Research Section > Western Desert Project

Diffusion of section terminology in the Western Desert of Australia

(the current on line version below only contains the introduction, the methodology, a general overview of the results and the conclusion)

Laurent DOUSSET, copyright 2001-2003

please do not quote without permission

The social organisation of Australian Aboriginal society is often associated with the existence of social category systems, such as moieties, sections and subsections, that is the division of society into a limited number of groups or named classes whose relations are organised by rules of marriage and filiation. Thus, the Kariera, for example, a "tribe" on the west coast, possess four sections (Panaka, Purungu, Karimarra and Palyarri), whose relations are such that a person of the Panaka section marries a person of the Purungu section and a person of the Karimarra section marries a person of the Palyarri section. The children of a Panaka women are Karimarra and vice versa, and the children of a Purungu women are Palyarri and vice versa. Those social categories are, thus, ascribed by birth, function as an overlay compatible with kin categories and function basically as labelling devices.

What is less often associated with section systems is the fact that their diffusion into some areas is relatively recent. In the Western Desert cultural bloc (Map 1), the spread of the system was completed in the thirties or forties, superimposing itself onto a social organisation that hitherto was dominated by socio-centric or ego-centric generational moieties. Another characteristic rarely mentioned is that the indigenous names for those sections, for which I have given those of the Kariera, above, are not identical among all groups possessing this form of social organisation. In other words, although the structural principles of this form of social organisation diffused over thousands of kilometres, the terminology associated with it has not always been adopted "as is", and has sometimes been considerably modified.

In this study, I analyse how section system nomenclature has been transformed during its diffusion into the Western Desert cultural bloc, and attempt to locate possible regions of origin or entry into the desert of the diverse terminologies associated with the section system. I also suggest possible routes of diffusion of the system itself.

Researchers such as McConvell and Brandenstein have provided crucial answers to the question of origins and diffusion: one point of emergence of the four section system is located in the Pilbara region, on the west coast of the continent. The method used by McConvell is principally linguistic and articulates the transformation of a term in the system (a section name) in space and time.

Yet the existence of a greater number of terms used in the Western Desert than formally necessary for the functioning of this kind of social organisation adds a supplementary difficulty to attempts to understand the principles of diffusion. It thus may be necessary to add to the analysis of linguistic transformations one focused on the rules of combination and mutual substitution of the section terms.

Is it possible to extract a "logic" or general rule inherent in the diffusion and substitution of sections names? Are there one or more "concepts of sections" progressing in space and reflecting the history of combinations – and transformation of combinations – to which those sections were subjected while being adopted by different groups? These are the general questions of central interest to this study.

It is first useful to define the "concept of a section". In fact, this notion is not to be confused with what Brandenstein called the "substance of a section", built around the signification of section names in their dualistic relation. The author‘s hypothesis suggests that the ?] division of sections follows two contrasting groups: a dynamic group possessing the characters "active" and "passive" and a "blood" group with the characters "cold" and "warm". The four section system of the Kariera would be part of "physiognomical typology", where each section is the combination of one element of each of the contrasted groups. Thus, Pannaga is "cold-blooded" and "active", Paltjarri is "warm-blooded" and "passive", etc. The section system is also a natural and totemic classification such as, for example, reptiles are of "cold blood" and therefore associated to Pannaga and Purungu; fire is classed Karimarra, dew Purungu, and so on. Yet Brandenstein goes further in proposing a thesis according to which the Aborigines would have used the biological laws of heredity, such that the "human temperaments" attributed to each section are passed through the generations according to the rules that organise the relations between the sections. Thus, a Purungu man would be "passive" and of "cold blood", and, marrying a Pannaga women with "same blood" but "active", would have children of Karimarra section whose temperaments would be "active" and of "warm blood" (see Brandenstein, 1970 and 1972).

The aim here is neither to confirm nor disconfirm the theses formulated by this author, although I will allow myself a brief comment in the conclusion, but to eliminate possible confusion between the "substance of a section" (Brandenstein’s humours) and the "concept of a section" used here. The latter is the "underlying logic" of the combination of a section with others accompanying it during its diffusion, i.e. its structural position. This "logic" is the result of the internal coherence of the system, i.e. the sequence of necessary events or causalities determined in concordance with those structural relations or positions. In other words, the main assumption made here is that the four section system diffuses along with the associated terminology, but that this terminology may vary, as when a new term replaces one of the existing terms. Both terms, the new one and the replaced one, constitute part of the same "concept of a section".

Thus the notion, and therefore the subject of this study, is far less "substantial" than were Brandenstein’s preoccupations. It is not the essence of the terms that is of interest here, but their existence and coexistence, i.e. their phenomenological and structural aspects. The logic or "truth" sought and the necessary verifications it will have to undergo are those formulated by Russel as being of the fourth type, that is,

"The correspondence theory of truth, according to which the truth of basic propositions depends upon their relation to some occurrence, and the truth of other propositions depends upon their syntactical relations to basic propositions "(1997 [1950] : 289).

What I do not examine is the deep significance, metaphoric or symbolic, of the terminology of social organisation, but the relational logic that accompanies the elements of the terminology during the diffusion of the four section system.

 

Four section systems

Hammel (1960: 15) suggests that section systems respond to three basic assumptions: first, section organization results from the permutation of lineal kinship affiliations; second, all groups of affiliates are exogamous; and third, the enrire model is endogamous.

Thus, formally and without being ethnographically correct, and certainly not for the Western Desert, society is divided into two matrilineal and two patrilineal moieties which are divided up into sections. Following what seems to be the most comprehensible representation, there are two matrilineal moieties, A and B, and two patrilineal moieties, 1 and 2. The combination of the moieties gives rise to four sections, e.g. A1, A2, B1 and B2. The relations between sections are expressed through rules of marriage and filiation; thus, a potential wife for male Ego is of the same section as the children of the sister of his father and the children of the brother of his mother, i.e. in the section combined with Ego’s opposite matri- and patrimoiety (Figure 1).

It is important to recall that, as Radcliffe-Brown (1931) and Scheffler (1978) have stressed, the section system is not at the basis of rules of marriage and descent, but is a ceremonial, and sometimes totemic, grouping compatible with the classification of kin. For example, some categories of female kin in the section of a male Ego’s intermarrying section (such as MBD, FZD, DD and FM) are not classified as "wives".

Figure 1 : Formal structure of the four section system

A woman of section A1 marries a man of section B2 who is related to her as "spouse", and their children are A2. If the woman is B2 and the man A1, then the children are B1. Thus, equations (or sections in the same row) link intermarrying sections, and vertical arrows (or sections in the same column) connect mother to child. This will be the general representation used in this study.

Despite similarities between a certain number of the indigenous names for sections used by different tribes along the coast and groups in the Western Desert, some 40 more or less distinct terms and, more important, their variation in structural position (i.e. swapping of terms between rows and columns) call into question a strict identity between terms, even those that are quite similar.

To illustrate, two groups, named 1 and 2, possess sections, the names for which are, in Group 1, A, B, C and D. Group 2 may use terms which are similar or identical to some of the names used by Group 1, say A’, B’ and C’. But Group 2 uses one term which is completely different from all terms used by Group 1, say E. In addition, the names used by Group 2 may have different structural positions in the system (Figure 2).

Figure 2 : Formal example of differences of section names and structural positions

From a global point of view, a man B’ in Group 2 marries what would be his daughter (C) in Group 1. Thus, even if A and A’, B and B’, C and C’ are linguistically close terms, the transformation of the structure of the system in its diffusion from Group 1 to Group 2 demands a global approach to the potential and observed combinations of section names. In other words, it is not possible to describe a direct historical "filiation" between the systems of those two groups without understanding the nature of the substitutions and permutations produced.

 

Homogeneity and heterogeneity in the Western Desert

The definition of the "Western Desert cultural bloc", proposed by Berndt (1959), finds its origin in what Elkin’s (1938-40) typology is called the "Aluridja", and was subsequently reinforced, in spite of opposition from Tindale and Birdsell, by various researchers working among groups in this region. Indeed, the bloc includes more or less distinct dialectal groups, and not political or territorial units, characterised by their common cultural traits, in particular their sharing of elements of the Aluridja-type kinship system.

CLICK TO SEE MAP IN SEPARATE WINDOW

Map 1 : Approximate extension of the Western Desert cultural bloc and groups/tribes mentioned in this study (following Tindale, 1974 and the classification used by AIATSIS)

Key to Map 1

Despite this cultural and linguistic homogeneity, there are variations in some elements of Western Desert social organisation and kinship. Thus, though most groups possess sections, some use six terms (Ngaatjatjarra, Ngaanyatjarra) or eight subsections (Pintupi). There are also groups not using the section system at all (Pitjantjatjara, Yankunytjatjara). While none of the Western Desert groups name matri-moieties, some have named patri-moieties (e.g. Mardu and Pintupi). The main social category classification used by all Western Desert groups are generational moieties (cf. White, op.cit.). Marriage is between cross-cousins (and some persons of identical alternate generational moiety, but different generation), yet here again some variations exist. The Mardu allow marriage between some people related as "cross-cousin", the Ngaatjatjarra allow marriage only between cross-cousins who are genealogically and spatially distant (at least of the third degree). Another characteristic varying among the Western Desert groups is the nomenclature used for sections.

The method

In order to analyse the variations in the nomenclature and to determine possible routes of diffusion into the region and among Western Desert groups, the terminology of some 24 tribes or groups for which reference to section names is explicit has been taken into account.

Three of those groups or places (Ngaatjatjarra, Ngaanyatjarra and Ooldea) have been divided into two sub-groups each (Ngaatjatjarra 1, Ngaatjatjarra 2, Ooldea 1, etc.) because two nomenclatures meet there, producing systems with 6 section names (not, however, inevitably six section systems). Thus, the corpus consists of 27 units with four section names each.

The first step is to group together section names that may without hesitation be identified as deduced from each other, even identical, into categories of sections. Indeed, in view of the general questions formulated, it is not relevant here to treat, for example, the section name Karimarra differently from the section name Karimera, a variation for which the explanation is a linguistic task and not relevant in the study of the concept of a section. Although this grouping may be undertaken without much "scientific knowledge", because, as with the example of Karimarra and Karimera, these two terms clearly must have something in common, I intend to formulate the steps leading to the attempted elimination of mistakes.

This grouping is carried out in two steps. The first step is to compare joint appearances of section names among identical groups. If two section names are used by a single group, they are not "identical" and may therefore not be grouped together. In addition to this step, some general linguistic "rules" will be used that allow interpretation of the different transcriptions of the section names made by various authors. Once the categories are established, they will be subject to three degrees in complexity of analysis:

1. Findings of the first order or simple findings: The aim will be to draw in space the categories in which section names have been grouped. Simple findings on the absence or presence of a category in a specific geographical area will be presented and approximate routes of diffusion of these categories will be suggested.

2. Findings of the second order or semi-complex findings: In this second part, the combination of the categories will be examined, with the aim of studying the distribution of categories as pairs. This part will allow me to formulate a first hypothesis on mutual substitutions of categories.

3. Findings of the third order or complex findings: In this part, the internal structure of the systems will be taken into account, i.e. the structural position of categories in the systems used by the different groups.

Further, it is hoped that the analysis of some specific cases will allow me to clarify and illustrate some of the hypotheses formulated in the preceding parts and facilitate a global view of the diffusion and substitution of section names in the Western Desert.

 /*.................................

THE DEMONSTRATION HAS NOT BEEN INCLUDED, YET.

 .................................*/

The diffusion of section names: an overview

The elements outlined above are now sufficiently developed for me to summarize the essential points on a map. Two presentations have to be elaborated. The first one shows the diffusion of section names themselves, i.e. without substitution and independently of their combinations, excluding also linguistic variations, which have been eliminated while defining the categories of sections; this map shows the routes of diffusion without substitution (Map 8). The second map, complementary to the first one, presents the substitutions between categories of sections (Map 9).

Christensen, in an appendix to his doctoral thesis (1981: 368, annex 12), proposed rules of substitution of sections. However, some of the routes seem doubtful. Indeed, the author describes substitutions from group to group, or from place to place, without elaborating a general view of the whole of the Western Desert. Some routes described by the author are in correspondence with what has been shown here, while others are divergent. Christensen suggests that Karimarra becomes Tjarurru when one moves from the Mardu to the Ngaatjatjarra (whereas my hypothesis is that Tjarurru came from the south-west and has been diffused in tandem with Karimarra). Christensen suggests that Karimarra becomes Burgulu when one moves from the Ngaatjatjarra to the Pintupi (I propose that Burgulu came to the Pintupi from the Yulbaridya area to the north-west, and that it was not necessary to replace Karimarra; rather, the section Karimarra is identical by assimilation with Milangka among the Ngaatjatjarra, and it was therefore not possible that it moved on its own). Christensen suggests that Yiparrka becomes Panaka when one moves from the centre of the desert to the southern limit (I conclude that Yiparrka and Panaka have been diffused in this region together, even if, or perhaps because, they are identical by relation. All these questions and differences of opinion indicate that the problem of the diffusion of section names into and in the Western Desert remains open to further consideration. Results of course differ according to the method of analysis one adopts: linguistic or logic, locally or regionally focused or general view.

Map 8 : Routes of diffusion of sections without substitution into the Western Desert

Map 9 : Routes of diffusion of substituted and substituting sections into the Western Desert

 

Conclusion

The principal questions addressed in this paper have been directed to uncovering a logic of diffusion and substitution of names of section from the Pilbara-Kimberley area into the Western Desert. We have seen that the logic of combination of sections adapts itself well to situations in which different terminological systems meet, and that adjustments and transformations are in some cases necessary and inevitable if the structural integrity of these systems is to be maintained over time and space.

We have also seen that there is a certain logic in the rules of substitution, such that, for example, in a given region if two sections may substitute each other, they are characteristic of a relational identify, and, therefore do not have a relation of filiation or marriage. Yet we have also seen that the concept of a section, its underlying logic, may vary along routes of diffusion. Comparing two sections whose originally identical concept has been transformed following two routes becomes very difficult. Recall the example of Paljeri which, on the one hand, has a logic of diffusion Milangka and Tjarurru, but is replaced by Burgulu following another route. The logic is such that the two concepts of Paljeri which meet again are treated as different, as not having the same logic any more.

I now return briefly to von Brandenstein’s thesis, according to which the names of sections translate psychological states and inherited humours. If the meaning of the section names are effectively humours transmitted from generation to generation, when we see how actors juggle those names and transform and change their structural position, one must conclude that either the genetic laws on which Brandenstein makes reference are different from region to region, or, simply, and others have argued similarly, that the theory of Brandenstein has only a vague and local foundation.

This is not a reason for not being without suspicion or critique about the present study. If I think that the principles formulated in introduction have been followed and that they have proved themselves efficient, the same may not be said for the accuracy of the principles themselves. In other words, it is not certain and demonstrated that the diffusion of names of sections accompanying the diffusion of the system itself has, effectively, followed whatever logic. It is possible that the adoption of terms and their substitution by others followed, historically, another principle than the one worried about the internal and global coherence of the system and between the systems. Though, if one accepts that the terms travelled at the same time as, and in common with, the systems itself, this last hypothesis seems relatively improbable, because supposes that groups would have juggled with the terminology before adopting the system.

Another problem is related to the temporal disparity of the used sources. There are, between Bates and Christensen, some 50 years or more and it is possible (and probable) that systems changed and moved further on replacing other terms and structural positions, as families and groups migrated especielly since contact with settlers. The time-span in which the diffusion has taken place is another question mark. We know that the Ngaatjatjarra, for example, received the terms they use today for the section system in the thirties. It might be that those terms replaced other terms already present in this area. And we do not know when the diffusion of the terms studied into the Western Desert "started".

The section systems (or, better, the analysed terms) seems to have followed two major routes of diffusion into the Western Desert. Those two routes are from the Pilbara with the sections Karimarra, Panaka, Paljeri, Purungu, and, south of it, Milangka; and the second route, from the south-west into the desert, with a possible origin or at least the place of entry into the Western Desert of Tjarurru and Yiparrka. This region, as a possible origin of a section system, was also indicated my McConvell (1990). And in a later paper, McConvell (1996) writes that there might have been an original western section "system" in the south-west, moving northwards and then into the desert. This means that the Pilbara is not the place of origin of the section system, but only a place of transit, from where it moved further on. I have been trying to analyse routes of diffusion in the Western Desert and have, may be erroneously, taken as granted that at least some terms moved into the desert from the Pilbara area. They might have, in an earlier period, as McConvell suggests, first have moved north, before penetrating the desert.

The "north-western system" followed two concepts, one along the coast, the other one south-east to the Yulbaridya and to the Pintupi. The "south-western system" went straight into the desert in common with elements of the north-western system". It is, at this stage, not possible to determine whether the "south-western system" is effectively a "system" or whether it is only the region where some terms of the "north-western systems" were replaced with new regional names.

Before penetrating the Western Desert itself, the "north-western system" moved along the coast to a point in the south where it encountered new terms or another system. As if this had enforced its diffusional capacity, given it new strenght, probably because of migratory mouvements and cultural affinities, the section system moved into the desert and installed itself, creating assimilatory identities which seem to slow down the progression of diffusion and substitution of sections, and being put out of race by the sub-section system, as the Pintupi case illustrates.

References cited

 

BATES D., 1925, "Organisation sociale des Biroungoumat et des Djouamat (Australie Occidentale)", Revue d'Ethnographie et des Traditions populaires, 21, p. 27-48.

BATES D. (édité par I. White), 1985, The Native Tribes of Western Australia, Canberra: National Library of Australia.

BERNDT R.M., 1959, "The concept of ‘The Tribe’ in the Western Desert of Australia", Oceania, 30(2), p. 81-107.

BERNDT R.M. & BERNDT C.H., 1942-45, "A preliminary report of field work in the Ooldea region, western South Australia", Oceania, 11, 12, 13, et 15.

BERNDT R.M. & BERNDT C.H., 1992, The world of the first Australians. Aboriginal Traditional Life: Past and Present, London: Angus & Robertson, [1964].

BIRDSELL J.B., 1976, "Realities and transformations: the tribes of the Western Desert of Australia", in N. Peterson (ed.), Tribes and Boundaries in Australia, Canberra: AIAS, p. 95-120.

BRANDENSTEIN C.G. von, 1970, "The Meaning of Section and Section Names", Oceania, 41(1), p. 39-49.

BRANDENSTEIN C.G. von, 1972, "The Phoenix Totemism", Anthropos, 67, p. 586-594.

BRANDENSTEIN C.G. von, 1982, Names and Substance of the Australian Subsection System, Chicago and London: The University of Chicago Press.

CHRISTENSEN W., 1981, The Wangkayi Way: Tradition and Change in a Reserve Setting, Perth: Unpublished PhD Thesis, University of Western Australia.

DIXON R.M.W., 1980, The languages of Australia, Cambridge: Cambridge University Press.

DOUGLAS W.H., 1977, An introduction to the Western Desert Language, Sydney: University of Sydney, Oceania Linguistic Monographs, 4, [1964].

DOUGLAS W.H., 1980, "The Desert Experience : Language", in R.M. Berndt & C.H. Berndt (eds), Aborigines of the West : Their Past and Their Present, Nedlands (WA): University of Western Australia Press, p. 108-118, [1979].

ELKIN A.P., 1931, "The social organisation of South Australian tribes", Oceania, 2(1), p. 44-73.

ELKIN A.P., 1933, "Totemism in north-western Australia (The Kimberley Division)", Oceania, 3(3), p. 256-296 et 3(4), p. 435-481.

ELKIN A.P., 1934, "Cult-totemism and mythology in Northern South Australia", Oceania, 5(2), p. 171-92.

ELKIN A.P., 1938-40, Kinship in South Australia, Oceania, 8(4);9(1);10(2);10(3);10(4), p. 419-452; 41-78; 198-234; 295-349; 369-89.

ELKIN A.P., 1967, Les Aborigènes d'Australie, Traduction de 1967 de la 2eme edition. 1ere edition en 1938, Paris: Gallimard, [1954].

FRY H.K., 1934, "Kinship in Western Central Australia", Oceania, 4(4), p. 472-478.

GLASS A., 1988, Ngaanyatjarra Word List, Alice Springs: Ngaanyatjarra Bible Project, Summer Institute of Linguistics.

GODDARD C., 1985, A Grammar of Yankunytjatjara, Alice Springs: IAD.

GRAAF de M., 1995, Do six section systems exist in Aboriginal Australia?, Darwin: Unpublished manuscript (ms).

HAMMEL E.A., 1960, "Some Models for the Analysis of Marriage-Section Systems", Oceania, 31(1), p. 14-30.

KUPKA K., 1975, "Les systèmes des sous-sections matrimoniales dans la famille aborigène d’Australie", Journal de la Société des Océanistes, 49, p. 435-466.

McCONNEL U., 1930, "The Wik-Munkan Tribe of Cape York Peninsula", Oceania, 1(2), p. 181-205.

McCONVELL P., 1985a, "The Origin of Subsections in Northern Australia", Oceania, 56(1), p. 1-33.

McCONVELL P., 1985b, "Time perspective in Aboriginal Australian culture: two approaches to the origin of subsections", Aboriginal History, 9 (1), p. 53-80.

McCONVELL P., 1990, "The linguistic prehistory of Australia : Opportunities for Dialogue with Archaeology", Australian Archaeology, 31, p. 3-27.

McCONVELL P., 1996, "Backtracking to Babel: the chronology of Pama-Nyungan expansion in Australia", Archaeology in Oceania, 31(3), p. 125-44.

McKELSON K., 1980, "Nadya Nadya Country", in Berndt R.M. & Berndt C.H. (eds), Aborigenes of the West, Their Past and Their Present, Perth: The University of Western Australia Press, p. 214-223, [1979].

MEGGITT M.J., 1986, Desert People: A study of the Warlpiri of central Australia, Sydney: Angus and Robertson, [1962].

MYERS F., 1986, Pintupi country, Pintupi self. Sentiment, Place and Politics among Western Desert Aborigenes, Washington & Londres, Canberra: Smithsonian Institution Press, AIAS.

O’GRADY G.N. & MOONEY K., 1973, "Nyangumarda Kinship Terminology", Anthropological Linguistics, 15(1), p. 1-23.

RADCLIFFE-BROWN A.R., 1913, "Three tribes of Western Australia", Journal of the Royal Anthropological Institute, 42, p. 143-194.

RADCLIFFE-BROWN A.R., 1931, "The social organisation of Australian tribes", Oceania, 1(4), p. 426-56.

RUSSEL B., 1997, An inquiry into meaning and truth, London, New York: Routledge, [1950].

SCHEFFLER H.W., 1978, Australian Kin Classification, Cambridge: Cambridge University Press, Cambridge Studies in Social Anthropology.

SPENCER B. & GILLEN F.J., 1927, The Arunta, London: Macmillan.

TINDALE N.B., 1974, Aboriginal Tribes of Australia, Berkeley: University of California Press.

TINDALE N.B., 1988, The Pitjandjara, in M.G. Bicchieri (ed.), Hunters and Gatherers today. A socioeconomic Study of eleven such cultures in the twentieth Century, New York, London: Waveland Press, p. 217-268, [1972].

TONKINSON R., 1974, The Jigalong Mob : Aboriginal Victors of the Desert Crusade, Meulo Park, California: Cummings Publishing Company.

TONKINSON R., 1991, The Mardu Aborigines. Living the Dream in Australia’s Desert,New York: Holt, Rinehart & Winston, Case Studies in cultural Anthropology, [1978].

TURNER D.H., 1976, "Levels of organisation and communication in Aboriginal Australia", in N. Peterson (ed.), Tribes and Boundaries in Australia, Canberra: AIAS, p. 180-191.

TURNER D.H., 1980, Australian Aboriginal Social Organisation, Canberra, Atlantic Highlands: AIAS, Humanities Press.

WHITE I.M., 1981, "Generation Moieties in Australia : Structural, Social and Ritual Implications", Oceania, 52(1), p. 6-27.

WURM S.A., 1988, "Australian Aborigines : Languages", in The Australian Encyclopaedia, vol. 1, 5ème édition, Terrey Hills NSW: Australian Geographic Society, p. 252-255.

YALLOP C., 1969, "The Alyawara and Their Territory", Oceania, 39(3), p. 187-97.