CHAPTER 14

LANGUAGE AND TERRITORIALITY IN ABORIGINAL AUSTRALIA

Alan Rumsey

INTRODUCTION

Linguists who have studied Australian Aboriginal languages (myself included) often get asked how to translate this or that English expression into ‘Aboriginal’. As anyone who has got very far in reading this book will know, there is no such language. Rather, there are, even now, at least 100 quite distinct Aboriginal languages, and 200 years ago there were many more. What relation is there between these language differences and other kinds of Aboriginal social differentiation?

Those who are aware of the diversity of Aboriginal languages generally think of this as a matter of tribal differentiation: each language is (or was) spoken by a distinct tribe. I am going to try to show in this chapter that that is a mistaken view — that language is actually related to the territorial differentiation in a quite different way. My evidence will come from two sources: from recent findings of the Aboriginal Land Commission in the Northern Territory, and from some other recent studies by anthropologists and anthropological linguists. First, I want to review some older ideas about this topic, which were widely held by anthropologists until recently, and probably still are held by most non-Aboriginal Australians.

LANGUAGE AND ‘TRIBE’

One of the preconceptions most people have about Aborigines (at least those living in central and northern Australia) is that they are ‘tribal’. What does that mean? Among other things, it means that they come in ‘tribes’. But what is a tribe? Some of the better-known examples of Aboriginal social groupings which are generally referred as to ‘tribes’ include Pitjantjatjara, Arrernte, Warlpiri, Kamilaroi and Wiradjuri. But what do we mean when we refer to these as ‘tribes’? If we look at the ways in which that word has been defined or used in relation to Aborigines, we find that it generally involves one or more of the following characteristics:

1. most or all members of the tribe live, or used to live, in a single, clearly bounded region;
2. they speak a common language or dialect which is unique to them;
3. they tend to marry within the tribe;
4. they share certain customs or cultural traits which are unique to their tribe.

Where do we get all these ideas about the nature of tribes? One thing to notice right away about them is that they are also characteristics which we attribute to the kind of social grouping which non-tribal peoples are thought to have instead of tribes, namely the nation-state. Indeed, what are now called Aboriginal ‘tribes’ in Australia were also commonly referred to in the nineteenth century as ‘nations’. Even as recently as 1976, an eminent linguist claimed of northeast Queensland that ’the only major difference‘ between ’the (so-called) “tribes”...and what are called nations in Europe and other parts of the world...was in terms of population size’ (Dixon 1976, 219-20).

But the idea of the nation-state is a very recent and historically specific one. Despite the fact that they are now everywhere, very few nation-states existed 200 years ago, and 400 hundred years ago they were unheard of. Since Australian Aboriginal cultures developed largely independently of outside influences for at least 40,000 years, we should be very cautious about assuming that a kind of institution developed here which just happens to include a good many of the characteristics of the nation-state. This is not to say that it could not have happened but, rather, that any such claim must be examined carefully. We cannot assume that any particular attribute will necessarily be found in combination with any other: so, each of the presumed ones must be considered separately in light of the available evidence. I will now do this for each of the four characteristics listed above, in reverse order.

The third and fourth characteristics may be quickly disposed of, as even those few scholars who have included them in their definition of tribe, have generally seen them as tendencies only. The rates of in-marriage (‘endogamy’) vary greatly from one ‘tribe’ to another, and are often lower than they are in a given area among ‘countrymen’ from more than one ‘tribe’. And there may be important cultural differences within named ‘tribes’. Circumcision, for instance, was practised by the southern Jawoyn people (east of Katherine, Northern Territory) in common with neighbouring ‘tribes’ to the east, south and west, but not by the northern Jawoyn. Thus, traits (3) and (4) cannot be taken as necessary attributes of the ‘tribe’.

In the end, those who have seriously tried to define the ‘tribe’ have generally fallen back on only the first two traits mentioned above: common links to a definite territory and common language. See, for example Howitt (1904, 41), Spencer (1921, lxiii), and Woodward (1974, 142) to be discussed below.

But this apparently simple definition has proven difficult to apply, for several reasons. First, regarding territoriality, the correlation between ‘tribal’ affiliation and residence is only a very loose one. This is true even in outback regions such as the Northern Territory. A very thorough statistical study which was based on data gathered there in the 1950s (Milliken 1976) showed that most members of most tribes did indeed live within a single identifiable region of the Territory, but that none of these regions was clearly bounded off from the others. Rather, there was an enormous amount of overlap among them, such that the membership of every ‘tribe’ was residentially dispersed among that of many others. The degree of dispersal has no doubt increased since European colonisation and the wholesale movement of Aboriginal people onto large government settlements and into towns. But it would be a mistake to read backwards from this to a situation in which we assume that tribes were sharply bounded residential groups. For, as far as we can tell from the available details of life history and genealogy going back to the pre-European past, a significant portion of Aboriginal marriages have always been between people of different ‘tribes’, especially in areas where differences at the level of ‘tribe’ are of greatest political importance. In such cases, one of the marriage partners usually ends up spending most of his/her time in the ‘tribe’ territory of the other, and the children may end up living in either, or yet another. Moreover, the life histories of many men show a high degree of residential mobility, within and beyond their ‘tribal’ territories’ (see Myers 1986, 77-102; Warner 1964, 467-90 for examples from the Western Desert and northeast Arnhem Land).

Map 8: Aboriginal group distribution according to residential patterns (from Milliken 1976, inside back cover)

This is not to say that there is no connection at all between tribal identity and territoriality. On the contrary, despite all the complications regarding residence, it is clear from what Aboriginal people say all around Australia that they think of the land as divided up into more-or-less clearly bounded regions, each associated with a label such as Warlpiri, Wiradjuri, and so forth. In this respect, there is a striking contrast between the residential map resulting from Milliken’s study, cited above, which is full of crossing lines and multiple overlaps, and that of Tindale (1974), which shows the whole of Australia neatly divided into a jigsaw puzzle of named tribal territories (see Maps 8 and 9). While the precise location of some of the boundaries drawn by Tindale has been disputed, no one has seriously challenged the idea that each ‘tribal’ name is associated in principle with a more-or-less clearly bounded region. Confusion arises only when we try to think of the tribal name as referring, in the first instance, to a group of people, and then to delimit its territory according to where those people live. Rather, it refers, in the first instance, to a piece of land.

Map 9: An example of Aboriginal group distribution according to Tindale (after Tindale 1974).

Almost every such label (Warlpiri, Wiradjuri, et cetera) also refers to a language. Why should this be so? Presumably because there is some kind of link between language and territory. Most people who have offered definitions of ‘tribe’ in Australia have assumed that that link is established by the fact that there is a distinct group of people, the tribe, who both occupy (and/or ‘own’) the territory and speak the language. Hence, the second of the two basic definitional features of the tribe: that its members are distinguished from others by speaking a common language.

But this assumption has now proven untenable. For, as more and more studies are made of Australian Aboriginal languages and their patterns of use, it becomes ever more obvious that almost no one speaks only one of them. As one linguist has put it: Australian Aborigines are ‘the leading contenders for being the most multilingual people in the world’ (Laycock 1979, 82). To this day, in mainland areas where Aboriginal languages are spoken, almost every fluent speaker is fluent in at least two of them, and it is not uncommon for one person to speak four or five, even in areas where the languages differ greatly in grammar and vocabulary. Nor does speakership cluster in such a way that, for example, languages A and B are spoken by all and only the members of some particular ‘tribe’ (Sutton 1978). Unlike in most parts of the world, language boundaries in Aboriginal Australia are not significant communicative boundaries, because people tend to be able to speak the languages of neighbouring regions as well as their own.

Here we are faced with an apparent paradox: the names for Aboriginal languages seem to identify them clearly with tribal territories, but it is not possible to delimit any clear-cut, non-overlapping groups of people on the basis of the languages they speak. To show how this paradox can be resolved, I turn now to some developments that have taken place during the course of recent land claim hearings in the Northern Territory.

THE ABORIGINAL LAND COMMISSION

During the past twenty years, especially in the Northern Territory, traditional Aboriginal forms of land tenure have been the subject of more intensive expert investigation than ever before. Much of this work has been done in connection with the passage and implementation of the Aboriginal Land Rights (Northern Territory) Act 1976, which for the first time gave some groups of Aboriginal Australians secure legal title to some or all of their traditional territories. This was done in two different ways. First, areas of the Northern Territory that had already been set aside as Aboriginal reserves (Arnhem Land, for example) were immediately converted to Aboriginal Land. Second, certain other kinds of land, chiefly vacant crown land, could become the subject of Aboriginal land claims, to be presented before a kind of royal commission — the Aboriginal Land Commission — which was specifically created to hear such claims and, after deliberating upon them, to make recommendations to the federal Minister for Aboriginal Affairs.

These land claim hearings have proven far more arduous and protracted than anyone had expected. Masses of data have had to be gathered for them concerning Aboriginal people’s relation to land, and thousands of hours of oral evidence from Aboriginal and other expert witnesses have been heard by the commission, resulting in hundreds of volumes of transcript. By contrast, among anthropologists specialising in Australian Aboriginal cultures, traditional land tenure had not been one of the major topics of investigation before the 1960s, and little had been published on it in comparison with topics such as kinship and ritual.

Under those circumstances, the work of the Aboriginal Land Commission was bound to produce some surprises, and it did. Most of them are beyond the scope of this chapter, which is specifically concerned with the role of language in relation to land and people. But in that regard, there is much to learn from the way Aboriginal people have chosen to present their claims.

The 1973-74 Royal Commission on whose inquiry the Land Rights Act was based — the Woodward Commission — clearly did not expect ‘tribes’ or ‘language groups’ to put themselves forward as claimants for land. Their finding was that the kind of Aboriginal groups that customarily held land were a kind of subdivision of the language group rather than the whole language group. In some areas the land-owning groups were thought to be ‘dialect groups’. More commonly, they were what it called ‘clans’. As for the more inclusive groups, which it called ‘language groups’ or ’tribes‘, the commission, adopting the minimal definition I have discussed above, saw each of them as having ’a common language, a commonly used name for that language and thus for the people speaking it, and an identifiable tract of country where those people live or used to live‘ (Woodward 1974, 142). Tribes were in this view dismissed as of little or no relevance for land tenure. In the words of the report: ’In no sense can the tribe be regarded as the basis of Australian social organization‘ (142). In all these respects, the Woodward Commission Report carries over a very influential view of Australian land tenure propounded by the first Australian Professor of Anthropology, A.R. Radcliffe-Brown (1930-31), who saw the ‘clan’ or ‘horde’ as having exclusive proprietary rights in land.

The Land Rights Act itself would seem to have incorporated the Woodward Commission’s views when it defined ‘traditional Aboriginal owner’ in such a way as to require Aborigines claiming land to show (among other things) that they comprise a ‘local descent group’. For language groups, if their membership is distinguished solely by their speaking a common language, cannot be ‘descent groups’, whereas clans clearly are. Where clans exist — in northeast Arnhem Land, for example — everyone belongs to one of them, namely that of his or her father (in other words, they are what anthropologists call ‘patrilineal’ clans). That is why such groups have always been readily accepted, in terms of the Act’s definition of ‘traditional Aboriginal owner’, as descent groups. And since each such clan is also associated with a particular set of sites or an ‘estate’, they are ‘local’ descent groups.

In the first nine land claims heard by the Aboriginal Land Commissioner (during 1977-80) the claimants presented their case for traditional ownership in groupings which were readily recognisable as patrilineal clans of the orthodox Radcliffe-Brownian sort. There was extensive testimony and deliberation concerning issues such as: Is this particular claimant group a clan? Does this particular clan qualify as traditional owner for this particular bit of claimable land? Can people with links to the clan other than through their father (for example, through their mother or grandmother) be included among the traditional owners of its estate? But there was no serious challenge to the assumption that the estates in question were clan estates, to which the points of possible linkage were provided by a core of kinsmen related through the male line, even if links to that ‘core’ might sometimes be through women.

The first real challenges to this assumption came in the early 1980s, when the commissioner first began hearing claims over land within the area of earliest and greatest impact from colonisation in the Northern Territory: the western half of the Top End. At the Finniss River hearing in 1980, the claimants presented themselves in groupings which, in terms of the Woodward Report, were not ‘clans’ but ’tribes‘: Maranunggu, Kungarakany and Warai (Aboriginal Land Commissioner 1981). On the evidence presented at the hearing, it is not clear that anything like the clan ever existed in this area. Certainly by 1980, there was no clear evidence for even a vestigial system of patrilineal clans or discrete clan estates. The claimants related themselves to land at the level of what they themselves called ‘tribes’, each of which was in principle associated with a well-bounded region, a language, and a name for that language.

In the following year a similar grouping, called the Malak-Malak /Madngele, presented themselves as claimants in the Daly River region, about 100 kilometres south of the Finniss River claim area. This claim differed from the Finniss River one in that, here, there was also at least a vestigial system of clan-like groupings, each of which was associated with certain sites within the larger Malak-Malak/Madngele region. Some of these groups had no surviving members. Sites or estates formerly associated with those extinct groups were said to have been ‘taken over’ by the Malak-Malak/Madngele group as a whole (Sutton and Palmer 1981; Aboriginal Land Commissioner 1982).

Further up the Daly River system, a claim was heard in 1983-84 over crown lands in the vicinity of Katherine, where a tradition also survives (among older people at least) of clan-like groupings, albeit with less clear-cut territorial associations than among the Malak-Malak/Madngele. The whole area under claim was said in Aboriginal testimony to have belonged to the Jawoyn people (Aboriginal Land Commissioner 1988) — a grouping which, again, would have been considered by Radcliffe-Brown to be a ‘tribe’, of which the various ‘clans’ were component parts.

In each of these three claims, as in some subsequent ones, the ‘tribal’ groupings not only chose to present themselves as such but were in the end accepted by the Aboriginal Land Commissioner as traditional owners of at least some of the land in question (Aboriginal Land Commissioner 1981, 1982, 1988). Indeed, in the Katherine claim (and perhaps in the Malak-Malak one, though the judge’s report is ambiguous on the matter) they were acknowledged in preference to the clan-like units even for areas where the latter were clearly of some relevance as local descent groups. This is in spite of the Woodward Commission’s conclusion of a few years before that ‘the language group was never a social or political unit, and so never a land-holding group’ (Woodward 1974, 145).

What are we to make of the apparent disparity between those earlier conclusions and the outcome of the land claim process in the Northern Territory? Part of the answer is that language groups have probably become more important as clans have become less viable as a result of European invasion and settlement. Consider, for example, the history of the Alligator Rivers region, just to the west of Arnhem Land. It has been estimated that, as a direct result of colonisation which began there in the 1880s, the Aboriginal population within the next two generations was reduced to about 5 per cent of what it had been (Keen 1980). As a result, many clans died out altogether, and whole language groups were reduced to about the same size as some single clans had been before.

A hundred miles to the south, in and east of Katherine, a similar, if somewhat less drastic decline was being experienced by the Jawoyn people, to whom I have referred above in my review of land claim proceedings. By 1940, only a small fraction of the Jawoyn population were living on or near country associated with their particular clan, and by 1980 it was impossible to find out, for most of the clans, where that country had been. But there was a very clear understanding where Jawoyn country was, and most Jawoyn people were still living on it (as indeed they still are, along with many other non-Jawoyn Aboriginal people). Most of the younger people had no clear idea what clan they belonged to, but their identity as Jawoyn was known by all, and of very great importance to them.

But what exactly is the basis of that identity? In our review of the notion of ‘tribe’ above, we were left with only two features — common territory and common language — and it seemed impossible to relate even those two to each other, since almost no one speaks only one Aboriginal language.

Here is where the evidence from land claim hearings has proven very enlightening. For, in land claims where language groups have been recognised by the Aboriginal Land Commissioner, the ability to speak the language in question has proven to be neither a necessary nor a sufficient condition for inclusion within the group. If it had been a necessary condition, many of the younger claimants — including almost all of them in the Finniss River hearing — would have to have been excluded, as their main linguistic competence is in various forms of English. Conversely, if it had been a sufficient condition, some people would have been included among the claimants — in the Katherine claim and probably elsewhere as well — who had had little or no association with the area in question, and who were not identified with it by themselves or anyone else. More generally, although many Aboriginal people in this area speak three or four Aboriginal languages, no one is equally identified with that many language names or ‘identifiable tracts of country’: not everyone who speaks Jawoyn, even fluently, feels entitled to say ‘I am Jawoyn’ or ‘Jawoyn is my language’ (Merlan and Rumsey 1982). The relevant relationship to language is not one of speakership, but one which is better glossed as language ownership (as in Sutton 1978; Sutton and Palmer 1981).

LAND, LANGUAGE AND PEOPLE

In order to see how that relationship is constructed in Aboriginal terms, let us reconsider the interrelationships among the three terms I mentioned above: land, language, people. I said that Westerners tend to assume that any link between territory and language will be established by the fact that there is a distinct group of people, the tribe, who both occupy (and/or ‘own’) the territory and speak the language. But in the land claim process, the links between language and land which have proven most relevant have not been of the kind which are mediated by links between language and people (as in Western ideologies of ‘tribe’ and nation) but, rather, are direct links between particular languages and particular tracts of country. Thus, it is not the case that, for example, Jawoyn country is called that because it is or was occupied by people who speak the Jawoyn language. Rather, it is called Jawoyn country because it is the region in which that language was directly installed or ‘planted’ in the landscape by Nabilil ‘Crocodile’, a Dreamtime creator figure who moved up the Katherine River, establishing sites and leaving names for them in the Jawoyn language (Merlan and Rumsey 1982). In this formulation, language and country are directly linked, and the mediated link is between language and people: Jawoyn people are Jawoyn not because they speak Jawoyn, but because they are linked to places to which the Jawoyn language is also linked.

What is the basis of that linkage? In all the land claims hearings where this matter has been taken up, it turns out to be a matter of what anthropologists call filiation — links through one or both parents. Thus, for example, when helping the Jawoyn claimants to prepare their case, the assisting anthropologists, Francesca Merlan and I, compiled a list of all living persons — approximately 400 of them — who were identified by Aboriginal people in the area as Jawoyn. We recorded the family trees of all these people, including, among other information, the clan and language group membership of their parents, and what languages they could speak. The latter did not correlate closely with language group membership. But parentage did. In every single case, it turned out that one or both of the Jawoyn claimant’s parents were also identified as Jawoyn. (In a very few cases these were what we would call ‘adoptive’ parents instead of ‘biological’ ones, but in all such cases, they were the ones who had actually reared the person.) From this and from claimants’ explanations about why they were considered Jawoyn, we concluded that filiation provided the basis for language group membership, and that ability to speak the language did not.

This conclusion was well-supported in the claimants’ evidence at the land claim hearing. When witness after witness was asked why he, she, or somebody else was considered to be Jawoyn (or of some other language group), the answer was never ‘Because I speak Jawoyn’, but almost always ‘Because my father was’ or ‘Because my mother was’.

Thus, contrary to Justice Woodward’s conclusions concerning ‘language groups’, it became possible to regard them as local descent groups. They were ‘descent groups’ in that their membership was determined not, as he had supposed, by speaking a common language, but by filiation from a member of the group. And they were local in that each is clearly identified with a particular region, with which the language is also identified.

It is in the latter respect — in the postulation of direct links between land and language — that Jawoyn and European world views are most different. But it is clear that the Jawoyn are not unusual in this respect among Aboriginal groups. The more one comes to understand the principle involved, the more evident it becomes just how widespread is its application. Let us consider some other examples, from anthropological sources rather than from Aboriginal Land Claim hearings.

Plate 5: Claimant Peter Jatbula gives evidence in Jawoyn (Gimbat area) land claim hearing, October 1992 (photograph by Alan Rumsey).

OTHER EXAMPLES OF THE LINK BETWEENLANGUAGE AND LAND

Southwest of Jawoyn country, in the Roper River region, the Mangarrayi people tell a story of the Dreamtime, in which the landscape in the eastern part of their territory was created by two Olive Pythons. As the story was told to Francesca Merlan, the Pythons speak in the Mangarrayi language, until they get to a place called Jambarlin. The place must be steep, for they climb slowly. When they get there, they start speaking a different language, Alawa. As the narrator tells the story, she herself switches from Mangarrayi to Alawa, and quotes them in that language as saying ‘Let’s you and I go quickly’. The narrator comments on the switch of language and says that now, from that point on [to the east] ‘people always talk Alawa’ (Merlan 1981, 144).

A thousand kilometres to the west, in the Kimberley district of Western Australia, the Ngarinyin people have told me of how their language originated at a place called Gulemen, ‘Beverly Springs’, where it was first spoken in the Dreamtime by Possum. From there he carried it all over present-day Ngarinyin country, and that is why the language is there today.

Far to the east, at Doomadgee in northwestern Queensland, the territories of the Ganggalida, Garawa, and Waanyi people are:

where particular languages are said to ‘belong’, implying that they fit there appropriately with other features of the landscape. When in that area using bush resources, and certainly when formally dealing with many totemic and other extra-human features of the landscape, it is appropriate to speak the language which belongs there. Other Aboriginal languages would not be effective in ritual matters; indeed use of another language may well bring forth hostility from totemic forces.

The concept of language as a fundamental characteristic of landscape is also evident from mythic accounts where travelling totemic figures change their language on reaching the boundary of a linguistic territory — as, for example, in the story of a snake who switches from Jingalu to Waanyi at the present border between the two. (Trigger 1987, 217-19).

In central Australia, T.G.H. Strehlow, who grew up among Arrernte (Aranda) people at Hermannsburg Mission, was told of how a Dreamtime horde of native cats:

after travelling from Port Augusta in South Australia through the territory of the Jankuntjatjara and Matuntara, entered the Aranda area at Ilbirla — a series of springs... As soon as they crossed the Palmer River their ears were deafened by the chirping of crickets in the river grass; ...in their confusion, they began to speak in a mixture of Aranda and ‘Loritja’ [Luritja, Western Desert Language] after speaking only ‘Loritja’ during their previous travels over many hundreds of miles of Western Desert Country.

From this point on they...began to address each other by [class-] names. And after they had gone on, night overspread the land behind their sterns, while they went forward in broad daylight.

And they laid down [as a barrier] that great expanse of sandhills,...those sandhills covered with stands of desert oaks. (Strehlow 1965, 133)

The ‘class names’ to which Strehlow refers are part of a system which divides the whole of society into four or eight ‘skins’, and specifies who may or may not marry each other. (This system is similar to that described earlier by Bavin for Warlpiri in chapter 6.) Strehlow reports that the Luritja narrator of this myth concluded by saying that the native cats:

through raising a sandhill barrier between the Aranda and themselves...and through causing night to fall on the people living south of this border, had authorized [them]...to marry indiscriminately...prohibitions on marriage operated only for the Aranda. (Strehlow 1965, 134)

Strehlow’s conclusion from this is that, in the Aboriginal view, these systems of marriage classes:

were based on the land itself [emphasis in the original]: for it was the same wandering horde of ancestral beings which had validated the ‘classless’ kin-grouping system of the Western Desert groups south of the sandhill barrier...and which had instituted the rule that all groups living north of this barrier had to address one other by [class-] names. (Strehlow 1965, 134)

This conclusion was very sound as far as it went, and at the time it represented a real advance over most Europeans’ understanding of the Aboriginal world view, in its emphasis on meaningful features of the landscape as the basis of the world order. But in light of what we have learned since Strehlow’s time about the links between Aboriginal languages and land, it seems likely that, in the Western Desert view which is represented in this myth, it is not only the distinctive Arrernte (Aranda) marriage system that is ‘based on the land itself’, but also their language. Indeed, this myth seems to show this at an even greater level of specificity than the myths from other areas I referred to above.

To this day, Western Arrernte country is linguistically ‘mixed’ between the two languages referred to in the myth: all Western Arrernte speakers also understand the nearby Luritja dialects, and those whose clan countries are near the Palmer River speak Luritja as well, and often mix in Luritja words when speaking Arrernte (Diane Austin-Broos, personal communication). The myth would seem to be establishing a connection between this linguistic mix and the landscape. As in the other myths, the travelling Dreamtime hero who creates the landscape switches language at a certain point, and by doing so creates a socially significant boundary (which in this case he also marks by laying down a series of sandhills). But rather than switching from one language to another, he switches from one language to a mixture of two — the same mixture which is still characteristic of that area.

Moreover, the myth at least indirectly (through a chain of cause and effect) relates specific features of the landscape at Ilbirla to its ‘mixed’ linguistic identity. Arrernte country, which begins there, is much more well-watered than the neighbouring Western Desert (‘Luritja’) area, and its spring-fed rivers grow the grass that feeds the crickets whose deafening chirps produced the confusion which caused the native cats to produce a mixed language.

I have dwelt at some length on this central Australian myth partly because it provides a particularly clear example of the direct link between land and language in Aboriginal Australia, and of how different it is from the Western notion of a ‘national language’. In the latter, the nation is thought of as, first of all, a group of people with a common history, common language, and common territory with which the language is thereby associated. The association between territory and language is a contingent one, subject to change as one people conquers or assimilates another.

By contrast, in the Aboriginal myths which associate language and land, no account at all is taken of people, or peoples. Languages, or even mixtures of them, are directly placed in the landscape by the founding acts of Dreamtime heroes. From that point on, the relation between language and territory is a necessary rather than a contingent one. People too, or their immortal souls, are similarly grounded in the landscape, in the form of spirit children (or ‘conception spirits’) associated with specific sites, and via links through their parents to more extensive regions. But the languages were already placed in those regions before any people came on the scene. The links between peoples and languages are secondary links, established through the grounding of both in the landscape.

ACKNOWLEDGEMENTS AND CONCLUSION

I wish to thank Diane Austin-Broos and Harold Koch for their advice on aspects of this chapter; Wilma Merlan and Francesca Merlan for their comments on a draft of it; and Francesca Merlan for the benefit of our many conversations about land, language and social identity, and for bringing to my attention the passage from Strehlow cited here. Portions of this chapter have previously appeared in Anthropological Forum 6 (1), and we thank its editors for permission to reprint them here. Thanks also to the many Aboriginal people in Katherine and the Kimberleys who have discussed these matters with me. I am particularly indebted to David Mowaljarlai, who in 1976 concisely summed up the point of this chapter by explaining to me that ‘everything goes back to the land’.

FOR DISCUSSION

This chapter is about the way Aborigines formulate the relationships among land, language and people.

1. To what other aspects of Aboriginal social life might that kind of formulation be related?
2. Why do you think there are so many different Aboriginal languages?
3. If the idea of the nation-state has led to false assumptions about Aboriginal social life, what other common Western preconceptions do you think might have to be overcome in order to understand it better? Do you think it is ever possible to start trying to understand another culture without any preconceived ideas? If so, how? If not, is the task a hopeless one? Why or why not?

REFERENCES

Aboriginal Land Commissioner
1981	Finniss River Land Claim, Australian Government Publishing Service, Canberra.
1982	Daly River (Malak Malak) Land Claim, Australian Government Publishing Service, Canberra.
1988	Jawoyn (Katherine Area) Land Claim Australian Government Publishing Service, Canberra.
Dixon, R.M.W.
1976	Tribes, Languages and other Boundaries in Northeast Queensland. In N. Peterson (ed), Tribes and Boundaries in Australia, Australian Institute of Aboriginal Studies, Canberra, 207-38.
Howitt, A.W.
1904	The Native Tribes of Southeast Australia, Macmillan, London.
Keen, I.
1980	Alligator Rivers Stage II Land Claim, Northern Land Council, Darwin.
Laycock, D.
1979	Linguistic Boundaries and Unsolved Problems in Papua New Guinea. In S. Wurm (ed), New Guinea and Neighbouring Areas: A Sociolinguistic Laboratory, Mouton, The Hague.
Merlan, F.
1981	Land, Language and Social Identity in Aboriginal Australia, Mankind 13, 133-48.
Merlan, F. and A. Rumsey
1982	The Jawoyn (Katherine Area) Land Claim, Northern Land Council, Darwin.
Milliken, E.P.
1976	Aboriginal Language Distribution in the Northern Territory. In N. Peterson (ed), Tribes and Boundaries in Australia, Australian Institute of Aboriginal Studies, Canberra, 239-42.
Myers, F.
1986	Pintubi Country, Pintubi Self, Smithsonian Institution Press, Washington, and Australian Institute of Aboriginal Studies, Canberra.
Radcliffe-Brown, A.
1930-31	The Social Organization of Australian Tribes, Oceania Monographs, Sydney.
Spencer, B.
1921	Presidential Address. In Report of the Fifteenth Meeting of the Australian Association for the Advancement of Science.
Strehlow, T.G.H.
1965	Culture, Social Structure and Environment in Aboriginal Central Australia. In R.M. and C.H. Berndt (eds), Aboriginal Man in Australia Angus and Robertson, Sydney.
Sutton, P.
1978	Wik: Aboriginal Society, Territory and Language at Cape Keerweer, Cape York Peninsula, Australia, PhD thesis, University of Queensland.
Sutton, P. and A. Palmer
1981	Daly River (Malak Malak) Land Claim, Northern Land Council, Darwin.
Tindale, N.
1974	Aboriginal Tribes of Australia, University of California Press, Los Angeles.
Trigger, D.
1987	Languages, Linguistic Groups and Status Relations at Doomadgee, an Aboriginal Settlement in North-West Queensland, Australia, Oceania 57, 217-38.
Warner, W. L.
1964	A Black Civilization, Harper and Row, New York.
Woodward, J.
1974	Aboriginal Land Rights Commission Second Report, Australian Government Publishing Service, Canberra.