Somewhat more than two hundred Aboriginals living at Numbulwar Mission, and smaller numbers on Groote Eylandt and at Ngukurr (Roper River) settlement speak this language. At the time of my fieldwork, most people older than thirty (especially the women) had only a limited knowledge of English and there were some virtual monolinguals. Most of the children speak Nunggubuyu but are also learning English. English spoken by the older people tends to be in large part based on the creole spoken at Ngukurr and elsewhere in the area; that spoken by the younger people is mainly standard English learned from Australian schoolteachers and mission staff.
Local patrilineal clans are associated with particular territories (estates) as owners; this territorial association also involves association with particular totems and ritual or (public) song performances. Core Nunggubuyu-speaking clans are these:
| Mandha:yung moiety | Yirija moiety |
| Murungun | Nunggargalug |
| Ngalmi | Nun-dhiribala |
The two (patrilineal) moieties thus subsume the four clans shown, and the moieties can be extended to categorise persons from neighboring "tribes." Moreover, because of the "brotherhood" relationships among clans linked by the travels of particular totemic beings, each of the four clans can be associated with a number of "brother" clans in neighboring "tribes." To some extent the four clan terms can themselves be used in a wide sense to cover such assemblages of brother clans; this is more often done with the clan name Murungun than with the others.
Each clan probably had around forty or fifty people prior to white contact, and if anything the average clan population is increasing in recent decades due to improved medical care. Most of the clans were themselves divided into smaller subclans, often without names (or with special names which were used only rarely); see text 72.
Major rituals were circumcision (Mandiwala), mortuary (Ya:di), Gunabibi, and Madayin (Ru:l). All were and are male dominated and all include secret episodes performed and seen only by initiated males. However, much of the Mandiwala and Ya:di are public. Nowadays these two rituals typically involve the didjeridu-accompanied wungubal type of singing, though in the past the Mandiwala involved the simpler type of singing practised by the Warndarang and Mara peoples to the south. Gunabibi is in large part a fertility ritual; the Madayin is more solemn.
Both the wungubal songs, and specific portions of the various rituals, are owned by particular clans or their headmen. Men whose mothers were in a given clan are the junggayi (we may translate this as 'managers') of the clan, a status distinct from but about as important as that of the 'owners' (bungawa) of the clan itself. See also text 49.
In addition to the major communal rituals (which not only involve virtually all mature Nunggubuyu men, but also usually attract visitors from other communities), there were minor rituals such as the increase rites of texts 58 and 59.
In addition to the four core Nunggubuyu clans, it appears that the Manggura clan just to the north on the coast of Blue Mud Bay was fully or partly Nunggubuyu-speaking, though the older people in the clan also speak Ritharngu and the clan has close marital and ritual relations to Ritharngu clans. Moreover, in recent times, due to the extinction of Warndarang and the sharply reduced vitality of Ngandi, the Nunggubuyu language has spread to several clans whose territory is near the mouths of the Rose and Roper rivers, south of the core Nunggubuyu territory. Many of these people are now classified by anthropologists and others as "Nunggubuyu," but their clan ancestors were not. Such clans as the Nunggangulgu, Nung-gumajbar, and Nunggayinybalany (Yirija moiety), and the inland and coastal Numamudidi (Mandha:yung), are in this category. As indicated in the notes to text 72, a formerly Dhay ' yi-speaking clan found well to the north has also been largely swallowed and absorbed by the Nunggubuyu.
In this region the primary linguistic division is between the Yuulngu and non-Yuulngu languages. Yuulngu is a tightly-knit family in the northeast of the Arnhem Land Reserve; it includes the Ritharngu and Dhay'yi just north of the Nunggubuyu. Non-Yuulngu languages of the area form another genetic group, hut one which is much more internally diverse. It has not yet been adequately classified genetically and has been referred to simply as the "prefixing" languages pending a more authoritative classification. It includes the Nunggubuyu and the immediately adjoining languages in the west, south, and east.
I have previously published a volume on Ngandi and have some unpublished grammatical and textual material on Anindhilyagwa; these are to the west and east of Nunggubuyu, respectively. It is clear that these go back to a common ancestor, and it is possible to reconstruct many specific forms (including affixes) for the proto-language. It is not yet clear whether any other languages such as Ngalakan (farther west) will have to be added to this genetic group.
To the south we find a group consisting of Warndarang, Mara, Alawa, and perhaps other languages like Mangarayi. Present indications are that this group is not only typologically but also genetically a unit. It is also clear that this group is genetically related to the group containing Nunggubuyu, and some concrete verb forms can be reconstructed, but the genetic distance at this level seems considerable.
At a much greater depth still, it is probable that all of these languages (including the Yuulngu group) are genetically related. However, the Yuulngu languages are more closely related genetically to languages well to the south in Central Australia than they are to the immediately adjoining non-Yuulngu (prefixing) languages.
Nunggubuyu and all adjoining languages, including Ngandi and Anindhilyagwa, are definitely mutually unintelligible. Nunggubuyu people who have not actually learned neighboring languages cannot understand them; in the cases of Ngandi and Anindhilyagwa even cognate forms have often been disguised by phonological or analogical changes. Because Nunggubuyu does not grade into these neighboring languages, there is absolutely no code-switching among them in the texts, though an occasional direct quotation is given in Anindhilyagwa when appropriate.
Nunggubuyu has borrowed some words (either directly or via other Aboriginal languages) from the Macassan (Indonesian) sailors and hunters who used to work seasonally in the Gulf (texts 166 to l68). To my knowledge the lexical borrowings are limited to nautical terms and a few others (almost all nouns) and are not extensive.
On the other hand, a fair amount of English (including creole) borrowings have already taken root, including about twenty important verb stems (typically with thematic suffix -dha-, as in =wirindha- 'to clothe' from creole wirim 'to wear'). Transitive stems, moreover, usually contain the creole transitive suffix -im as in this example. In addition to assimilated borrowings of this type there are other English borrowings, including particles like too ('also'), jaldu ('only', from creole that'll do), and now. However, some passages simply involve partial code-switching as the narrator attempted to explain a point to me in English. In this event there might be a Nunggubuyu expression, followed by the same thing in English for my benefit, as a kind of aside. On the other hand, sometimes the speaker would use an English word (or English loanword with Nunggubuyu affixes), then realise that an intrusive English word had crept in, and finally he would give the closest true Nunggubuyu equivalent.
We will not present a grammatical sketch of the language here. However, we should note that this is a language with an elaborate noun-class system, usually marked in prefixes (on nouns and verbs). For humans, we have MSg, MDu, FSg, FDu, and Pl (three or more regardless of gender). For nonhuman nouns we have several classes which are labeled by a capitalised form of one of the actual prefixes (though other prefix variants occur as well): NA, NgARA, ANA, MANA, WARA. WARA can also be used for humans, especially for indefinite plural. Subtypes of ANA, ANA(ø) and ANA(wu), are distinguished only by different patterns of agreement in verb forms.
Nunggubuyu is extremely complex in its nominal and verbal morphology. Unfortunately, it is difficult even to give meaningful glosses to individual morphemes, since often the meaning of the whole set of affixes on a word is not merely the sum of the individual meanings but rather is a special configurational meaning.