree of morphi)l()gical disergence
within a "relatively homogeneous race." the American Indian languages are a
clear illustration of the non-congruence of language, race and culture.' "lliis
means that American Indian linguistics stands as a silent refutation of those who
try to establish an innate psychological rapport between cultural and linguistic
forms" (p. 3)."'

Serving as a "test field" for "solid linguistic thinking." the data obser\ed m
American Indian languages should prevent the linguist from unjustified gener-
alizations or universalizations, or other unwarranted statements (e.g., concerning
the world-wide attestation of nominal classifications, of specific grammatical
categories or historical processes). What Sapir wants to make clear here is that
the study of how "basic concepts" are grammatic(al)ized (i.e. integrated in the
formal system of grammar) should be based on an extensive (typological) in-
spection of the world's languages.

The paper ends with remarks on the practical (and personal) value o\ field
work; Sapir insists on the indispensable experience of familiarizing oneself,
inductively (p. 4), with native languages. American Indian languages can thus ful-
fil the role of an eye-opener and may help the linguist in getting "a thoroughly
realistic idea of what language is" (p. 4).

Pierre Swku.i rs

''' Sec also l.iinfiua^c. o.c. pp. 21 1-21.1.

''^ Sec also E. Sapir - Morris Swadesh, 'American Indian drammatical Categories". Word 2 (1946). 1().V|I2
[reprinted in riw Collccicil Works of Edwuni Sapir. vol. V, pp. 1.1.^-142: this paper was begun by Sapir around

1^)24|.

""■ This is also the bottom-line of chapter 10 of Sapirs Laniiiuim' (New York. 1^21 ): this chapter was reprinted
in Victor Francis Calverton ed.. llw Xtukinfi of Man (New York. l^?>\ ). pp. 142-l.'>ft.

'^ Sapir uses examples taken from the North-Wcst Coast, from the Plains culture and the Pueblo culture. In his
l-dniiiidf^e. o.c. pp. 22H-22^). he had given examples taken from Californian tritvs. and from the different cul-
tures to which speakers of Athabaskan languages belong.

484 General Linguistics I

ANALYSIS 21

THE CONCEPT OF PHONETIC LAW AS TESTED IN PRIMITIVE
LANGUAGES BY LEONARD BLOOMFIELD^

By EDWARD SAPIR'
University of Chicago

A large part of the scientific study of language consists in the formula-
tion and application of phonetic laws. These phonetic laws are by no
means comparable to the laws of physics or chemistry or any other of
the natural sciences. They are merely general statements of series of
clianges characteristic of a given language at a particular time. Thus,
a phonetic law applying to a particular sound in the history of English
applies only to that sound within a given period of time and by no means
commits itself to the development of the same sound at another period
in the history of English, nor has it anything to say about the treatment
of the same sound in other languages. Experience has shown that the
sound system of any language tends to vary slightly from time to time.
These shifts in pronunciation, however, have been found to work accord-
ing to regular laws or formulas. Thus, the / of the English word father
can be shown by comparison with such related languages as Latin, Greek,
and Sanskrit to go back to an original p. The change of p in the original
Indo-European word for "father," reflected in the Latin pater and the
Greek pater, is not, however, an isolated phenomenon but is paralleled
by a great many other examples of the same process. Thus, foot corre-
sponds to Greek pous, genitive podos; five corresponds to Greek pente;
full corresponds to Latin plenus; and for is closely related to Latin prd.
A comparison of English with certain other languages, such as German,
Swedish, Danish, Old Icelandic, and Gothic, proves that these languages
share with English the use of the consonant / where other languages of
the same family which are less closely related to English than these have
a p.

Inasmuch as such languages as Latin, Greek, Sanskrit, and Slavic differ

A Set of Postulates for the Science of Language," Language: Journal of the
Linguistic Society of America, II (1926), 153-64; "On the Sound-System of Central
Algonquian," ibid., I (1925), 130-56; "A Note on Sound-Change," ibid., IV (1928),
99-100; also E. Sapir, MS materials on Athabaskan languages.

[In preparing this analysis. Professor Sapir was invited to discuss his own work
at length because of its similarity to the work of Bloomfield.— Editor.]

297

Six: LdHi^nd^c in Rclaiion to History uml Socwtx 485

298

among themselves about as much as any one of them diflcrs from the
Germanic grouj) to which English belongs, it is a fair assumption that
llicir concordance is an archaic feature and not a parallel development,
and that the / of English and its more closely related languages is a
secondary sound derived from an original p. This inference is put in the
form of a phonetic law, which reads: "Indo-European p becomes Ger-
manic /." The change cannot be dated, but obviously belongs to at least
the period immediately preceding the earliest contact of the Germanic
tribes with the Romans, for in all the Germanic words and names that
have come down to us from the classic authors this change is already
manifest. It is important to realize that two distinct historic facts may
be inferred from such evidence as we have given, which is naturally but
a small part of the total evidence available. In the first place, the change
of /) to / is regular. In other words, we do not find that in one corre-
spondence / is related to p while in another correspondence / seems to
parallel w or b or some other sound. In the second place, the general
consensus of the Indo-European languages indicates that the change has
been from p to f and not from / to p. Incidentally, this is in accord with
general linguistic experience, for stopped consonants more often become
spirants (continuous "rubbed" consonants) than the reverse.

Such phonetic laws have been worked out in great number for many
Indo-European and Semitic languages. There are obviously many other
historical factors that contribute their share to the phenomena of change
in language, but phonetic law is justly considered by the linguist by far
the most important single factor that he has to deal with. Inasmuch as
all sound change in language tends to be regular, the linguist is not
satisfied with random resemblances in languages that are suspected of
being related but insists on working out as best he can the phonetic
formulas which tie up related words. Until such formulas are discovered,
there may be some evidence for considering distinct languages related —
for example, the general form of their grammar may seem to provide
such evidence — but the final demonstration can never be said to be given
until comparable words can be shown to be but reflexes of one and
the same prototype by the operation of dialectic phonetic laws.

Is there any reason to believe that the process of regular phonetic
change is any less applicable to the languages of primitive peoples than
to the languages of the more civilized nations? This question must be
answered in the negative. Rapidly accumulating evidence shows that
this process is just as easily and abundantly illustrated in the languages
of the American Indian or of the Nejiro tribes as in Latin or Greek or

486

General Linguistics I

299

Eno'lish. If these laws are more difficult to discover in primitive lan-
"uaf^es, this is not due to any special characteristic which these languages
possess but merely to the inadequate technique of some who have tried
to study them.

An excellent test case of phonetic law in a group of primitive languages
is afforded by the Algonkian linguistic stock of North America. This
stock includes a large number of distinct languages which are, however,
obviously related in both grammar and vocabulary. Bloomfield has taken
four of the more important of the languages that belong to the central
division of the stock and has worked out a complete system of vocalic
and consonantal phonetic laws. We have selected in Table I five of
these phonetic laws in order to give an idea of the nature of the corre-
spondences.

TABLE I

PC A

Fox

Ojibwa

Plains Cre€

Menomini

1. tck

2. ck

3. xk

4. hk

5. nk

hk
'ck
hk
hk
g

ck
ck

hk
hk

sk
sk
sk

hk
hk

tsk
sk

hk
hk

Table I shows how five different consonantal combinations in which the
second element is k were respectively developed in Fox, Ojibwa, Plains
Cree, and Menomini. The Primitive Central Algonkian prototype (PCA)
is, of course, a theoretical reconstruction on the basis of the actual dia-
lectic forms.

Observe that this table does not say that a particular k combination
of one dialect corresponds uniquely to a particular k combination of an-
other dialect, but merely that certain definite dialectic correspondences
are found which lead to such reconstructive inferences as are symbolized
in the first column of the table. Thus, the Plains Cree sk does not al-
ways correspond to the Fox hk but may just as well correspond to the
Fox ck} The Cree sk that corresponds to the Fox hk, however, is obviously
not the same original sound as the Cree sk which corresponds to the Fox
ck, as is indicated by the fact that in Menomini the former corresponds
partly to tsk, partly to hk, while the latter regularly corresponds to sk.
None of the four dialects exactly reflects the old phonetic pattern, which
must be constructed from series of dialectic correspondences.

'C indicates the sound of sh in ship; tc indicates the sound of ch in chip; x indi-
cates the sound of German ch in ach.

.SV.v; I.iini^ndi^i' in Rclniion to History mid Socictv

487

300

The methodology of this table is precisely the same as the methodology
which is used in Indo-European linguistics. The modern German ei of
rnein corresponds to the diphthong i of English mine, but it does not
follow that every modern German ei corresponds to the English diphthong.
As a matter of fact, a large number of German words with ei have Eng-
lish correspondents in o, as in home. Thus, while mine corresponds to
German mein, thine to German dein, and nine to German U ein, the
English home corresponds to German Heim, soap to German Seife, and

TABLE II

PCA

Fox

Ojibwa

Plains Crce

Mcnomini

1 * -alakatckw- .

■inagacku-

■ayakask

■indkatsku-

"palate"
*ketckyawa . .
"he is old"

2. *ickutdwi ...

"fire"

3. *maxkesini ...

"moccasin"

kehkyawa

ackutdwi

mahkasdhi
(dim. form)

kotsklw

ickudd
mahkizin

iskutdw
maskisin

iskotdw
mahkdsin

4. *n6hkuma . . .

*nohkumehsa
"my grand-
mother"

nohkuma
"my mother-
in-law"

nohkumesa
"my grand-
mother"

nohkumis

nohkum

ndhkumeh

5. * tankeckawdwa
"he kicks
him"

tageckawdwa

tangickawad

tahhiskawdw

tahkdskawew

loaf (of bread) to German Lmb. We have to conclude that the modem
German ei represents two historically distinct sounds. In this particular
case we have the documentary evidence with which to check up a neces-
sary or, at least, a highly plausible inference. The type illustrated by
English mine : : German mein corresponds to Old High German l and
Anglo-Saxon i, while the type illustrated by English home :: German
Heim corresponds to Old High German ei and Anglo-Saxon a. We can
briefly summarize all the relevant facts by saving that Early Germanic
I has become a diphthong in Modern English and a practically identical
diphthong in modern German, while an Early Germanic sound which
we may reconstruct as ai (cf. Gothic ai in such words as haims, "home")

488

General Linguistics I

301

has developed to a in Anglo-Saxon, whence modern English o, and ei in
Old Hif^h German, whence the diphthong in modern German. The im-
portant thing to observe about the English and the German examples is
that even in the absence of historical evidence it would have been possible

TABLE III

re A

Fox

Ojibwa

Plains Cree

Menomini

1. tcp

9 ri)

hp
hp
hp
P

hp
mb

sp
sp

tsp
sp

hp
hp
hp

^ XD

4 hn

5. mp

to infer the existence in Early Germanic of two distinct sounds from the
nature of the correspondences in English and German.

Table II gives examples of actual words illustrating the five phonetic
laws in question. The examples given are not isolated examples but are,

TABLE IV

rcA

Fox

Ojibwa

Plains Cree

Menomini

5. *mcqkiisiwa .
"he is red"

meckusiwa

michuzi

mihkusiw

(mehkon)

for the most part, representative of whole classes. The true general-
ity of the phonetic laws illustrated in Table I goes even farther than
there indicated, as is shown by the set of correspondences in Table III.

TABLE V

PCA

Fox

Ojibwa

Plains Cree

Menomini

6. ck

It will be observed that in this table p takes the place, for the most part,
of k of Table II.

Bloomfield found, however, that there was one Algonkian stem evi-
dently involving a k combination which did not correspond to any of the
five series given above. This is the stem for "red" illustrated in Table IV.

For this series of correspondences Bloomfield has constructed a sixth
phonetic law, which is expressed in Table V. It should be understood
that the symbol q is not a phonetic symbol in the ordinary sense of the

.SV.v; Lani^Kd^c in Kclution to Hi\t(>r\ and Society 4K9

302

word. It is merely a formuh or tag which is inlended to hold down
a place, as it were, in a pattern. It may represent a sound similar to
the ch of the German ich, or it may represent some other sound or com-
bination of sounds. Its chief purpose is to warn us that the ch or hk «)f
the Central Algonkian dialects is not to be historically equated with other
examples of ck or hk in these dialects.

The justification for setting up a special phonflic law on the basis of
one set of correspondences is given by Bloomficld himself. He says,
"Since there appeared to be no point of contact fur analogic substitution
of hk for ck, or vice versa, in any of the languages, and since borrowing
of the stem for red seemed unlikely, it was necessary to suppose that the
parent speech had in this stem for red a difTerfnt j)honetic unit."'

Sometimes one is in a position to check ujj a phonetic reconstruction
such as is implied in the use of the symbol ck. A related dialect may
turn up in which the theoretical phonetic prototype is represented by a
distinctive sound or sound combination. As a matter of fact, exactly this
proved to be the case for Central Algonkian. Some time after Bloom-
field set up the sixth phonetic law, he had the opportunity to study the
Swampy Cree dialect of Manitoba. Interestingly enough, this Cree dia-
lect had the consonant combination htk in forms based on the stem for
"red,'* e.g., mihtkusiiv, "he is red"; and in no other stem did this com-
bination of sounds occur. In other words, the added evidence obtained
from this dialect entirely justified the isolation for Primitive Central Al-
gonkian of a particular phonetic-sound group, symbolized by qk. The
setting-up of phonetic law No. 6 was, by implication, a theoretically
possible prediction of a distinct and discoverable phonetic pattern. The
prediction was based essentially on the assumption of the regularity of
sound changes in language.

Bloomfield's experience with the Central Algonkian dialects is entirely
parallel to my own with the Athabaskan languages. These constitute
an important linguistic stock which is irregularly distributed in North
America. The northern group occupies a vast territory stretching all the
way from near the west coast of Hudson Bay west into the interior of
Alaska. To it belong such languages as Anvik (in Alaska), Carrier
(in British Columbia), Chipewyan, Hare, Loucheux, Kutchin, Beaver, and
Sarcee. We shall take Chipewyan and Sarcee as representatives of this
group. The geographically isolated Pacific division of Athabaskan con-
sists of a number of languages in southwestern Oregon and northwestern
California. We shall take Hupa as representative. The southern division
of Athabaskan is in New Mexico and Arizona and adjoining regions, and

490

General Linguistics I

303

is represented by Navaho, Apache, and Lipan. We shall take Navaho as
representative of the group. In spite of the tremendous geographical dis-
tances that separate the Athabaskan languages from each other, it is per-
fectly possible to set up definite phonetic laws which connect them ac-
cordin*^ to consistent phonetic patterns. Navaho, Hupa, and Chipewyan
are spoken by Indians who belong to entirely distinct culture horizons,
vet the languages themselves are as easily derivable from a common source
on the basis of regular phonetic law as are German, Dutch, and Swedish.

TABLE VI

Ath.

Hupa

Chipewyan

Navaho

Sarcce

LL s

2. z

3. dz

4. ts

5. ts

ILL c

2. j

3. dj

4. tc

5. tc'

IILL ?

2. y

3. gy

4. ky

5. ky'

S
S

dz
ts
ts'

tew

to'

ky'

d8
td
tO'

ts
ts'

c
y

dj
tc
tc'

z
dz
ts
ts'

tc
tc'

ts
ts'

5
z
dz
ts
ts'

ts'

dj
tc
tc'

Table VI shows the distribution in Hupa, Chipewyan, Navaho, and
Sarcee of three initial consonantal sets, each of which consists of five con-
sonants. In other words, the table summarizes the developments of fifteen
originally distinct Athabaskan initial consonants in four selected dialects.
Each of the entries must be considered as a summary statement applying
to a whole class of examples.*

The table merits study because of its many implications. It will be ob-
served that no one dialect exactly reproduces the reconstructed Atha-
baskan forms given in the first column. Series I is preserved intact in
Navaho and Sarcee and very nearly so in Hupa, but has been shifted to

* The apostrophe symbolizes a peculiar type of consonantal articulation, charac-
terized by simultaneous closure of the glottis and point of contact in the mouth, with
glottal release preceding oral release. / is the French / of jour; dj is the / of English
just; z is the ch of German ich; W is approximately the wh of English what; d is
the th of English thick; 6 is the th of English then.

.SV.v; I.ii/ii^iKii^r in Rclaiion lo lli\U)i\ mid SocU'lv

491

304

another series in Chipewyan. Series II is preserved intact in Navaho, hut
has been shifted in Sarcee to identity with the series that corresponds
to original I, while Hupa has introduced several peculiar dialectic devel-
opments and Chipewyan has shifted it to the original form of I. Series III
is nowhere kept entirely intact but nearly so in Hupa, while in Chipewyan
and in Sarcee it has moved to the original form of Scries 11, in Navaho
to a form which is identical with the original and the Navaho form of
Series I. It is clear from the table that a Sarcee s is ambiguous as to
origin, for it may go back either to Athabaskan s or Alhabaskan c. On the
other hand, a Sarcee s which is supported by either Navaho or Hupa 5
must be the representative of an original Athabaskan s. Sarcee tc is, in
the main, unambiguous as to origin, for it corresponds to the original
Athabaskan ky. It is curious and instructive to note that, of the four

TABLE VII

Ath.

Hupa

Chipewyant

Xavahot

Sarceet

* ley an

tcq

nltsq

"there's a
rainfall"

tCQ

"rain"

t Q represents nasalized a, as in French an. Sarcee 9 is a peculiar a with velar resonance,
regularly developed from Athabaskan a.

languages given in the table, Hupa and Chipewyan are the two that most
nearly correspond as to pattern but never as to actual sound except in the
one instance of y (III, 2).

Let us take a practical example of prediction on the basis of the table.
If we have a Sarcee form with tc, a corresponding Navaho form with ts,
and a Chipewyan form with tc, what ought to be the Hupa correspondent?
According to the table it ought to be ky.

Table VII shows the distribution in three dialects of the Athabaskan
sound ky (III, 4) in the word for "rain." When I first constructed the
Athabaskan prototype, I assumed an initial Ay, in spite of the absence of
the test form in Hupa, on the basis of the dialectic correspondences. Nei-
ther an original ts nor tc could be assumed in spite of the fact that these
sounds were actually illustrated in known dialects, whereas ky was not.
The Hupa column had to remain empty because the cognate word, if still
preserved, was not available in the material that had been recorded by
P. E. Goddard.

In the summer of 1927, however, I carried on independent researches
on Hupa and secured the form kyaij-kyoh,^ meaning "hailstorm." The sec-

* (7 is the ng of English sing.

492 General Linguistics I

305

ond element of the compound means "big" and the first is obviously the
missin*^ Hupa term corresponding to the old Athabaskan word for "rain."
In other words, an old compound meaning "rain-big" has taken on the
special meaning of "hailstorm" in Hupa. The Hupa form of the old word
for "rain" is exactly what it should be according to the correspondences
that had been worked out, and the reconstruction of the primitive Atha-
baskan form on the basis of the existing forms was therefore justified by
I he event.

Table VHI gives the chief dialectic forms that were available for the
reconstruction of the Athabaskan word for "rain." Observe that not one
of these has the original sound hy which must be assumed as the initial
of the word. This is due to the fact that the old Athabaskan ky and re-
lated sounds shifted in most dialects to sibilants but were preserved in

TABLE VIII
Dialectic Forms for "Rain"

Anvik (Alaska) tcoN*

Carrier (B.C.) tcan

Chipe^vyan tea

Hare teg

Loucheux tcien

Kutchin tscin

Beaver teg

Sarcee tea

Navaho n-l-tsq

• 3 represents open o, as in German voll; o is nasalized o. N is voiceless n.

Hupa and a small number of other dialects, some of which are spoken
at a great remove from Hupa. In other words, in working out linguistic
reconstructions we must be guided not merely by the overt statistical evi-
dence but by the way in which the available material is patterned.

For those interested in a summary statement of the concepts and assump-
tions involved in the foregoing, the following quotations from Bloomfield's
"A Set of Postulates for the Science of Language " may prove of interest :

Def. — A minimum same of vocal feature is a phoneme or distinetive sound.

Assumption. — The number of different phonemes in a language is a small
sub-multiple of the number of forms.

Assumption. — Every form is made up wholly of phonemes Such a

thing as a "small difference of sound" does not exist in a language.

Assumption. — The number of orders of phonemes in the morphemes (i.e.,
"minimum forms") and words of a language is a sub-multiple of the number
of possible orders.

Six: I.tini^iuii^i' in Rchiiion lo History uml Sncietv 4^)3

306

Assumption. — Every language changes at a rate which leaves contemporary
persons free to communicate without disturbance.

Assumption. — Among persons, linguistic change is uniform in ratio wiili
the amount of communication between them.

Assumption. — Phonemes or classes of phonemes may gradually change.

Def. — Such change is sound-change.

Assumption. — Sound-change may affect phonemes or classes of phonemes in
the environment of certain other phonemes or classes of phonemes.

Def. — This change is conditioned sound-change.

At the end of "A Note on Sound-Change," in which the Swampy Crce
forms in htk are discussed, Bloomfield remarks:

The postulate of sound-change without exceptions will probably always re-
main a mere assumption, since the other types of linguistic change (analogic
change, borrowing) are bound to affect all our data. As an assumption, how-
ever, this postulate yields, as a matter of mere routine, predictions which
otherwise would be impossible. In other words, the statement that phonemes
change (sound-changes have no exceptions) is a tested hypothesis: in so far as
one may speak of such a thing, it is a proved truth.

It may be pointed out in conclusion that the value to social science of
such comparative study of languages as is illustrated in the present paper
is that it emphasizes the extraordinary persistence in certain cases of com-
plex patterns of cultural behavior regardless of the extreme variability of
the content of such patterns. It is in virtue of pattern conservatism that it
is often possible to foretell the exact form of a specific cultural phe-
nomenon.®

' [This analysis was first written in December, 1928, and revised by the analyst in
February, 1929. — Editor.]

Editorial Note

In: Stuart A. Rice cd.. Methods in Social Science: A O/.vc Book (Chicagi):
University of Chicago Press, 1931), pp. 297-306. |Rcprintcd in: Hdward Sapir.
Selected Writini^s in Lani^uage, Culnirc, and Personality, f-.ditcd by Dasid (i.
Mandelbaum. Berkeley: University oi ("alirornia Press, 1949, pp. 73-S2: an
excerpt of this article (corresponding to pp. 3()2-3()(i) has been reprinted in The
Collected Works of Edward Sapir, vol. \'I. pp. 1^)9 -201 ]

494

General Linguistics I

COMMUNICATION. It is obvious that for
the building up of society, its units and sub-
divisions, and the understandings which prevail
between its members some processes of com-
munication are needed. While we often speak of
society as though it were a static structure de-
fined by tradition, it is, in the more intimate
sense, nothing of the kind, but a highly intricate
network of partial or complete understandings
between the members of organii^tional units of
every degree of size and complexity, ranging
from a pair of lovers or a family to a league of
nations or that ever increasing portion of hu-
manity which can be reached by the press through
all its transnational ramifications. It is only ap-
parently a static sum of social institutions; actu-
ally it is beiiig reanimated or creatively reaf-
firmed from day to day by particular acts of a
communicative nature which obtain among indi-
viduals participating in it. Thus the Republican
party cannot be said to exist as such, but only to
the extent that its tradition is being constantly
added to and upheld by such simple acts of com-
munication as that John Doe votes the Republi-
can ticket, thereby communicating a certain kind
of message, or that a half dozen individuals
meet at a certain time and place, formally or in-
formally, in order to communicate ideas to one
another and eventually to decide what points of
national interest, real or supposed, are to be
allowed to come up many months later for dis-
cussion in a gathering of members of the party.
The Republican party as a historic entity is
merely abstracted from thousands upon thou-
sands of such single acts of communication,
which have in common certain persistent fea-
tures of reference. If we extend this example

into every conceivable field in which communi-
cation has a place we soon realize that every
cultural pattern and every single act of social
behavior involve communication in either an
explicit or an implicit sense.

One may conveniently distinguish between
certain fundamental techniques, or primary
processes, which are communicative in character
and certain secondary techniques which facili-
tate the process of communication. The distinc-
tion is perhaps of no great psychological im-
portance but has a very real historical and
sociological significance, inasmuch as the funda-
mental processes are common to all mankind,
while the secondary techniques emerge only at
relatively sophisticated levels of civilization.
Among the primary communicative processes of
society may be rnentioned: language; gesture, in
its widest sense; the imitation of overt behavior;
and a large and ill defined group of implicit
processes which grow out of overt behavior and
which may be rather vaguely referred to as
"social suggestion."

Language is the most explicit type of com-
municative behavior that we know of. It need
not here be defined beyond pointing out that it
consists in every case known to us of an abso-
lutely complete referential apparatus of phonetic
symbols which have the property of locating
every known social referent, including all the
recognized data of perception which the society
that it serves carries in its tradition. Language is
the communicative process par excellence in*
every known society, and it is exceedingly im-
portant to observe that whatever may be the
shortcomings of a primitive socie.ty judged from
the vantage point of civilization its language
inevitably forms as sure, complete and poten-
tially creative an apparatus of referential sym-
bolism as the most sophisticated language that
we know of. What this m«ans for a theory of
communication is that the mechanics of signifi-
cant understanding between human beings are
as sure and complex and rich in overtones in one
society as in another, priririitive or sophisticated.
Gesture includes much more than the manip-
ulation of the hands and other visible and mov-
able parts of the organism. Intonations of the
voice may register attitudes and feelings quite as
sigruficantly as the clenched fist, the wave of the
hand, the shrugging of the shoulders or the lift-
ing of the eyebrows. The field of gesture inter-
plays constantly with that of language proper,
but there are many facts of a psychological and
historical order which show that there are subtle

Six: I.dni^Uiiiic in Ri-laiion lo Uisiorv an, I S(h i(t\

495

yet firm lines of demarcation bttwttn them.
Thus, to give hut one example, the t^onsistcnt
message deiivereil by language symholisin in the
narrow sense, whether by speech or by writing,
mav llatly contradict the mesiuige cx)mmunicated
by the synchronous system of gestures, consist-
ing of movements of the hands and hc*ad, intona-
tions of the voice and breathing symbolisms.
The former system may be entirely conscious,
the latter entirely uncx)nscious. Linguistic, as
opj^xised to gesture, communication tends to be
the official and socially accredited one; hence one
may intuitively interpret the relatively uncon-
scious symbolisms of gesture as psychologically
more significant in a given context than the
words actually used. In such cases as these we
have a conflict between explicit and implicit
communications in the growth of the individ-
ual's social experience.

The primary condition for the consolidation
of society is the imitation of overt behavior. Such
imitation, while not communicative in intent,
has always the retroactive value of a communica-
tion, for in the process of falling in with the
ways of society one in effect acquiesces in the
meanings that inhere in these ways. When one
learns to go to church, for instance, because
other members of the community set the pace
for this kind of activity, it is as though a com-
munication had been received and acted upon.
It is the function of language to articulate and
rationalize the full content of these informal
communications in the growth of the individual's
social experience.

Even less dirertly communicative in character
than overt behavior and its imitation is "social
suggestion" as the sum total of new acts and new
meanings that arc implicitly made possible by
these types of social behavior. Thus, the partic-
ular methotl of revolting against the habit of
church going in a given society, while contra-
dictory, on the surface, of the conventional
meanings of that society, may nevertheless re-
ceive all its social significance from hundreds of
existing prior communications that belong to the
culture of the group as a w hole. The importance
of the unformulated and unverbalized commu-
nications of society is so great that one w ho is not
intuitively familiar with them is likely to be
hartlcd by the significance of certain kinds of
behavior, even if he is thoroughly aware of their
external forms and of the vertial symbols that
accompany them. It is largely the function of
the artist to make articulate these more subtle
intentions of society.

Communicative proccuses do not merely «p-
ply to s<x:icty a.s such; they arc indefinitely varied
as to furtn and meaning for the various t>(>cs uf
pcrsoiul relationships into which •ocicty re-
solves iusclf. 'ihua, a fued type of conduct or a
linguistic symlx>l has not by any means neces-
sarily the same c-ommunicativc significance with-
in the coniincs of the family, among the mem-
bers of an economic group and in the nation at
large. Generally speaking, the smaller the circle
and the more complex the understandings al-
ready arrived at within it, the more economical
can the act of communication afford to become.
A single word passed between members of an
intimate group, in spite of its apparent vagueness
and ambiguity, may constitute a far more pre-
cise communication than volumes of carefully
prepared corrcsp)ondence interchanged between
two governments.

There seem to be three main classes of tech-
niques which have for their object the facilita-
tion of the primary communicative processes of
society. These may be referred to as: language
transfers; symbolisms arising from special tech-
nical situations; and the creation of physical
conditions favorable for the communicative act.
Of language transfers the best known example is
writing. The Morse telegraph code is another
example. These and many other communicative
techniques have this in common, that while they
are overtly not at all like one another their
organization is based on the primary symbolic
organization which has arisen in the domain of
speech. Psychologically, therefore, they extend
the communicative charaaer of speech to situa-
tions in which for one reason or another speech
is not possible.

In the more special class of communicative
symbolism one cannot make a word to word
translation, as it were, back to speech but can
only paraphrase in speech the intent of the
communication. Here belong such symbolic
systems as wigAvagging, the use of railroad
lights, bugle calls in the army and smoke signals.
It is interesting to observe that while they arc
late in developing in the history of society they
are very much less complex in structure than
language itself. They are of value partly in
helping out a situation where neither language
nor any form of Lingiiagc transfer cun be ap-
plied, partly where it is desired to encourage the
automatic nature of the desired response. 'Ilius,
because language is extraordinarily rich in
meaning it sometimes becomes a little annoying
or even dangerous lo rely upon it where only a

496

General Linguistics I

simple this or that, or yes or no, is expected to be
the response.

The importance of extending the physical
conditions allowing for communication is obvi-
ous. The railroad, the telegraph, the telephone,
the radio and the airplane are among the best
examples. It is to be noted that such instru-
ments as the railroad and the radio are not com-
municative in character as such; they become so
only because they facilitate the presentation of
types of stimuli which act as symbols of com-
munication or which contain implications of
communicative significance. Thus, a telephone
is of no use unJess the party at the other end
understands the language of the person calling
up. Again, the fact that a railroad runs me to a
certain point is of no real communicative im-
portance unless there are fixed bonds of interest
which connect me with the inhabitants of the
place. The failure to bear in mind these obvious
points has tended to make some writers exag-
gerate the importance of the spread in modem
times of such inventions as the railroad and the
telephone.

The history of civilization has been marked by
a progressive increase in the radius of communi-
cation. In a typically primitive society commu-
nication is reserved for the members of the tribe
and at best a small number of surrounding tribes
wth whom relations are intermittent rather than
continuous and who act as a kind of buffer be-
tween the significant psychological world — the
world of one's own tribal culture — and the great
unknown or unreal that lies beyond. Today, in
our own civilization, the appearance of a new
fashion in Paris is linked by a series of rapid and
necessary events with the appearance of the same
fashion in such distant places as Berlin, Lxjndon,
New York, San Francisco and Yokohama. The
underlying reason for this remarkable change in
the radius and rapidity of communication is the
gradual diffusion of cultural traits or, in other
words, of meaningful cultural reactions. Among
the various types of cultural diffusion that of
language itself is of paramount importance.
Secondary technical devices making for ease of
communication are also, of course, of great
importance.

The multiplication of far-reaching techniques
of communication has two important results.
In the first place, it increases the sheer radius of
communication, so that for certain purposes the
whole civilized world is made the psychological
equivalent of a primitive tribe. In the second
place, it lessens the importance of mere geo-

graphical contiguity. Owing to the technical
nature of these sophisticated communicative
devices, parts of the world that are geographi-
cally remote may, in terms of behavior, be actu-
ally much closer to one another than adjoining
regions, which, from the historical standpoint,
are supposed to share a larger body of common
understandings. This means, of course, a tend-
ency to remap the world both sociologically
and psychologically. Even now it is possible to
say that the scattered "scientific world" is a
social unity which has no clear cut geographical
location. Further, the world of urban under-
standing in America contrasts rather sharply
with the rural world. The weakening of the
geographical factor in social organization must
in the long run profoundly modify our attitude
toward the meaning of personal relations and of
social classes and even nationalities.

The increasing ease of communication is pur-
chased at a price, for it is becoming increasingly
difficult to keep an intended communication
within the desired bounds. A humble example of
this new problem is the inadvisability of making
certain kinds of statement on the telephone.
Another example is the insidious cheapening of
literary and artistic values due to the foreseen
and economically advantageous "widening of
the appeal." All effects which demand a certain
intimacy of understanding tend to become diffi-
cult and are therefore avoided. It is a question
whether the obvious increase of overt communi-
cation is not constantly being corrected, as it
were, by the creation of new obstacles to com-
munication. The fear of being too easily under-
stood may, in n^any cases, be more aptly defined
as the fear of being understood by too many — so
many, indeed, as to endanger the psychological
reality of the image of the enlarged self confront-
ing the not-self.

On the whole, however, it is rather the ob-
stacles to communication that are felt as an-
noying or ominous. The most important of these
obstacles in the modem world is undoubtedly
the great diversity of languages. The enormous
amount of energy put into the task of translation
implies a passionate desire to make as light of the
language difficulty as possible. In the long run it
seems almost unavoidable that the civilized
world will adopt some one language of inter-
communication, say English or Esperanto,
which can be set aside for denotive purposes
pure and simple.

Edward Sapk
See: Society; Social Process; Cxji-turb; Tradition;

.SV.v; I.iinguage in Relation to l/istorv and Society 497

Symhomsm; Collective BrHAVioR; Imiiaiion; Con-
iiNt II Y, S()( ial; Lanci'/Vii ; WKiiiNci; Pri:ss; Public
Opinion.

Editorial Note

In: Encyclopaedia of the Social Sciences, vol. 4 (New York: Macmillan, 1931 ). pp.
78-81. [Reprinted in: Edward Sapir. Selected Writini^s in Lanf^uai^e, Culture, and
Personality. Edited by David G. Mandelbaum. Berkeiev: University of California
Press, 1949, pp. 104-109]

498 General Linguistics I

CONCEPTUAL CATEGORIES IN PRIMITIVE
LANGUAGES*

Conceptual categories in primitive languages: Edward Sapir (introduced by C.
Wissler). The relation between language and experience is often misunderstood.
Language is not merely a more or less systematic inventory of the various items
of experience which seem relevant to the individual, as is so often naively as-
sumed, but is also a self-contained, creative symbolic organization, which not only
refers to experience largely acquired without its help but actually defines experi-
ence for us by reason of its formal completeness and because of our unconscious
projection of its implicit expectations into the field of experience. In this respect
language is very much like a mathematical system, which, also, records experi-
ence, in the true sense of the word, only in its crudest beginnings but, as time goes
on. becomes elaborated into a self-contained conceptual system which previsages
all possible experience in accordance with certain accepted formal limitations.
Such categories as number, gender, case, tense, mode, voice, "aspect" and a host
of others, many of which are not recognized systematically in our Indo-European
languages, are, of course, derivative of experience at last analysis, but, once
abstracted from experience, they are systematically elaborated in language and
are not so much discovered in experience as imposed upon it because of the
tyrannical hold that hnguistic form has upon our orientation in the world.
Inasmuch as languages differ very widely in their systematization of fundamen-
tal concepts, they tend to be only loosely equivalent to each other as symbolic
devices and are, as a matter of fact, incommensurable in the sense in which two
systems of points in a plane are, on the whole, incommensurable to each other if
they are plotted out with references to differing systems of coordinates. The point
of view urged in this paper becomes entirely clear only when one compares lan-
guages of extremely different structures, as in the case of our Indo-European lan-
guages, native American Indian languages and native languages of Africa.

Editorial Note

Science 74 (1931), 578. [Reprinted in: Language in Culture and Society. A reader
in linguistics and anthropology, ed. by Dell H. Hymes. New York: Harper & Row,
1964, p. 128]

.SV.v; Ldni^iKii^i' in Rchuion i<> Hisioty mul Society

A^N

DIALECT. This term has a connotation in
technical linguistic usage which is somewhat
ditlerent from its ordinary meaning. To the
linguist there is no real difference between a
dialect and a language which can be shown to
be related, however remotely, to another lan-
guage. By preference the term is restricted to
a form of speech which does not differ suffi-
ciently from another form of speech to be
unintelligible to the speakers of the latter. Thus
Great Russian and White Russian are said to
be dialects of the same language. Similarly,
.Alsatian, Swabian and Swiss German are dia-
lects or groups of dialects of a common folk
speech. Literal mutual intelligibility, however,
is not a criterion of great interest to the tech-
nical linguist, who is more concerned with the
fact and order of historical relationships in
speech. To him Venetian and Sicilian are equally
dialects of Italian, although as far as mutual
intelligibility is concerned these two might as
well be called independent languages. Russian,
Polish, Czech, Bulgarian and Serbian, conven-
tionally considered as independent languages
because of their national affiliations, are no less
truly dialects of a common Slavic speech or
linguistic prototype than Venetian and Sicilian
are dialects of a supposedly common Italian
language. If two obviously related forms of
speech are spoken at the same time, the linguist
does not say that one of them is a dialect of
the other but that both are sister dialects of
some common prototype, known or inferred.
When they diverge so far as not only to be
mutually unintelligible but no longer to be too
obviously related to each other, the term lan-
guage is more freely used than dialect, but in
principle there is no difference between the
two. Thus in a sense all Romance languages,
all Celtic languages, all Germanic languages,
all Slavic languages and all Indo-Ar)an ver-

naculars arc merely dialect grr>ups of a common
An.'an or Indo-Kuropcan language.

A group of dialects is merely the socialized
form of the universal tendency to individual
variation in speech. Thc.ic variations affect the
phonetic form of the language, its fom\al char-
acteristics, its vocabulary and such pn>sodic
features as intonation and .stress. No known
language, unless it be artificially preserved for
liturgical or other non-popular uses, has ever
been known to resist the tendency to split up
into dialects, any one of which may in the long
run assume the status of an indeixndcnt lan-
guage. From dialects formed by inherent differ-
entiation one may distinguish dialects which
owe their origin to speech transfers. A com-
munity which takes on a language that is
different from the one to which it has originally
been accustomed will unconsciously carrv' over
into the adopted language peculiarities of its
own form of speech which are pronounced
enough to give its use of the foreign language
a dialectic tinge. Many linguists attach much
importance to the influence of superseded lan-
guages in the formation of dialects. Thus some
of the distinctive peculiarities of both Celtic
and Germanic are supposed to be due to the
retention of phonetic peculiarities of pre-.'\r)an
languages.

In less technical or frankly popular usage the
term dialect has somewhat different connota-
tions. Human speech is supposed to be differ-
entiated and standardized in a number of
approved forms known as languages, and each
of these in turn has a number of subvarietics
of lesser value known as dialects. A dialect is
looked upon as a departure from the standard
norm, in many cases even as a corruption of it.
Historically this view is unsound, because the
vast majority of so-called dialects are merely
the regular, differentiated development of ear-
lier forms of speech which antedate the recog-
nized languages. Popular confusion on the
subject is chiefly due to the fact that the ques-
tion of language has become secondarily identi-
fied with that of natinnalitv in the larger cultural
and ethnic group which in course of time ab-
sorbs the local tradition. The language of such
a nationality is generally based on a local dialect
and spreads at the expense of other dialects
which were origiKally of as great prestige as
the culturally more powerful one.

Of the large number of dialects spoken in
Gennany, German Switzerland and .Austria,
for example, N'^ry few, if any, can be considered

5()0

General Linguistics I

124

as modified forms of the culturally accepted
lloiliJeutsch of literature, the pulpit, the stage
and general cultural activity. The dialects of
the German speaking folk go back unbrokenly
to the Old High German of early mediaeval
times, a German which was even then richly
ditferentiated into dialects. The present stand-
ardized German of the schools arose compara-
tively late in the history of German speech as
a result of the fixing of one of the Upper Saxon
dialects as the recognized medium of otTicial
communication within the German speaking
donu'nions. Luther's Bible helped considerably
in the diffusion of this form of German as the
recognized standard. It has taken a long time,
however, for Hochdeutsch to take on a recog-
nized phonetic form and to be looked upon as
a well standardized form of oral communica-
tion, and to this day a large proportion of
Germans, including the educated ranks, are
bilingual in the sense that they use the stand-
ardized German for formal purposes but em-
ploy the local dialect for more familiar uses.

The history of German is paralleled more or
less by the history of all the other national
languages of Eurof)e and of other parts of the
world. As a result of cultural reasons of one
kind or andther a local dialect becomes accepted
as the favored or desirable form of speech
within a linguistic community that is cut up
into a large number of dialects. This approved
local dialect becomes the symbol of cultural
values and spreads at the expense of other local
forms of speech. The standardized form of
speech becomes more and more set in its vo-
cabulary, its form and eventually its pronun-
ciation. The speakers of local dialects begin to
be ashamed of their peculiar forms of speech
because these have not the prestige value of
the standardized language; and finally the illu-
sion is created of a primary language, belonging
to the large area which is the territory of a
nation or nationality, and of the many local
forms of speech as uncultured or degenerated
variants of the primary norm. As is well known,
these variations from the norm are sometimes
much more archaic, historically speaking, than
the norm from which they are supposed to
depart.

Local dialects are in a sense minority lan-
guages, but the latter term should be reserved
for a completely distinct form of speech that is
used by a minority nationality living within
the political framework of a nation. An example
of such a minority language would be the

Basque of southwestern France and northern
Spain or the Breton of Brittany. These lan-
guages are not dialects of French and Spanish
but historically distinct languages that have
come to occupy culturally secondary positions.

There is naturally no hard and fast line
between a dialect and a local variation of a
minor nature, such as New England English
or middle western English. In the case of the
older dialects the connection with the stand-
ardized speech is quite secondary, while in the
case of such local variations as New England
and middle western American speech standard
English, however loosely defined, is present in
the minds of all as the natural background for
these variations, which are thus psychologically,
if not altogether historically, variations from
the primary or standard norm. It would be
possible for the speaker of a local Swiss dialect
or of Yorkshire English to build up a nation-
alistic gospel around his local dialect in oppo-
sition to the accepted speech of the cultured
group, but the attempt to do this for middle
western English in America would have some-
thing intrinsically absurd about it because of
the feeling that this form of English is at best
but a belated departure from an earlier norm.
As usual in social phenomena, however, it is
the symbolism of attitude that counts in these
matters rather than the objective facts of history.

Ever since the formation of the great national
languages of Europe toward the end of the
mediaeval period there have been many social
and political influences at work to imperil the
status of the local dialects. As the power of
the sovereign grew, the language of the court
gained in prestige and tended to diffuse through
all the ramifications of the official world. Mean-
while, although the Roman Catholic and Greek
churches with their sacred liturgical languages
were little interested in the question of folk
versus standardized speech, the Protestant sects
with their concern for a more direct relation
between God and His worshipers emphasized
the dignity of folk speech and lent their aid
to the diffusion of a selected form of folk
speech over a larger area. The influence of such
documents as Luther's Bible and King James'
authorized version in the standardization of
German and English has often been referred to.
In more recent days the increase of popular
education and the growing demand for ready
intelligibility in the business world have given
a tremendous impetus to the spread of stand-
ardized forms of speech.

Six: l.aniiuaf^i' in Relation to History and Society

501

Ir. !^pitc of all thc-sc standardizinp influences,
however, local dialects, particularly in Murope,
have persisted with a vitality that is little short
of amazing. Obviously the question of the con-
jcTvatism of dialect is not altogether a negative
matter of the inertia of speech and of the failure
oi overriding cultural influences to permeate
into all corners of a given territory. It is to a
verv significant degree a }x>sitivc matter of the
risistance of the local dialects to something
which is vaguely felt as hostile. This is easily
understood if we look upon languages and dia-
lects not as intrinsically good or bad forms of
speech but as symbols of social attitudes. Be-
fore the growth of modern industrialism culture
tended to be intensely local in character in spite
of the uniformizing influences of government,
religion, education and business. The culture
that gradually seeped in from the great urban
centers was felt as something alien and super-
ricial in spite of the prestige that unavoidably
attached to it. The home speech was associated
with kinship ties and with the earliest emotional
experiences of the individual. Hence the learn-
ing of a standardized language could hardly
seem natural except in the few centers in which
the higher culture seemed properly at home,
and even in these there generally developed a
hiatus between the standardized language of
the cultured classes and the folk speech of the
local residents. Hence cockney is as far removed
psychologically from standard British English
as is a peasant dialect of Yorkshire or Devon.
On the continent of Europe, particularly in
Germany and Italy, the culture represented,
for example, by standardized German or stand-
ardized Italian was until very recent days an
exceedingly thin psychological structure, and
its orticial speech could hardly take on the task
of adequately symbolizing the highly ditTeren-
tiated folk cultures of German speaking and
Itjlian speaking regions.

The .Age of Enlightenment in the eighteenth
century was, on the whole, hostile to the f)er-
sistence of dialects, but the romantic movement
which followed it gave to folk speech a glamour
which has probably had something to do with
the idealization of localized languages as sym-
bols of national solidarity and territorial integ-
rity. Few writers of the seventeenth and eight-
eenth centuries would have taken seriously the
use of dialect in literature. It was only later
that Lowland Scotch could be romantically
restored in the lyrics of Robert Burns, that
I-ntz Reuter could strive to establish a Low

German (Plaitdcutach) literary bngujgc and
that Mistral could attempt to revive the long
lost glory of Provencal. One nruy suspect that
this renewed emphasis on linguistic ditfcrcncM
is but a passing phase in the history of nnxJern
man. Be that as it may, it has had much to do
with the emergence of new nationalisms in
recent times. It is doubtful if such countries
as Lithuania, Estonia and C/xchoslovakia could
have so easily proved their right to exist 1/ it
had not come to be felt that just as every
nationality needs its language, so everv unat-
tached language needs its nationality and terri-
torial independence to fulfil its inherent mission.
Perhaps the best example of what might be
called linguistic romanticism is the attempt of
the Irish nationalists to renew the viulity of
Gaelic, a form of speech which has never been
standardized for literary, let alone folk, pur-
poses and which is profoundly alien to the
majority of the more articulate of Irish na-
tionalists.

No doubt the respect for local forms of
speech has received assistance from scientific
linguistics and its tendency to view all lan-
guages and dialects as of equal historical im-
portance. It is very doubtful, however, if lin-
guistic localism can win out in the long run.
The modern mind is increasingly realistic and
pragmatic in the world of atnion and concep-
tualistic or normative in the world of thought.
Both of these attitudes are intrinsically hostile
to linguistic localism of any sort and necessarily
therefore to dialeaic conservatism. Compulsory
education, compulsory military service, modem
means of communication and urbanization are
some of the more obvious factors in the spread
of these attitudes, which, so far as language is
concerned, may be defined by the thesis that
words should cither lead to unambiguous action
among the members of as large a group as is
held together culturally or in the domain of
thought should aim to attach themselves to
concepts which are less and less purely local
in their application. In the long run therefore
it seems fairly safe to hazard the guess that
such movements as the Gaelic revival in Ireland
and the attempt to s;ive as many minority
languages and dialects from cultural extinction
as possible will come to be looked upon as little
more than eddies in the more powerful stream
of standardization of speech that set in at the
close of the mediaeval jx-riod. The modem
problem is more complex than the classical or the
mediaeval problem, because the modern mind

502

General Linguistics I

126

insists on having the process of standardization
take the form of a democratic rather than an
aristocratic process.

A word may be added in regard to the social
psychology of dialectic forms of speech. In the
main, markedly dialectic peculiarities have been
looked upon as symbols of inferiority of status,
but if local sentiment is strongly marked and
if the significance of the local group for the
larger life of the nation as a whole allows, a
local dialect may become tJie symbol of a kind
of inverted pride. We thus have the singular
specucle of Lowland Scotch as an approved
and beautiful linguistic instrument and of cock-
ney as an undesirable and ugly one. These
judgments are extrinsic to the facts of language
themselves but they are none the less decisive
in the world of cultural symbolisms.

If an individual is brought up in a community
that has its characteristic dialect and if he
becomes identified later in life with another
community which has a second mode of speech,
some very interesting personality problems arise
which involve the status symbolism or affec-
tional symbolism of these differing forms of

speech. Individuals who vacillate somewhat in
their conception of their own role in society
may often be detected unconsciously betraying
this feeling of insecurity in a vacillating pro-
nunciation or intonation or choice of words.
When under the influence of an emotional crisis
such individuals are thrown back upon their
earliest emotional experiences— "regress," in
short — they are likely to relapse into early dia-
lectic habits of speech. It is suggested that the
question of the relation of the individual to
the various dialects and languages to which he
has been subjected from time to time is of far
more than anecdotal interest, that it constitutes
as a matter of fact a very important approach
to the problem of personality subjected to the
strains of cultural change.

Eow.yiD Sapir

See: Language; Localism; Centralization; Custom-
Consult: Jespersen, J. O. H., Language: Its Nature,
Development and Origin (London 1922); Bally, Charles,
Le langage et la vie (Paris 1926); Vendryes, Joseph,
he langage: introduction linguistique d Vhistoire (Paris
1921), tr. by Paul Radin, History of Civilisation
series (London 1925); Meillet, Antoine, Les langues
dans I'Europe nouvelle (2nd ed. Paris 1928).

Editorial Note

In: Encyclopaedia of the Social Sciences, vol. 5 (New York: Macmillan, 1931), pp.
123-126. [Reprinted in: Edward Sapir, Selected Writings in Language, Culture,
and Personality. Edited by David G. Mandelbaum. Berkeley: University of
California Press, 1949, pp. 83-88]

The following errors in the originally published version have been corrected
directly into the text printed here (page references are to the original):
p. 126, bibliography: Vendryes (correct: Vendryes)
p. 126, bibliography: Veiirope (correct: V Europe)

.SV.v; /.(iniiiKim' in Rclnfinn i,> //istorv and S,nut\

503

l^ANCUAGE. The gift of speech and a well
ordered language are chnracteristic of every
l;no\\'n group of human beings. No tribe has
ever l)een found which is without language and
.ill statements to the contrary may be dismissed
as mere folklore. There seems to be no warrant
whitever for the sutcment which is sometimes
nude that there are ccrt-iin peoples whose vo-
c-abuiary is so limited that they cannot get on
without the supplcmcnLiry use of gesture, so
that intelligible communication iKtwcen mem-
bers of such a group becomes impossible in the
dark. The truth of the matter is that language
is an essentially perfect means of expression and
communication among every known people. Of
all aspects of culture it is a fair guess that lan-
guage was the first to receive a highly developed
form and that its essential perfection is a pre-
requisite to the development of culture as a
whole.

There are some general characteristics which
apply to all languages, living or extinct, written
or unwritten. In the first place language is pri-
marily a system of phonetic symbols for the
expression of communicable thought and feel-
ing. In other words, the symbols of language are
dilTerentiated products of the vocal behavior
which is associated with the larynx of the higher
ifLmirnals. As a mere matter of theory it is con-
ceivable that something like a linguistic struc-
ture could have been evolved out of gc-sture or
other forms of bodily behavior. The fact that
at an advanced stage in the history of the human
race writing emerged in close imitation of the
patterns of spoken language proves that lan-
guage as a purely instrumental and logical device
is not dependent on the use of articulate sounds.
Nevertheless, the actual history of man and a
wealth of anthropological evidence indicate with
overwhelming certainty that phonetic language
tikes precedence over all other kinds of com-
municative symbolism, which are by comparison
either substitutive, like writing, or merely sup-
plementary, like the gesture accompanying
speech. The speech apparatus which is used in
the articulation of bnguage is the same for all
known peoples. It consists of the laryruc, with
its delicately adjustable glottal chords, the nose,
the tongue, the hard and soft palate, the teeth
and the lips. While the original impulses leading
to speech may be thought of as localized in the
larynx, the finer phonetic articulations arc due
chiefly to the muscubr activity of the tong\]c,
an organ whose primary function h;is of course
nothing whatever to do with sound production

but which in actual speech behavior ii indis-
pensable for the development of cmotiorully ex-
pressive sound into what wc call Un|{uage. It i«
so indispensable in fact that one of the mo»t com-
mon tcnns for language or speech is "tongue."
Language is thus not a simple bi<.ilogicj| func-
tion even as regards the simple nutter of tound
production, for primary laryngeal patterns of
behavior have had to be oirnpletcly u%er(uuled
by the interference of lingual, labial and nasal
modifiGitions before a "speech organ" was
ready for work. Perhaps it is because this »|)ecch
organ is a diffused and secondary network of
physiological activities which do not correspond
to the primary functions of the organs involved
that language has been enabled to free itself from
direct bodily expressiveness.

Not only are all languages phonetic in char-
acter; they are also "phonemic." Between the
articuhtion of the voice into the phonetic se-
quence, which is immediately audible as a mere
sensation, and the complicated patterning of
phonetic sequences into such symbolically sig-
nificant entities as words, phrases and sentences
there is a very interesting process of phonetic
selection and generalization which is easily over-
looked but which is crucial for the development
of the specifically symbolic aspect of language.
Language is not merely articulated sound; its
significant structure is dependent upwn the un-
conscious selection of a fixed number of "pho-
netic stations," or sound units. These are in
actual behavior individually modifiable; but the
essential point is that through the unconscious
selection of sounds as phonemes definite psy-
chological barriers are erected between various
phonetic stations, so that speech ceases to be
an expressive flow of sound and becomes a sym-
bolic composition with limited materials or
units. The analogy with musical theory seems
quite fair. Even the most resplendent and dy-
namic symphony is built up of tangibly distinct
musical entities or notes which in the physical
world flow into each other in an indefinite con-
tinuum but which in the world of aesthetic
composition and appreciation arc definitely
bounded off against each other, so that they iiuy
enter into an intricate nuihematics of significant
relatipnships. The phonemes of a langu.igc arc
in principle a distinct system peculiar to the
given language, and its words must be made up,
in unconscious theory if not al\va)-s in actualizcrd
behiivior, of these phonemes. I jnguages differ
very widelv in their phonemic structure. But
whatever the details of these structures may be,

504

General Linguistics I

,56

the important fact remains that there is no
knouTi language which has not a perfectly defi-
nite phonemic sj'stem. The difference between
a sound and a phoneme can be illustrated by a
simple example in English. If the word matter
is pronounced in a slovenly fashion, as in the
phrase "What's the matter?" the t sound, not
being pronounced with the full energy required
to bring out its proper physical characteristics,
tends to slip into a d. Nevertheless, this phonetic
d will not be felt as a functional d but as a
variety of < of a particular type of expressiveness.
Obviously the functional relation between the
proper t sound of such a word as matter and
its d variant is quite other than the relation of
the t of such a word as town and the d of down.
In every known language it is possible to dis-
tinguish merely phonetic variations, whether
expressive or not, from s^Tnbolically functional
ones of a phonemic order.

In all known languages phonemes are built
up into distinct and arbitrary sequences which
are at once recognized by the speakers as mean-
ingful symbols of reference. In English, for
instance, the sequence g plus o in the word go
is an unanalyzable unit and the meaning attach-
ing to the symbol cannot be derived by relating
to each other values which might be imputed-to
the ^ and to the independently. In other words,
while the mechanical fonctional units of lan-
guage are phonemes, the true units of language
as symbolism are conventional groupings of such
phonemes. The size of these units and the laws
of their mechanical structure vary widely in the
different languages and their limiting conditions
may be said to constitute the phonemic mechan-
ics, or phonology, of a particular language. But
the fundamental theory of sound symbolism
remains the same everywhere. The formal behav-
ior of the irreducible symbol also varies within
wide limits in the languages of the world. Such
a unit may be either a complete word, as in the
English example already given, or a significant
element, like the suffix ness of goodness. Be-
tween the meaningful and unanalyzable word or
word element and the integrated meaning of
continuous discourse lies the whole complicated
field of the formal procedures which are intui-
tively employed by the speakers of a language
in order to build up aesthetically and function-
ally satisfying symbol sequences out of the the-
oretically isolable units. These procedures con-
stitute grammar, which may be defined as the
sum total of formal economies intuitively recog-
Tiized by the speakers of a language. There seem

to be no types of cuJtural patterns which vary
more surprisingly and with a greater exuberance
of detail than the morphologies of the known
languages. In spite of endless differences of de-
tail, however, it may justly be said that all
grammars have the same degree of fixity. One
language may be more complex or difficult
grammatically than another, but there is no
meaning whatever in the statement which is
sometimes made that one language is more
grammatical, or form bound, than another. Our
rationalizations of the structure of our own lan-
guage lead to a self-consciousness of speech and
of academic discipline which are of course inter-
esting psychological and social phenomena in
themselves but have very little to do with the
question of form in language.

Besides these general formal characteristics
language, has certain psychological qualities
which make it peculiarly important for the stu-
dent of social science. In the first place, language
is felt to be a perfect symbolic system, in a per-
fectly homogeneous medium, for the handling
of all references and meanings that a given cul-
ture is capable of, whether these be in the form
of actual communications or in that of such
ideal substitutes of communication as thinking.
The content of every culture is expressible in
its language and there are no linguistic materials
whether as to content or form which are not felt
to symbolize actual meanings, whatever may be
the attitude of those who belong to other cul-
tures. New cultural experiences frequently make
it necessary to enlarge the resources of a lan-
guage, but such enlargement is never an arbi-
trary addition to the materials and forms already
present; it is merely a further application of
principles already in use and in many cases little
more than a metaphorical extension of old terms
and meanings. It is highly important to realize
that once the form of a language is established
it can discover meanings for its speakers which
are not simply traceable to the given quality of
experience itself but must be explained to a large
extent as the projection of potential meanings
into the raw material of experience. If a man
who has never seen more than a single elephant
in the course of his life nevertheless speaks with-
out the slightest hesitation of ten elephants or
a million elephants or a herd of elephants or of
elephants walking two by two or three by three
or of generations of elephants, it is obvious that
language has the power to analyze experience
into theoretically dissociable elements and to
create that world of the potential intergrading

.SV.v; I.(in}^H(ii^c in Rrlmion to Hisioiy ami Smietv

505

u-ith t'^*^ actual which enables human beings to
.j.jf^>;cenil the immediately given in their individ-
ual experiences and to j(5in in a larger common
u'lJerstanding. This common understanding
■vinstitutes culture, which cannot be adequately
Jctined bv a description of those more colorful
patterns of behavior in society which lie open
to observation. Language is heuristic, not merely
in the simple sense which this example suggests
but in the much more far reaching sense that its
forms predetermine for us certain modes of ob-
s<rnation and interpret;ition. This means of
course that as our scientific experience grows we
nnist learn to fight the implications of language.
"The grass waves in the wind" is showT\ by its
!in<niistic form to be a member of the s;ime
relational class of experiences as "The man
uorks in the house." As an interim solution of
the problem of expressing the experience re-
ferred to in this sentence it is clear that the
language has proved useful, for it has made
sieniticant use of certain symbols of conceptual
relation, such as agency and location. If we feel
the sentence to be poetic or metaphorical, it is
largely because other more complex types of
experience with their appropriate symbolisms of
reference enable us to reinterpret the situation
and to say, for instance, "The grass is waved by
the wind" or "The wind causes the grass to
wave." The point is that no matter how sophis-
ticated our modes of interpretation become, we
never really get beyond the projection and con-
tinuous transfer of relations suggested by the
forms of our speech. After all, to say that "Fric-
tion causes such and such a result" is not very
different from saying that "The grass waves in
the wind." Language is at one and the same
time helping and retarding us in our exploration
of experience, and the details of these processes
of help and hindrance are deposited in the
subtler meanings of different cultures.

A further psychological characteristic of lan-
g\iage is the fact that while it may be looked
upon as a symbolic system which reports or
refers to or otherwise substitutes for direct ex-
pt-rience, it docs not as a rruitter of actual behav-
ior stand apart from or run parallel to direct
experience but completely interpenetrates with
it. This is indicated by the widespread feeling,
particularly among primitive people, of that
virtual identity or close correspondence of word
and thing which leads to the magic of spells.
On our own level it is generally difficuh to make
a complete divorce between objective reality and
our linguistic symbols of reference to it; and

things, qualities and events are on the whole felt
to be what they arc called. For the nonruU per-
son every experience, reaJ or potential, is sat-
urated with verbalism. This explains why »
many lovers of nature, for instance, do not feel
that they are truly in touch with it until they
have ma-stered the names of a great many flowers
and trees, as though the prinruirv world of reality
were a verbal one and as though one could not
get close to nature unless one first mastered the
terminology which s<jmehow iiugicaJlv expresses
it. It is this constant interplay between language
and experience which removes language from
the cold status of such purely and simply sjTn-
bolic systems as mathematical symbolism or flag
signaling. This interpcnctration is not only an
intimate ;issociative fact; it is also a contextual
one. It is im{"K)rtant to realize that language may
rtot only refer to experience or even mold, inter-
pret and discover experience but that it also
substitutes for it in the sense that in those
sequences of interpicrsonal behavior which form
the greater part of our daily lives speech and
action supplement each other and do each
other's work in a web of unbroken pattern. If
one says to me "I^cnd me a dollar," I may hand
over the money without a word or I may give
it with an accompanying "Here it is" or I may
say "I haven't got it. I'll give it to you tomor-
row." Each of these respxjnses is structurally
equivalent, if one thinks of the larger beha\ior
pattern. It is clear that if language is in its
analyzed form a s)'mbolic system of reference
it is far from being merely that if we consider
the psychological part that it plays in continuous
behavior. The re;ison for this almost unique po-
sition of intimacy which language holds among
all known svTnbolisms is probably the fact that
it is learned in the earliest years of childhood.

It is because it is learned early and piecemeal,
in constant association with the color and the
requirements of actual conte.xts, that lang\iage
in spite of its quasi-mathematical form is rarely
a purely referential organiz-ation. It tends to be
so only in scientific discourse, and even there
it may be seriously doubted whether the ideal
of pure reference is ever attained by language.
Ordinary speech is directly expressive and the
purely forn^al piitterns of stjunds, words, gram-
matical forms, phra-scs and sentences arc ;dwa>T
to be thought of as a)mpounded with intended
or unintended svmlxjlisms of expression, if they
are to be understood fully from the standpoint
of behavior. The choice of words in a particular
context may convey the opposite of what they

506

General Linguistics I

158

mean on the surface. The same external message
is differently interpreted according to whether
the speaker has this or that psychological status
in his personal relations, or whether such pri-
mary expressions as those of affection or anger
or fear may inform the spoken words with a
significance which completely transcends their
normal value. On the whole, however, there is
no danger that the expressive character of lan-
guage will he overlooked. It is too obvious a fact
to call for much emphasis. What is often over-
looked and is, as a matter of fact, not altogether
easy to understand is that the quasi-mathemat-
ical patterns, as we have called them, of the
grammarian's language, unreal as these are in a
contextual sense, have nevertheless a tremen-
dous intuitional vitality; and that these patterns,
never divorced in experience from the expres-
sive ones, are nevertheless easily separated from
them by the normal individual. The fact that
almost any word or phrase can be made to take
on an infinite variety of meanings seems to indi-
cate that in all language behavior there are
intertwined in enormously complex patterns
isolable patterns of two distinct orders. These
may be roughly defined as patterns of reference
and patterns of expression.

That language is a perfect symbolism of expe-
rience, that in the actual contexts of behavior it
cannot be divorced from action and that it is
the carrier of an infinitely nuanced expressive-
ness are universally valid psychological facts.
There is a fourth general psychological peculi-
arity which applies more particularly to the
languages of sophisticated peoples. This is the
fact that the referential form systems which are
actualized in language behavior do not need
speech in its literal sense in order to preserve
their substantial integrity. The history of writing
is in essence the long attempt to develop an
independent svTnboHsm on the basis of graphic
representation, followed by the slow and be-
grudging realization that spoken language is a
more powerful symbolism than any graphic one
can possibly be and that true progress in the art
of writing lay in the virtual abandonment of the
principle with which it originally started. Effec-
tive systems of writing, whether alphabetic or
not, are more 6r less exact transfers of speech.
The original language system may maintain it-
self in other and remoter transfers, one of the
best examples of these being the Morse tele-
graph code. It is a very interesting fact that the
principle of linguistic transfer is not entirely
absent even among the unlettered peoples of the

world. Some at least of the drum signal and
horn signal systems of the west African natives
are in principle transfers of the organizations of
speech, often in minute phonetic detail.

Many attempts have been made to unravel
the origin of language but most of these are
hardly more than exercises of the speculative
imagination. Linguists as a whole have lost in-
terest in the problem and this for two reasons.
In the first place, it has come to be realized that
there exist no truly primitive languages in a
psychological sense, that modem researches in
archaeology have indefinitely extended the time
of man's cultural past and that it is therefore
vain to go much beyond the perspective opened
up by the study of actual languages. In the
second place, our knowledge of psychology,
particularly of the symbolic processes in general,
is not felt to be sound enough or far reaching
enough to help materially with the problem of
the emergence of speech. It is probable that the
origin of language is not a problem that can be
solved out of the resources of linguistics alone
but that it is essentially a particular case of a
much wider problem of the genesis of symbolic
behavior and of the specialization of such behav-
ior in the laryngeal region, which may be pre-
sumed to have had only expressive functions to
begin with. Perhaps a close study of the behavior
of very young children under controlled condi-
tions may provide some valuable hints, but it
seems dangerous to reason from such experi-
ments to the behavior of precultural man. It is
more likely that the kinds of studies which are
now in progress of the behavior of the higher
apes will help supply some idea of the genesis of
speech.

The most popular earlier theories were the
interjectional and onomatopoetic theories. The
former derived speech from involuntary cries of
an expressive nature, while the latter maintained
that the words of actual language are conven-
tionalized forms of imitation of the sounds of
nature. Both of these theories suffer from two
fatal defects. While it is true that both inter-
jectional and onomatopoetic elements are found
in most languages, they are always relatively
unimportant and tend to contrast somewhat with
the more normal materials of language. The
very fact that they are constantly being formed
anew seems to indicate that they belong rather
to the directly expressive layer of speech which
intercrosses with the main level of referential
symbolism. The second diflficulty is even more
serious. The essential problem of the origin of

Six: I.dn^iKii^i' in Kr/mion to llistorx and Societv

507

pccch is not to attempt to discover the kinds
f vocal elements whicli constitute the historical
ucleus of language. It is rather to point out
u,,\s vocal articulations of any sort could become
Jis.-.ociated from their original expressive value.
\bout all that can be said at present is that while
jneech as a finished organization is a distinrtly
hunun achievement, its roots probably lie in the
power of the higher apes to solve specific prob-
lems bv abstracting general forms or schemata
from the details of given situations; that the
hjbit of interpreting certain selected elements
in a situation as sigi\s of a desired total one
rraJuallv led in early man to a dim feeling for
svmbolism; and that in the long run and for
rcas<3ns which can hardly be guessed at the ele-
ments of experience which were most often
interpreted in a symbolic sense came to be the
liT^ely useless or supplementary vocal behavior
that must have often attended significant action.
According to this point of view language is not
so much directly developed out of vocal expres-
sion as it is an actualization in terms of vocal
expression of the tendency to master reality, not
bv direct and ad hoc handling of its elements
but bv the reduction of experience to familiar
forms. Vocal expression is only superficially the
same as language. The tendency to derive
speech from emotional expression has not led
to jn\thing tangible in the way of scientific
theor\' and the attempt must now be made to
see in language the slowly evolved product of a
peculiar technique or tendency which may be
called the symbolic one, and to see the relatively
meaningless or incomplete part as a sign of the
whole. Language then is what it is essentially
not because of its admirable expressive power
but in spite of it. Speech as behavior is a won-
derfully complex blend of two pattern systems,
the symbolic and the expressive, neither of which
could have developed to its present perfection
without the interference of the other.

It is difficult to see adequately the functions
of language, because it is so deeply rooted in
the whole of human behavior that it may be
suspected that there is little in the functional
side of our conscious behavior in which hin-
guage does not play its part. The primary func-
tion of language is generally said to be communi-
cation. There can l)e no quarrel with this so long
as it is distinctly understood that there may be
cfTective communication without overt speech
and that language is highly relevant to situations
which are not obviously of a conamunicative
sort. To say that thought, which is hardly pos-

sible in any sustained sense without the sym-
bolic organization brought by languaf^e, is that
fonn of anninunication in which the speak .t
and the person addressed arc identified in one
person is not far from begging the ijuestion. The
autistic speech of children seems to show that
the purely communicative as|>ect of language has
been exaggerated. It is best to adnut that lan-
guage is primarily a vocal actualization of the
tendency to see reality symbolically, that it is
precisely this quality which renders it a fit
instrument for communication and that it is in
the actual give and t:ike of social intercourse
that it has been complicated and refined into the
form in which it is known today. Besides the
very general function which language fulfils in
the spheres of thought, communication and ex-
pression which are implicit in its very nature
there may be pointed out a number of special
derivatives of these which are of particular in-
terest to students of society.

Language is a great force of socialization,
probably the greatest that exists. By this is meant
not merely the obvious fact that significant social
intercourse is hardly possible without language
but that the mere fact of a common speech
serves as a peculiarly potent s)Tnbol of the social
solidarity of those who speak the language. The
psychological significance of this goes far be-
yond the association of particular languages with
nationalities, political entities or smaller local
groups. In between the recognized dialect or
language as a whole and the individualized
speech of a given individual lies a kind of lin-
guistic unit which is not often discussed by the
linguist but which is of the greatest importance
to social psychology. This is the subform of a
language which is current among a group of
people who are held together by ties of common
interest. Such a group may be a family, the
undergraduates of a college, a labor union, the
undenvorld in a large city, the members of a
club, a group of four or five friends who hold
together through life in spite of ditferenccs of
professional interest, and untold thousands of
other kinds of groups. Each of these tends to
develop peculiarities of sfx-ech which have the
symlx)lic function of somehow distinguishing
the group from the larger group into which in
members might be too completely absorbed.
The complete absence of linguistic indices of
such small groups is obscurely felt as a defect
or sign of emotional poverty. Within the con-
fines of a particular family, for instance, the
name Georgy, having once been mispronounced

508

General Linguistics I

1 60

Doody in childhood, may take on the latter form
forever after; and this unofficial pronunciation
of a familiar name as applied to a particular
person becomes a very important symbol indeed
of the solidarity of a particular family and of the
continuance of the sentiment that keeps its
members together. A stranger cannot lightly
take on the privilege of saying Doody if the
members of the family feel that he is not entitled
to go beyond the degree of familiarity sym-
bolized by the use of Georgy or George. Again,
no one is entitled to say "trig" or "math" who
has not gone through certain familiar and pain-
ful experiences as a high school or undergradu-
ate student. The use of such words at once
declares the speaker a member of an unorganized
but psychologically real group. A self-made
mathematician has hardly the right to use the
word "math" in referring to his own interests
because the student overtones of the word do
not properly apply to him. The extraordinary
importance of minute linguistic differences for
the symbolization of psychologically real as con-
trasted with politically or sociologically official
groups is intuitively felt by most people. "He
talks like us" is equivalent to saying "He is one
of us."

There is another important sense in which
language is a socializer beyond its literal use as
a means of communication. This is in the estab-
lishment of rapp>ort between the members of a
physical group, such as a house party. It is not
what is said that matters so much as that some-
thing is said. Particularly where cultural under-
standings of an intimate sort are somewhat lack-
ing among the members of a physical group it
is felt to be important that the lack be made good
by a constant supply of small talk. This caressing
or reassuring quality of speech in general, even
where no one has anything of moment to com-
municate, reminds us how much more language
is than a mere technique of communication.
Nothing better shows how completely the life
of man as an animal made over by culture is
dominated by the verbal substitutes for the
physiqal world.

The use of language in cultural accumulation
and historical transmission is obvious and im-
portant. This applies not only to sophisticated
levels but to primitive ones as well. A great deal
of the cultural stock in trade of a primitive
society is presented in a more or less well de-
fined linguistic form. Proverbs, medicine for-
mulae, standardized prayers, folk tales, stand-
ardized speeches, song texts, genealogies, are

some of the more overt forms which language
takes as a culture preserving instrument. The
pragmatic ideal of education, which aims to re-
duce the influence of standardized lore to a
minimum and to get the individual to educate
himself through as direct a contact as possible
with the facts of his environment, is certainly
not realized among the primitives, who are often
as word bound as the hunianistic tradition itself.
Few cultures perhaps have gone to the length
of the classical Chinese culture or of rabbinical
Jewish culture in making the word do duty for
the thing or the personal experience as the ulti-
mate unit of reality. Modem civilization as a
whole, with its schools, its libraries and its end-
less stores of knowledge, opinion and sentiment
stored up in verbalized form, would be unthink-
able without language made eternal as document.
On the whole, we probably tend to exaggerate
the differences between "high" and "low" cul-
tures or saturated and emergent cultures in the
matter of traditionally conserved verbal author-
ity. The enormous differences that seem to exist
are rather differences in the outward form and
content of the cultures themselves than in the
psychological relation which obtains between
the individual and his culture.

In spite of the fact that language acts as a
socializing and uniformizing force it is at the
same time the most potent single known factor
for the growth of individuality. The fundamen-
tal quality of one's voice, the phonetic patterns
of sf)eech, the speed and relative smoothness of
articulation, the length and build of the sen-
tences, the character and range of the vocabu-
lary, the stylistic consistency of the words used,
the readiness with which words respond to the
requirements of the social environment, in par-
ticular the suitability of one's language to the
language habits of the person addressed — all
these are so many complex indicators of the
personality. "Actions speak louder than words"
may be an excellent maxim from the pragmatic
point of view but betrays little insight into the
nature of speech. The language habits of people
are by no means irrelevant as unconscious indi-
cators of the more important traits of their per-
sonalities, and the folk is psychologically wiser
than the adage in paying a great deal of attention
willingly or not to the psychological significance
of a man's language. The normal person is never
convinced by the mere content of speech but is
very sensitive to many of the implications of
language behavior, however feebly (if at all)
these may have been consciously analyzed. All

.SVv; /.(///t,'//</,i,'(' /// Rclitiion [,, /Iisiorv and Society

5()9

in :iil. ■' '^ "°^ ^°° much to say that one of the
rcallv iniport.int functions of language is to be
conAt.mtly declaring to society the psychological
place held by all of its tncmbers. Besides this
niore general type of [xrsonality expression or
julfilinent there is to be kept in mind the imjxjr-
tant role which language plaj-s as a substitutive
nieans oi expression for those individiuils who
\\x\x a greater than normal difficulty in adjusting
thenxsehes to the environment in terms of pri-
marv action patterns. Even in the most primitive
cultures the strategic word is likely to be more
povserful th;in the direct blow. It is unwise to
»ie;ik too blithely of "mere" words, for to do so
nuv be to imperil the value and perhaps the very
existence of civilization and personality.

The l.inguages of the world may be classified
either structurally or genetically. \n. adequate
structural analysis is an intricate matter and no
classification seems to have been suggested
which does justice to the bewildering variety of
known forms. It is useful to recognize three
distinct criteria of classification: the relative
degree of synthesis or elaboration of the words
of the language; the degree to which the various
parts of a word are welded together; and the
extent to which the fundamental relational con-
cepts of the language are directly expressed as
such. As regards sjTithcsis languages range all
the way from the isolating type, in which the
single word is essentially unanalyzable, to the
tvpe represented by many American Indian
languages, in which the single word is function-
all v often the equivalent of a sentence with many
concrete references that would in most languages
require the use of a number of words. Four
stages of synthesis may be conveniently recog-
nized; the isolating type, the weakly synthetic
type, the fully synthetic typ^e and the polys)Ti-
ttietic type. The classical example of the first
type is Chinese, which does not allow the words
of the language to be modified by internal
changes or the addition of prefixed or suffixed
elements to express such concepts as those of
number, tense, mode, case relation and the like.
This seems to be one of the more uncommon
types of huiguage and is best represented by a
number of languages in eastern Asia. Besides
Chinese itself Siamese, Burmese, modern Ti-
betan, .Aruiamiteand Khmer, or Cambodian, may
be given as examples. The older view, which
regarded such languages as representing a pecul-
iarly primitive stage in the evolution of lan-
piage, inay now be dismi.sscd as antiquated. .Ml
e\idence points to the contrary hypothesis that

ibi

such I^iguagci are the lofpcaJly cxtrcioc aju-
lytic developments of more tynihclic lani^uagcs
which betausc of prtxTisc* of pi ' in-

tegration lu\e had to rccxprcM • -a I

means combinations of idratorigtrull) lxj ra*ed
within the fnmcwtirk of the single word. The
weakly synthetic type of Ungiugc is bc»t rrp-
rcsented by the most familiar nuxjcm languaj^ca
of Europe, such as English, French, Spanish,
Italian, German, Dutch and Danuh Such lan-
guages modify words to acme extent but have
only a moderate fornial elaboration of the word.
The plural formations of [-English and French,
for instance, are relatively simple and the tense
and modal systems of all the lai\guagc3 of this
type tend to use analvtic methods as supple-
mentary to the older synxhetic one. The third
group of languages is represented by such lan-
guages as .\rabic and earlier Indo-European
languages, like Sanskrit, Latin and Greek. These
are all languages of great formal comple.xity, in
which classificatory ideas, such as sex gender,
number, cise relations, tense and mood, are
expressed with considerable nicety and in a great
variety of ways. Because of the rich formal im-
plications of the single word the sentence tends
not to be so highly energized and ordered as in
the first mentioned types. Lastly, the pol)-5j'n-
thetic languages add to the formal c»mplcxity
of the treatment of fundamental relational ideas
the power to arrange a number of logically dis-
tinct, concrete ideas into an ordered whole
within the confines of a single word. Eskimo and
Algonquin are chissical examples of this type.

From the standpoint of the mechanical co-
hesivencss with which the elements of words arc
united languages may be conveniently grouped
into four types. The first of these, in which there
is no such process of combination, is the iso-
lating tvpve already referred to. To the second
group of languages belong all those in which the
word can be adequately analyzed into a mechan-
ical sum of elements, each of which has its more
or less clearly establuihed meaning and each of
which is regularly used in all other words into
which the associated notion enters. These are
the so-called agglutinative languages. The nui-
jority of languages seem to use the agglutinaUVT
technii|ue. which has the great advanUge of
combining logical an.jl>-sis with economy of
means. The .Mtaic langujgcs. of which Turkish
is a good example, and the Bantu laiiguagcs of
Africa are agglutinative in form. In the third
type, the s*i-callcd inflcctivr languages, the de-
gree of union between the radical clen^nt or

510

General Linguistics I

162

stem of the word and the modifying prefixes or
suffixes is greater than in the agglutinative lan-
guages, so that it becomes difficult in many cases
to isolate the stem and set it off against the
accreted elements. More important than this,
however, is the fact that there is less of a one
to one correspondence between the linguistic
element and the notion referred to than in the
agglutinative languages. In Latin, for instance,
the notion of plurality is expressed in a great
variety of ways which seem to have little pho-
netic connection with each other. For example,
the final vowel or diphthong of equi (horses),
dona (gifts), mensae (tables) and the final vowel
and consonant of hostes (enemies) are function-
ally equivalent elements the distribution of
which is dependent on purely formal and his-
torical factors that have no logical relevance.
Furthermore in the verb the notion of plurality
is quite differently expressed, as in the last two
consonants of amant (they love). It used to be
fashionable to contrast in a favorable sense the
"chemical" qualities of such inflective languages
as Latin and Greek with the soberly mechanical
quality of such languages as Turkish. But these
evaluations may now be dismissed as antiquated
and subjective. They were obviously due to the
fact that scholars who wrote in English, French
and German were not above rationalizing the
linguistic structures with which they were most
familiar into a position of ideal advantage. As
an offshoot of the inflective languages may be
considered a fourth group, those in which the
processes of welding, due to the operation of
complex phonetic laws, have gone so far as to
result in the creation of patterns of internal
change of the nuclear elements of speech. Such
familiar English examples as the words sing,
sang, sung, song will serve to give some idea of
the nature of these structures, which may be
termed symbolistig. The kinds of internal change
which may be recognized are changes in vocalic
quality, changes in consonants, changes in quan-
tity, various types of reduplication or repetition,
changes in stress accent and, as in Chinese and
many African languages, changes in pitch. The
classical example of this type of language is
Arabic, in which as in the other Semitic lan-
guages nuclear meanings are expressed by se-
quences of consonants, which have, however, to
be connected by significant vowels whose se-
quence patterns establish fixed functions inde-
pendent of the meanings conveyed by the
CX)risonantal framework.

Elaboration and technique of word analysis

are perhaps of less logical and psychological
significance than the selection and treatment of
fundamental relational concepts for grammatical
treatment. It would be very difficult, however,
to devise a satisfactory conceptual classification
of languages because of the extraordinary diver-
sity of the concepts and classifications of ideas
which are illustrated in linguistic form. In the
Indo-European and Semitic languages, for in-
stance, noun classification on the basis of gender
is a vital principle of structure; but in most of
the other languages of the world this principle
is absent, although other methods of noun classi-
fication are found. Again, tense or case relations
may be formally important in one language, for
example, Latin, but of relatively little grammat-
ical importance in another, although the logical
references implied by such forms must naturally
be taken care of in the economy of the language,
as, for instance, by the use of specific words
within the framework of the sentence. Perhaps
the most fundamental conceptual basis of classi-
fication is that of the expression of fundamental
syntactic relations as such versus their expres-
sion in necessary combination with notions of a
concrete order. In Latin, for example, the notion
of the subject of a predicate is never purely
expressed in a formal sense, because there is no
distinctive symbol for this relation. It is impos-
sible to render it without at the same time
defining the number and gender of the subject
of the sentence. There are languages, however,
in which syntactic relations are expressed purely,
without admixture of implications of a non-
relational sort. We may speak therefore of pure
relational languages as contrasted with mixed
relational languages. Most of the languages with
which we are familiar belong to the latter cate-
gory. It goes without saying that such a con-
ceptual classification has no direct relation to the
other tvvo types of classification which we have
mentioned.

The genetic classification of languages is one
which attempts to arrange the languages of the
world in groups and subgroups in accordance
with the main lines of historical cormection,
which can be worked out on the basis either of
documentary- evidence or of a careful compari-
son of the languages studied. Because of the fai
reaching effect of slow phonetic changes and of
other causes languages which were originally
nothing but dialects of the same form of speech
have diverged so widely that it is not apparent
that they are but specialized developments of a
single prototype. An enormous amount of work

Six: /.(ini^uiii^c in RcUiUon t,> //istnrv u/ul Socit'lv

511

has ^<^'^'" Jonc in the genetic classification and
^jt,cl.issiticalion of the languages of the world,
hut ^'^^y niany problems still await rcse:irch ;uid
^liition. At the present time it is known defi-
nitely that there are certain ven,' large linguistic
oroups. or families, as they are often called, the
p^^-nibers of which may, roughly speaking, be
Kxjked upon as lineally descended from lan-
OTjUizes which can be theoretically reconstructed
in their main phonetic and structural outlines.
It is obvious, however, that languages may so
Ji\cr|;e as to leave little trace of their original
rcI.ition--^hip. It is therefore very dangerous to
assume that languages are not at last analysis
divergent members of a single genetic group
merelv because the evidence is negative. The
onlv contrast that is legitimate is between lan-
guages known to be historically related and
iangu.iges not known to be so related. Languages
tnown to be related cannot be legitimately con-
tra.>^tcd with languages known not to be related.
Because of the fact that languages have dif-
ferentiated at different rates and because of the
important effects of cultural diffusion, which
ha\e brought it about that strategically placed
languages, such as Arabic, Latin and English,
have spread over large parts of the earth at the
expense of others, very varied conditions are
found to prevail in regard to the distribution of
linguistic families. In Europe, for instance, there
are only two linguistic families of importance
represented today, the Indo-European languages
and the L^gro-Finnic languages, of which Fin-
nish and Hungarian are examples. The Basque
dialects of southern France and northern Spain
are the survivors of another and apparently iso-
lated group. On the other hand, in aboriginal
.America the linguistic differentiation is extreme
and a surprisingly large number of essentially
unrelated linguistic families must be recognized.
Some of the families occupy very small areas,
while others, such as the Algonquin and the
.\thabaskan languages of North .America, are
spread over a large territory. The technique of
establishing linguistic families and of working
out the precise relationship of the languages
included in these families is too difficult to be
gone into here. It suffices to say that random
word comparisons are of little importance. Ex-
p^'rience shows that very precise phonetic rela-
tions can be worked out between the languages
of a group and that on the whole fundamental
morphological features tend to preserve them-
selves over exceedingly long periods of time.
ITius modem Lithuanian is in structure, vocab-

.63

ulary and, to a large extent, even ph-xicnuc
pattern \ery much the kind of a ianguajjc which
must be a.«umcd as the prototvpc for the ln*V>-
Europcan language* aa a wliole. In tpile of the

fact that structural clasaificationa a/c in theory
unrelated to genetic one* and in apHe of the fact
tliat languages can l>c shown to have influmcrd
each other, not only in phonetics and vucabulary
but also to an appreciable cvtcnt in Mructure.
it is not often found that the languages of a ge-
netic group exhibit utterly irrcci>ncilablc struc-
tures. Thus even English, which is one of the
least conser\ativc of Indo-European languages,
has many far reaching points of structure in
common with as remote a language as Sanskrit
in contrast, say, to Basque or Firuiish. .Again,
different as are .Assyrian, modem .Arabic and
the Semitic languages of .Abyssinia they exhibit
numerous points of resemblance in phonctica,
vocabulary and structure which set them off at
once from, say, Turkish or the Negro lajiguagcs
of the Nile headwaters.

The complete rationale of linguistic change,
involving as it does many of the most complex
processes of psychology and sociology, has not
yet been satisfactorily worked out, but there are
a number of general processes that emerge with
sufficient clarity. For practical purposes inherent
changes may be distinguished from changes due
to contact with other linguistic communities.
There can be no hard line of division between
these two group>s of changes because every indi-
vidual's language is a distinct psychological
entity in itself, so that all inherent ch.mges arc
likely at last analysis to be peailiarly remote or
subtle forms of change due to contact. The dis-
tinction, however, has great practical value, all
the more so as there is a tendency among anthro-
pologists and sociologists to operate far loo
hastily with wholesale linguistic changes due to
external ethnic and cultural influences. The
enormous amount of study that has been lav-
ished on the history of particular languages ar>d
groups of languages shows very clearly that the
most powerful ditTerenliating factors arc not
outside influences, as onlinarily uiulerstood. but
rather the very slow but powerful unconscious
ch.angcs in certain dircctiorus which seem to be
implicit in the phonemic s)'stems and nvirpholo-
gies of the languages thenviclvcs. ITieae "drifts"
are jwwerfully comlitioncd by unconscious for-
mal feelings and are made necessary by the
inability of human beings to actualize ideal pat-
terns in a pcnnanently .set fashion.

Linguistic changes may be anal)-xed into pbo-

512

General Linguistics I

164

netic changes, changes in fonn and changes in
vocabulary. Of these the phonetic changes seem
to be the most im|x>rtant and the most removed
from direct observation. The factors which lead
to these phonetic changes are probably exceed-
ingly complex and no doubt include the opera-
tion of obscure symbolisms which define the
relation of various age groups to one another.
Not all phonetic changes, however, can be ex-
plained in terms of social symbolism. It seems
that many of them are due to the operation of
unconscious economies in actualizing sounds or
combinations of sounds. The most impressive
thing about internal phonetic change is its high
degree of regularity. It is this regularity, what-
ever its ultimate cause, that is more responsible
than any other single factor for the enviable
degree of exactness which linguistics has at-
tained as a historical discipline. Changes in
grammatical form often follow in the wake of
destructive phonetic changes. In many cases it
can be seen how irregularities produced by the
disintegrating effect of phonetic change are
ironed out by the analogical spread of more
regular forms. The cumulative effect of these
corrective changes is quite sensibly to modify
the structure of the language in many details
and sometimes even in its fundamental features.
Changes in vocabular\- are due to a great variety
of causes, most of which are of a cultural rather
than of a strictly linguistic nature. The too fre-
quent use of a word, for instance, may reduce
it to a commonplace term, so that it needs to be
replaced by a new word. On the other hand,
changes of attitude may make certain words with
their traditional overtones of meaning unaccept-
able to the younger generation, so that they tend
to become obsolete. Probably the most impor-
tant single source of change in vocabulary is the
creation of new words on analogies which have
spread from a few specific words.

Of the linguistic changes due to the more
obvious types of contact the one which seems
to have played the most important part in the
history of language is the "borrowing" of words
across linguistic frontiers. This borrowing nat-
urally goes hand in hand with cultural diffusion.
An analysis of the provenience of the words of
a given language is frequently an important in-
dex of the direction of cultural influence. Our
English vocabulary, for instance, is very richly
stratified in a cultural sense. The various layers
of early Latin, mediaeval French, humanistic
Latin and Greek and modem French borrow-
ings constitute a fairly accurate gauge of the

time, extent and nature of the various foreign
cultural influences which have helped to mold
English civilization. The notable lack of German
loan words in English until a very recent period,
as contrasted with the large number of Italian
words which were adopted at the time of the
Renaissance and later, is again a historically sig-
nificant fact. By the diffusion of culturally im-
portant words, such as those referring to art,
literature, the church, military affairs, sf)ort and
business, there have grown up important trans-
national vocabularies which do something to
combat the isolating effect of the large number
of languages which are still spoken in the mod-
em world. Such borrowings have taken place in
all directions, but the number of truly important
source languages is surprisingly small. Among
the more imfxjrtant of them are Chinese, which
has saturated the vocabularies of Korean, Japa-
nese and Annamite; Sanskrit, whose influence
on the cultural vocabulary' of central Asia, India
and Indo-China has been enormous; Arabic,
Greek, Latin and French. English, Spanish and
Italian have also been of great importance as
agencies of cultural transmission, but their in-
fluence seems less far reaching than that of the
languages mentioned above. The cultural influ-
ence of a language is not alwaj's in direct pro-
portion to its intrinsic literary interest or to the
cultural place which its speakers have held in
the history of the world. For example, while
Hebrew is the carrier of a peculiarly significant
culture, actually it has not had as important an
influence on other languages of Asia as Aramaic,
a sister language of the Semitic stock.

The phonetic influence exerted by a foreign
language may be very considerable, and there
is a great deal of eridence to show that dialectic
peculiarities have often originated as a result of
the unconscious transfer of phonetic habits from
the language in which one was brought up to
that which has been adopted later in life. Apart,
however, from such complete changes in speech
is the remarkable fact that distinctive phonetic
features tend to be distributed over wide areas
regardless of the vocabularies and structures of
the languages involved. One of the most striking
examples of this tvpe of distribution is found
among the Indian languages of the Pacific coast
of California, Oregon, Washington, British Co-
lumbia and southern Alaska. Here are a large
number of absolutely distinct languages, be-
longing to a number of genetica'ly unrelated
stocks, so far as we are able to tell, which never-
theless have many important and distinctive

.SVv; /.(ini^udi^c in Relation lo Hisu>t\ und Society

513

165

phonetic features in common. An analogous fact
is the distribution of certain peculiar phonetic
fc-atures in both the Slavic languages and the
I'gro-F"innic languages, which are unrelated to
them. Such processes of phonetic diflFusion must
be due to the influence exerted by bilingual
speakers, who act as imconscious agents for the
spread of phonetic habits over wide areas. Prim-
iti\e man is not isolated, and bilingualism is
probably as important a factor in the a^nUct of
primitive groups as it is on more sophisticated
levels.

Opinions difTer as to the importance of the
purely morphological influence exerted by one
language on another in contrast with the more
external types of phonetic and lexical influence.
Undoubtedly such influences must he taken into
account, but so f;u- they ha\e not been shown to
operate on any great scale. In spite of the cen-
turies of contact, for instance, between Semitic
and Indo-Euro{x:an languages we know of no
language which is definitely a blend of the struc-
tures of these two stocks. Similarly, while Japa-
nese is flooded with Chinese loan words, there
seems to be no structural influence of the latter
on the former. A type of influence which is
neither one of vocabulary nor of linguistic form,
in the ordinary sense of the word, and to which
insufficient attention has so far been called, is
that of meaning pattern. It is a remarkable fact
of modem European culture, for instance, that
while the actual terms used for certain ideas
vary enormously from language to language, the
range of significance of these equivalent terms
tends to be verj- similar, so that to a large extent
the vocabulary of one language tends to be a
psychological and cultural translation of the vo-
cabulary of another. A simple example of this
sort would be the translation of such terms as
Your Excellency to equis alcnt but etymologically
unrelated terms in Russian, .\nother instance of
this kind would be the interesting parallelism
in nomenclature between the kinship terms of
affinity in English, French and German. Such
terms as muther-in-larv, bclk-mere and Schwiegrr-
mutter are not, strictly speaking, equivalent
either as to etymology or literal meaning but
they are patterned in exactly the same manner.
Thus mother -in -line and father -in-hnv are panil-
lel in nomenclature to belle-mtre and beau-pere
and to Schttnegermutter and Schfcifgcrt>ater.
Tliese terms clearly illustrate the diffusion of a
lexical pattern which in turn probably expresses
a growing feeling of the sentimental equivalence
of blood relatives and relatives by marriage.

The imp<irLancc of language M a whole for
the definition, expression and tranMniuion ol
culture is undoubted. The relevance of linf^uistic
deuils, in Ixjth content and fonn, for the pro-
founder understanding of culture is also clear.
It dofs not follow, however, that there i» a
simple correspondence between the form of a
language and the fonn of the culture of those
who speak it. The tendency to see linguistic
categories as directly expressive of overt cultural
outlines, which seems to have come into fashion
among certain sociologists and anthrojvjiogists,
should be resisted as in no way warranted by
the actual facts. There is no general correlation
between cultural type and linguistic structure
So far as can be seen, isolating or agglutinative
or inflective types of speech are possible on any
level of civilization. Nor does the presence or
absence of grammatical gender, for example,
seem to have any relevance for the understand-
ing of the social organization or religion or folk-
lore of the associated peoples. If there were any
such parallelism as has sometimes been main-
tained, it would be quite impossible to under-
stand the rapidity with which culture diffuses in
spite of profound linguistic difl^erences between
the borrowing and giving communities. The
cultural significance of linguistic form, in other
words, lies on a much more submerged level
than on the overt one of definite cultural pattern.
It is only very rarely, as a matter of fact, that
it can be pointed out how a cultural trait has
had some influence on the fundamental structure
of a language. To a certain extent this lack of
correspondence may be due to the fact that
linguistic changes do not proceed at the same
rate as most cultural changes, which arc on the
whole far more rapid. Short of yielding to an-
other language which takes its place, linguistic
organization, largely because it is unconscious,
tends to maintain itself indcfmitely and does not
allow its fundamental formal categories to be
seriously influenced by changing cultural needs.
If the forms of culture and language were then
in complete ct>rrespondence with one another,
the nature of the pnxxsses making for linguistic
and ailtural changes respectively would soon
bring about a Lack of necessan,' corresptmdcnce.
This is exactly what is found to be the ca.<4e.
Ix)gically it is indefensible that the m.u»c>iline,
feminine and neuter genders of (Irrman and
Russian should be allowetl to continue their
sway in the modem world; but aiiy intellec-
tiKiIist attempt to weed out these unnecessary
genders would obviously be fruitless, for the

514

General Linguistics I

166

normal speaker does not actually feel the clash
which the logician requires.

It is another matter when we pass from gen-
eral form to the detailed content of a language.
Vocabulary is a very sensitive index of the cul-
ture of a people and changes of meaning, loss of
old words, the creation and borrowing of new
ones are all dependent on the history of culture
itself. Languages differ widely in the nature of
their vocabularies. Distinctions which seem in-
evitable to us may be utterly ignored in lan-
guages which reflect an entirely different type
of culture, while these in turn insist on distinc-
tions which are all but unintelligible to us. Such
differences of vocabular)' go far beyond the
names of cultural objects, such as arrow point,
coat of armor or gunboat. They apply just as
well to the mental world. It would be difficult
in some languages, for instance, to express the
distinction which we feel between "to kill" and
"to murder" for the simple reason that the
underlying legal philosophy which determines
our use of these words does not seem natural
to all societies. Abstract terms, which are so
necessarv- to our thinking, may be infrequent
in a language whose speakers formulate their
behavior on more pragmatic lines. On the other
hand, the question of tne presence or absence
of abstract nouns may be bound up with the
fundamental form of the language; and there
exist a large number of primitive languages
whose structure allows of the verj- ready creation
and use of abstract nouns of quality or action.

There are nnany language patterns of a special
sort which -are of interest to the social scientist.
One of these is the tendency to create tabus for
certain words or names. A very widespread cus-
tom among primitive peoples, for instance, is the
tabu which is placed not only on the use .of
the name of a person recently deceased but of
any word that is etymologically connected in the
feeling of the speakers with such a name. This
means that ideas have often to be expressed by
circumlocutions or that terms must be borrowed
from neighboring dialects. Sometimes certain
names or words are too holy to be pronounced
except under very special conditions, and curi-
ous patterns of behavior develop which are de-
signed to prevent one from making use of such
interdirted tenns. An example of this is the
Jewish custom of pronouncing the Hebrew name
for God, not as Yah we or Jehovah but as Adonai,
My Lord. Such customs seem strange to us but
equally strange to many primitive communities
would be our extraordinary reluctance to pro-

nounce obscene words under normal social con-
ditions. Another class of special linguistic phe-
nomena is the use of esoteric language devices,
such as passwords or technical terminologies for
ceremonial attitudes or practises. Among the
Eskimo, for example, the medicine man has a
peculiar vocabulary which is not understood by
those who are not members of his guild. Special
dialectic forms or otherwise peculiar linguistic
patterns are common among primitive peoples
for the tex-ts of songs. Sometimes, as in Mela-
nesia, such song texts are due to the influence of
neighboring dialects. This is strangely analogous
to the practise among ourselves of singing songs
in Italian, French or German rather than in
English, and it is likely that the historical proc-
esses which have led to the parallel custom are
of a similar nature. Thieves' jargons and secret
languages of children may also be mentioned.
These lead over into special sign and gesture
languages, many of w'hich are based directly on
spoken or written speech; they seem to e.xist on
many levels of culture. The sign language of the
Plains Indians of North America arose in re-
sponse to the need for some medium of com-
munication between tribes speaking mutually
unintelligible languages. Within the Christian
church may be noted the elaboration of gesture
languages by orders of monks vowed to silence.
Not only a language or a terminolog\- but the
mere external form in which it is WTitten may
become important as a symbol of sentimental or
social distinction. Thus Croatian and Serbian
are essentially the same lang-aage but they are
presented in very different outward forms, the
former being written in Latin characters, the
latter in the Cyrillic character of the Greek
Orthodox church. This external difference, as-
sociated with a difference of religion, has of
course the important function of preventing
people who speak closely related languages or
dialects but who wish for reasons of sentiment
not to confound themselves in a larger unity
from becoming too keenly aware of how much
they actually resemble each other.

The relation of language to nationalism and
internationalism presents a number of interest-
ing sociological problems. Anthropolog)- makes
a rigid distinction between ethnic units based
on race, on culture and on language. It points
out that these do not need to coincide in the
least — that they do not, as a matter of fact, often
coincide in reality. But with the increased em-
phasis on nationalism in modem times the ques-
tion of the symbolic meaning of race and Ian-

.S7\. Iiini^nai^i' in Rf/mion [,, Uisiorv and Socictv

515

(tu-iCJC has tJccn on a new significance and,
wh.itfver the scientist may say, the layman is
j.\tr inclined to see culture, language and race
^^ !nit different facets of a single social unity,
which he tends in turn to identify with such a
ii)|::ical entity as England or France or Gcr-
iru'iv- To fx)int out, as the anthropologist easily
c-in, that cultural distributions and nationalities
override hinguage and race groups docs not end
tlje matter for the sociologist, because he feels
that the concept of nation or nationality must
be integrally imaged by the non-analjlical per-
son as carrying with it the connotation, real
or supposed, of both race and language. From
this ^t.mdpoint it really makes little difference
whether history and aiUhrojx)log\' support the
popular identification of natioruility, langiiage
and race. The important thing to hold on to is
that a particular language tends to become the
titting expression of a self-conscious nationality
and that such a group will construct for itself
in spite of all that the physical anthropologist
can do a race to which is to be attributed the
mvstic power of creating a language and a cul-
ture as twin expressions of its psychic peculi-
arities.

So far as language and race are concerned, it
is true that the major races of man have tended
in the past to be set off against each other by
important differences of language. There is less
point to this, however, than might be imagined,
because the linguistic differentiations within any
given race are just as far reaching as those which
can be pointed out across racial lines, yet they
do not at all correspond to subracial units. Even
the major races are not always clearly sundered
by language. This is notably the case with the
Malayo-Polyncsian languages, which are spoken
by peoples as racially distinct as the Malays,
the Polynesians and the Negroes of Mela-
nesia. Not one of the great languages of modem
man follows racial lines. French, for example,
is s(X)kcn by a highly mixed population, which
is largely Nordic in the north, .Alpine in the
center and Medi(crr;mean in the south, each of
these subraces being liberally represented in the
rest of Europe.

While language differences have always been
important symbwls of cultural difference, it is
only in comparatively recent times, with the
exaggerated development of the ideal of the sov-
ereign nation and with the resulting eagerness
to discover linguistic symbols for this ideal of
^vcreignty, that language differences ha\e taken
on an implication of antagonism. In ancient

T67

Rome and all through mcdije\-al Kunif^ tlicic
were plenty of cultural difference* mnninK wde
by side with lingui-Hic one*. «nd the political
st.it us of Roman citizen or the fact of adherence
to the Roman Catholic church wiu of \-vn\y
greater significance a* a »ymbr)| of ihc individ-
ual's place in the world than the lan|{ua({c or
dialect which he happened to S()<-ak. It ti prob-
ably altogether incorrccn to maintain that lan-
guage differences arc responsible for national
antagonisms. It would seem to be much more
reasonable to suppose that a political and ru-
tional unit, once definitely formed, uses a pre-
vailing language as a symbol of its identity,
whence gmdually emerges the peculiarly mod-
em feeling that every language should properly
be the expression of a distinctive nationality. In
earlier times there seems to have been little
systematic attempt to impose the language of a
conquering people on the subject people, al-
though it happened frequently as a result of the
processes implicit in the spread of culture that
such a conqueror's language was gradually taken
over by the dispossessed population. Witness
the spread of the Romance langu.iges and of the
modem .\rabic dialects. On the other hand, it
seems to have happened about as frequently that
the conquering group was culturally and lin-
guistically abs<3rbcd and that their own language
disappeared without necessary danger to their
privileged status. Thus foreign dynasties in
China have alwavs submitted to the superior
culture of the Chinese and have taken on their
language. In the same way the Moslem Moguls
of India, while true to their religion, uhich was
adopted by millions in northern India, made
one of the Hindu vernaculars the Ixisis of the
great litcrar\' language of .Moslem India, Hin-
dustani. Definitely repressive attitudes toward
the languages and dialects of subject peoples
seem to be distinctive only of European political
policy in comparatively recent times, llie at-
tempt of czarist Russia to stamp out Polush by
forbiilding its te.iching in the sch(x)ls and the
similarly repressive policy of contemporary luly
in its attempt to wipe out German fmm the
territory recently acquired from Austria are
illuminating examples of the heightened empha-
sis on language as a symbol of political allegianoe
in the modem world.

To match these repressive mcx^ures there it
the oft repeated attempt of minority groups to
erect their language into the status of « fully
accredited meilium of cultural and literary ex-
pression. Many of these restored or semimanu-

516

General Linguistics I

i68

factured languages have come in on the wave of
resistance to exterior political or cultural hos-
tility. Such are the Gaelic of Ireland, the Lithu-
anian of a recently created republic and the
Hebrew of the Zionists. In other cases such
languages have come in more peacefully because
of a sentimental interest in local culture. Such
are the modem Proven9al of southern France,
the Plattdeutsch of northern Germany, Frisian
ajid the Norsvegian landsmaal. It is doubtful
whether these persistent attempts to make true
culture languages of local dialects that have long
ceased to be of primar)- literary importance can
succeed in the long run. The failure of modern
Proven9al to hold its own and the ver)' dubious
success of Gaelic make it seem probable that
following the recent tendency to resurrect minor
languages will come a renewed leveling of speech
more suitably expressing the internationalism
which is slowly emerging.

The logical necessity of an international lan-
guage in modern times is in strange contrast to
the indifference and even opposition with which
most people consider its possibility. The at-
tempts so far made to solve this problem, of
which Esperanto has probably had the greatest
measure of practical success, have not affected
more than a very small proportion of the people
whose international interests and needs might
have led to a desire for a simple and uniform
means of international expression, at least for
certain purposes. It is in the less important
countries of Europe, such as Czechoslovakia,
that Esperanto has been moderately successful,
and for obvious reasons. The opposition to an
international language has little logic or psy-
chology in its favor. The supposed artificiality
of such a language as Esperanto or of any of
the equivalent languages that have been pro-
txjsed has been absurdly exaggerated, for in
sober truth there is practically nothing in these
languages that is not taken from the common
stock of words and forms which have gradually
developed in Europe. Such an international lan-
guage could of course have only the status of a
secondary form of speech for distinctly limited
purposes. Thus considered the learning of a
constructed international language offers no fur-
ther psychological problem than the learning of
any other language which is acquired after child-
hood through the medium of books and with
the conscious application of grammatical rules.
The lack of interest in the international language
problem in spite of the manifest need for one
is an excellent example of how little logic or

intellectual necessity has to do with the acquire-
ment of language habits. Even the acquiring of
the barest smattering of a foreign national lan-
guage is imaginatively equivalent to some meas-
ure of identification with a people or a culture.
The purely instrumental value of such knowl-
edge is frequently nil. .\ny consciously con-
structed international language has to deal with
the great difficulty of not being felt to represent
a distinctive people or culture. Hence the learn-
ing of it is of very little symbolic significance for
the average person, who remains blind to the
fact that such a language, easy and regular as it
inevitably must be, would solve many of his
educational and practical difficulties at a single
blow. The future alone will tell whether the
logical advantages and theoretical necessity of
an international language can overcome the
largely sjTnbolic opposition which it has to meet.
In any event it is at least conceivable that one
of the great national langtiages of modem times,
such as English or Spanish or Russian, may in
due course find itself in the position of a de facto
international language without any conscious
attempt having been made to put it there.

Edward Sapir

See: Writing; CoMMT.fNiCATiON; Symbolism; Cul-
tl're; Anthropology; Race; Nationalism; Dlvlect;
Isolation; Standardization; Civilization.

Consult: Sapir, E., Language, an Introduction to the
Study of Speech (New York 1921), "Language and
Environment" in American Anthropologist, n.s., vol.
xiv (19 1 2) 226-42, "The Histor>- and Varieries of
Human Speech" in Popular Science Monthly, vol.
Ixxix (19 11) 45-67, and "Sound Patterns in Lan-
guage" in Language, vol. i (1925) 37-5 1; Paget, Rich-
ard, Human Speech (Lxjndon 1930); Ogden, C. K.,
and Richards, I. A., The Meaning of Meaning (2nd rev.
ed. New York 1927); Markey, John F., The Symbolic
Process and Its Integration in Children (London 1928);
Piaget, Jean, Le langage et la pensee chez I'enfant
(Paris 1924), tr. by M. Gabain (2nd ed. New York
1932); Vendryes, J., Le langage: introduction lift-
guistiqiie a I'histoire (Paris 1921), tr. by Paul Radin,
Histor>- of Civilization series (London 1925); Jesper-
sen. Otto, Language, Its Nature, Development and
Origin (London I922\'and The Philosophy of Gram-
mar (London 1924); Whitney, W. D., Language and
the Study of Language (7th ed. New York 1910), and
The Life and Grotvth of Language (New York 1875);
Paul, H., Primipien der Sprachgeschichte (5th ed.
Halle 1920); Meillet, A., and Cohen, M., Les langues
du monde (Paris 1924); Kantor, J. R., "An Analj-sis
of Psychological Language Data" in Psychological
Rei-ien., vol. x.\ix (1922) 267-309; Pillsbury, W. B.,
and Meader, C. L., The Psychology of Language (New
York 1928); Boas, P., Introduction to Handbook of
Atnerican Indian LangtMges, Smithsonian Institution,
Bureau of American Ethnology, Bulletin no. 40, 2
vols. (Washington 1911-22); Miiller, M., Lectures on

.S"/v; !.tini^inii^i' in Rrhiuon to fhstorv and Society 517

169

the Sa'fnte of Language, 3 vols. (7th ed. Ix>ndon 1873);
Schnxidt, W., Die Sprachfamilien und Sprnchmkreiifn
der F.rJe (I Icidelberg 1926); nioomficld, I, , .-In Intro-
duition to the Study of l.aitt^agf (New York 1914);
Vosslcr, K., (Jeist unJ Kuttur in Jer Sprachr (Hei-
delberg 1925), tr. as Tlie Spirit of l.nut^aue in Cii-
JiiJtton (Londt)n 1932); Meillct, A., l.ti laugufs Juru
FFurope nouvrlle (md ed. F'aris 1928); Bally, Charles,
Le langitgf et la fie (Paris 1926); Gardiner, Alan H.,
The Theory of Speech and iMnguage (Oxford 1932).

Editorial Note

In: Encxclopacdia of the Social Sciences, vol. 9 (New York: Macniillan, 1^33). pp.
155-169. (Reprinted in: Edward Sapir, Selected Writings in Lani^iiaf^e, Culture.
and Personality. Edited by David G. Mandelbaum. Berkeley: IJniversilv of
California Press, 1949, pp. 7-32]

The following errors in the originall\ published \ersion ha\e been eorreeted
directK into the text printed here (page references are to the original):
p. 168. bibliography: Vendryes (correct: Vendryes)
p. 169. bibliography: I'europe (correct: V Europe)

518 General Linguistics I

LA KEALITE PSYCHOLOGIQUE
DES PHONEMES

Le concept de « phoneme » (unite qui a une signification fonction-
nelle dans la forme ou le systeme rigidement determines des sons
propres a une langue), oppose au concept de « son » ou « element
phonelique » en soi (enlite que Ion peuldefinir objeclivement dans
la tolalile du langage parle ou enlendu), devient de plus en plus
familier aux linguisles. La difficulte que beaucoup de ces derniers
paraissenl encore eprouver a les distinguer I'un de I'autre est
appelee h. disparaitre, quand on aura comprisqu'il n'est pas d'entite,
dans I'experience humaine, susceptible d'etre definie exacteinent
comme la somme ou le produit mecanique de ses proprietes
physiques, Certes, ces proprietes physiques sont necessaires pour
nous fournir, en quelque sorte, i'indice qui nous permettra d'iden-
lilier I'entile donnee comme un point qui a une signification fonc-
lioanelle dans un systeme complexe de rapports; mais Ton sait que,
dans un texte donne, il est possible de negliger comme accessoires
nombre de ces proprietes physiques, et Ton sait aussi qu'une
propriete parliculiere, possedant momentanement oupar convention
sociale une valeur sigoificative inaccouturaee, peut determiner la
definition de celte entite dans une mesure hors de toute proportion
avec son « poids physique », Pourtant, si Ion admet que, dans
I'experience, toutes ces entiles qui ont une signification peuvent
Sire ainsi modifiees a partir du donne physique par leur passage a
Iravers le tillre de la signification fonclionnelle, si Ton admet qu'il
est impossible d'etablir une echelle de sens nouveaux ou modifies
qui corresponde uniquemeo t a I'echelle des accroissemenls physiques,
on fait iraplicitement, consciemmenl ou inconsciemment, une

.SVv. I.(ini;nni^c in Ritation ii> Hisu>r\ and Society 519

24H

distioction enlre le phoneme el le son dans le cadre parlirtii. r li--
I'experience desipne sous le num de langage el aclualise par la
parole. Dire (juun phoneme donne ne se delinit pas coniplelofnenl en
termes articulaloires ou acousliques, mais doil s encadrer daos
tout 1 cnsomhle du systeme des rapports sonores propres a une
langue, ce n'esl pas.'au fond, plus elranpe que d'aflirmer qu»' Ton
ne nous a pas sullisarament delini un gourdin quaud on nous a dil
qu'il est en bois, qu'il a telle ou telle forme, telles ou lelles ditiieii-
sions. Nous devons saisir pour(]ui)i un ohjet a pen prts pareil a ce
gourdin, pen different d'aspecl, n'en est pas un, el pourquoi un
troisieme objet. Ires different de couleur, beaucoup plus long el
beaucoup plus lourd que le premier, est pourtant un gourdin.

Quelques linguisfes semblent estimer que le concept de phoneme
peut etre utile dans une discussion linguistique abslraite (dans la
presentation theoriqile de la forme dune langue ou dans la compa-
raison entre les langues apparentees), mais qu'il est peu adequai aux
realit^s de la parole. Cette facon de voir me parail lout a fail oun-
traire aux fails. De meme que seul un physicien ou un philosophe
delinit un objet en termes de concepts abstrails comme la masse, le
volume, la structure chimique, la position, seul un linguisle abslrail.
un phoneticien pur el simple, reduit la parole arlii-ulee a de simples
processus physiques. Four le physicien, les troisobjets de bois donl
il vient d'etre question sont tous les trois egalemenl dissemblables :
les ft gourdins » ne sont qu'intrusions romanliques dans la conli-
nuite severe de la nature. Mais Ihomme ignorant est beaucoup plus
siir de ses « gourdins » et de ses <« perches » que des objels sans
nom qu'il v aura Ii*'U de dctinir en lermfs physiques Hf u)t''U)e.
dans la parole, seule lobservalion attentive peul abstraire les posi-
tions phoneliques exacles et cela souvent aux depons de nos
intuitions phoneliques, phonemiques devrionsnous dire. Dans le
monde physique, le sujct parlanl et Taudileurpeu iiislruils ♦•mellent
des sons el les pergoivenl, mais ce qu'eux-mrmes sentt'iil lors(ju lU
parlenl ou entendent, ce sont des phonemes, lis oiganisenl les
elements fondamentaux de leur experience linguistique dans des
formes fonctionnelles et cstheliques delermin«'es. rharune d elles
etantdecoupee.dans la totalilecomplexede tons les rapporls sonores
possibles, par ses lois de rapports propres. Tour le sujet parlanl ou

520 General Linguistics I

249

pour 1 auditeur peu inslruits, les sons, c'est-k dire les phonemes, ne
ditrereni pas dans la mesure ou dilTerent des entiles de cinq ou six
polices, maisdans la mesure oil different les gourdins et les perches.
Si le phont'lifien decouvre dans le courantdu langage parle quelque
chnst' qui n est ni un c gourdin » ni une « perche », il a le droit, en
Liril fjue plioneticien, d'elablir une entite « a mi-cliemin enlre le
;^'ourdin el la perche ». Au point de vue fonclionnel, toutefois, cette
enliffiest une fiction, et le sujet parlant ou lauditeur peu iostruils
ne sonl pas seulement amenes a la classer, d'apres lesressemblances
qu'elle presente, soit dans les gourdins, soil dans les perches, mais
encore I'entendent et la pergoivent ainsi.

Si I'allitude plionemique est plus essenlielle, au point de vue
psychologique, que I'attitude plus strictemenl phonelique, on devrait
pouvoir la decouvrir dans les jugemenls spontanes enonces par des
sujels qui possedent complelement leur langue au point de vue
pratique, niais qui n'en ont pas une connaissance ralionnelle ou
consciemmcint sysleiiialique. On peut s atlendre a voir se produire
ties !t erreurs » d'analyse, au jugement du moins de Tobservaleur
« sopliistique », erreurs qui seront caracte'risees par I'inexactitude el
rinconsistance phoneliques, mais qui Irahiront un penchant pour
Texactilude phonemique. Ges « erreurs », souvent negligees par le
linguiste qui opere sur le concret, peuvent fournir des temoignages
probants quant a la realite dynamique de la structure phonemique
du langage.

Au cours dune longue experience dans la notation et I'analyse de
Ungues non ecrites, indo-americaines ou africaines, je suis arrive a
reunir des preuves concrete.sdu fait que le sujet parlant peu instruit
n enlend pas des elements phoneliques mais des phonemes. Ce pro-
bleme prend la forme d'une experience pratique, lorsque Ton desire
apprendre a ecrire sa propre langue a un indigene intelligent, c'est-a-
dire a un indigene capable de comprendre et d'ecrire assez bien
I anglais et doue, en outre, de quelque curiosite intellectuelle. La
dilTiculte de la tache varie, naturellement, avec rintelligence de
1 indigene, avec la dilficulte intrinseque de sa langue, mais elle
vane egalement avec le degre d' « intuition phonemique » du
iiiailre. Beaucuup de linguistes bien intentionnes ont fait, a cet
egard, des experiences decevantes, avec des indigenes tout a fait

Six: lAini^iiaj^c in Kclatioti to Hi\i(H\ ami Sa ui\ 521

i5u

iolelligeiits, .sans jiimais soupv'oruier <iii.- la fault; crrii t-lail pii^ a
rindigene ma-s a eux-iiuiiies. II lsI exlr6memenl diflicile, sin<»n
iiiipo!?sil)le, d'aftprendre a un indigene a Ifiiir coniple d*- vana.
lions phonetiques puremeni niecaniques, varialiuns qui n onl,
pour liii, aucune realile pliun»'niique. Le mallre, qui arrive avec
uiie gatnme Louie prele de possibililes plioneliques absolues el
qui, inconscienimeol, en depil de sun apprentissage, lend k
projeter les evaluations plionemiques de sa propre langne dans
ce qu'il enlend et note de la langue etrangere, peul aiseinenl
derouler un indigene. L'indigene s'apervoit que ce qu'on lui
apprend « tinle » coinme ce que ses inluiliuns phonologiques lui onl
deja appris, inais il se sent mal a I'aise quand on lui nionlre des
dislinctioDS puremeni phonetiques, distinctions qui lui SHjuljIenl
assez reelles quand il fixe son allenlion sur piles, rnais qui dispa-
raissenl conlinuellement de sa conscience, parce que ses (( inluiliuns
phonologiques » ne confirment pas leur realite objeclive.

Parmi les nombreux fails d'audilion el d'ecrilure plionemiques que
j'ai pu observer, au cours de mon experience avec des indigenes el
des etudiants, j'ai choisi cinq exemples que jeludierai brievemenl
et que j'opposerai a I'audition et a I ecrilure phonetiques. On ot»ser-
vera que nous avons, dans chacun de ces cas, la preuve nelle d'un^
reinterprelation inconsciente des fails objeclifs, reinlerpretation
causee par une disposition phonologique perturbalrice qui se Irouvc
mal adaplee a ces fails.

Quand je Iravaillais sur la langue paiule meridionale, langue
du sud-ouest de I Ulah et du nord-ouesl de lArizona, j'ai pas"»<?
quelque temps a essayer d apprendre a ecrire phont'liquemenl sa
langue a mon inlerprele indigene, jeune h(»inme d'lntelligence
moyt^nne. Le paiule meridional est. au poinl de viit- ptniiinlngiqur.
une langue dune rare complication, el comme a lepoque jiosislais
beaucoup plus sur la correction phonetique que sur I'exarlilude
plionemique, je ne crois pas que je serais arrive a la lui ap[)rrndrc
assez bien pour satisfaire mes exigences, m«''me si j'avais roosacre
k eel elTort beaucoup plus de te[ii[ts qur je ne I ai fail, (iocntn*-

522

General Linguistics I

25 1

pxemple de mot comparativement simple, je cho\s\s pd- ^{i^a' « a I'eau »
(plosive labiale sourde ; a long accentue ; spirante bilabtale sonore;
a hre'" non aoceolu^; aspiration finale). J'appris k Tony a diviser le
mol en svllahes et 4 decouvrir, par une audition attentive, quels sons
cntraient dans la composition de ahacune de ces syllabes et dans
riiifl ordre ils y entraient, puis a essayer d'ecrire le symbole exact
pour cliacun des elements phonetiques decouverls. A mon grand
etonnement, Tony divisa alors en syllabes pa-, repos, pa'. Je dis
f^rand elonnement, car le paradoxe m'apparut lout de suite : Tony
n' « entendait » pas d'apres les sons reels, la bilabiale sonore etant
objeclivemenl Ires differente de la plosive initiale, mais d'apres une
reconstruction etymologique : pa-- « eau » plus la postposition *-pa'
« a ». Le leger repos qui venait apres la racine avait suffi pour
e^arler Tony de la forme phonetique exacte de la postposition et
pour I'amener a une forme possible en theorie, mais n'exislant pas
dans le cas present.

Pour expliquer le comporlement de Tony, comportement qui
n'elait pas du a la negligence, ni a une tendance des sujels parlant
cettii lannuc « a confondre les sons » — pour reprendre une vieille
foiinulc — , iiuus dcvons avoir recours a la phonologie du pamte
indriditjnal.

Le Iraiteinent des consonnes plosives peut se resumer dans le
tableau suivant :

Lubiales

Deulales

(iullurales . . . .

OuUurales labiali-
st'es '. . .

I.NITIALES

P
t

k
kio

1. Spu-an-
tisees.

POSTVOCALIQUES

2. Naaali-
sees.

mp
nt
yk

rjkw

i. Geminiea.

a. A pecs

voyelle

sonore.

A-
k ic

b. Apres

voyelle

sourde.

P
t

k
kw

Los formes posLvocaliques des plosives des types I, i elo subissent

Six: I.iini^inii^c in Rchiiion lo HiMorv mul Society 523

25 •>

une nouvelle modification dcvaiit une vcjyelle sourde : les spirantes
sonores deviennent spirantes sourdt'S (<l>. R, x. xH'; el les plosives
nasalisees el geminees {inp' , jr-\ nr. ( • ; tjk- . A'; ijkW , k H')
deviennent aspirees. II est impossible de (loniifi i( i une idee syslema-
tique des processus phonologiques qui ameneni les echanges de son
a I'inlerieur d'une serie arliculatoire donnee, mais il est important de
savoir que les plosives nasalisees el geminees ne peuvenl se Irouver
qu'en position postvocalique el sonl largemenl delerminees par la
nature de I'elemenl, racine on sufti.xe, qui les precede, element qui
peut ^Ire considere comme ayant un pouvoir inherent de spirantisa-
tion, de nasalisalion ou de gemination. La racine;>rt - est une racine
spirantisanle el la spiranlisation d"un *-pa' theorique « ^ « en -^a'
est parallele k la spiranlisation de po- « chemin » en -^j - dans un
compose comme pa'-^o - « chemin deau ». En d'autres lermes, la
forme de celle langue est telle que des exemples du type pj - : -^o -
conduisenl au rapport */>«' : -prt'-et que, tandis que *pa'- « a » na
pas d'existence reelle comme element independanl mais doit
toujours etre aclualise dans Tune des Irois formes postvocaliqnes
possibles, son existence Iheorique apparail soudain quand le
probl^me dediviser lentemenl un mot en syllabes est pose pour la
premiere fois a un indigene. II apparail alors que le -^a' de la langue
parlee est, en tant qu'enlile syllabique independante sans syllabe la
precedant immedialemenl, pergu comme un pa' phonologique, dont
ildiffere a deux egards phoneliques importanU (consonne sonore
au lieude consonne sourde, consonne spirante au lieu de consonne
plosive).

Tout ceci a une influence importanle sur la construction dune
graphie aussi correcle que possible du paiute meridional, si par
« aussi correcte que possible » nous entendon.s non pas la p>lus
exacle phoneliquemenl, mais la plus conforme au systeme phonolo-
gique de celle langue. En fail, il y a des raisons de supposer. a la
fois d'apres les preuves internes el les donnees comparatives, que la
forme spiranlisee d'une consonne est sa formf normair ft pnmaire

\. Le signe j repr^sente le son 7 (dans Tail, ach) : VV reprtsenle le u- sourd
2. Ce 'pa' tht'orique. apparaissiinl seulenionl sous la fortne 'ui', -mpa , -pa'

en position postvocaliqut, ne doit pus <^tre ronlondu avec Ic -pa' serondair.-

(type 3 b) <i-p a' (type 3 a).

524 General Linguistics I

253

;ipres une voyelle et que ses formes nasalisee et geminee sont dues
a la reapparilion d'anciennes consonnes, nasales ou autres, qui
avainnl disparu dans la forme archaique de I'element qui las
pra'ctidail'. II s'onsuit que la poslvocalique -^- est plus etroitement
;i|>|);irt;ni<'<', fonctionnellemenl, a la simple initiate jd- queue Test la
postvoo.ilique -p- (apres une voyelle sourde), qui doit toujours elre
iiitt'r|irt'lee comme une forme secondaire de -p--. Ces rapports sont
hn^voinent indiques dans le tableau suivant des formes theoriques
nun finales :

(iUAl'HIK I'HONKTIUUK GRAPHIE PHO.XOLOGIQUE

i . pa- pa-

i. pa|5a- papa-

3. pa*A-- papa-

4. papa- papa-

5. pApa- papa-
t). pap-A- pap a-

l,a j^rapliie plioneiique est plus complexe et, en un sens, plus
t'xaole, iiiais olle va a Tenconlre de la nature de la langue sur un
point capital, car elie identifie le second j9 du typeSavec \e p initial,
re (jui est incorrect au point de vue phonologique. D'autre part, la
graphie phonologique ne sert a rien k celui qui ne possede pas la
phonologiede la langue, car elle aboulit, ou parait aboutir, a une
prononciation incorrecte qui finirait, a la longue, par rendre celte
langue, ainsi lue, tout a fait inintelligible a un indigene. Toutefois,
pour I'indigene un peu iiistruit, ['equivoque n'est pas grave, car les
formes phonetiques ne decoulent des formes phonologiques que par
I'application de lois phonetiques purement mecaniques: la spiranti-
sation, le changement des accents, la perte de la sonorite. 11 n'est
pas necessaire d'etudier ces lois^ ici, mais on peul indiquer leurs
elTets dans le tableau suivant des formes finales theoriques :

1. Lanalogie avcc la liaison frani^aise et, plus encore, avec les trois types de
Iraileinenl des consonnes en vieil irlandais (consonnes spirantist5es ou « aspi-
rties «, consonnes nasalisees ou « eclipsdes », consonnes geminees) est apparenle.

-. A rcpresente le a sourd.

3. EUos sont exposees en detail dans E. Sapir; The Soulhern Paiute Language
{I'roceedings of the Amei-ican Academy of Arts and Sciences, vol. LXIV, I, 1931).

.S7.\v l.iini^iiiii^c in Rclntion lo //istorv mui Sociclv

525

r.KAI'HIE I'HO.NKT1i.iI;K

pd4>A

pa''^d*

p(i*/l

p« iia' n a I'eau «

r> .

pdpA

pApd'

7 .

pd'pA

pdpa^

vin'^A'VA

iu.

ma'^jd'^a'

11.

T7jfl,:;« 't'A

12.

inald''^a'

13.

maftdpA

vxa'^ap'w

Id.

maid- p- A

16.

ma'^dApa'

17.

MApd<PA »

18.

MApd'i^a'

19.

MApd *A

ao.

MApd-'^a'

21.

MApdpA

22.

MApdp-a-

23.

MApd'pA

24.

MApdApa'

(.MAFHIK fHoNOI.fXilyfl

papa

papa-

pop- a

pnpa'

pap'a

pa-pa-

viapapa

mapapn'

viapapa

mapa pa

viapapa

viapap-a

mapa pa

mapa pa-

mapapa

viap'apa'

map'a'pa

map' a pa'

map'ap-a

map-ap-Q-

map'a-p'a

map'ap a-

Evidemmenl, dans une langue comme celle-ci, les spiranles.
sonores ou sourdes, et les voyelles sourdesnt^ sont pas des ptioneme.s.
mais siiDplenient le.s reflexes phon(^tiques de plo.sives ou de voyelles
sonores dans des conditions dynamiques delermint^es. Les consonnes
longues et les voyelles longues sont des sous-phoDemes Les pre-
mieres resuUent de phonemes simples (plosives) et soni l.-i inisc en
(Kuvre de certaines possibilites phonologiques el morphologiques
dans des syliabes donnees, possibilites qui existent on ont exisle
jadis. Les dernieres se decomposent phonologiquemeul en voyelle
br^ve pins voyelle breve, c'esl-a-dire en deux syllabes dont chacune
a la longueur de liuiile de longueur (mora) et dont la secouilf ruiw-
mence par une consonne zero.

Le paiute meridional est done une iangur dans laffuellc une

1. .*/ reprise nle le m sourd.

y,i(^ General Linguistics I

2sr»

structure phooemique particulierement simple est aclualis^e par
une structure phonetique particulierement compliquee. L' « erreur »
(Je Tonv marque, k son insu. cette opposition..

Quand je travaillais sur le sarsi, langue athabaskeenne d'Al-
berla, Canada, je cherchai a resoudre le probleme suivant : cer-
tains mots qui semblaient homonymes I'etaient-ils reellement ou
presentaient-ils quelque legere difference, inappreciable immedia-
tement? Deux de ces homonymes — en apparence du moins —
^taient d'lni « celui-ci » et dlni^ « cela fait du brujt ». Au debut de
notre travail, je demandai a mon interprete, John Whitney, si ces
deux mots lui paraissaient avoir le meme son, et il me repondit
sans hesiter qu'ils etaient totalement differents. Toutefois, cette
affirmation ne prouvait pas qu'il etit objectivement raison, car il
e3t possible que des mots parfaitement homonymes donnenl au
sujel parlant rillusion de differences phonetiques, k cause des diffe-
rents contextes qui les encadrent ou k cause de leur position diff^-
rente dans leur sysleme paradigmatique respectif-, Quand je lui
demandai en quoi consislait cette difference, il eut du mal a me
repondre, etplus il se repetait les mots, moins il percevait neltement
leuis differences phonetiques, Cependantil paraissail, tout le temps,
parfaitement certain qu'il existaitune difference. A diverses reprises,
je crus percevoir une legere difference phonetique, par exemple :

i. Lacoeot grave repr^sente un ton bas, I'accent aigu un ton haul. Le sarsi
lisl uno langue a tons.

2. .Ainsi, en anglais, le mot led (de : to lead, men(!r) (ei : I led him away) est
per.;u cooiine ayant une voyelle deriv^e de la voyelle de lead (ex. : I lead him
away) et n'est par consequent pas homonyme, psychologiquement, aveo le mot
tfud (plomb) dans lequel la voyelle est pergue comme primaire, non comme
derivee (cf. en outre : « the leading of the windowpane », « the leaded glass »,
« the different leads now recognized by chemists »). L'homonymie de led
iincnais) ot lead (plomb) est done d'un autre ordre, au point de vue psycholo-
gique. que l'homonymie de yard (cour, garage : ex. : « He plays in my yard »)
el de yard (mesure de longueur : ex. : « I want a yard of silk »), car ces der-
niers inots entrent dans des systemes paradigmatiques sensiblement parall^les
(ox. : « Their yards were too small to play in » : « I want two yards of silk » :
« yard upon yard of railroad tracks » : « yard upon yard of lovely fabrics »). II
est probableraent plus facile au sujet parlant peu instruit qui ne salt «5peler qi
ieU, ni lead de se convaincre quil y a une diff»5rence phonetique entrc ces deux
mots que de penser qu'ily en a une entre les deux mots yard.

Six: Lnni^uaf^i- in RcUttioti to Hisiorv u/ul Soiifiv 527

256

l** le -;i/ de « celui <"i » avail iin ton lt'g» rcineiil plus ba.s (jue le ni
de « cela fail du Lruil » ; "2" il y avail iin It-gtr accent sur le d(- de
« celui-ci )) (analyse : racine di- <» celui » plus suftixe-nj « uiu- p«'r-
sonne ») et, de meme, un leger accent sur le ni de « cela fait du
bruit » (analyse : prelixe di- plus racine verbale-n*]; 3" le -/jj de
(( celui-ci w se terminait par une voyelle pure, suivie ou non d une
breve expiration, landis que le ni de « cela fail du bruit » etait
suivid'une expiration plus marquee el etait proprement -Tii'. John
exanaina ces suppositions el les accepta parfois a contre-ccEur, rnais
on voyait facilement qu'il n'elail pas intinneinenlconvaincu. Laseule
supposition tangible qu'il fit lui-meme elail evidemment erronee : le
-ni de (( cela faildu bruit » se serai Iter niine par un « / ». John aflirrnait
qu'il K sentail un / » dans la syllabe, el pourtanl, quand il eut refail
I'experience plusieurs fois, il dul admetlre qu il ne puuvait pas
entendre de « / » el qu'il ne senlait pas sa langue en prononcer un.
Nous diimes abandonner le problenie, et j'en conclus, a part moi,
qu'il n'y avail, ci vrai dire, aucune difference phonelique enlre ces
mots et que Jolin essayail de se convaincre qu'il y en avail une.
simplenDent parce que ces mots elaienl Ires differents, lant par leur
fornie grammalicale que par leur function, et que pour lui celle
dissemblance devail necessairemeni enlrainer une difference phone-
lique.

Je no connaissais pas assez, alors, la phonologic sarsi pour
comprendre la myslerieuse theorie du « / » Plus lard, il mapparut
qu'il existe en sarsi des types de voyelles iinales phonologiquemenl
dislincls, voyelles deuces ou simples el voyelles a possibilile couso-
nantique, c'esl-a-dire voyelles ^uivies a lorigine dune consonne
disparue dans la forme pausale du mot, mais qui reapparail quand
ce mot a un suflixe commen(^anl par une voyelle ou rionl on devine
la presence dans d'aulres phenomenes. de phonelique combinalnire.
Une de ces consurines en voie de disparilion est le /. donl en jieul
considerer le -' coiunie une fonne allaildie Cr il se Imuve que I'du
prononce loutes les voyelles tinales avec une expiration dans la
forme pausale du mot et qu'il n'y a pas de difference objective enlre

-' secondaire que Ion peul symboliser par -Cj, zero phonologi
quemenl, et le -', elyinologiquement organique. qui pent affecter
d'elemenls suffixes cerlaines consonnes qui le suivenl, ou. dans

528 General Linguistics I

257

c»Ttains cas, devenir une autre consonae comrae le t'. Le -ni de
« ct'lui ci ». plioneliquement -ni\ dans la forme pausale du mot, est
phonoIouMqiiernent un simple -ni; le -ni de « cela fait du bruit »,
phoiKlitiniment -ni' dans la forme pausale du mot, peut se repre-
«ii>nti"r plionologiquement comme -ni' (-nit'-). Nous comprendrons
iiueiix CCS fails si nous etudions la nature de ces deux syllabes et si
nous vovons comment elles se comportent lorsqu'elles sont suivies
du siiflixe relatif -i « celui qui... » et du suffixe inferenliel -la^ « il

apparait que... » :

plus —i plus — la

dini n celui-ci » dind '^^ dlnila

dini « cela fait

du bruit)) dinWi dinlia^

Nous vovons tout de suite que dini « celui-ci « se comporte comme
un mot termine par une voyelle douce (temoin la contraction de
/ + i en une voyelle prolongee et 17 non modifie de -la), tandis que
dhii « cela fait du bruit » se comporte comme si la voyelle finale
avait une possibilile consonanlique sourde, qui s'inscrit en partie
comme -' {-'-la se changeant comme toujours en -la), en partie
comme -t'-.

II estclair que, bien que John ne fut qu'un phonelicien amateur,
sa phonologie elait raffinee et exacte. Sa reaction indiquait son
intuition que d\ni « celui-ci » = dini, que dini « cela fait du
bruit » =:p d\ni' et que ce dernier -ni' = -nit'. La certitude qu'avait
John d'une difference en presence de I'identite objective est analogue
au sentiment qu'aurait I'Anglais moyen que des mots tels que sailed
et soared ne sont pas identiques phoneliquement. II est vrai que
sawed el soared se represenlent phoneliquement I'un et i'autre par
sj-d'*, mais les formes en -ing de ces deux verbes {sawing et

1. Labsence d'accent indique que cette syllabe a un ton moyen.

2. a" est un a prolong^ qui consiste en un a- long suivi d'un a faiblement
rearticul^. Les syllabes de cette nature proviennent, en sarsi, de la contraction
d"aticiennes voyelles finales avec la voyelle ajout^e qui les suit.Le changement
do ([ualitti de -i en -d-" est du a des facteurs historiques. -ni a une personne »

estun ancien *-ne (avec voyelle r»?duite),le relatif-iest rancien*-e; deux voyelles
r«i(Iuites se contractent en une voyelle ouverte longue *-e'' ; de meme que I'e
ulhabask^en devientl'a sarsi, cet ancien *-£•' devient en sarsi -a-'.

3. I est la spirante sourde I, comme dans le gallois 11.

4. Ces remarques sappliquent a I'usage anglais normal et non a Tusage am^-

Six: lAifii^iiiii^c in Relation to llistorx lunl Society 529

258

soaving), phoncliquemeiil so'-itj el <».>T-iy, el des foriucfe df |)lira5t>
comme « Saw on, my boy! ))-el « Soar rulo llicsl<\ I », concoureni u
<produire lionpression que le sxd de aancil ^^ >j*-J, inai-. qut- |c
so'd>de soared = sj'V'd. Dans le preniifr oas, zero = zero, (Liii.s le
second, zero =r. Les gens inslruiLsqui disculenl ces quewlious suns
-avoir fail de linguislique t;onsidt}reul luujours la grapliie loimnc
responsable de ces didiTonces d'approcitilion. C'esl une erreur, huns
nul doule, citez la plupurl dcs gens du inuios, el c'cht -mi.lin- la
charrue avanl les Lauifs.

Si I'anglais n'elail pas une kingue dcrile, la difference phonolo
gique, delerminee par les sysl(>ines fonclionnuls dcs sons, cnlrc de.s
doublets lels que saicfd el soured se percevrail qiiand in(>nic coinim-
une illusion coHeclive, coinme une verilable dilTorencc p|joncHi(iue

III

L'eleve indo-americain le plus brillant que jaie €u en phone-
liqne pratique est Alex Thomas. Alex Thomas ecrit sa langTje maler-
nelle, le nootka', avec une alsance remarquable et une preci-
sion admirable. La graphic d'Alex est, uaturellement, toujours
phoDologique dans son essence, et e'est surtout d'apres I'elude de
ses textes que j'ai nppris a estimer, a sa juste valeur, la diflerence
psychologique entre un son et un phoneme. Quiconque connail le
tnecanisme phou^tique du noolka pent facilemenl reproduire sa
graphic. Ainsi hi, phonologiquement parallele a si ou ni, se
pTononce reellement hi, avec une voyelle qui est beaucoup plu.s
pres de Ye de I'anglais met que de celle de sit. Ceci vient de la
nature purticuliere des consonnes laryngales qui favorisout un tim
bre da et transforment les voyelles suirantes i el k respeclivemenl
en I el o. Les grapliies hi et hii sont Ires claires, car il ne pent .'\is
ter phonologiquement de syllabediFlincle du type hi ou ho-.

Une autre pa-licularite mecanique du noolka. cesl ralloogemenl
des consonnes apres une voyelle breve suivio d urip aulro voyelle.
Cette longueur purement mecanique n'a aui une signilicalitm im>rj»lio-

i. Cotle languc est parlee sur la cdle occidenlale dc Tile do Vancouver.
Colombie Britannique.
2. h est une spiranle larsngalc sourdc, presquc ideiilique a larabe ha.

530 General Linguistics I

259

logiijue oa plionuiogique et la graphie d'Alex I'ignore. Ses hisi'k et
/ii.sa- d<jivi?nt done oormalement se pronoocer kis;i-k' et htS'a-. II
arriv.' parHjis pourlant qu'uoe coosoQue longue, en particulier s* et
S-, naisse de la reocoBtre de deux eonsonnes morphologiquement
di>liu< h's (par exemple : s+ s>s- ou s -f- s > S" ou, plus rarement:
S -r- s oa s -j- s> S'). Dans ces cas-loi, on n'a pas ^impression queia
<:on>oime longue soil rallongenient mecanique de la conson^ne
simple'; un a le sentiment qu'elle est an grouf)€ de deux consanjies
idenliques. Ainsi Alex ecrit, par exemple, ts^vqsitlassatlni^ cc nous
n'v sommes alles que pour parler », ce qui se decompose en tsiq-
sitl-'as-sa-Ca}tlni. Le s de -as « aller en vue de » et le § de -sa
« simplement M, « seulement ))^ gardent leurindependance phono-
logi({ue, et i'inlervocalique normale -5- de -'as at I s'interpr^tecomme
■SS'. De meme, kwissila « agir differemment » se decompose en
k(cis-sila Pourlant, il ne semble pas y avoir de veritable difference
phonetique enlre V-s-, phonologiquement -s-, de mots lels que tlasatl
'< le baton se dresse sur la plage )){tla-satl), qu'on prononce tlas'all,
el i'-s-- de -'assail ci-dessus. Nous avons de nouveau ici des pheno-
inenes plioneliques idenliques qui regoivent des interpretations pho-
nologiqucs difTerentes.

Dans le premier systeme de graphie qu'apprit Alex, les plosives
el les afTriquees n'etaient pas traitees comme les nasaks glotla-
lis.'es ou les semi-voyelles. On representait les premieres par
/>!, t!. k!. k'.w, q!,q!w, ts!, tc! i=ls), et L! [=tl) ; les autrespar
')}}, n.'u, ct '?/.'. Ce fart s'explique parla Iradilion. Les plosives glolta-
lisees el les affriquees glottaiisees, en tant que types particuliers de
eonsonnes, avaienl ete de bonne heure decouvertes dans differentes
langues par M. F. Boas et deeriles comme des xc fortes », c'est-a-dire
des plosives el des affriquees « prononcees avec une intensite parti:
culierement forte d'articulation ». Les types 'm, 'n, 7, 'y,ei\vne
furenl decouverts que beaucoup plus tard parM. Boas, d'abord dans
lo kwakiull, et decomposees en nasales, sonores lalerales et semi-

- J ai legiremcnt modifl^ la graphie d'Alex pour qu'elle s'accorde avec ma
Krapliie actuellc, raais ces changemenls sont de pures substitutions mecaniqaes
ct nairoaenton rien le probl^me. q est la v^laire k arabe, tl est une aUriquee
lalcrale. tL en est la forme glottalis^e.

Six: Luni^uui^c in Rchuion to //istorv uml Sinict\ 531

260

Toytlles precedees immedialement dune orclu<iuu glollale. La
graphie de ces cousonnes (decouverles plus lard en Isimshian, eu
noolka, en haVda el eu plusieurs aulres langucs, rnais pas aussi
repandues que celles que Ton appelle les « forles ») rappelle
comment elles ont ete formees, mais la graphic des plosives glolla-
lisees est puremenl conventionnelle el n'indique en aucune fa<^on
leur formation, sauf en ce qu'elle monlre qu'il a fallu uue energie
plus grande pour les prononcer^ Au poinl de vue plionelique, alors
qu'en noolka la formation des affriquees el des plosives glollalisees
est approximalivement parallele a celle des consonnes sonoantes
glollalisees, elle ne Test ni ne peut I'elre entierement. En ce qui
concerne, par exemple, le p glollalise, noire p actuel et ancien p !, il
se produit une occlusion synchronique des levres et des cordes
vocales, unechambre k air fermee etant ainsi obtenue, puis il y a
explosion brusque de Tocclusion labiale, pause el enfin ouverture de
I'occlusion gloltale, C'est I'ouverlure de I'occlusion labiale (ou de toule
autre occlusion orale) avant celle de I'occlusion gloltale qui donne a
ces consonnes leur caractere apparent de « dies »-. D'aulre part, en
ce qui concerne Ym glottalise, notre 'w, alors que les occlusions
labiales et glotlales sont synchroniques, comme dans le cas prece-
deDt^ Tocclusion glottale doit cesser au point initial de vocalisation

1. Ceci, soil dit en passant, n'est pas n^cessairemcnt vrai. Dans certaines
langues, les plosives et les affriquees gloltalist^es semblent demander une inten-
site plus grande d'articulation que les consonnes non-glotlalis^es correspon-
dantes; dans d'autres langues, 11 n*y a pas de difference notable en ce qui con-
cerne « rintensite d'articulation ». Dans les langues alhabaskeennes que j"ai
entendues (sarsi, kutchio, hupa, navabo). les plosives et les affriqu*?es sourJes
aspirt^es (types t' , k' , ts') sont beaucoup plus « forles » par nature que les
consonnes g!ottalis(5es 'correspondantes (par exennple : t', k', I's). 11 n'y a pas
nticessairement corrtilation entre le type laryngal d'articulation (sonore, sourde,
glottalisee, ou ces differents types avec aspiration) et linlensilt* d'articulation
(fortis, lenis). En ce qui concerne le noolka, il ne ni'a pas seniblt- que les plcv
sives et les affriquees glottalist^es (les n forles > de M. Boa$) fussent Ir^s dilTt'-
rentes en inlensitt^ des plosives et des affriquees ordinaires. Dans les langues qui
reconnaissent une difference phonologique entre I'emphatique etie non-empha-
tique et posstident, en nieme temps, des consonnes gloltalis^es, il n'y a pas de
raison pour qut ces derni^res n apparaissent pas a la fois dans les formes
emphaliiiues et dans les formes non emphatiqucs. Comme la monlre le prince
Tribetzkov, quelques-uncs des langues du Caucase septentrional ont, en fait,
des plosives el des affriquees glollalisees. emphaliques el non-emphatiqucs.

2. Ces consonnes sont, semblet-il, identiques aux i< ejeclives » de Daniel
Jones. II existe un autre type de plosive ou daffriqut'e glottalisee, moins com-
mun, dans lequel I'explosion orale el I'explosion glottale sont synchroniques.

3. La prononciation de 'm, '«, 'n et 'y comme une occlusive gIolla!e (') suivic
de tn, n, u- et y est repoussee par loreille noolka comme etant incorrecte.

532 General Linguistics I

261

de Yin. En gros, par consequent, p peut se decomposer en ;? -f',
landis que m peul se decomposer en '-f m. Ainsi, une difference de
erjinhic t''lle que p! s'opposant i 'm, herilee par moi de la tradition
aincri' anisic, n'olait pas injustifiee du point de vue purement

pliont'tiqiio.

Nous (;n arrivons maintenant a une experience phonologique
inluilivc. qui nous permettra de savoir si p et 'm sont ou ne sont
p;is (les consonnes de meme type. Alex a appris a ecrire tres facile-
inent les consonnes de type p et ts (^nos p! et ts! primitifs), par
example : papi' v oreille » {p vim'iVw emeni p!ap!i)^ tsaak (.<■ ruisseauw
(primilivement tslaak). A mon grand etonnement, Alex risqua m!
dans des mots tels que 'ma 'mi qsu « le frere ou la soeur ainee »,
qu'W ccriyait ml a77i!iqsu- En d'autres termes, nous avons ici une
preuve evidente de la realite phonologique d'uneclassede consonnes
glotlalisees comprenant a la fois le type p (avec explosion initiale de
rocclusion orale) el le type 'm (avec explosion initiale de I'occlusion
glotlale). Unegraphie toujours d'accord avec la phonologic exigerait
/» et m (ou;;/et wi/j. Repetons-le encore, 1' « ignorance » phonetique
d*un indigene sans instruction se montrait plus exacte, du point de
vue phonologique, que la « science » des savants. II est aise de
justifier phonetiquement V « erreur » d'Alex. Les consonnes du type
p sont exactement analogues aux consonnes du type 'm, parce que :

i° On les (rouve toutes au debut d'une syllabe, et, puisqu'aucun
mot ne peut commencer par un groupe de consonnes, ceux qui
parlent nootkareconnaissent que les sons p et 'm ne sont ni Tun ni
I'autre analysables en unites phonologiques. En d'aulres termes, on
ne peut pas davantage isoler I'occlusive glottale dans 'm que dansp.
De meme, les affriquees et les affriquees glottalisees ne se decora-
posent pas en unites phonologiques.

^2" Toutes les consonnes peuvent apparaitre en fin de syllabe, sauf
les plosives glottalisees, les afTriquees glottalisees, les sonnantes
glottalisees {'m, 'n, '?/, 'w), les semi-voyelles {y, w), les nasales
('«, n'). I'occlusive glottale ('j, etl'A. Cette regie range de maniere

1. "» etn peuvent etre suivis d'une voyelle murmur(5e de timbre i qui est une
lorine rcduite de a, u. ou i. Les syllabes ou demi-syllabes de types W ou «'
sont prcc^d^es pari, produit assimil^ de a, u ou i ; in' et im' rdsultent done, en
partiede series de types ama, umi, anu. Les simples -am ou -an deviennent
-ap, -at. ^

Six: Luni^udi^c in Rclaiion to History and Society 533

262

plus precise ]es consonnes du lype 'in avec Ics consonnrs du type p.

3° De nombreux suflixesonL pour effcl de « durcir »• la consonne
qui les precede, en d'aulres termes de raffecler d'arliculalioa glol-
tale Sous rinfluence du processus de « durcissemenl », ji. t, k
deviennent //, /', k\ tandis que in et n devienoenl 'tn el ';*.

Par exemple, de raeme que les suffixes '-a'a- i'-a-'a) « sur les
rochers » el '-ahs « daas un recipient » changenl les racines winap-
« demeurer, resler » en ivi nap'- (ex. : un nap'a'a « resler sur les
rochers ») et wik- « ne pas ^tre « en laik'- (ex. : wik\ihs « tie pas
Sire dans un recipienl, la pirogue esl vide »), — de raSnne t'lum-
(alternant avec flup) « avoir chaud, elre chaud » devient t'Lu'vi-
(alternant avec r/?</)'-) (ex. : Vliima-'a « avoir chaud sur les rochers »,
t'lu'77iahs « elre chaud dans un recipient, il y a de I'eau chaude » ;
cf. t'lupH tsh « ete, saison chaude » = I (up- -{- '-i'tch) el kan-
« s'agenouiller » (ainsi : kanil « s'agenouiller dans la raaison »)
devient kan- (ainsi : ka iiahs « s'agenouiller dans une pirogue »).
Comme il ne sernble pas exister de racines terminees par h ou '.
le groupe '771, 'n, \o, 'rj^ reste comme fonctionnellemenl appa-
rente au groupe m, n, iv, y, dans la mesure oii le groupe du
type p s'apparenle au groupe du type p. En d'autres termes, la
morphologic confirms de maniere decisive le rapport phonologique
p : p z= m : 'm. C'esl, me semble-til, ce sysl^me phonologique
implicite qui fail que le son 'in a paru a Alex suffisamment semblable
au son p pour justifier une graphie analogue de ces deux sons. Dans
d'autres laugnes, dans lesquelies lessons onl des relations morpho-
logiques et phonologiques differentes, un tel parallelismede graphie
ne sejustifierait pas et la difference phonetique reellemenl exislanle
entre 'm et ;; aurait une importance psychologique de luut aulre
portee.

Dans des conferences de phonetique pratique quo j'ai failes
durant plusieurs annees, j'ai si souvenl reniarqu»'' rillu.si««n phon^-

1. Ce terme esl enipi unLc a la description que donne M. Boas d un ph^oomine
k'wakiull Equivalent.

2. Le signe ' indique le » durcissement » provoqui^ par le suTlixe.

3. Les dt^tiili phonologiques iaipliquaot tu el y el leurs rapfwrts avec u- el y
et les aulres consonnes sonl trop compliqufis pour ^tre bri'''vemenl cxpos«?s ici.

534 General Linguistics I

263

tiaue line je vais cxposer qu'il m'est impossible de ne pas formuler,
pour r.-Npliqiier, une theorie phonologique genorale. J'ai note que,
lorsfiin; Ics "Jladiants ont appris a recorrnaiire I'occlusive glottiale
coinui'' inio unite phonelique, beaucoup d'eoire eux ont tendance a
Tcnteii'lri} apres un mot termine par une voyelle br^Ye aceentuee de
timbi'^ clair (par exemple, a, e, e, i). Cetle illusion ne se produitpas
aussj souvent en ce qui concerne les mots termines par une voyelle
Jori'Mio OH par une voyelle obscure de qualile assez mal definie (p)
ou par une voyelle inaccenluee. Ainsi, lorsqu'on dicte un mot
n'avanl pas de sens comme sms ou />j7a, il arrive qu'il soil parfois
inal saisi el ecril smz' et pild', mais il semble que la tendance a
enlondre une occlusive glottale finale soit moins nette dans des
raols commQ pila ou pild\ Comment expliquer ce type etrange de
(( surperceplion » auditive? Suffit-il de dire que les etudiants qui
viennenl d'apprendre un son nouveau aiment a s'en servir et que
leiir atlenle de ce son a pour effet de le leur faire inlroduire dans le
cours dos slimuli acousUques qu'on leur deraande d'observer?
Sans (loute, une explication aussigenerale est-elle, dans une certaine
nie«iure, une formule dynamique correcte; elle n'est pas assex
pre'ci'^e pour le phonologiste, parce qu'elle ne tient pas suffisamment
compte deslimites de ''illusion,

II faut se rappeler que la langue de mes etudiants estrl'anglais.
Nous pouvons supposer que I'illusion d'une occlusive glottale finale
est due a quelque caractere de la structure pbonologique de i'anglais.
Mais, I'anglais n'a pas d'occlusive glottale. Comment, par consequent,
la phonologie anglaise pourrait-elle expliquer la « surperceplioo »
d'une consonne etrangere des I'abord au genie de la langue? Je crois
pourtant que les etudiants qui projetaient une occlusive glottale finale
dans les mots dictes eraployaient un element pbonetique etranger,
Tocclusive glultale, suivant un systeme pbonologique fermeraeot
elabli, mais totalement inconscient. Pour expliquer Tillusion, il faut
faire appel a la fois au processus d'apprentissage avec la tendance
toujours en eveil qui en decoule a reconnaitre ce qui a ete appris,
«?l a la phonologie anglaise. Si nous eludions les espeees de syl"
liibes qui, en anglais, peuvenl normalement constituer un mono:
syllabe accenlue ou une syllabe finale aecenLuee (ou k accent secon-
daire). nous nous apercevrons qu'on pent Us classer en trois^ types .

Six: I .(ini^udi^c in Kclntmn to Hi\i(>r\ ntul Sm tct\ 535

2G4

A. Mols qui se lorininenL par uiie voyellc lougue ou unr- diitli-
tongue. par exemple : sea, /low, shoe, review, apply :

B. Mols qui se Iprniinenl par uiu* voyellc lorii^iu- ou unc diph
tongue suivie dune ou do plusieurs consonnes, par exemple : hall,
cease, dream, alcove, amount ;

C. Mots qui se lerminent par une voyelle breve suivie dune on d*.-
plusieurs consonnes, par exemple : back, fill, come, remit, object ;

La quatri^me classe possible theoriquement :

D. Mols qui se lerminent par une voyelle breve, par exemple : les
mots fran^ais ami, fait, le mot russe xardsD,

n'exisle pas en anglais. Les sujets de langue anglaise tendenl a pro-
noncer les mots du type I) d'ane maniere « trainante » qui les fail
passer au type A (par exemple, ami pour a)7ii}. Remarquons que la
possibilile, en apparence non fondee, dune syllobe accenlut^e n<in
finale se terminanl par une voyelle breve (par exemple : fiddle,
butter, double, pheasant) se justifie par la theorie anglaise de la
syllabe, theorie qui place le point de division dans la consonne
suivante.(rf, t, b, z, dans les exemples donnes), si bien que la syl-
labe accentuee de ces mots appartient reellement, au point de vue
phonologique, au type C, non au type D. Des consonnes intervo-
caliques comme le d de fiddle ou le :; Aq pheasant, quniqu elles ne
soient pas longues, sonl, au point de vue plionoloi^ique, « mi-
toyennes » ou k double face, par le fail qu'elles lerminent une
syllabe el en commencenl une autre en meme temps. Si le point
de division en syllabes est reporte avant la consonne, la voyelle
precedente s'allongera immediatemenl, malgre sa qualile de « breve »
(type A), el nous aboulirons ainsi aux j>iononciations dialeclales
americaines de mots c;omme fiddle el pheasant, prononcialions dans
lesquelles la voyelle accentuee garde sa qualile primitive, mais a
ete allongee jusqu'^ I'unile de longueur des « voyelles longues » du
type feeble, reason, ladle.

Nous voici done mainlenant prepares a comprendre I'lllusion qui
a^lt^ noire point de depart. Des mols comme. s//n el ;n7rt sonl incons-
ciemment essayes comme membres possibles de la classe .\ ou de la
classe C. Deux illusions sont possibles si I'auditeur doit 6lre
viclime de son systeme phonologique. I'uisqu'une voyelle finale
breve el accentuee est une enlile peu connue. eiie pout elre « legi-

536 General Linguistics I

265

limee », soil par une projeclion de longueur {snif et pild' mal
entendus sont places dans la classe A), soit par la projection d'une
consonne finale apres cetle voyelle (classe Cj. Nousappellerons celte
consonoe imaginaire « x » et nous ecrirons smtx et pildx. Or, le
fait d'avoir ajoute Focclusive gloltale a notre materiel consonantique
nous conduit souvent a tenter deresoudre le problemephonologique
symbolise par smtx et pildx en termes d'occlusive glottale et k
entendre swu' et pila . L'occlusive glottale est la consonne la plus
irreelle, la plus nuUe en valeur, pour une oreille anglaise ou ameri-
caine, et elle est admirablement adaplee, une fois son existence
admise, pour servir comme projection actualisee d'une consonne
finale phonologiquement necessaire, mais aussi peu sonore que
possible. L'illusion d'une occlusive glottale finale est, essentielle-
ment, Tillusion dune consonne finale generalisee (« x ») necessaire
pour classer les mols dicles dans une categorie connue (type C), ou,
pour parleren termes plus analytiques: la phonologic anglaise cree
le fond (-x) de l'illusion synthetique, tandis que le processus
d'apprentissage le colore sous la forme de — '. L'erreur qui consiste
a entendre une occlusive glottale la oii il n'y en a pas^ dans les mols
du type D, est, au fond, une forme plus savante de l'erreur qui
consiste a entendre a la dictee une occlusive glottale finale sous la
forme;?, t, k, erreur qui se produit frequemment au premier slade
d'acquisilion d'une technique phonetique.

Le danger d'entendre une occlusive glottale quand le mot dicte se
termine par une voyelle longue ou une diphlongue est nalurelleraent
araoindri par le fait que ces mots se conferment a un modele anglais
commun (type A). La raison pour laquelle cette erreur ne se produit
pas si facilement quand on entend des mots dictes termines par une
voyelle breve inaccentuee (par exemple : dm, su li)^ c'est que ces
mots se conforment egalement a un modele anglais, bien que I'echelle
de qualites attribuees a la voyelle dans cette position ne soit pas
aussi elendue que dans le cas oii la voyelle est couverte par une
consonne (par exemple : idea, very, follow).

{Sew-Haven, 1932.

Traduxt par M"« Daliuieb).

Edward Sapir.

Six: Ijifii^udj^c in Ri'lciiion (o Ifistorv (iml Society 537

' Editorial Note

I I Journal de Psvcholoiiic nornuilc ct paih(>U>\:,umc 31) ( \^)^}>), 247-2(i>. [KcprimccI

' in: Essais sur Ic lani^ai^c, prcscnlcs par Jcan-( laiidc Paricnlc. Paris: r.ditinns dc

Miiuiit, l%9,pp. 167-188]

Corrections to the French version pubhshed in 1933

The offprint with Edward Sapir's annotations (in PhiHp Sapir's archives)
contains the following relevant corrections (page references are to the original
publication):

p. 252 1. 3-4: deviennent aspirees is marked to precede the parenthesis,
p. 256 1. I: ///'(twice)
p. 257 I. 1 : consonne. comme
p. 257 1. 8: -/'
p. 257 1. 13: dinila
p. 257 n. 3: / is voiceless
p. 257 n. 4: americain
p. 258 n. 2: lu'i
p. 2591. l:lmi-k
p. 259 1. 2: his-i-k'
p. 259 1. 25: (= t's), et L! (= r7)
p. 262 1. 8: wi-nap'a'a

Although the offprint has a separate title page with Sapir's name and the title
of the article, it seems to reflect the stage of a page proof: as a matter of fact, some
of the handwritten corrections by Sapir concern errors that do not (/no longer)
appear in the published version of the journal issue: e.g.,
p. 257 n. 3: / is voiceless
p. 257 n. 4: americain
p. 259 I hhisi-k
p. 259 1. 2: his-i'k'
p. 262 1. 8: wi-nap'a'a

Apparently, the correction (by Sapir or another proofreader) concerning p. 259
I. 25 was misunderstood by the printer, since the version published in the journal
issue has the diacritic sign ' (for glottalization) before the t in both cases.

It may thus be that the annotations for p. 252 1. 3-4 and p. 257 1. 1 reflect proof
corrections by Sapir which were not taken into account by the editor/printer of
the journal issue.

Pierre Swiggers

.S7v; fani^iuii^i' in Rvlnnoti to Ihsiorv and Society 539

THE PSYCIIOLOGICAL REALTTV OF PIIOXKMES*

The concept of the "phoneme" (a functionally significant unit in the
rigidly defined pattern or configuration of sounds peculiar to a language),
as distinct from that of the "sound" or "phonetic clement" i\s such (an
objectively definable entity in the articulated and perceived totality of
speech), is becoming more and more familiar to linguists. The difficulty
that many still seem to feel in distinguishing between the two must
eventually disappear as the realization grows that no entity in human
experience can be adequately defined as the mechanical sum or prcxluct
of its physical properties. These physical properties are needed of course
to give us the signal, as it were, for the identification of the given entity
as a functionally significant point in a complex system of relatednesses;
but for any given context it is notorious how many of these physical
properties are, or may be, overlooked as irrelevant, how one particular
property, possessing for the moment or b}' social understanding an
unusual sign value, may have a determinedness in the definition of the
entity that is out of all proportion to its "physical weight."

As soon, however, as we admit that all significant entities in experience
are thus revised from the physically given by passing through the filter
of the functionally or relatedly meaningful, as soon as we see that we
can never set up a scale of added or changed meanings that is simply
congruent to the scale of physical increments, we implicitly make a dis-
tinction, whether we know it or not, between the phoneme and the
sound in that particular framework of experience which is known as
language (actualized as speech). To say that a given phoneme is not
sufficiently defined in articulatory or acoustic terms but needs to be
fitted into the total system of sound relations peculiar to the language is,
at bottom, no more mysterious than to say that a club is not defined for
us when it is said to be made of wood and to have such and such a shape
and such and such dimensions. We must understand why a roughly
similar object, not so different to the eye, is no club at all, and why a
third object, of very different color and much longer and heavier than the
first, is for all that very much of a club.

Some linguists seem to feel that the phoneme is a useful enough concept
in an abstract lingui.stic discussion — in the theoretical pre.>^entation of
the form of a language or in the comparison of related languages - but
that it has small relevance for the actualities of speech. This point of

* Published originally in French under the title "La Hi'alittf psychologiquc do«
phonemes," Journal de Psychologic Xormale et Pathologique, 30 (1933): 217-265.

[46]

540 General Linguistics I

view seems the reverse of realistic to the present writer. Just as it takes a
physicist or philosopher to define an object in terms of such abstract
concepts as mass, volume, chemical structure, and location, so it takes
very much of a linguistic abstractionist, a phonetician pure and sim-
ple, to reduce articulate speech to simple physical processes. To the
physicist, the three wooden objects are equally distinct from each other,
"clubs" are romantic intrusions into the austere continuities of nature.
But the naive human being is much surer of his clubs and poles than of
unnamed objects to be hereinafter defined in physical terms. So, in
speech, precise phonetic stations can be abstracted only by patient ob-
servation and frequently at the expense of a direct flouting of one's
phonetic (one should say "phonemic") intuitions. In the physical world
the naive speaker and hearer actualize and are sensitive to sounds, but
what they feel themselves to be pronouncing and hearing are "pho-
nemes." They order the fundamental elements of linguistic experience
into functionally and aesthetically determinate shapes, each of which is
carved out by its exclusive laws of relationship within the complex total
of all possible sound relationships. To the naive speaker and hearer,
sounds (i.e., phonemes) do not differ as five-inch or six-inch entities
differ, but as clubs and poles differ. If the phonetician discovers in the
flow of actual speech something that is neither "club" nor "pole," he, as
phonetician, has the right to set up a "halfway between club and pole"
entity. Functionally, however, such an entity is a fiction, and the naive
speaker or hearer is not only driven by its relational behavior to classify
it as a "club" or a "pole," but actually hears and feels it to be such.

If the phonemic attitude is more basic, psychologically speaking, than
the more strictly phonetic one, it should be possible to detect it in the
unguarded speech judgments of naive speakers who have a complete
control of their language in a practical sense but have no rationalized or
consciously systematic knowledge of it. "Errors" of analysis, or what the
sophisticated onlooker is liable to consider such, may be expected to
occur which have the characteristic of being phonetically unsound or
inconsistent but which at the same time register a feeling for what is
phonemically accurate. Such "errors," generally overlooked by the prac-
tical field linguist, may constitute valuable evidence for the dynamic
reality of the phonemic structure of the language.

In the course of many years of experience in the recording and analysis
of unwritten languages, American Indian and African, I have come to
the practical realization that what the naive speaker hears is not pho-
netic elements but phonemes. The problem reaches the stage of a prac-
tical test when one wishes to teach an intelligent native, say one who can

Six: /.(uii^iKii^c in Rclntion i<, l/iMorv and Society 541

read and write English reasonably well and has some intellectual curi-
osity besides, how to write his own lanRuage. The difficulty of Kuch a
task varies, of course, with the intelligence of the native and the intrinsic
difficulty of his language, but it varies also with the "phonemic intuitive-
ness" of the teacher. Many well-meaning linguists have had disiippoint-
ing experiences in this regard with quite intelligent natives without ever
suspecting that the trouble lay, not with the native, but with themselves.
It is exceedingly difficult, if not impossible, to teach a native to take
account of purely mechanical phonetic variations which have no pho-
nemic reality for him. The teacher who comes prepared with a gamut of
absolute phonetic possibilities and who unconsciously, in .spite of all his
training, tends to project the phonemic valuations of his own language
into what he hears and records of the exotic one may easily befuddh^ a
native. The native realizes when what he is taught "clicks" with what
his phonological intuitions have already taught him; but he is made un-
comfortable when purely phonetic distinctions are pointed out to him
which seem real enough when he focuses his attention on them but which
are always fading out of his consciousness because their objective reality
is not confirmed by these intuitions.

I have selected for brief discussion five examples of phonemic versus
phonetic hearing and writing out of many which have come to me in the
course of my experience with natives and students. In each of these, it
will be observed, we have clear evidence of the unconscious reinterpreta-
tion of objective facts because of a disturbing phonological preparedness
not precisely adjusted to these facts.

I. When working on the Southern Paiute language of southwestern
Utah and northwestern Arizona I spent a little time in trying to teach my
native interpreter, a young man of average intelligence, how to write his
language phonetically. Southern Paiute is an unusually involved lan-
guage from the phonological standpoint and, as my point of view at that
time stressed phonetic accuracy rather than phonemic adequacy, I doubt
if I could have succeeded in teaching him well enough to satisfy my
standard even if I had devoted far more time to the effort than I did.
As an example of a comparatively simple word I selected pd/3a' "at the
water" (voiceless labial stop; stressed long a; voiced bilal)ial spirant;
unstressed short a; final aspiration). I instructed Tony to divide the word
into its syllables and to discover by careful hearing what sounds entered
into the composition of each of the syllables, and in what order, then to
attempt to write down the proper symbol for each of the discovered
phonetic elements. To my astonishment Tony then syllabified: pa-, pause,
pa\ I say "astonishment" because I at once recognized the paradox that

542

General Linguistics I

Tony was not "hearing" in terms of the actual sounds (the voiced bilabial
/3 was objectively very different from the initial stop) but in terms of an
etymological reconstruction: pa-: "water" plus postposition *-pa' "at."
The slight pause which intervened after the stem was enough to divert
Tony from the phonetically proper form of the postposition to a theo-
retically real but actually nonexistent form.

To understand Tony's behavior, which was not in the least due to mere
carelessness nor to a tendency of the speakers of this language "to con-
fuse sounds," to quote the time-worn shibboleth, we must have recourse
to the phonology of Southern Paiute. The treatment of the stopped
consonants may be summarized in the following table :

iNITIAi

POSTVOCALIC

1. Spirantized

2. Nasalized

3. Geminated

a. After
voiced vowel

b. After un-
voiced vowel

Labial
Dental
Guttural
Labialized guttural

k
kw

7
yw

mp
nt
Tjk
Tjkw

t-
k-
k-w

P
t
k
kw

The postvocalic forms of the stops of types 1, 2, and 3a are further
modified before an unvoiced vowel, the voiced spirants becoming un-
voiced spirants (d, R, x, x^),^ and the nasalized and geminated stops
becoming aspirated {mp', p''; nt', t'; rjk', k-'; rjkW, k-W). It is impossible
here to give a systematic idea of the phonologic processes which bring
about the sound interchanges within a given articulatory series, but it is
important to know that the spirantized, nasalized, and geminated stops
can occur only in postvocalic position and that they are largely deter-
mined by the nature of the element (stem or suffix) which precedes them
and which may be said to have an inherently spirantizing, nasalizing, or
geminating force. The stem pa-- is a spirantizing stem, and the spirantiz-
ing of a theoretical *-pa' "at" to -^a' is parallel to the spirantizing of po--
"trail" to -^0-- in such a compound as pa--^o--, "water-trail." In other
words, the language is so patterned that examples of type po--'. -^o-- lead
to the proportion *pa': -/Ja'^ and, while *pa' "at" does not actually exist
as an independent element but must always be actualized in one of the
three possible postvocalic forms, its theoretical existence suddenly comes

' W represents voiceless w.

^ This theoretical *-pa' , occurring only as -|3a', -mpa', -p-a' in postvocalic
position, is not to be confused with secondary -pa' (type 36) < -p-a' (type 3a).

Six: LdHi^udi^i' in Rclmion (,, History and Sinwiv 543

to the light of day when the problem of slowly sylluhifyinR a word is
presented to a native speaker for the first time. It tlien appears that the
-/3a' of speech behavior, as a self-contained syllabic entity without im-
mediately preceding syllable, is actually felt as a phonologic pa', from
which it differs in two important phonetic respects (voiced, not voiceless,
consonant; spirant, not stop).

All this has an important bearing on the construction of a maximally
correct orthography of Southern Paiute, if by "maximally correct" we
mean, not most adequate phonetically, but most true to the sound pat-
terning of the language. As it happens, there is reason to believe from
both internal and comparative evidence that the spirantized form of a
consonant is its normal or primary form after a vowel and that the
nasalized and geminated forms are due to the emergence of old nasal and
other consonants that had disappeared in the obsolete form of the precetl-
ing element.' It follows that the postvocalic -/3- is more closely related
functionally to a simple initial p- than is the postvocalic -p- (after un-
voiced vowel), which must always be interpreted as a secondary form of
-p-. These relations are summarized in the following table of theoretical
nonfinal forms.

Phonetic Oethochaphy

Phonolocic Obthocxapby

1. pa-

pa-

2. pa/3a-

papa-

3. psidA-*

papa-

4. papa-

papa-

5. pApa-

papa-

6. pap-A-

papa-

The phonetic orthography is more complex and, in a sense, more ade-
quate, but it goes against the grain of the language in one important
respect, for it identifies the second p in type 5 with the initial p, which is
phonologically unsound. The phonologic orthography, on the other hand,
is useless for one who has not mastered the phonology of the language, as
it leads, or seems to lead, to incorrect pronounciations which would have
the cumulative effect of making the language, so read, entirely unintelli-
gible to a native. To a slightly schooled native, however, there can be
no serious ambiguity, for the phonetic forms result from the phonologic
only by the application of absolutely mechanical phonetic laws of spiran-

* The analogy to Fronch liaison and, still more, to the three types of consonan-
tal treatment in Old Irish (spirantized or "aspirated," nasalized or "eclipsed,"
and geminated) is obvious.

* A represents voiceless a.

544

General Linguistics I

tizing, alternating stresses, and unvoicing. It is not necessary to deal
with these laws here^ but we can indicate their operation by the following
table of theoretical final forms:

Phonetic Orthography

Phonologic Orthography

1. v^eA

papa

2. pa/3<i'

papa*

3. p6,-<i>A

pa-pa

4. pd-^a' "water-at"

papa'

5. pdpA

pap-a

6. pApd'

pap-a-

7. pd-pA

pap-a

8. pdpa'

pa-p-a-

9. mal3d4>A

mapapa

10. ma^d^a'

mapapa-

11. mafid-eA

7napa-pa

12. ma^d-^a'

jnapa-pa*

13. ma^dp-A

mapap'a

14. ma^dp-a'

mapap-a'

15. maPd-p-A

mapa-p-a

16. ma^dApa'

mapa-p-a'

17. MApd<pA<^

map'apa

18. MApd^a'

map-apa'

19. MApd<pA

map-a-pa

20. MApd-0a'

map-apa'

21. MApdp-A

map'ap'a

22. MApdp-a'

map-ap-a-

23. MApd-p-A

map-a-pa

24. MApdApa"

map-a-p-a-

Obviously, in such a language as this, spirants, whether voiced or voice-
less, and voiceless vowels are not phonemes but are merely phonetic
reflexes of stopped consonants and voiced vowels under fixed dynamic
conditions. Long consonants and long vowels are sub-phonemes. The
former are the resultants of simple phonemes (stopped consonants) and
the operation of certain phonologic (and morphologic) latencies in given
syllables, present or formerly present. The latter are phonologically
resolvable into short vowel plus short vowel, i.e., into two syllables of
unit length (moras), of which the second begins with a zero consonant.

Southern Paiute, then, is a language in which an unusually simple

* They are described in detail in E. Sapir, The Southern Paiute Language,
Proceedings of the American Academy of Arts and Sciences, 65 (1930).

* M is voiceless m.

Six: f.dni^iKii^i' in Rrlaiion lo History und Society 545

phonemic structure is actualized by a more than ordinarily complex
phonetic one. Tony's "error" unconsciously registered this contnust.

II. When working on Sarcee, an Athahaskan language of AllxTta,
Canada, I was concerned with the problem of deciding whether certain
words that seemed homonjonous were actually so or differed in .some
subtle phonetic respect that was not immediately obvious. (Jne su<"h
homonymous, or apparently homonymous, pair of words was d\nV "this
one" and d\ni "it makes a sound." In the early stage of our work I jiskni
my interpreter, John Whitney, whether the two words sounded alike lo
him and he answered without hesitation that they were quite different.
This statement, however, did not prove that he was objectively correct,
as it is possible for perfectly homonymous words to give the speaker the
illusion of phonetic difference because of the different contexts in which
they appear or because of the different positions they occupy in their
respective form systems.^ When I asked him what the difference was,
he found it difficult to say, and the more often he pronounced the words
over to himself the more confused he became as to their phonetic dif-
ference. Yet all the time he seemed perfectly sure that there was a dif-
ference. At various moments I thought I could catch a slight phonetic
difference, for instance, (1) that the -ni of "this one" was on a slightly
lower tone than the -ni of "it makes a sound"; (2) that there was a
slight stress on the di- of "this one" (analysis: stem di- "this" plus suffix
-ni "person") and a similarly slight stress on the -ni of "it makes a sound"
(analysis: prefix di- plus verb stem -ni); (3) that the -ni of "this one"
ended in a pure vowel with little or no breath release, while the -ni of
"it makes a sound" had a more audible breath release, was properly -nC.
These suggestions were considered and halfheartedly accepted at various
times by John, but it was easy to see that he was not intuitively con-
vinced. The one tangible suggestion that he himself made was obviously
incorrect, namely, that the -ni of "it makes a sound" ended in a "t."

^ The grave accent represents a low tone, the acute accent a high one. Sarcee
is a tone language.

* Thus, in English, the word led (e.g., "I led him away") is fell as having
a vowel which has been deflected from the vowel of lead (e.g., "I lead him away")
and is therefore not psychologically homonymous with the word for a metal,
lead, in which the vowel is felt to be primary, not deflected (cf. further, "the
leading of the windowpane," "the leaded glass," "the dilTerent leads now recog-
nized by chemists"). The homonymy of led ami lead (metal) i.s therefore of a
different psychological order from the homonymy of yard ("He plays in my
yard") and yard ("I want a yard of silk"), for the la.st two word."* enter
into roughly parallel form systems (e.g., "Their yards were loo .small to play in"
"I want two yards of silk"; "yard upon yard of railroad tracks" "yard upon
yard of lovely fabrics"). It is probably easier for the naive spp.-iker. who does not
know how to spell either led or lead (metal), to convince himself that there is •
phonetic difference between these two words than between the two worda yard

546 General Linguistics I

John claimed that he "felt a <" in the syllable, yet when he tested it
over and over to himself, he had to admit that he could neither hear a
'T' nor feel his tongue articulating one. We had to give up the problem,
and I silently concluded that there simply was no phonetic difference
between the words and that John was trying to convince himself there
was one merely because they were so different in grammatical form and
function that he felt there ought to be a difference.

I did not then know enough about Sarcee phonology to understand the
mysterious '7" theory. Later on it developed that there are phonologic-
ally distinct types of final vowels in Sarcee: smooth or simple vowels; and
vowels with a consonantal latency, i.e., vowels originally followed by a
consonant which disappears in the absolute form of the word but which
reappears when the word has a suffix beginning with a vowel or which
makes its former presence felt in other sandhi phenomena. One of these
disappearing consonants is -t\ of which -' may be considered a weakened
form. Now it happens that all final vowels are pronounced with a breath
release in the absolute form of the word and that there is no objective
difference between this secondary -', which may be symbolized as -('),
phonologically zero, and the etymologically organic -', which may affect
certain following consonants of suffixed elements or, in some cases, pass
over to one of certain other consonants, such as V. The -ni of "this one,"
phonetically -ni' in absolute form, is phonologically simple -ni; the -ni of
"it makes a sound," phonetically -nV in absolute form, can be phono-
logically represented as -ni' (-nit'-). We can best understand the facts if
we test the nature of these two syllables by seeing how they behave if
immediately followed by suffixed relative -i "the one who ..." and
inferential -la^ "it turns out that."

plus -{ plus -la

dlni "this one" dlnd""' dlnlla

d\ni "it makes a sound" dlniVi dlniia"

We see at once that dini "this one" behaves like a word ending in a
smooth vowel (witness contraction of i + i to an over-long vowel and

' The lack of a tone mark indicates that this syllable is pronounced on the
middle tone.

'" a* is an over-long a, consisting of a long a followed by a weak rearticulated
a. Syllables of this type result in Sarcee from contraction of old final vowels with
following suffixed vowels. The change in quality from -i to -4* is due to historical
factors, -ni "person" is an old '■ -ne (with pepet vowel), relative -i is old *-e; two
pepet vowels contract to long open *-«*; as Athabaskan e becomes Sarcee a, this
older *-€* passes into Sarcee -4".

" \ is voiceless spirantal I, as in Welsh II.

Six: l.cmi^mii^e in Relation u> llistorx and Socieiy 54'

unaffected I of -la), while dini "it makes a sound" acta a« though the final
vowel had a voiceless consonantal latency, which rcRistors jjartlv n»
-' {-'-la passing, as always, to -la), partly as -t'-.

It is clear that, while John was phonetically amateurish, he was phono-
logically subtle and accurate. His response amounted to an index of the
feeling that d)nl "this one" = dini, that dinl "it makes a sound" -
dinV, and that this -fiV = -nlV. John's certainty of difference in the face
of objective identity is quite parallel to the feeling that the average
Englishman would have that such words as sawed and soared are not
phonetically identical. It is true that both sawed and soared can l>e pho-
netically represented as sa-d,'^ but the -ing forms of the two verbs (saw-
ing, soaring), phonetically SD--ir) and so-r-irj, and such sentence siindhi
forms as "Saw on, my boy!" and "Soar into the sky!" combine to produce
the feeling that the sod of sawed = sD--d but that the sod of soared =
sj-r-d. In the one case zero = zero, in the other case zero = r. Among
educated but linguistically untrained people who discuss such matters
differences of orthography are always held responsible for these differ-
ences of feeling. This is undoubtedly a fallacy, at least for the great mass
of people, and puts the cart before the horse. Were English not a written
language, the configuratively determined phonologic difference between
such doublets as sawed and soared would still be "heard," as a collective
illusion, as a true phonetic difference.

III. The most successful American Indian pupil that I have had in
practical phonetics is Alex Thomas, who writes his native language,
Nootka,^' with the utmost fluency and with admirable accuracy. .Vlex's
orthography, as is natural, is phonologic in spirit throughout and it is
largely from a study of his texts that I have learned to estimate at its
true value the psychological difference between a sound and a phoneme.
Anyone who knows the phonetic mechanics of Nootka can easily actualize
his orthography. Thus, AzV phonologically parallel to si or tii, is actually
pronounced he, with a vowel which is much nearer to the e of English
met than to that of sit. This is due to the peculiar nature of the laryngeal
consonants, which favor an a-timbre and cause the following vowels j and
u to drop to c and o respectively. The orthographies hi and hii are entirely
unambiguous because there can be no phonologically distinct syllables of
type he and ho.

Ajiother mechanical peculiarity of Nootka is the lengthening of con-
sonants after a short vowel when followed by a vowel. This purely

"These remarks apply to British, not to normiil .Xnioricati. U8U(?c.

'' This is spoken on the west coast of Vancouver Island. H. C'.

'*/i is a voiceless laryngeal spirant, almost identical with the Arabic Ad.

548 General Linguistics I

mechanical length has no morphological or phonological significance and
is ignored in Alex's orthography. His hisi'k and hisa- are, then, to be
normally pronounced his-rk' and hes-a-. It sometimes happens, however,
that a long consonant, particularly s- and s*, arises from the meeting of
two morphologically distinct consonants (e.g., s + s > s- or 5 + 5 > 1
or, less frequently, s + s or s + s > s-). In such cases the long consonant
is not felt to be a mechanical lengthening of the simple consonant but
as a cluster of two identical consonants, and so we find Alex writing, for
example, tsrqsit'lassatlni^^ "we went there only to speak," to be analyzed
into tsrqsitl-'as-sa-{'a)tl-ni. The s of -'as "to go in order to" and the s of
-sa "just, only" keep their phonologic independence and the normal
intervocalic -s*- of -'as-atl is interpreted as -ss-. Similarly, kwis-sila "to
do differently," to be analyzed into kwis-sila. It does not seem, however,
that there is an actual phonetic difference between the -s- (phonologically
-S-) of such words as tlasatl "the stick takes an upright position on the
beach" ( = tla-sall), pronounced tlas-atl, and the -s-- of -'assatl above. Here
again we have objectively identical phonetic phenomena which receive
different phonologic interpretations.

IV. In the earlier system of orthography, which Alex was taught, the
glottalized stops and affricatives were treated differently from the
glottalized nasals and semivowels. The former were symbolized as p!, t!,
k!, k!w, q!, q!w, tsl, tc! (= Is), and L! (= tl)\ the latter as 'm, 'n, 'y, and
'ly. The reason for this was traditional. The glottalized stops and affrica-
tives, as a distinctive type of consonants, had been early recognized by
Dr. F. Boas in many American Indian languages and described as
"fortes," that is, as stops and affricatives "pronounced with increased
stress of articulation." The type 'm, 'n, 'Z, 'y, and 'w was not recognized
by Dr. Boas until much later, first in Kwakiutl, and described as consist-
ing of nasal, voiced lateral, or semivowel immediately preceded by a
glottal closure. The orthography for these consonants (later discovered
in Tsimshian, Nootka, Haida, and a number of other languages, but not
as widely distributed as the so-called "fortes") suggested their manner
of formation, but the orthography for the glottalized stops and affrica-
tives was purely conventional and did not in any way analyze their for-
mation except to suggest that more energy was needed for their pro-
nunciation.^^ As a pure matter of phonetics, while the Nootka glottalized

'* I have slightly modified Alex's orthography to correspond to my present
orthography, but these changes are merely mechanical substitutions, such as
ii for L, and in no way affect the argument, q is velar k (Arabic /:), tl is a lateral
affricative, ll its glottalized form.

^^ This, incidentally, is not necessarily true. In some languages the glottalized
stops and affricatives seem to be somewhat more energetic in articulation than
the corresponding unglottalized consonants, in others there is no noticeable

Six: /.uni^uiii^i' in Kilation to Hisiors unJ Society 549

stops and affricatives are roughly parallel in formation with the glottal-
ized sonantic consonants, they are not and cannot l>o entirfly ho. In a
glottalized p, for instance, our present p and former p.', there is a syn-
chronous closure of lips and glottal cords, a closed air chamber is thus
produced between the two, there is a sudden release of the lip closure, a
moment of pause, and then the release of the glottal closure. It is the
release of the lip (or other oral) closure in advance of the glottal closure
that gives consonants of this type their superficial "click-like" char-
acter.'^ On the other hand, in a glottalized m, our 'm, while the lip cUxsure
and glottal closure are synchronous as before,'* the glottal closure must
be released at the point of initial sonancy of the m. Roughly speaking,
therefore, p may be analyzed into p + ', while 'm may be analyzed into
' + m. Such an orthographic difTerence as p! versus 'm, therefore, which
I had inherited from the Americanist tradition, was not unjustified on
purely phonetic grounds.

We now come to the intuitive phonologic test whether p and 'm are
consonants of the same type or not. Alex learned to write consonants of
tA-pe p and ts very readily (our earlier p! and ts!), e.g., papr "ear"
(earlier plap.'i), tsa'ak "stream" (earlier ts.'a'ak). To my surprise Ale.x
volunteered m! in such words as 'ma-'mrqsu "the older [brother or
sister]," which he wrote m!dm!iqsu. In other words, we had valuable
evidence here for the phonologic reality of a glottalized class of conso-
nants which included both type p (with prior release of oral closure) and
type 'm (with prior release of glottal closure). A phonologically consistent
orthography would require p and m (or p! and m!). Once more, a naive
native's phonetic "ignorance" proved phonologically more accurate than
the scientist's "knowledge." The phonologic justification for Alex's

difference so far as "stress of articulation" is concerned. In the Athabuskan
lanpua^^es that I have heard (Sarcee, Kutchin, Hupa, Navahoi the ji5i>irated
voiceless stops and affricatives (of type t\ A', ts') are far more "foriis" in char-
acter than the corresponding glottalized consonants (e p., I, k\ is). There is
no necessary correlation between laryngeal type of articulation (voiced, voice-
less, glottalized; or any of these with aspiration) and force of articulation (foriis,
lenis). So far as Xootka is concerne(i, it did not seem to me that the nlottali/ed
stops and affricatives (Boas' "fortes") were significantly different in en.phajis
from the ordinary stops and affricatives. In such hinguagcs as recognize a phono-
logical difference of emphatic and nonemphatic and, at the sanje tiin>' ti..>~.>in
glottalized consonants, there is no reason why the glottalized con.^ona; •(

appear in both emnhatic and nonemphatic form. .-\s i'rince Trubt .s

shown, some of the North Caucusic languages, as a matter of fact, pinwrjw bfiih
empliatic and nonemphatic glottalized stops and affricatives.

" These consonants are apparently identical with the "ejectives" of Dnnicl
Jones. There is another, apparently less common, type of glottalized 5top or
affricative in which the oral and glottal releases are synrhronouiJ.

" The pronunciation of 'm, 'n, 'w, and ';/ as a simple secjuence of glottal slop
(') plus m, n, u', and y is rejected by the Nootkaear as incorrect.

550 General Linguistics I

"error" is not difficult. Consonants of type -p are entirely analogous to
consonants of type 'm for the following reasons.

1. Each occurs at the beginning of a syllable and, since no word can
begin with a cluster of consonants, both p and 'm are felt by Nootka
speakers to be unanalyzable phonologic units. In other words, the glottal
stop can no more easily be abstracted from 'm than from p. Similarly,
the affricatives and glottalized affricatives are phonologically unanalyz-
able units.

2. All consonants can occur at the end of a syllable except glottalized
stops and affricatives, glottalized sonantic consonants ('m, 'n, 'y, 'ly),
semivowels (y, w), nasals (m, n),^® the glottal stop ('), and h. This rule
throws consonants of type 'm more definitely together with consonants
of type p.

3. Many suffixes which begin with a vowel have the effect of "harden-
ing"2o the preceding consonant, in other words, of glottally affecting it.
Under the influence of this "hardening" process p, t, k become p, t, k,
while m and n become 'm and 'n. For example, just as the suffixes '-a'a^^
i'-a-'a) "on the rocks" and '-ahs "in a receptacle" change the stem
m-nap- "to stay, dwell" to wi-nap- (e.g., wvnapa'a "so stay on the
rocks") and wik- "to be not" to mk'- (e.g., wikahs "to be not in a re-
ceptacle, a canoe is empty"), so Vlum- (alternating with Vlup-) "to be
hot" becomes t'lu'm- (alternating with t'lup-)(e.g., tlu'ma-'a "to be hot
on the rocks" and tlu'mahs "to be hot in a receptacle, there is hot water";
compare tluprtsh "summer, hot season" = parallel tlup- + '-i-tch
"season") and kan- "to kneel" (e.g., kanil "to kneel in the house") be-
comes ka'n- (e.g., ka'nahs "to kneel in a canoe"). As there seem to be no
stems ending in /i or ', the group 'm, 'n, 'w, 'y"^^ is left over as functionally
related to the group m, n, w, y in the same sense as the group exemplified
by p is related to the group exemplified by p. Morphology, in other
words, convincingly supports the phonologic proportion p'.p = m:'m. It
is maintained that it was this underlying phonologic configuration that
made Alex hear 'm as sufficiently similar to p to justify its being written
in an analogous fashion. In other languages, with different phonologic
and morphologic understandings, such a parallelism of orthography
might not be justified at all and the phonetic differences that actually

'' m and n may be followed by a murmured vowel of t-timbre which is a re-
duced form of a, w, or i. Syllables or half-syllables of type m' or n' are preceded
by i, an assimilated product of a, u or i; in} and im} result therefore, in part, from
sequences of type ama, umi, anu. Simple -em or -an become -ap, -at.

" A term borrowed from Boas' equivalent Kwakiutl phenomenon.

'' The symbol ' indicates the "hardening" effect of a suffix.

" The phonologic details involving 'lo and 'y and their relation to w and y
and other consonants are too intricate for a summary statement in this place.

Six: LdNi^iKii^c in Rchiutm m //istorv and Stniclv 551

obtain between 'm and p would have a significantly different psycholoRic
weighting.

V. In a course in practical phonetics which 1 have been giving for a
number of years I have so often remarked the following illusion of hear-
ing on the part of students that there seems no way of avoiding a general
phonologic theory to explain it. I tind that, after the students have been
taught to recognize the glottal stop as a phonetic unit, many of them
tend to hear it after a word ending in an accented short vowel of clear
timbre (e.g., a, e, e, i). This illusion does not seem to apply so often to
words ending in a long vowel or an obscure vowel of relatively undefined
quality (a) or an unaccented vowel. Thus, a dictated nonsense word like
sme or pild would occasionally be misheard and written as smt' and
pild' but there seems far less tendency to hear a hnal glottal stop in words
like pila or pild-. What is the reason for this singular type of "overhear-
ing?" Is it enough to say that students who have learned a new sound
like to play with it and that their preparedness for it tends to make them
project its usage into the stream of acoustic stimuli to which they are
asked to attend? No doubt such a general explanation is a correct dy-
namic formula so far as it goes but it is not precise enough for a phonolo-
gist because it does not take sufficient account of the limitations of the
illusion.

It must be remembered that the language of my students is English.
We may therefore suspect that the illusion of a final glottal stop is due
to some feature in the phonologic structure of English. But English has
no glottal stop. How, then, can English phonology explain the overhear-
ing of a consonant which is alien to its genius to begin with? Nevertheless,
I befieve that the students who projected a final glottal stop into the
dictated words were handling an exotic phonetic element, the glott^il
stop, according to a firmly established but quite unconscious phonologic
pattern. It requires both the learning process, with its conse(}uent alert
preparedness to recognize what has been learned, and English phonolog}-
to explain the illusion. If we study the kinds of syllaliles in English which
may normally constitute an accented monos^'Uabic word or an accented
(or secondarily accented) final syllable of a word, we find that they may
be classified into three types:

A. Words ending in a long vowel or diphthong, e.g., sea, flow, shoe, revieir,
apply.

B. Words ending in a long vowel or diphthong plus one or more consoDants,
e.g., ball, cease, dream, alcove, amount.

C. Words ending in a short vowel plus one or more consonants, e.g., back,
fill, come, remit, object.

552 General Linguistics I

The theoretically possible fourth class:

D. Words ending in a short vowel, e.g, French /ai7, ami; Russian x^r^so'

does not exist in English. English-speaking people tend to pronounce
words of type D in a "drawling" fashion which transfers them to type A
(e.g., ami- for ami). Observe that the apparently inconsistent possibility
of a nonfinal accented syllable ending in a short vowel (e.g., fiddle, butter,
double, pheasant) is justified by the English theory of syllabification,
which feels the point of the syllabic division to he in the following con-
sonant {d, t, b, z, in the examples cited), so that the accented syllables of
these words really belong phonologically to type C, not to type D. Inter-
vocalic consonants like the d oi fiddle or z of pheasant, in spite of the fact
that they are not phonetically long, are phonologically "flanking" or two-
faced, in that they at one and the same time complete one syllable and
begin another. Should the point of syllabic division shift back of the
consonant, the preceding vowel at once lengthens in spite of its "short"
quality (type A), and we thus get dialectic American pronunciations of
words like fiddle and pheasant in which the accented vowel keeps its
original quality but has been lengthened to the unit length of "long
vowels" of type feeble, reason, and ladle.

We are now prepared to understand the illusion we started with. Such
words as sme and pild are unconsciously tested as possible members of
class A or class C. Two illusions are possible, if the hearer is to be a victim
of his phonologic system. Inasmuch as a final accented short vowel is an
unfamiliar entity, it can be "legitimized" either by projecting length into
it (misheard sme- and pild- fall into class A) or by projecting a final
consonant after it (class C). We shall call this imaginary consonant "x"
and write smex and pildx. Now the fact that one has added the glottal
stop to his kit of consonantal tools leads often to the temptation to solve
the phonologic problem symbolized as smex and pildx in terms of the
glottal stop and to hear sme' and pild\ The glottal stop is the most unreal
or zerolike of consonants to an English or American ear and is admirably
fitted, once its existence has been discovered, to serve as the projected
actualization of a phonologically required final consonant of minimum
sonority. The illusion of the final glottal stop is essentially the illusion of
a generalized final consonant ("x") needed to classify the dictated words
into a known category (type C). Or, to speak more analytically, English
phonology creates the groundwork {-x) of the synthetic illusion, while the
learning process colors it to the shape of -'. The error of hearing a glottal
stop where there is none, in words of type D, is fundamentally a more
sophisticated form of the same error as hearing a dictated final glottal

Six: Lani^mif^c in Rchitiun to llisiorv and Socielv 553

stop as p or < or k, which occurs frequently in an earlier stage of the
acquiring of a phonetic techni(}ue.

The danger of liearing a glottal stop when th«- dictated word ends in a
long vowel or diphthong is of course rendered very unlikely by the fact
that such words conform to a common English pattern (type A). The
reason wh\' the error does not so easily occur in hearing dictatetl words
ending in an unaccented short vowel (e.g., ont, sii-li) is that such words,
too, conform to an English pattern, though the range of the cjualities
allowed a vowel in this position is not as great as when the vowel is
covered by a following consonant (e.g., idea, very, follow).

Editorial Note

In: Edward Sapir, Selected Writini^s in Lani^uai^e, Culture, and Personality. Edited
by David G. Mandelbaum. Berkeley: University of California Press. 194M. pp.
46-60. [Reprinted in: V. Becker Makkai ed., Phonoloi^icul iheorx: Evolution and
Current Practice. New York: Holt, Rinehart & Winston. 1972. pp. 22-.^ 1 ]

554 General Linguistics I

Editorial notes on the English version published in

1949

Of the text printed in the Selected Writings (Berkeley, 1949), pp. 46-60, a
typescript with handwritten corrections by Edward Sapir has been preserved.
There are a number of differences between this typescript version and the text
published by David G. Mandelbaum in 1949. A number of divergences in the
pubUshed version correspond to the American Indian forms as they are given
in the pubHshed French version of 1933. The typescript version also contains a
few (self-correcting) errors not corrected by Sapir (but absent from the 1949
published version).

The following divergences between the posthumously published text of
1949 and the typescript version (= TS) may be of interest to the reader; except for
cases of stylistic changes, the reading of the typescript as corrected by Sapir
should be followed, although in three cases the reading of the typescript is still
not the correct one, because of diacritics that are missing. Page references are to
the 1949 published text reprinted here.

p. 47 1. 12-13: In the physical world the naive speaker and hearer
TS: The naive speaker and hearer, in the physical world

p. 48 1. 36-38: I instructed Tony to divide the word into its syllables and to dis-
cover by careful hearing what sounds entered into the composition of each of the
syllables, and in what order

TS: I instructed Tony to divide the word into its syllables and to discover by
careful hearing what sounds, and in what order, entered into the composition of

each of the syllables,

p. 49 1. 3 below table: (9, R, x, xW)
TS:($,R,x,xW)

p. 50 1. 14: obsolete form

TS (handwritten correction): absolute form

p. 50 table, no. 3: paOA-
TS: paf A-

p. 50 1. 6 below the table: pronounciations
TS (handwritten correction): pronunciations

p. 50 1. 9-10 below the table: phonologic only
TS (handwritten correction): phonologic ones

Six: Laniiuagc in Relation m lliMorx and Society 555

p. 51 tabic, no. 1: piiHA
rS: paf A

p. 53 I. 7; to be a diricrciicc.
TS: to be one.

p. 55 1. 6: 5 + s or s + 5 > s-
TS: 5 + s > s- or s + .s'

p. 55 1. ^): tsiqsit'lassatlni
TS: tsiqsit"! as sallni

p. 55 1. 10: tsi-qsit}-'as-sa-('a)tl-ni

TS: tsi- q-sitl-'-as-sa-(a)tl-ni [to be corrected as: tsiq-sitl-'-as-sa-('n)tl-ni ]

p. 56:n. 16 1.3 fortis
TS: fortes

p. 56 n. 16 1. 1 1-12: why the glottalized consonants may not appear
TS: why the latter may not appear

p. 56 1. 12: p
TS:p'
also: p. 561. 16J. 18. 1. 23J. 25: p. 57 1. 1.1.4.1.6.1. 13

p. 57 1. 22: tlu'ma'a

TS: tlumaa [to be corrected as: l"lu"nia"a]

p. 57 1.24: tlup- + -itch

TS: tlup- + -itsh [to be corrected as: ilsh]

p. 58 1. 22: of my students
TS: of the students

p. 59 1. 5: ami-
TS: ami'

p. 59 1. 3 from below: shape of -'
TS: shape o\' -'

p. 60 1. 7: our,
TS: one

Pierre Swic.diiRS

556 General Linguistics I

THE RELATION OF AMERICAN INDIAN LINGUISTICS
TO GENERAL LINGUISTICS*

EDWARD SAPIR

THE IMPORTANCE of American Indian linguistics to the general linguist
is so obvious that it does not need to be stressed. A few indications, however,
of the light thrown by American Indian languages on general problems of lin-
guistics may be welcome. Many indications of the general importance of the
Americanist's field have been given by Dr F. Boas in the "Introduction" to his
Handbook of American Indian Languages}

One of the problems which the general linguist has to deal with is the distribu-
tion of phonetic elements. It has become well known that sounds and groups of
sounds have their geographical distribution like any other cultural trait, and
from the stricdy anthropological point of view there is no reason to suppose that
these distributions would necessarily follow the lines given by a genetic classifica-
tion of languages. It so happens that the languages of native America are a
particularly fruitful field for this type of research. It is remarkable, for instance,
that the phonetic systems of the languages spoken along the Pacific coast south
of the Eskimo area have many characteristics in common in spite of the fact that
they are far from being members of the same genetic group. The phonetics of
Tsimshian, for instance, agrees in numerous peculiar respects with that of suc(i
languages as Kwakiutl and Nootka, yet it is almost certain that the generic affilia-
tions of Tsimshian are with languages far to the south and that its genetic
relationship to Kwakiutl and Nootka is, at best, exceedingly remote and in all
probability non-existent. Analogous phenomena have from time to time been

* This posthumous manuscript from the Boas Collection of the American Philosophical
Society is made available through the generosity of the Society. C. F. Voegelin ami Morris
Swadesh kindly brought our attention to it — Editor.

' Bulletin, Bureau of American Ethnology, no. 40, part I, 1911.

Six: l.iini^niii^i' in Rcluiian lo Ui\i<>r\ and Society 557

pointed out in other parts of the world. Thus, Armenian owes certain of iu
phonetic peculiarities to contact with Caucasic languages in spite of the fact that
it is itself a typical Indo-European language. The American Indian field n\ay
very appropriately be thought of as an objea lesson of the greatest importance
for a general theory of historical phonetics, for the irregular distributions of
many of the linguistic stocks on the continent has been especially favorable for
the spread of phonetic features far beyond the confines of a single genetic group.
Experience gained from a study of American Indian data must have important
methodological consequences for judgments on phonetic history in such fields as
Indo-European and Hami tic-Semi tic

What applies to phonetics is, to a considerable extent, also true of morpho-
logical features. In general, we may operate with the hypothesis that a given
type of linguistic structure tends to maintain itself for exceedingly long periods
of time. But it cannot be denied that important re-formations can be, and often
have been, due to contact between fundamentally alien languages. The Ameri-
can Indian languages that we have sufficient knowledge of seem to behave very
differently in this respect. Thus, the languages of the Athapaskan group are
singularly conservative in form as well as in phonetics and vocabulary in spite of
their enormous ?nd irregular spread. There is far less difference in form
between, say, Chipewyan, Hupa, and Navaho, three languages selected from the
Northern, Pacific, and Southern divisions of Athapaskan, respectively, than
between Baltic and Slavic within the Indo-European group. Here we have lan-
guages that seem to have been significantly resistant to exotic influences. On
the other hand, there are important morphological characteristics which seem
to have diffused over a continuous territory occupied by languages of alien
stocks. A good example of such a distribution is the presence of instrumental
prefixes in the verb in the Maidu (north central California) and Takelma
(southwestern Oregon) languages of the Penutian group, in which such prefixes
are not ordinarily founcL It seems very probable that we have here an influence
exerted by the Hokan languages such as Shasta and Karok (northwestern Cali-
fornia) on neighboring languages. These instrumental prefixes are further
found in Shoshonean, which adjoins the area under discussion to the east. A
careful study of a distribution of this sort should help materially in clarifying our
ideas about the relative persistence or non-persistence of grammatical features.
Even if the importance of diffusion of formal linguistic features as an explana-
tion of linguistic resemblances may not be as far-reaching as some diffu:;ionists
suppose, it nevertheless remains true that the intercrossing influence of diff^usion
must be taken into account very much more seriously than is done by students

558 General Linguistics I

o( compararivc and historical grammar in the Old World. Here again American
Indian linguisrics seems destined to become an important object lesson in lin-
guistic methodology.

Far more important, however, than the suggested importance of American
Indian linguistics for problems of diffusion is the intrinsic analysis of these lan-
guages. As is well known, they are unusually variant in form from each other,
and it is perhaps not too much to say that there is hardly a morphological type
which is not illustrated in the American field. It has often been pointed out
that many of these languages are highly synthetic or polysynthetic in form, but
on the other hand there are not a few languages in native America which are
highly analytic in structure. In view of the confusion which still prevails in
regard to the relation of linguistic form to race and cultural backgrounds, it is
peculiarly important to survey the American Indian field, for within it we find
maximal morphological divergences within a relatively homogeneous race and
with complete lack of correspondence with the cultural groupings of the ethnolo-
gist. This means that American Indian linguisrics stands as a silent rcfutarion
of those who try to establish an innate psychological rapport between cultural
and linguistic forms. Surely, the content of language reflects culture with
painstaking accuracy but its morphological outlines seem to be essentially inde-
pendent of such cultural influence. Just what this means in a psychological
sense it is for the future to determine. For the present it is obviously important
to gather the abundant materials on this point. The American Indian languages
are in a peculiarly favorable posirion to give us the required data. It is possible
to find areas in native America in which a relatively uniform culture is shared
by peoples who speak languages that present the very widest possible contrasts
of form. Such entities as the West [Northwest] Coast culture area, the Plains
culture area, and the Pueblo culture area are in flat contradiction to the lingubtic
affiliations of the languages spoken within them. If, therefor, there arc fimda-
mental relations between cultural and formal linguistic phenomena, they cannot
be of the type which so many linguistic philosophers and social scientists are in
the habit of discovering. Here again American Indian linguistics is an invaluable
test field for solid linguistic thinking.

A lingubt who is familiar with the forms of only one drcimiscribed group of
languages, such as Indo-European or Semitic, necessarily runs the risk of uni-
versalizing formal features which are after all local in their distribution, or of
rejecting as unlikely conceptual peculiarities which are abundantly attested out-
side of his special group of languages. American Indian languages give abun-
dant opportunity to correct both of these possible misconceptions. The Indo-

.S7.v; l.nni^udi^i' in Rchiiion to HiMorv unci StuielY >>'')

Europeanist, for instance, will find nominal classifications based on sex grndcr
sparely represented in America and this negative fact cannot but give him a new
respect for the possible genetic value of the presence of $cx gender in Indo-
European and Hamitic-Semitic Again, the rarity in America of formal com-
paratives and superlatives in the adjective gives this formal feature of Indo-
European a contrastive emphasis that it might not otherwise seem to possess.
On the other hand, the development of secondary cases from postposed locative
particles, ordinarily felt to be a rather strange feature of certain IndoFuropean
languages, finds numerous parallels in other languages all over the world, includ-
ing those of America. The conceptual separarion of aspect and tense comes out
very clearly in many American languages, whereas the two categories tend to
be blended in various ways in Indo-European and Semitic Tliese are but a
few out of hundreds of examples of what may be learned from American Indian
languages of basic linguisric concepts, or rather of the grammatical treatment of
basic concepts. There is hardly a classificatory {peculiarity which does not receive
a wealth of illumination from American Indian languages. It is safe to say that
no sound general treatment of language is possible without constant recourse
to these materials.

A word may C>e said on the value of field work in American Indian lingubtics
for those not planning to occupy themselves professionally with them. Modem
training in linguistics must emphasize more and more the importance of direct
contact with sf)eech rather than the conventionally recorded language. It is of
great pedagogical importance for a young Indo-Europeanist or Semitist to try
to work out inductively the phonetic system and morphology of some language
which is of an utterly different structure from those that he has been studying.
Such an experience frees him from numerous misconceptions and gives him the
very best evidence that he could wish for the phonetic and grammatical consist-
ency of a language that is handed down entirely by word of mouth. One may go
so far as to say that only students who have had this type of experience have a
thoroughly realistic idea of what language is. For this type of training nothing
more suitable could be thought of than an investigation into one or more of the
languages of native America. They are readily accessible and competent inter-
preters can be found in most cases. It may be pointed out that E. Sapir's lan-
guage and L. Bloomfield's Language owe not a little to the personal experience
of the writers in the inductive study of American Indian languages.

Editorial NdIc

Soiithwc.slcni Journal oj Anlhropoloiiy 3 (1947). 1 4. [ llus icM is also reprinted
in The Collected Works of Edward Sapir, vol. V, pp. 1 4.^146]

Acknowlcducnicnls

The papers •Ilcidcr's I rspmiig dcr Spnichc." 'i.anuiiaec as a Form dI Hiirnan
Behavior." •'The (\)nccpl ol Phonetic law as icsicd in Pnmilivc Languages bv
Leonard Bloomfield," and *'(iradinu. a Sludy in Semanlics" arc reprinted wilh
permission of Hie University of ("hieauo Press. Itie review of O. Jespcrsen's
Mankind, Nation, and Individual from a Liniiuisiii Point i)j View and portions
from S. Newman's re\ iew of Edward Sapir: Selected Wriiini^s in Lanfiua^v. Culture
and Personality are reset here with permission of Ilie I 'nivcrsity of Chicago
Press.

ITie papers "Sound Patterns in Language,"* " The Status of Linguistics as a
Science." Totality, and The Expression of the Endini^-l'oini Relation in En\^lish.
Erench, and German and Z. Harris's review of Edward Sapir: .Selected Wntmiis m
Language, Culture and Personality are reprinted with permission of'Ilie linguistic
Society of America.

The articles '"Communication." "Dialect" and "Language" are reprinted \Mth
permission of Macmillan Publishing Companv.

The papers "The Grammarian and his Language" and "Wanted, a World
Language" are reprinted with permission of The American Mercury and ilie
University of California Press.

The papers "Language and Environment." "The Function oi an International
Auxiliary Language" and "Tlie Psychological Reality of Phonemes " are reprinted
with permission of The University of California Press and of the American
Anthropological Association.

The paper "The History and Varieties of Human Speech" is reprmted with
permission of Science Press.

The paper "An Approach to Symbolism" is reset here with permission of ihe
Tree num.

The paper "Memorandum on the Problem oi an InternatioUiil .XuMliar)
Language" is reprinted with permission of Columbia University I^ress.

The review of A. Meillet and M. Cohen's Les langues dii monde is reset here
with permission of The Johns Hopkins Uni\ersity Press.

TTie article "F^hilology" is reset here with permissit>n oi the Encyclopaedia
Britannica.

TTie review of R.G. Kent. Language and Philology, is reset here with permis-
sion t)f the Editor of Lhe Classical World (formerly ///<• Classical Weekly) and of
the Classical Association of the Atlantic Slates.

Tlie paper "A Studv in Pln)netic Symbolism" is reprmled uith permission ol
the American Psychological Association.

The paper "The Case for a Constructed lnkrnatu)n il 1 anguagc" is reprinted
with permission of the Comile international permanent ile linguisies.

Ilie paper "The Relation i)f American Indian Linguistics to (ieneral Linguistics"
is reprinted with permission of The I niversity of New Mexict> Press.

Philip S AIMK

Index ot Personal Names

Ahbt. Thomas 68
Adicr. Herman M. 227
Albrcchl. Erhard >')
Andersen, Henning 25
Anderson, Stephen R. 47fi
Ardencr. Ldwin 17
Ai\kiini. Slefano 106
Asakawa. Kaniehi 288
Aulran. Charles 177

Rally. Charles 293.502.517

Barriga \ illanueva. Rebeea 104

Basehart. Harry 47

Basso. Keith 1 1

Baudouin de Courtenay. Jan 247-248

Bausani. Alessandro 247

Beek.Anke 13

Beeker Makkai. Valerie 194.553

Bellavitis. Ciiusto 247

Berger, Ric 247

Berlin. Isaiah 295

Bernier. Helene 103

Berlhoff.Ann E. 17

Bibovie. Ljiljana 23

Bloeh. Jules 177

Bloomfield. Leonard 10. 15. 21. 24.

28-29. 34. 45. 48-49, 53-54. 104-105. l.M.

160-161, 191, 220, 245, 263. 474-475. 482.

484, 486, 488-489, 492-493. 5 1 7. 559

Boas. Franz 7, 44-45, 57-58, 107. 109. 153.

158, 197, 245, 263, 473, 482. 516, 530-531,

533, 548-550, 556

Boek. Philip K. 12.47

Bopp. Fran/ 96

Bright. William 11-12

Brondal. Viiiuo 473

Brugmann. Karl 248
Brunol. I crdinand 293
Buekeburg. Counlcvs of 93
Buhler. Karl 473
liurns. Robert 501
Burton. J W. 12
B\ lion, Iheodora 105

Cain. Michael 158

Cal\ erton. Victor F. 483

Carr. Har\ey A.227

Carroll. John B. 250

Casado Velarde, Manuel KKt

Cassirer. Frnst 196.473

Catiline (= Lucius Sergiiis ( .iiilin.ii

217

(Miale. Wallace 61

Chappellier. Paul 248

Chaucer, Cicoffrey 123

Chavee, Honore 63

Cicero. Marcus Tullius 175, 197,217

Cohen. Marcel 9. 15. 157, 177. 200. 473.

516

Coleman. Algernon 327

(\)llmson. William E. 248. 287 2SS. 2^.

3(MK^()1.386.470

(\>iuiillae. Ilienne Boniioi lie «>o.'>^.

( onlueius 1 W

Conklin. Harold 11

Couturat. Louis 247 248

Cowan. William 1 1-12, 17.25.47. 103-104.

474,481

Croce. Benedetto 50

C urme. (Icoruc 356. 375-377. 379-380

* Tlic Index of Personal Names liocs mu iikIiuJc ihc names Imm litcriitA -••"■
Aihillcs.oT those ligurinj: in the examples (.|uoteil In lexiet)graphical sources

of the Fndinu-P»»inl Relation' References are to the papes i>f the prcM.'nl \»'.

oriiiinal texts.

.1 l>\ S.ll-M

564

Index of Personal Names

Dalimier 536

Darnell, Regna 1 1-12. 17. 47. 57. 103-105,

158,245

Delafosse (Jules ?)' 196

Delafosse, Maurice 177

Desmel. Piet 63. 479

Diebold. Richard A. 108

Diodorus of Sicily 74

Dixon. Robert B. 200

Doroszevvski, Witold 473

Drechsel, Emanuel J. 58

Dressier. Wolfgang 476

Dummer, Ethel S. 475

Dynes, Wayne 57

Edgerton. Franklin 7
Einstein. Albert 176
Elisseev, Serge 177
Embleton. Sheila 104
Emeneau, Murray B. 12
Erickson, Jon 58

Faegren, Frances 327

Fenton, William N. 103

Ferguson. Charles A. 246

Fernandez Casas, Maria Xose 17, 106, 475

Ferrand. Gabriel 177

Fife, Robert H. 288

Finck, Franz Nikolaus 106, 131, 157, 178

[Formey, Jean-Henri-Samuel] 67

Foster, Michael K. 1 1-12, 17, 25, 47, 103-105,

474, 481

Fought, John 25

Fowler, Catherine S. 481

Fowler, Don D. 481

Franck, Luanne 59

Fraustadter, Werner 247

Freire Llamas. Antonio 106

Freud, Sigmund 50

Friedrich, Paul 11
Fromkin, Victoria 11

Gaier, Ulrich 59, 62

Garcia Fajardo, Josef ina 104

Gardiner, Alan H. 517

Gerig, John L. 245, 263

Gessinger, Joachim 59

Gilissen, Emmanuel 60

Ginneken, Jacques van 473

Goddard, Ives 104, 474

Goddard, Pliny Earle 491

Goethe, Johann W. 69, 96

Golla, Victor 1 1-12, 23, 104, 480

Gordon, Terrence 154

Gregoire, Antoine 473

Grierson, George A. 202

Grimm, August-Theodor von 247

Grimm, Jacob 59, 96, 179, 353, 356,

375-384

Grimm, Wilhelm 353, 356, 375-384

Guerard, Albert L. 247

Gumperz, John 160

Giinther, Gottfried 58

Guthrie, Edwin Ray-^ 34

Gymnich, Marion 58

Hadzsits, George Depue 217

Hagboldt, Peter 327

Hamann, Johann G. 63, 93-95

Hamp, Eric P. 12

Handler, Richard 12

Harrington, John Peabody 129

Harris, Zellig S. 7, 9, 19, 21, 23^5, 47, 296,

478

Hartknoch, Johann Friedrich 66, 94

Haupenthal, Reinhard 247

Haym, Rudolf 58, 63, 66-68, 97

Heath, Shirley Brice 246

^ In his encyclopedia article '-Philology" Sapir mentions "Cassirer, Delafosse and Ogden and Richards", as
"philosophers of standing" who have studied linguistic problems. It is rather unlikely that Jules Delafosse is
meant here, and one may wonder whether in fact Sapir was not thinking of Henri Delacroix, a student of
Bergson.

In Z. Harris's review of Edward Sapir: Selected Writings in Language. Culture, and Personality, one should
correct "E.T. Guthrie" into "E.R. Guthrie".

Index of Personal \ames

565

Heidegger, Martin 47^^

Heinz, Marian 62

Henry, VictcM" IM

Herder, joliann (i. '). 15. 17. .^7 6."^, dS w

71-9S

Herodotus 197

I lewes, Gordon W. 59

Hirabayashi, Mikio 17

Hoekett. Charles F. 24

Hoenigswald, Henry M. 27.47, 157

Hoijer, Harry 475

Homburger, Lilias 177

Homer 73,80. 112

Hornstein, Norbert 293

Hsii Tsan Hwa 245

Humboldt. (Karl) Wilhelm von 5.S-59, 63.

95-98

Hunt, George 57

Hunt, Joseph McVicker 34

Hymes, Dell H. 1 1-12, 23, 25, 105, 498

Irvine, Judith 11-12, 17,23
Isaac, Glynn L. 60

Jacob, Henry 247

Jacobi, Wilhelm H.59

Jakobson, Roman 477

Jenness, Arthur 34

Jespersen, Otto 9, 15, 41, 43, 63, 157-158,

165, 195, 202-203, 245, 248-249, 275.

287-288,473,502,516

Johnston. Harry Hamilton 201-202

Jones, Daniel 531,549

Joos, Martin 35, 194

Joseph, John E. 104. 160.480

Jucquois, Guy 60

Jung, Carl Gustav 176

Juul, Arne 245

Kant, Immanuel 171
Kantor, Jacob R. 158. 195.516
Kast. Ludwig 327
Kautman, lerrence 246
Kaye,AlanS. 12, 104
Kent. Roland G. 9. 15, 1.56.217
Key, Mary Ritchie 47, 157
Kieffer, Bruce 60

Koerncr.Konrad 12 17 2"^ 25 47 U)1 104

154. 16().474.4«1

Kohler. Wolfganu 35

Krapp. (icorgc IMulip 245. 2^3

Krauss, Michael h. 104.474

Kfepinsky, Max 104

Kroeber, Alfrcil 1 II. l(>-4 ,"'"" «^" «^'

Lacombe. George 177

Lambert. Heinrich 68

Langdon. Margaret 1 1

Lauchert, Friedrich 63. 97

Lauwers. Peter 479

Leau. Leopold 247

Lee. Pen nv 160,250.294

Lenin (^ \ladimir llych riyano\ ) 175

Leskien. August 24.S

Levinson. Stephen l6(i

Liebe-Harkort. Marie-Louise 12-13

Littre, Emile 356. 369-375

Liver, Ricarda 478

Locke, John 160

Lodge, R. Anthony 480

Lopez Pereira.Jose E. 106

Lewie, Robert H. 1 1

Lucy, John A. 160.294

Ludwig. Albert 248

Luschul/ky. Hans 476

Luther. Martin 5(K)

Lyons. John 10. 292.2W

Malinowski. Hromsla\s !.•>>. iv7

.Malkiel.^akov 11.25. 108. 157.478

Manaster R.imer. Alexis KM

McGary. Jane 12

Mandelbaum. Da\id G. 11. 16. 21. 4<>. 54.

149. 176. 194. 226. 241. 275. 2%. 470. Am,

493. 497. 502. 517. 553-554

Mans/yny. Heinrich von ."^47

Markey.John F516

Markworth.Ilni) 58

Marolta. Gii>vanna 25

Marr. Nikolaj 107.159.201

Marshack. Alexander W)

M.iiA. Karl 34

Matthews. Peter H. 23

M.ituszewski. Jozcf 59

566

Index of Personal Names

Maupcriuis, Pierre-Louis-Moreau 60, 74

Mauthner, Fritz 69, 94-95, 154, 164

Max Muller, Friedrich 63-64. 98, 516

Maxwell, Dan 247

Meader, Clarence Linton 516

Meillet. Antoine 9, 15, 157-158, 160, 177,

200, 473, 478-480, 502, 516-517

Mendelssohn, Moses 67, 77

Meriggi, Piero 473

Meyer, Gustav 247

Meyer-Lubke.Wilhelm 104

Meysmans, Jules 247

Michelson. Truman 474

Mistral. Frederic 501

Mithun, Marianne 104

Moller. Hermann 159,201

Morf, Heinrich 478

Morris, Alice V. 245, 248, 300-301, 303, 325,

328. 409. 446. 470

Muller, Friedrich 177

Muller, Henri F. 327

Muller, Johann von 58, 67

Muret, Eduard 356, 382

Murray. Stephen O. 57, 103, 158

Nevin. Bruce 23

Nevinson. Henry W. 63, 67, 94-96
Newman, Stanley 9, 19, 21, 27, 47-54, 157
Nicolai, Christoph Friedrich 67, 93
Niederehe, Hans-Josef 104, 154
Nielsen, Hans F. 245
Nielsen. J0rgen Erik 245
Nunning, Ansgar 58

Ogden. Charles K. 9. 17, 153-156, 158-159,
161, 164-165, 174, 196, 202, 477, 516
Orwell, George 43
Osmundsen, Lita 11

Paget, Richard 516
Palmer, Frank 105
Palmer, Harold E. 203
Pariente, Jean-Claude 537
Pascual, Jose 1. 106
Paul, Hermann 105,516
Peano, Giuseppe 246, 252, 262
Pei, Mario 250

Pena, Jesus 106
Pfeiffer, Oskar 476
Pfister, Oskar 50
Piaget, Jean 516
Pike, Kenneth L. 475
Pillsbury, Walter Bowers 516
Plato 206

Polome, Edgar C. 104
Powell, John Wesley 105
Preston, Richard J. 103
Przyluski, Jean 177
Putz, Martin 160

Quine,WillardV.0.61

Radin, Paul 200. 502. 516

Rahden, Wolfert von 59

Reichenbach, Hans 293

Rennison, John 476

Reuter, Fritz 501

Rice, Stuart A. 15,493

Richards, Ivor A. 9, 17, 153-156, 158-159,

161, 164-165, 174, 196, 202, 477, 516

Richter, Elise 104

Rivet, Paul 157, 177-178, 200

Robins, Robert H. 12

Robinson, David Moore 217

Rosiello, Luigi 12

Rousseau, Jean-Jacques 60, 67-69

Salmon, Paul 61

Sanders, Daniel 356, 381-383

Sapir,David J. 11,17-18

Sapir, Edward 7-11, 13, 15-19, 23-54,

57-64, 98, 103-109, 111, 149, 153-162, 165,

176, 178-179, 194-195, 199-200, 202-204,

218-219, 226-228, 236, 241, 245-250, 263,

275-276, 284, 287-288, 291-300, 446-447,

470, 473-484. 493, 496-498, 502, 516-517,

524, 536, 538, 544, 553-554, 556, 559

Sapir, Midge 18

Sapir, Philip 9, 13, 18, 21, 248, 295, 538, 561

Saussure, Ferdinand de 26, 157, 203, 297

Sauvageot, Andre 177

Scerba, Lev 247

Schaafhausen, Reimar von 327

Scheel, Heinrich 58

Index of Personal .\iimes

567

Schcltncr. Johaiiii ( icorg h7

Schclling, r ricdrich 5<S

Schin/. Albert 24S

Schleicher. August (\\')(\ 107

Schmidt, Wilhelm 200,517

Schneider, Ciisela 474

Schrader,Ollo 108

Schubert, Klaus 247

Schuchardt. Hugo 247-249

Sebeok,'niomas A. 64

Sechehaye, Albert 24S. 473

Seifert, Siegfried 5S

Seldeslachts, Herman 247

Shakespeare, William 112

Shapiro, Michael 25

Shenton, Herbert N. 248-249, 275

Sherzer, Joel 1 1

Shipley, William 58

Siegel, Carl 58-59

Silverstein. Michael 17, 23, 25, 47

Smith Stark, Thomas 104

Sommerfelt,Alf473

Spiller, Gustav 107

Stam, James H. 59-60, 63

Steinthal, Heymann 58-59, 63. 95-97, 106

Suphan. Bernhard 58, 67, 71-72

Sussmilch, Johann P. 59-60, 67-68, 73-74.

77, 86, 88, 92

Swadesh, Morris 15, 30. 42. 45. 293. 295.

446, 483, 556

Swanton, John Reed 200

Swiggers. Pierre 9. 10, 12, 17-18,21,23,25,

60, 64, 106, 109, 162, 247, 293, 295, 478^79,

483, 538, 555

Szilagyi, Denes 248

rcsnierc. Lucicn 293. 478. 4«H

Iliomas. Alex 529-530. 532-533. 547-549

nioniason. Sarah f Wcn 246
lilU)hash. Ion\ 4s I v\\ <<2(y U2 M5
<Tourney>' 67
Trahanl. JUrgcn <i<). (>4
Irager. (icorgc 1.. 48
Irombolli, Alfredo 157. 177-178. 200
lr(o)ubet/koy, Nikolaj S. 177. 247. 473.
53 1 . 54^'
Twaddell. William 1 . 4M

Vendryes, Joseph 17, l-^S. '"~

473, .502. 5 16

Verspoor. Marjolijn 160

Vitruvius 74

\\)egelin. Charles F. 30, 556

Volgina, Albina 58

Vossler. Karl 517

Waltz, Heidi 12

Watson. John Hroadus 195

Webster. Noah 69

Weigand, Hermann J. 327

Weisbart. Josef 248

Werlen. I war 47S

Westermann. Diedrich 1.58. 197

Whitney. John 526-528, .545-.547

Whitney. W illiam D. 63. 98. 516

Whorl. Ikniamin Lee 160.250.291.294

Wiener. Ninbert 42

Wilbur. Terence H.474

Winter. Werner 104

Wissler. Clark 4WS

Wuiulerli. IVler 478

Taboada, Manuel 106
Taylor, Allan R. 246

-^This is an error in Sapirs text: the name ol tlic secretary i.f the H. .I.n u.ulc ms sluuild he read as (Jean-Mcnn-
Samuel) Fonncw

Index of Concepts

abnormal types of speech 49
abstract(ion)78,85,514
accent 129-130, 142
acoustic (quality) 180, 237-238

act

functional -210
adequacy of language 33
adoption -^ borrowing
affirmation and negation 291,301
affixing 128-129
agentive 126

agglutination/agglutinative 106, 131-132,
199.509-510,513
aggregate 306-310, 312

reasserted - 306-307

simple - 306

whole - 307
allophone 30
alphabet 73, 91

also -^ orthography; writing
alternation 186-187
American spirit 168
analogical leveling 160, 220
analogy 105,123-124,171
analytic language 255, 257, 259, 261-262,
509

also -^ isolating

animal cries 98

anthropological linguistics 48, 153

anthropology 17, 21, 23, 47, 158, 162, 177,

197, 219, 221-222, 226, 297, 514

anthropomorphism 92

antonymy 298

apperception 76-77

arbitrary sign(s)/symbol(s) 70-71, 78, 227

archeology

linguistic- 108,139
area

phonetic- 125
areal

-diffusion 105

- relationship(s) 17
articulation(s) 121
artificial language 25,285
aspect 271, 299

aspiration 192

association(s) 35, 81, 229, 238
attention 75
attribute(s) 78

auxiliary language -^ international auxiliary language

behaviour (of speakers) 25, 35, 157, 169,
195, 494-496

group - 52

(human) - 34, 39, 50-51, 54, 153, 155,

164, 204-216, 220, 225-226

individual - 52

relational - 32

social - 34, 36, 160, 197, 205, 226, 505
also -^ socialized conduct
behaviourism/behaviouristic 154, 158, 161,
195

Besonnenheit 61, 76-77, 88, 97
bilingualism 48, 158, 478, 513
biology 64, 161, 225-226
borrowing(s) 44, 109, 125, 140, 220, 512

cartography

linguistic - 201
case 118, 128, 131, 144, 198, 254
category

conceptual - 498

covert - 250

grammatical - 80, 147, 196, 206, 245,

254,295,475^76,498,559

linguistic - 174, 196, 245, 475-476, 498

logical - 117-118, 126, 128, 143-144,

271

overt - 250

psychological - 143
causation (causative) 33, 171-172, 269, 279,
354-355
causing 353
change

- in language 43, 104-105, 115, 124,
146,148,168,197,511-512

-in culture 146, 148, 502

- of pronunciation 124
direction of - 45
factors of- 158, 198

external -198, 477, 512
internal - 198

Index of Concepts

569

graciiKil character ol 114

internal change 2S-2y. IW. 510

nH)rphok)gical - IW
clKiraclcrology [linguistic ] 1()<1
chri)iuil()gy

relative- 104. I 13

class(es)

closed - 36

t)pen - Mt
class indicator 314
classilication

- of experience 1%
-of reality 156
language -3(K3 1.477. 509
noun - 36, 559

also — ► distributional classilication; k)gical
classilication: structural classification

click(s) 199.531.549

cognatc(s) 44

combination(s)/combinalory 33.40. 190

communication 494—496
and expression 38

- of ideas 65, 69-70
method of -39
non-verbal - 39

comparative linguistics 44, 49-51, 64, 86,

195

comparison 292, 298

reference to terms of - 38
composite wholes 291 . 301
compounding 128, 191
concept 132, 143-144. 165, 195-196, 246,
315.498,559

logical -127, 131-132

type of - (as typological parameter)

159
concreteness of meaning 127
conligurational pressure 45
connotation(s) 156,391-392,400
constructed language 40, 52

also — • international auxiliary language

construction

- of a language 41
construclivity 33
conleiil

- of experience 170

- of expression 170, 389

- of language structures iod, iiiw,|4N.
558

- of morpht)logy 1 26

contrast (phonclic/symbolic -) 190,230 236.
238

contrastivc emphasis 559
correlation between language and

culture I4f>-I48
corruption

-of Idlers'

cosmopolitan linguiMics 4?y
counting 447
creativeness 272. 278. 2K1
cultural

hchaviour 475

- evolution 159

- history 36. .50. 219
setting 475

unit 143
culture 45.62, 135, 137. 140. 146-147.478
high - 508
low 508

dative (relation) 327

declarative 1 18

definiteness/indefinitencss 175. 292

definition of language 64

dcixis 296

demonstrative 127-128

denominating lerm(s) 104. 117

denotive (function of speech svmbo!) 2(IS.

210,212-214.216

derivation(al) derivative 12(> l2>i.2:>U l>i,

255-256. 259. 33S

description

phonetic lfil
descriptive

- function 31

- linguistics 17, 2" 4'? >1 10"
techniques 28

development

- of language 67,91
of speech 67

diachron\ 25

dialect 90. 113. 122. 177-178. 197.203.220.

479-480. 499-.502

differentiati(Mi 51 1

diffusion 558

of culture 142.513

phonetic - 513

morphoUigical - 557
direction(al) .^.^(V-331. .^2. .^46, 363. 368.
396. 398 39<;. 402. 407. 45H
directionality 474
directive .M9. 4 10. 4 14. 426
dissociation! s) 35

570

Index of Concepts

distance 329. 338, 342, 358, 394, 401, 409
distribution(al) 27-30, 32, 36, 39, 43

- of phonetic elements 556
distributional classitication 31

diversity

- of languages 58, 97
drift(s) 25, 45. 158, 198, 477^79, 511

ejective 531,549

elaboration (of word analysis) 510

ending-point relation' 37, 287, 291 , 293-299,

326-445

predicative - expressions 329-330,

333-335, 343. 357, 369, 375, 405, 431,

433

restrictive - expressions 409, 431, 433
envelopment 292,447
environment 29, 109, 135-149, 170

linguistic- 38

physical - 42, 108, 136-137, 139-143,

146-147

social/sociocultural - 107-108, 136,

139-140.143,146,173
equality 459^61

explicitly dynamic - 459^60

implicitly dynamic - 459-460
ethnology 51, 197

etymology 343, 397, 481, 513-514, 542
evolution

- of languages 288
-of mankind 59, 115

exclusive 144

existence 300

existent 292, 300, 304-308, 320, 322, 332,

447-448, 460

indefinitely massed - 308

individualized - 308

persistently whole - 306

reassertedly whole - 306

summated - 305-307

totalizable - 305-306

totalized - 307

whole - 305, 307
experience

- of reality 155, 476^77, 498. 505
analysis of -42, 176

elements of -173, 208, 453
grammar/language and - 34, 291,
539-540

growth of - 70

inventory of -300

linguistic -38

structuring of - 32, 250, 539

types of- 174
expression

linguistic -412
expressive (dimension) 159, 209, 212-214,
300

also -* symbolism

factor

linguistic - 236
family (as social institution) 89
family

language - 114, 159, 177-178, 196, 198,

511
feeling 82-83

- significance 230
language of - 72

field-work 483
flexibility 252, 257
forces

linguistic - 220
form 169

- and meaning 36, 165
-of culture 141,513

- of expression 175

- of speech 141

grammatical - 33, 106, 108, 111, 122,

135,146,154,224

overt - 208
form-feeling 155, 172
formal

- completeness 33, 170

- elaboration 172

- groundwork of a language 145, 148

- orientation 172
-process 132, 144

- relation(s) 36

- structure 33

- technique(s) 155, 172
function 31, 180, 477, 481

- of language 165, 169, 507, 509
grammatical - 79

fusion(al)106, 156, 199,218

gender 127, 144, 146, 175, 254

general linguistics 17, 57, 97, 103, 105, 158,

177, 245, 556-559

See the definition of this term given on p. 332.

Index of C'omcpis

571

genetic

- classification 31. i:(\ 15^^ 477.
509-5 1 1 . 556

- rclation(ship)(s) 17.30-31. 103 104.
1 13-1 14. 1 IS, 120. 139, 198, 2(H), 499

get)graphical variation 105
geography

language - 159
gesture! s) 394(). 116, 120, 150.209.
212-213. 237, 494-495, 503, 514
glottal stops 27-28.32.44. 142.551-552
giottali/alion

postglottalizalion 482

preglottalization 4(S2
glottalized consonants/continuants 27, 44,
49. 199,531-532,548-550
gradable

implicitly gradable 449
graded

- with reference to norm 450

- with reference to terms of compari-
son 450

explicitly -449

implicitly - 449

lower-graded 450

upper-graded 450
grading 37, 287, 291-292, 294-299, 301.
447-470

-judgment 460— 162

explicit dynamic - 460,462
implicit dynamic - 460. 462
nondynamic - 461-462

affect in - 462^66

- with decrease 463-^6.'^

- with increase 463-464
degrees of explicitness in - 44K— 449
linguistic - 451^57

explicit-' 455-457.468
implicit--^ 455-457.468

logical - 449-451 , 453-454, 457. 467

ordinary -469

polar -'468-470

psychological - 45 1 -455. 457. 468
closed-gamut grading 451-453
open-gamut grading 451-453.468

psychology of - 466
grammar 32, 85-86, 215, 504

genesis of - 89

gr.itntnan.in

and hiN language 167-176
grammatical

calcponcs • calcgi»r\
classilica(H)n 14"
complexity 86
-concepts 104.246.254
concfct '
dcrivai!

rclalioiul lMi.l>^-
-element 131-132

- proccss(es) KM. 117. I2S-130. 147.
224. 245, 249

- sense 86
gramniaticali/alion 483
grainiiiaticalness 24. 32
ground-plan of language 156, 2os

also — formal groundwork of .i i.iiuu.ui

growth

- of language 87-89

habits

language- 161. 197.206.221 222.265.

508,516
hearing 82-83

historical(-comparaii\e) linguistics i~,4>.
51.86. 103. 105. 153. 160, 195. 219. 474.4.S2.
558

history

- and variation/varieties of language
64. 103. 11M33

- of language 111 112. 120. 124^11^.
168

of linguistics 9''

hi)nu)[">lu)ny 482

iconicity 156

identity and difference 291.301

idiom(s)/idiomalic 269. 279

imitation 120, 122

iniiieralive 1 18

inanimate 175

inclusive 144

incommensurahje/incommensurabilily

176,476,498

indicalitni 291. 296. 300-301. 3ii5 3.vi/v 301

^ Further subdivided into ■abstract" and "spcciali/cd".

^ Further subdivided into "abstract" and •speciali/ed" (the l.iiu. unn uirthcr >ubdi%iMon " one - term sct»".

"two-term sets "."three-term sets", "four-term sets ')

572

Index of Concepts

indicative 331, 386-391

individual participation 31

inflection 246, 255

inflective/inflectional [type of language]

106. 131-133, 199, 509-510, 513

inslinct(ive) 72,75

interdisciplinarity 53

interjection(s) 72, 1 1 1 , 1 19, 213, 506

interlinguistics 247-248

internal vowel/consonant change 123, 126,

129,199

international (auxiliary) language 18, 37,

42, 50, 156-157, 203, 216, 245-288, 300, 326,

478,516

also -^ constructed language

International Auxiliary Language

Association 18, 43, 245, 287-288, 291, 296,

300. 326, 470

international understanding 265

interrogative 118

intonation(s) 39-40, 494

intransitive (verb) 270, 280, 353

invention of language 94-95

irregularity

grammatical- 124
isolating [type of language] 106, 123, 130,
132-133,198-199,509,513
item-and-arrangement 24
item-and-process 24

juxtaposing/juxtaposition 126, 128

kinaesthetic 237-238
kymograph records 193

language

- as a form of human behaviour
204-216,503-516

- as social activity 33

- as symbolic guide to culture 222
language, culture and race 1 7, 104, 109, 478,
482-483,515

language, culture, society and personality

21,26,34,107,154,476^77

language and environment 107, 135-149

language and thought/reason 50, 88, 98,
196

language family -^ family

languages of the world 177-178
langiie 157
laryngeal 44
limiting 291, 301

- condition(s) 26
lingua franca 198, 258, 265, 276
linguistics as a science 153, 168, 195, 473,
501

loanword(s) -^ borrowing(s)

locative 298, 327, 329-330, 333-338,
340-345, 347-350, 353, 358-359, 370, 376,
385, 387, 390-392, 394-399, 402-403,
405-406, 408^09, 412^14, 422, 431, 559
locution^ 329-330, 332, 335-336, 339-345,
347, 349-350, 357, 363-364, 374, 380, 390,
393,398,407,409,411,413
logic 37-38, 164, 271, 273-274, 278, 281,
283, 285, 458, 470

symbolic - 43, 272-273, 281

unconscious - 241
logical classification 136

male and female forms of speech 49

mathematics 37, 39, 43, 154, 170, 173, 176,

215-216, 266-267, 272, 277, 281, 285, 454,

458,461,476,498,505-506

meaning 37, 163-165, 196, 295, 304, 470,

513

elements of - 38, 385

factors of - 37

total - 37-38

also -^ concreteness of meaning
measuring 448
mechanism

grammatical - 120
metaphor(ization) 62, 84, 307-308, 315,
360, 393-394, 396, 504
method of description 27-28
methodology

linguistic - 482^83, 558
microlinguistics 47, 53
mind [human -] 76, 85
mixed-relational language 156

See the definition of this term given on p. 332.

huU'x of Concepts

573

mode 254

nionicnlancous (verbs) 272

monoiicncsis of lanizuagcs ^3. WI . 115. 1 IS.

157,:()()

morpheme 2f>-27. 2^). 3h, 3S. 40, 2^M. Alb,

492

morplioloiiy/iiu>rplu>U)gical 2(S 29, 107,

109. 117. 119. 122, 124-12.5. 12.S. 1.30.

136. 140-141. 143. 145. 190. 193. 199. 203,

217-21 9. 297. 4S2. 511.51 3, 550. 557-55S

morphophonemics 28-29. 44. 480-481

motion 340, 343, 346, 349. 353. 355, 38 1 . 392.

413—445 (symbt)li/ed by M)

movement' 332. 335, 340-342. 345. 3.50.

352-384, 395. 397. 402-404. 414-445

(symbolized by M)

moving 353. 355-384. 404. 414-445

(svmboli/ed by M)

nuiltiiingualism 48

music 51, 80, 176

mysticism 94

national language(s) 89, 264, 267, 271-272,
274, 277-278, 282, 284-286, 288. 326, 500.
502.515
natural

- language 96
-la\v(s)87

- sound(s) 74
naturalist(s) 70-71
naturalistic linguistics 63
negation 292.301
non-agentive (verb) 271.280
norm

reference to a - ?>H
normation 291.301
noun 117-119.268,278

\erhal origin of- 79
number 118. 127.253-255
numeral 145

object (of the verb) 131

objective 144

occurrence 300

occurrent 292. 3(K). 322, 448. 460

onomatopoetic 35. 1 1 9 1 20. 2 1 3, 506

operations

and operands 43
order 300

organicist vic\^ on lani.'uai'e 249 !*X^

oricnlalion Ih''

origin t)f 1.11 . ."^ vs. iiM. 4 \ MJr>

coi .r> of wt f^i

divine lhcor> i^

cxclumutury thcu: .

instinctive - 61

alMi -. inMinctivc

sensualist theory of - 60
origMi of speech M. 1 12. 1 IS
origmal language(s) 83. 'A)
orthography 187, 191. 260, 543-544. 547.
550

also — writing

overhearing 534. 551

paleontology 158

paradigm(atic) 36. 526

parallel deveU)pment 1 1>. r'> .i>^

parallelism 190

parole 157

part(s) of speech 79. 123

pattern! ing) 29-30. 32-34. 40. 43. 52.

153-156, r59-161, 179-193, 196. 203. 218.

295. 297. 473-476, 490. 492. 507. 510-511

- alignment IS^J
-concept IN9
-feeling l^Kl \')2

- group 189

- in/of culture 45, 157, l(M, 203. 221.
493_4i>4,5(M

- inof language 24.45.47. 153.224

- of behaviour 35, 50. 222. 493
-of elements 31

-of forms 25.30. P' '^'^

action - 39

grammatical - 31

lexical -512

meaning - 512

morphi>logical - 30

morphophonemic .V)

phonemic - M)

phonetic - 18.V184. 187. 190-193.486

social -47. 206. 211

symbolic - 1 54

unconscious 52

als«.> -• pouils in o( the ivillcm

perception 32

personal names 197

persiMialitv 23. .52. 284. 502

phdology philologist 157. 165. 167-168,

195-202.217-218

574

Index of Concepts

philosophy 155, 158, 161, 174, 196,219,221,
224,295

18th-century - 65, 68, 88, 92
philosophy of language 57, 96
phoneme(s)/phonemic 26, 28-29, 50, 153,
476, 481, 492^93, 503-504, 518-536,
539-553
phonemic

- elements 40
-hearing 32, 521, 541

phonetic

-change(s) 105, 121, 123-1 24, 139, 153,
192, 2 1 9-220. 225. 474-475. 493

- configuration 236
-definiteness 117

- difference 527
-feeling 186
-hearing 521, 541

- law 45. 104. 121, 123, 160, 474-^75,
484-^93, 524, 543

- process 179, 182
phonetic symbolism -^ symbolism

phonetics 199,225

experimental - 193
phonological analysis 50
phonology 28, 105, 140, 297, 482, 546, 551
phylum 105

place in the pattern -^ points in/of the pattern
placenamesl39, 158, 197
poetry 51

genesis of -80
point-' 332^45
point of reference'' 332^45
points in/of the pattern 32, 183-184, 187,
190-191,195
polar 292

polygenesis of languages 115
polysynthesis/polysynthetic 127, 132, 199,
509, 558

position 356-384, 396, 399, 402, 404, 407,
410,414,430,433
pragmatics 298
predicate 117-118,131
predicating/predicative term(s) 104, 117,
339

predication 333

prefixing 126, 128, 145,372,557

prestige 499

primacy

orders of - 449
primitive language(s) 73, 90, 108, 111, 170,
174, 200, 220, 255, 272, 281, 474, 486, 506
primitive people/tribe 73, 80, 85, 88, 90,
108, 111, 116, 137, 140, 147, 170, 195, 197,
222, 496, 514

primitiveness of Hebrew 68
process(-model; -type of statements) 24,
27-28, 106, 225
productivity 105
pronouns

indicative^ - 386-391, 406-408, 410,

434-435

relative** - 386-391, 406, 408, 411, 413,

440
properties of language 98, 124
proto-form(s) 474-475
prototype [linguistic -] 202, 491, 499
psychiatry 18,53

psychic condition(s)/difference(s) 71,74-75
pychic process(es) 173
psychoanalysis 35, 52-53
psycholinguistics 247
psychological

- classification 106, 126, 136

- influence 142

psychological reality of phonemes 18, 473,

480,518-536,539-553

psychologism 28

psychology 17-18, 47, 54, 92, 155, 158,

160-161, 164, 168-169, 173, 195-196,

219-226, 295, 457-458, 473, 51 1

- of a language 153, 179, 266, 277
-of sounds 179,183

cultural -21,23
experimental - 35
Gestalt- 53,158,161,195,224
personal -21,473

qualification 292
qualifier 322-323

See the definition of this term given on p. 332.
See the definition of this term given on p. 332.

^ Subdivided into independent indicative pronouns and referential indicative pronouns.
Subdivided into simple relative pronouns and compound relative pronouns.

Index of Concepts

575

qLiantiriahlc(s) 301 3()3.44.S
quantilicate(s) 301-303, 321-324
c|uantiticati()n 2W.3()1.44S

types i)f -" 2^)1-292
quanliticr 303, 313. 316-317. 32(K324. 44^)
c|uantity 291. 300-301. 316.3 IS. 447-44S
- iu)rni 447

ril.iiiMi\ 176

•(concepts 155, I7f>
Df the form of though i i
relevance 31
respiration 214
richness (of a lani>u.ii-«
rt)nianticisni W3

race 62. 90. 147

radical elemenl(s) 79, S4

ratio and proportion 291 . 301

rationalism 92-93. 95

reciprocal (verb) 270.280

reconstruction 4S. 49. 104. 112 113. \\>.

139,157.219,489

external - 104

internal - 104
reduplication 29, 129. 143-145. 199
reference 216, 309, 385-386, 390, 504-505

symbolic - 163-165

system of- 170
referent 164-165,300,385
referential

dimcnsit)!! 1.S9.479

- organizaticMi 34
also -^ symhuliMii

reflection — » Besonncnheil

reflexive (verb) 270, 280

regularity 271. 278, 281, 512

relater'" 327. 329-330. 332. 335-339,

341-343. 346-350, 357-358. 369. 375. 390,

392, 394-396, 398-399, 404-405, 43 1

relation 297-298, 326-327, 416-445

spatial - 293

temporal - 293

also — » cndinti-poinl rclalion

relational 127-128,199

- concept 1.^0. 300. 326, 328. .397. 477.
509

- statement 39

- svstcm 297

mixed-relational l.s6. 19W.21.S.510
pure-relational 1W,511)

relationships — areal reiatii)nship(s); ircnelic
relalion(ship)(s)
relativism'rclati\istic 153. 158

savage language KO

scalar 292

secret speech 198,514

selection 26

semantic relalion(s) 37

semantics (general -) 18. 37. 52, 154. 292.

294-295. 297-299. 470

-of t:nglish29l

component iai 2*'s

structural 297 24S
semasiology 62, 84
sensation 77

stream of - SI
sense(s) 82

also — feeling; iKMrnu- mi-Iu
sentence 294. 3(X)
setting of language 103

also — siKial

shorthand 253. 259-260. 262
sight 82-83
sign(s) 35. 156.223

theory of - 165
similarities

grammatical 114

lexical - 114

phonetic - 1S9
simplicity 268-269.271.278-280

grammatical - 270

phonetic - 262

slang 198.342

social

-bltKk4l

force \^5

need 41

reality 42

sciencc(s) 34. 52. 54. 161.225

setting of language 17
- Mluati'Vi ♦»<

For the different types of quantification, see p. 2^\ ( fn. 4). jind p M)\ Of these t>pc» only 'lolahty- » induded
in this Index, since the other types are not dealt with explicilly hv Sapir
' See the definition of this term given on p. 3.^2.

576

Index of Concepts

- solidarity 38

- structure(s) 18, 153
socialized conduct 205, 226, 507
society 24
sociolinguistics 48

sociology 17, 155, 158, 160-162, 168,
196-197, 219-222, 225-226, 473, 511
song 80, 518-536. 539-553

sound change -^ phonetic change(s)
sound law —>■ phonetic law
sound

-pattern(s) 179-193,236

- process 117
-symboI(s)119
-variation 117

sounding actions 79
space 291, 300

space-locative/space-location 330-331,340,
344-345, 349, 355, 385, 406-408, 410^11,
434, 440
speech

- organs 116, 119-120, 179, 182, 214,
225

development of > development

standard language 480, 500-501
status of linguistics 167, 219-226
stimulus 72, 161
stock

linguistic - 90, 105, 114-115, 125-126,

128-130,201-202,220
structural classification 31, 477, 509
structure 34, 43, 178, 268, 503, 511

-of (a) language 24, 26, 31, 33, 116, 169,

199,288,295,300,512

grammatical - 33, 91, 117, 141, 252, 271

linguistic - 40, 104, 193

morphological - 112, 117

phonemic - 503, 540

phonetic- 112, 117

semantic - 25

syntactic- 112

total/whole - 30-31
style(s)[ofE.Sapir]47,51
subject (of sentence)/subjective 36, 117-118,
130, 144

subjectivity 132
substratum

psychological - 1 18

suffix 127-129, 145, 191,259
local - 145

suffix -ness 272, 281

suffixation 29, 126, 128, 377, 382, 406

suggestiveness

symbolic -229-231
superlative 466-468

conditioned - 467-468

unconditioned - 467^68
symbol 141, 148, 163-165, 180,209,211-216,
223,391,495,502,515

- of solidarity 284
complex of- 136

symbolic (type of language process) 106
symbolic

- character/function of language 34,
156,207,477,515

- socialization 158

- system 155-156, 174, 207, 209, 212-216,
504-505

- of reference 24

symbolism(s) 18,35,43,70,76,107-108,111,
159, 163-165, 196, 221-224, 227-241, 266,
277, 281, 285, 288, 463-464, 495, 506-507

expressive - 35, 159, 192, 21 1-213, 215,

228

phonetic/sound -35, 43, 53, 160, 227-241,

503

referential - 35, 159, 210, 215-216, 224,

227-228, 504-505
symbolization 39, 104, 153-154, 156, 161,
165,214,412,479,508

primary - 156

secondary - 156, 214, 480
synaesthesis 62
synchrony 25
syntactic

- development 141
-relations 36,510

syntax 109, 119,136

- of pronouns 385
synthesis

degree of - 106, 159, 199, 477, 509
synthetic 257-258, 509, 558

-structure 171,199
system 32, 42, 507

grammatical - 147-148, 205

language - 43

morphologic - 132

phonemic - 504

phonetic - 117, 125, 141-143, 146, 185,

205,254,259,262-263,556

sound- 182,195

Index of Concepts

STJ

teaching (i)l languages) 274.282. 2K5
tcLhnit|Uc(s)

linguistic - 47, UK). I5W. IW, 2IK. 24h.

24^.285.477.509-51(1
tense IIS. 1 27. 253-255. 2^>3
lerm(s)

general - I3S

polar- 470
tcrniiiu)logy

scicntilic - 42
theoretical linguistics 292
theory of language 165

lhi>ui;hl and language — * language and Ihciuehl

time 300

time perspective(s) 52

tone(s) 19S-199.246

tone analysis 187

totality" 37, 287, 291-292, 294-325

totalizer'- 302-303. 309-325

transfer/transposition 479, 495

transference of thought 1 16

transitive 327. 353-354

translation 41-42. 257. 261-262. 265. 270.

276,280,328,464.496.513

infcMination loss in - 41^2
lype(s) 124

grammatical - 105

language - 30. 106-107. 12.S. 130-131.

509^

morphological -123-126. 130. 132. 146.

159
typological linguistics 64. 156.299
typology 156

formal - 36

linguistic -36. 245. 291

morphological - 159

Umlaut 122. 179. IW
uniti/cr 319
unity

gcnciic of lanftuago 1 16
universal (conceptual) grammar 2K7-288.
291-293.295.315

universal statements 302.314-315.323
universality of language 155.207

variation .5(K). 5()4

conJilional Is

individual 1K> >
verb 117-119, 268.278
Vermin ft 87

vocabulary 33. 78. 84. 108. 112. 136-140.
14h. 170-171. 215. 217-218. 246. 288. 478.
514

- change 45.512
growth of- 70
minimum - 257

NcnscI harmony 145

u7/-sounds 179 IM

whole/part relationships 292. 298. 302,

304-307

word 163-165,285,294

- formation 270. 280. 300

- mcaning(s) 24
writing 148, 197.2()8.50'<. ^(V^

phonetic - 524
phonologic -.*?> =
54.V544
systems i»f - 116

" See also the GUissary of technical terms" (p. M)}) concerning the con».cplual field uTlcHalily'
'^ For the general subdivision of totali/erv see S.tpir's scheme on p .^1^

Index of Languages

Achomawi 200

African* 7. 17. 107. 1 16. 157. 160. 199, 220,

498, 520, 540

Ainu 177

Algonki(o)n/Algonqiii(a)n^ 45, 50, 115,

140, 143, 160, 178, 199, 474-475, 484, 486,

488.509.511

Central Algonkian- 104, 219-220, 489
Algonquin-Wakashan 200
Alsatian 499
Altaic 509

American Indian*' 18, 30-31, 42, 49-51,
57-58. 86. 103-105, 107-108, 120, 125, 127,
129, 141-142, 144, 157, 160, 177-178, 186,
195, 220, 246, 474, 482, 498, 509, 520, 540,
548, 556-559

Andanian(ese) 157.177,202
Anglo-Saxon 103, 113, 123-124, 217-218,
487^88

Annamite-* 125, 177, 198-199,509,512
Anvik 489. 492
Apache 490

Arab(ic) 84, 143,253,509-512,515,529-530,
547-548
Aramaic 512
Armenian 557
Aryan -^ Indo-European
Asianic* 201
Assyrian 511

Athabascan^ 45. 50, 104, 128, 140. 158, 160,
178, 187, 192, 198-200, 219-220, 474-475,
483-484, 489-492, 511, 526, 528, 531,
545-546, 549. 557
Australian* 177, 200
Austric 198, 200
Aiistro- Asiatic 111
Aztek-Tanoan 200

-^ see also Uto-Aztekan

Baltic 557

Bantu 86, 129, 177, 199, 201-202, 509

Basque 115. 137, 177, 201-202, 500, 511

Beaver 489, 492

Breton 500

Bulgarian 499

Burmese 509

Burushaski 202

Bushman 177-178, 199

Cambodian -^ Khmer

Carian 159, 201

Carrier 489, 492

Caucasic^ 107, 141-142, 177, 199, 201. 557

North(em) Caucasia 177,531,549
Celtic 125, 499
Central Algonkian -^ Algonki(a)n

Chasta Costa 49
Cheremiss 125

* In this Index of Languages the names of (standard) languages, of dialects, and the generalizing names for
language continua (e.g., Chinese) and for complex historical-linguistic entities (e.g., Assyrian, Egyptian, Hittite)
are printed in normal type. Names of genetic clusters (subgroups, families, stocks and phyla: e.g.. Bantu. Indo-
European, Hokan-Siouan, Aztek-Tanoan), including the names of highly suspect regroupings (such as Japhetic)
are prmled in italics. Names of geographical/areal clusters are followed by an asterisk (e.g.. Mediterranean*).
For the term "Ostyak". see the entry below (and the reference to p. 178). The hyperonym "international auxili-
ary language" is not included here, but can be found in the Index of Concepts. - References are to the pages of
the present volume, not to the pagination of the original texts.

Sapir most frequently uses the spelling Algonkin or Algonkian. - None of the occurrences refers to the dialect
of Ojibwa called Algonquin.

"This is the spelling used by Sapir in the papers reprinted in this volume.

^ In Sapir's writings the term "(American) Indian languages" mostly refers to the North American languages.
^ This IS the term traditionally used for Vietnamese.
^ Other speUings: Athapaskan: Athabaskan. - Sapir also uses 'Dene' as synonym for 'Athasbaskan' (p. 198).

Or -Caucasian". The lemma 'Caucasic' covers the designations 'languages of the Caucasus" and 'languages spo-
ken in the Caucasus" used by Sapir (p. 141-142, 199).

Index III I (ini;ii(ij^es

579

ChiiiKiriki) 50

Chin 202

('hinanlcc200

("hincsc 6S. 91. 10(> 107, MS. 12.\ 125.

128-131. 133. 142. 175. 186-187. 197-lW.

217, 233, 245-246. 251. 253, 255-257. 262.

509-510.512-513.515

(iintoncsc ("liinesc 197

M.indarin Chinese 1^7
Chinook 49. 107,120,144.474

I ppcr Chinook 184-1S6
Chinook Jargon 246. 258
Chipewyan f98. 475. 489-492. 557
Chippewa 175
Chukchcc 177
Coahuiltecan 104
Com ox 49
Corcan' 177.512
Cree

Plains Cree 486-488

Swampy Cree 475. 489. 493
Croatian 514
Czech 499

Danish 107. 142.192.484.509
Dardic 202

Draviduin 177. 198,202
Dutch 490, 509

Egyptian 91
Elamitc 177

English 10, 15, 17, 29, 32-33. 36-37. 39
41.^52.73. 103. 107. 111-114. 119. 122-124
126-129.131.133.138-141.143,148.156-157
167. 172. 174. 184-185. 187. 190-193. 196
199. 203-204. 212. 217-218. 229, 231. 233
236. 246. 251. 253-255. 258. 262. 264-282
285. 288. 291. 293, 295, 299, 308, 312, 318
326-327, 329, 331, 334-337, 342-343, 345
349. 356-357. 385-387. 389-392. 394, 396
398. 400-403. 405^09, 412-445, 464. 468
484-488. 496. 500-501. 504. 509-514. 516
520. 526. 528-529. 534-536. 541. 545. 547
551-553

— sec also Anplt)-S;».\on; Old lvnj;lish

American l.ngluh 52K. 535. 547. 552

Hriiish I ngl»h 52M. 54"
liranum 202
Eskimo .57-58. 86. 107. III. 129-130. 142,

170 172. 186. IW-2()(). 207. :is. SW. 514
Espcranio2l6 '-' ^-i "^"258.261-264.
267.27.3-275.:^ 16

Etruscan 177. 197. 2oi
European* .^8.256
Ewe 107. 142. IW

Finnish 1 15. 5 11

I'ltmo- Uj^rian'' 1 77, 474. 511,513
Fox 486^488

French 10, 15. 37. 41. 52, 73. 104. .., .
125. 140-143, 167, 175. 186. 188. 193. 196.
198-199. 212. 231. 2.36. 246. 251. 257-258.
262, 266, 270-271. 273-274. 277. 280.
282-283, 285, 288. 293, 295. 299. 312.
326-327. 329-331. 3.'^4-335. 342-345. .347.
349. 356, 369, 387, 3^)0- .V>1. 397. 4(KM02.
405-409, 413-414. 416-445. 481. 490, 492.
500. 509-5 1 0. 5 1 2-5 1 5. 524. .535. 543. 552

mediaeval F iiiu h >]'^

Frisian 5 1 6
f'uci^iim liH)
F"ulbc 129

Ciaclic 125. 501. 5 Ih

Gallic (= Ciaulish) 125

Georgian 142

German 10. 15. 37. 41. 52. 113 11 :

171-172, 174-175, 184. 188. UW. 1}

246, 253, 257-258. 262, 268. 27('

278. 280-281. 283. 285. 288. 293. 29-

326-327. 329-331. 334-335. .V4;

345-350. 356. 375, 385-387. 390-391

397 40() 402. 405 4(W. 412-414. 4I(> Jl>.

4,S4. 4S6 487. 4SU 4«X). ^OO ^01. ■^<»*> 510.

512 514.523

. sec \ ;;■- -i». (HJ

High (icrman. Plalidculwh. S4ton. Sw«hi-
an;Swiv. {icrmnn
Low (icrman 501
North (icmian 1W»

^ Sapir uses the spellings (\)rcan (p. 177) and Korean (p "^ITl
^ Sapir also uses the term (J f^ro- Finnic (p. 51 1. .SI 3).

580

Index of Languages

Germanic 103 J 14, 192,485,488,499

Wesi Germanic 198
Gilyak 177
Gothic 484

Greek 68, 73, 112, 114, 123, 129-130, 167,
169, 172, 197, 206, 217-218, 258, 484-485,
492,509-510,512
Guinea (languages of-)* 177

Haida 50. 104. 107, 142. 186, 199-200, 531,

548

Hamitic 177,199,201

-^ see also Hamito-Semitic
Hamito-Semitic 157, 177-178, 482, 557,

559

-^ see also Semitic
Hare 489. 492
Hattic 159
Hebrew 57, 68, 73, 83, 190, 512, 514

Modern Hebrew 516
Hittite 51, 159. 177,201
Hochdeutsch 500
/yoA:«/7 50,200,483,557
Hokan-Siouan 31,200
Hottentot 107, 142, 144, 170-172, 177-178,
199,201,207

Hungarian 107,115,145,511
Hupa 103, 107, 143. 198, 475, 489-492, 531.
549,557
Hyperborean* 177

Icelandic -* Old Icelandic

Ido 252, 254, 257-258, 260-262

Indian -^ American Indian
Indo- Aryan 198,202,499

-^ see also Eranian; Sanskrit
Indo-Chinese 157, 177, 198

-^ Sin(it)ic; Sino-Tibetan
Indo-European 7, 17, 44-45, 49, 51-52, 104,
107-108, 114-115, 123, 127, 129-130, 139,
144-145, 156, 159-160, 177, 179, 186, 198,
200-202, 218-220, 474, 482, 484^85, 487,
498-499, 509-51 1, 513, 557-559
Indo-Germanic/Indogermanic -^ Indo-European
Irish -> Old Irish

Iroquois 107, 142-144

Italian 198, 251, 257, 499, 501, 509, 512, 514

-^ sec also Sicilian; Venetian
Italic 197

Japanese 177, 199, 246, 262, 512-513
Japhetic 201

Karen 202

Karokl03, 107, 143,557

Khasi 200

Khmer 509

Korean -^ Corean

Kott 178

Kutchin489,492,531,549

Kutenai 50

Kwakiutl 36, 44, 49, 57, 103, 107, 127, 129,

142, 144-145, 175, 186, 197, 530, 533, 548,

550,556

Laos 198

Latin 36, 68, 103, 114, 123, 125, 127, 132,

156, 174, 184, 186, 191, 196, 199, 217-218,

246, 258-259, 261, 266, 268, 274, 277-278,

283-284, 318, 337-338, 464, 484-485,

509-512,514

Humanistic Latin (= Neo-Latin) 512
Latino 257, 260

Latino sine flexione 246, 252, 258-259,
261-263

Latvian/Lettish 107, 142
Letto-Slavic 142
Lipan 490

Lithuanian 130, 511, 516
Loucheux 489, 492
Lycianl59,201
Lydian 159,177,201

Maidu 107, 145.557
Malayan 107, 129, 144
Malayo-Polynesian 111, 200, 515
Man 202
Maya 178
Mediterranean* 44
Melanesian 107, 144
Menomini 486^88

Sapir uses the terms Indo-Chinese, Sino-Tibetan, Sinic and Sinitic as synonyms (see p. 177).

///</(• X (}f /.iini^iia^es

581

Minoan 201

MixU'czapotcc 200

Mohave 200

Moni^ol Ml

Mon-K/mwr 107. 144, 177. 1W.200

MuiKln 177.200

Romanal 257-258. 2M>-?fil

Ronwncf 104.499.515

Russian 1 14. IIS. 28K. 499. .SI3. 516. 535.

5.S2

(ircat Ruvsian 4W

While Russian 4W

Nadene/Na-Dciii' 52. 104. 200

Nahua(tl)49. 107. 140

Navaho 30. 44, 49. 19S. 490-492. .531. .549.

557

Negro languages of the Nile* 51 1

Nicoharcse 200

Nootka 30. 33. 36. 44. 49-.5(), 107. 127. 129.

137. 143. 145-146, 172, 175-176, 186. 192.

199, 294. 481-482. 529, 531-532, 547-5.50,

556

Norwegian 516

—>■ see also Old Norse

Novial 43
Nubian 201

Ojibwa 486-488

Old English 113

Old High German 475.487-488.500

Old leelandie 484

Old Irish 190,481.524.543

Old Norse 192

Ostyak

Samoyed Ostyak 178

Ugro-Finnic Ostyak 178

Yenissei Ostyak 178
Otomi 200

Paiute49. 107.128

— > see also Southcin Paiutc; Ute

Papuan* 177

A7/;/m//;5(). 200

Phoenieian 91

Pisacfui 202

Plains Indian gesture language 1 16. 514

Plattdeulsch 516

Polish 499

Polynesian 157. 178. 199 200

Portuguese 107, 142

Proto-Malaccan 2(K)

Proven(;al 501

Modern Provcn(,-a! 5Hi

Sak 202

Siilisli 107. 142. 144 145

Sinnoyed 177

Sandawe/Zandawc 178.201

Sanskrit 114. 15^. 185,201.218.484.509.

511-512

Vcdic Sanskrit 202
Sareee 30. 50. 187, 489-492, 526-528. 531.
545-546, 549
Saxon

I'ppcT Saxon >(Ni
Seandinavian* 198
Scotch (Lowland-) 501
Semitic 1. 17.44-45.51-52.8(1. 107. \Z^, 144.
159-160. 177. 186. 198-199. 201. 474. 4X2.
485, 510-513, 558-.5.59
Serbian 499. 514
Shan 19S
Shasta 557
Shilluk 197. 199
Shoshonean 144.557
Siamese 125. 198. .509
Siberian* 1.57.178
Sicilian 499

SiniltiU- 177.198

• sec also lnilo-(liincsc:Sino-Tibcl«n

Sino-Tihi'hin 157.177-178

• \ec also lndi> ("hincNC. Stni(U)c
Sioiuui 107. 129, 140. 142. 144 I'S
Sioux 128

.S7(n7( 103. 114. 125.4K4.499.513..>.^:
.Southern Paiute 30 40 so t"" 140. 4,*^1.
521 -.525, 541 -.544

• sec alvo I'.iiuii- I Ic

Spanish 188. 192. 197. ISl. 257. 274. 283.

.5(K)..509.512.51h

Sudanese* 177.19^^201

Sumerian 177.202

Swabian 499

Swedish 142. 192.4S4.490

SvMss (ierman 4W-.50()

582 Index of Languages

jai 193 Venetian 499

Takelma 49-50, 57. 107, 129. 142. 144. 186,
200, 557 Wakoshon 44

Tasmanian 200 Wasco 49

Tewa 107, 142, 200 Welsh 528, 546

Tho 198 Wishram 49

Tibetan52, 192, 199

Classical Tibetan 192 Yahi 186

Modern (Central) Tibetan 192, 198, Yana 49-50, 103, 107, 127, 144-145, 186

509 Yokuts 27, 199

Tlingit 107,142,200 Yukagir 177

^'tf.'^l^'^Z^Z^^^ Yurok 103, 107, 143

Tocharian'M4, 52, 159, 201

Tsimshian 142, 145,531,548,556 Zandawe ^ Sandawe

Tungiisic ill

Turkish (languages) 177

Turkish 106,115,131,199,509-511

Tutelo 49

Ugro-Finnic -^ Finno-Ugrian

Ural-Altaic 103, 115, 125, 129
Ute 49, 144

-^ see also Paiute; Southern Paiute
Uto-Aztekan'' 49, 140

-^ see also Aztek-Tanoan

'" Sapir uses the spelling Tokharian (p. 201)

This is the spelling used by Sapir (see p. 140). Newman uses the spelling Uto-Aztecan.

L OOy bDJ ut) i