Printer Friendly

Indigenous modes of representing social relationships: a short critique of the 'genealogical concept'.

Abstract: The first part of the article discusses briefly the notions of genealogy and kinship within the Euro-American epistemological context and advocates the necessity for a sharp distinction between these two domains. While being a useful tool, especially in comparative approaches, the collection of genealogies is nevertheless the enactment of the genealogical concept, which in turn is a particular mode for legitimising status associated to a culturally specific iconography.

The second part of the article portrays ethnographic material illustrating, as an alternative to the genealogical concept, Indigenous modes for representing relationships between people in the Western Desert, and that do simultaneously include affiliations to and structuration of space. These Indigenous modes and iconography for representing 'genealogies' reflect a cultural schema that can be summarised as an unalienable link (or identity) between people and locales on the one hand, and between relationships and routes/tracks on the other. A genealogy in the understanding of a Ngaatjatjarra-speaking person of the Western Desert is a representation of both social and spatial affiliation and structure.


Genealogies and what will be called here the 'genealogical concept' formed one of the foundations of kinship studies and have long served as the implicit proof of well-conducted fieldwork. Following their partial rejection as ethnocentric tools from the 1970s onwards, they have, however, lately become again an unavoidable iconographic part of native title evidence and have therefore survived and maintained some scientific or jural validity, albeit in contrasting historic and epistemological contexts. This persistence of the genealogical concept in anthropology and related fields inevitably leads to an interrogation of its possibly opportunistic nature or uses, reflecting its adaptive capacity as well as the persistent cultural theme it conveys. In addition to apparently being a culturally (Euro-American) defined mode of representing 'natural evidences' of procreation, may we presume that it is also the 'natural' (i.e. intrinsic) expression of social relationships?

In this article, I reiterate the reasons why the 'genealogical concept' is not always an accurate model for depicting emic representations of social relationships and subsequent determinations of affiliations to land. In some cases, it is a framework imposed for the purpose of providing evidence relating to questions of Indigenous interests. I begin by briefly recapitulating some of the major criticisms directed towards the use, and more particularly the historical background, of genealogies in anthropology. Within this discussion, I endeavour to explain why such concepts as 'kinship system' or 'kin category' should be sharply distinguished from 'genealogies'.

In the second part of the article, I present and discuss some examples of Indigenous modes of depicting people and land, which I recorded among Ngaatjatjarra-speaking people in the eastern part of the Western Desert. These examples illustrate an Indigenous iconography for representing social relationships that accords with what anthropologists would call a genealogy or a pedigree. However, the meanings expressed reflect a particular cultural schema that is distinct from that conveyed by the genealogical concept. In this schema, social relationships, rather than being firmly based on genealogical associations, also include representations and elements of space-structuration.

The genealogical concept

The pedigree or genealogy, wrote Barnes (1967:101), 'is a statement of the way in which individuals are, or assert that they are, connected with one another through marriage and common parentage'. Later in the same paper, Barnes distinguishes such individuals and genealogical statements according to a very specific understanding of 'objectivity' or 'genealogical truth' that is associated with seemingly non-alterable biological conditions and consequences embedded in an absolute and universal definition of the basis of kinship: the recognition of procreation and filiation. Barnes refines his definition by marking with different names--'pedigree' and 'genealogy'--a distinction and widespread opposition that had already been underlined by Rivers (1910). 'Genealogy', Barnes (1967:102-3) writes, 'conforms to the logic of the system', resulting from a 'statement made by an ethnographer as part of his field record or of his analysis', while 'pedigree' is the emic statement, 'made orally, diagrammatically, or in writing by an actor or informant': a pedigree is a 'contemporary statement, making assertions about connexions between people'.

These citations eminently reflect one of the major criticisms of kinship studies and anthropology in general that have been advanced by strong relativists and adepts of interpretative anthropology: what is the basis for a judgement in which the etic or anthropological point of view takes precedence over the emic or cultural expression of relatedness? Moreover, Barnes asserts that genealogy conforms to the logic of the system, thereby conflating two epistemological categories: 'genealogies' are a tool for recording one aspect of social history, while 'the system' is largely a theoretical conclusion expressing structural particularities that may or may not be reflected in practice. Barnes does acknowledge the importance of culturally defined ways of engendering genealogies, or what he calls pedigrees. However, he also makes clear that 'the cultural milieu of the actor marks the method of construction of the pedigree, whereas the demands of science determine how the genealogy is recorded' (Barnes 1967:103, emphasis added). In so doing, the rather complex underlying criterion for distinguishing between genealogy and pedigree is brought down to a plainly categorical, and in some respects binary, opposition between science and non-science, depicted as products differentiated through their architects: the ethnographer and the informant.

One might wonder what kind of science genealogies are a part of, since the underlying semantic field is intimately related to historically well-established Western ideologies. (1) Indeed, their metaphorical layout in the form of a tree and their organising principle--descent, or in fact the recognition of ascent through direct lines, excluding persons more than including them--are found as far back as the Bible, where Christ's ancestry is traced back to Abraham and Adam. That anthropology (and other sciences such as medicine) have inverted the tree so that its roots are upside down bears little significance in our context. My aim is not a deconstruction of the genealogical concept itself. What concerns me is that, in contexts in which the Indigenous interest is the prime drive--that is, in what Barnes would probably call a non-scientific context--relying on 'classic genealogies' alone, to the neglect of Indigenous iconography and modes of representing social relationships, may not accurately serve the purpose.

Let me illustrate some of the claims I advanced above. Etymological dictionaries explain the origin of 'genealogy' as the combination of two Greek words--genea, 'race' or 'family', and logia, '-logy', 'discourse/science'--which undoubtedly points towards the ideological question underlying its use: what is the logic, the science that lies behind the family or the race? While the notion appears to have emerged in its present form only in the fourteenth century, the idea of genealogy, which I call the genealogical concept, seems to have operated in the European context since ancient times, being explicitly and eagerly used by the Greeks and Romans to trace their ancestry back to the gods. Genealogies later flourished in the Middle Ages, when they were applied, if not adjusted, to ascertain status and determine transmission of ownership in the feudal context. With the appearance of 'modern genealogy', developing in the seventeenth and eighteenth centuries and then legitimised by Mendel's laws on heredity in 1865 and Wilkins' discovery of DNA in 1952, (2) the genealogical concept gathered new momentum. Here, again, its primary non-scientific use was to demonstrate that one belonged to the nobility. The genealogical concept finds its origins in a specific iconographic mode of representing an extract of social or personal history used as a sociopolitical tool to claim and legitimise a status.

In recent years, the search for one's origins has become a widespread and popular activity, and is no longer limited to privileged strata of Western society. Indeed, an Internet search using the word 'genealogy' with the Google search engine returns more than 10 million web pages; and the number of books and software packages available for the collection and management of one's personal family history is growing impressively. (3) The disappearance of nobility from the political scene, and the at least theoretical possibility of engaging in social mobility, are certainly linked to, if not the reasons for, the popularisation of tracing genealogies. Additionally, the increasing use in the public discourse of such notions as identity and cultural roots has without doubt contributed to such movements.

The origins and historical uses of the genealogical concept depicted above reflect the commonly accepted and applied definitions of the notion. A genealogy is (1) both the register or diagram showing a line of ancestors and the study of this register; (2) an ancestral line itself, that is, the origin or history of something; and (3) a distinguished pure-bred ancestry.

Whether it is in the context of Greek or Roman society, the feudal Middle Ages or even modern genealogy, we find, in all cases, a meaning and usage that, embedded in rather pragmatic aspirations, refer to the notion of 'origins' and therefore to a historical depth justifying a non-negotiable status. It refers to the transmission of wealth, identifying substances or essences by way of a linear principle, in which intermediary steps or lateral branches are only mentioned if and when they satisfy the purpose of demonstrating the inalienable consubstantiality with divinities or, more generally speaking, status-legitimating origins. More particularly, within the Euro-American ideology, genealogies are considered to be scientifically justified representations of the transmission of 'blood', one of the metaphors for identity.

The cultural and historical background of genealogies, as outlined above, is reflected in the imagery used for portraying them. The analogy between the etymological and historical meanings and uses of the genealogical concept, on the one hand, and the origin of its iconographic visualisation in folklore, science and anthropology, on the other, are evident, as Bouquet (1996, 1997, 2000) has made clear. Indeed, genealogies are informed by the tree metaphor, with its roots grounded in earth and its emphasis on the central trunk, a metaphor in which 'earth becomes blood, which becomes sap, which is turned into seed' (Bouquet 1996:52). 'My conclusion', writes Bouquet (1997:376), 'was that we are obliged to consider the genealogical method and its diagram as ethnographic data and not just as analytical apparatus', making the layout of genealogical trees a subject of study itself. Using examples to trace the origins of the genealogical iconography employed by anthropologists back to the biblical pedigree and phylogenetic and philological trees, Bouquet (1996:60-2) shows how:
 the connecting lines of the genealogical diagram
 continue to evoke 'the social recognition of biological
 ties', and hence the kind of European biological notions
 of substance embodied in the examples discussed ...
 the fundamental vision of kinship remains arboreal.

However, that kinship and genealogy are not identical concepts, and not even concepts based on identical principles, emerges clearly in Parkin's (1997) answer to Bouquet's work. He qualifies as misleading Bouquet's (1996:43) claim that for 'anthropologists, the genealogical diagram is a piece of graphic shorthand whose theoretical status is rarely considered', and suggests that there have been 'repeated challenges to the use of both genealogical diagrams and genealogical modes of thought ... in one domain in particular, namely kinship terminology' (Parkin 1997:374).

It is, indeed, a confusing practice to illustrate kinship terminologies and kin categories using the same iconography and layout used to represent the 'biological' or social history of succeeding generations of people. In recent times, therefore, the tabular method (4) has in many cases replaced the tree metaphor in the domain of kinship terminology. What Bouquet seems to misunderstand is that kinship as an anthropological concept reflects emic social categories (and rules), while genealogies reproduce an extract of social history in ways and forms that are--and one must concede her this point--largely reflected in the Euro-American iconography and ideology. Kinship, of which terminology is one constituting element, is a conflation of a theoretical genealogy in accordance with Indigenous linguistic and behavioural patterns, and bears only an indirect relationship to the actual genealogical grid of a social group. In this sense, the Indigenous modes of representing social relationships illustrated below are to be interpreted as expressions in the domain of kinship, rather than as an enactment of the genealogical concept.

Discussions about the confusion of genealogy and kinship are not recent. The debates regarding the efficacy and accuracy of the componential analysis of terminologies in reproducing a genealogical grid are one example. (5) Others are provided by alliance theorists, especially Levi-Strauss (1967), who, as White (1981:21) explained, was not able (or not interested enough) to distinguish, on the one hand, between the linguistic and classificatory elements of a terminology, including its resulting structural prescriptions, and, on the other, the actual ethnography of genealogical ties:
 Again, as in much of Levi-Strauss's writing
 about Australia, we find him unable to comprehend
 that compliance with a prescribed marriage rule seldom
 entails marriage with a close relative. In nearly all
 Australian societies the rule is for a man
 to marry a certain type of cross-cousin, but not the
 actual one.

Both White's critique and Levi-Strauss' arguments directly derive from a conflation of the genealogical grid with categories of kin. While Levi-Strauss seems to think, wrongly, that cross-cousin marriage prescription has a direct impact on a genealogy, White assumes that categories, as defined by Levi-Strauss, are a representation or summary of the actual genealogical grid. The conceptual discrepancy between White's approach and Levi-Strauss' claims is best viewed in the latter's dismisive answer to Hiatt's chapter (1968) in Man the Hunter, where Levi-Strauss explicitly writes that he analyses how 'tribes have explained their system to men such as Spencer and Gillen, Strehlow' (emphasis added), and that he is in fact 'concerned with a different problem: to ascertain what was the meaning of rules, whether they are applied or not' (Levi-Strauss 1968:210-11). Not only does he dismiss practice, but he is also solely interested in Indigenous discourse and abstractions. A genealogical grid, however, is the consequence of practice and not discourse.

Similar comments about the confusion of kinship and genealogy could be made for Sackett's (1976) critiques of Levi-Strauss, when the former argues that direct exchange and reciprocal marriage are not practised, and, in fact, not even possible or desirable in the Australian Western Desert. It is not clear from Levi-Strauss' writing if he decided in the end that the Western Desert or Aluridja kinship system was of the direct or indirect exchange type. Whatever the answer, the divergence between Sackett's and Levi-Strauss' claims remains an epistemological one. Sackett addresses ethnographic (genealogical) data, while Levi-Strauss is arguing on the basis of abstract terminological and discursive categories.

Yet another example of the problems involved in blending genealogies and kinship systems, or notions and concepts defined for kinship systems, is the resurrection of the expression 'cognatic descent groups' in Australia (for example, Finlayson et al 1999:4; Sutton 1998:65-7). An enduring cognatic descent group based solely on genealogical principles is a practical impossibility, unless continuous endogamous marriage within the group is practised, or there are social mechanisms in place that prevent the proliferation of affiliations to all descent categories. Sutton (1998) was an important figure in the 'reintroduction' of the notion of 'cognatic descent groups' in Australia.

Interestingly, the way Sutton describes these groups conforms, rather, in my opinion, to a principle of ambilineal affiliation. (6) For example, that the 'increased freedom of choice over which of the descent groups into which they were born will become their primary group(s) of identification', or that 'individuals belong to several cognatic stocks by reason of their ancestry but only one or a few normally provide their main families or orientation' (Sutton 1998:64, emphasis added), are classic features of ambilineal systems (for example, Coult 1964; Fortes 1959; Leach 1962; also Fox's [1996:85] use of the notion of 'ambilocal', which he describes as a 'lack of firm rule'). Bernd (1966), another example, showed for the Northern Gilbert Islands that, in an ambilineal system, one may simultaneously be a member of several ramages, but that the most important rights and obligations depended on whose house sites one could settle in, a localised core therewith controlling the estate. Ramage is the notion that was (is) used to describe ambilineal groups in Polynesia (Firth 1957; Murdock 1960). (7)

These so-called 'cognatic descent groups' are sets of people for which the criteria of relatedness are not solely based on biological ties (filiation), but are also derived from common history, co-residence, economic cooperation, and so on. Only in a subsidiary manner do they have a direct connection with the 'genealogical concept'. Cognatic descent groups may be a set of persons for whom the 'genealogical concept' is the basis for decision making in questions of recruitment and affiliation, without constituting an actual corporate group. Here, the 'genealogical concept' may provide the capability of excluding others in a context in which ethnic distinctions have become primordial, although the cognatic recognition of affiliation does not seem to be a recent feature of Aboriginal culture. (8) It may well be that the definition of these groups in terms of genealogical connections is a fallacy, and that the description of them as 'cognatic descent groups' may be the uneasy application to the domain of genealogy and social practice of concepts defined in the realm of kinship systems.

The following example may further underline the necessity of differentiating practice and discourse relevant to the definition of kinship systems and actual genealogical patterns. In the Western Desert, among Ngaatjatjarra-speaking persons, prolonged co-residence is a creator of what anthropologists--not biologists or socio-biologists--would call consanguinity. Indeed, 'consanguinity', as Heritier (1981:13) and Malinowski (1963:182) explained, is a question of choice and social recognition, and cannot simply be defined in terms of blood relationships. The 'consanguinisation' of co-residents is demonstrated through terminological usages (the use of the term 'brother' or 'sister' for a classificatory cross-cousin) and marriage patterns (the prohibition against marrying such a person). Hence, prolonged co-residence will have immediate impacts on the genealogical grid, in the form of absence of 'biological' ties where they could have been expected, for example, or in the conflation of co-residents by future generations into a set of ancestral siblings. While these people and their descendants remain in distinct categories (such as cross- and parallel-cousins) from a genealogical point of view, they may also be brought closer, involving 'unexpected' behaviour and terminology, from an interrelational point of view (Dousset 2002).

Genealogies must be distinguished not only from categories and terminologies but also from the notion of 'kinship system' as a whole. Unless we accept Barnes' (1967:102-3) claim, quoted above, that genealogies conform to the logic of the system--in which case they already are an interpreted summary of the grid--genealogies are not, and cannot be, directly conflated with the kinship system. The sequence of births, the exhaustive enumeration of ascendants and descendants, and the narratives about marriages, separations, divorces, and so on have little direct connection with the systemic aspects of kinship. A kinship system and its definition may be discussed and modified by anthropologists (whether Indigenous or not) as a result of differential interpretative processes. A genealogy, on the other hand, may be modified only by Indigenous peoples themselves: not solely in the process of reinterpreting or readjusting principles, but mainly as the result of the exertion of memory.

In this first part of the article, I have briefly discussed the notions of genealogy and kinship within the Euro-American epistemological context, and have advocated the necessity for a sharp distinction between these two domains. I have also demonstrated that, while being a useful tool, especially in comparative approaches, the collection of genealogies is, nevertheless, the enactment of the genealogical concept, which in turn is a particular mode for legitimising status associated with a culturally specific iconography.

In the remaining part, I present ethnographic material without analyses other than those directly relating to the content of the iconography itself. As the reader will see, the Indigenous drawings reflect a conception that, in many ways, is an addition, if not a contradiction, to the 'classic' anthropological mode for representing social relationships through the genealogical tree, and therefore questions the supposed supremacy of the genealogical concept in questions relating to the representation of socio-biological relationships.

Indigenous modes of representing social relationships

While I was working on kinship and collecting genealogies in a Ngaatjatjarra community in the Australian Western Desert, people systematically explained matters to me through drawings in the sand. However, my insistence on and wish for clarifications made the whole procedure rather complex. Previous drawings that would explain the points I wanted clarified had been erased on the sand to give way to other explanations. It became advantageous, therefore, to grab my notebook and draw the relationships next to the genealogies I was writing down. With time, I became aware of the richness and complexity of what appeared at first sight to be simple figures and sketches. They seemed to reveal more about the kinship system, and more specifically its emic cognition, than my neat triangles and circles with interconnecting lines could ever do.

At the stage of analysis I decided to contextualise these sketches within a simple and elementary question: what are the representations reflecting cognitive schemas at work in the conception of kinship? Within this larger question, three subquestions constituted the analytical guidelines. The first question interrogated the cultural basis for the construction of what could be called 'the minimal social unit', that is, something that constitutes as a whole more than the sum of social interactions between two individuals. The second question related to the emic conception of ascent and asked how older or deceased persons and generations are incorporated into the construction of social distinctions. The third question concerned marriage and how it reflected the establishment and conception of networks of solidarity and dependency: how does the representation of alliance mirror embeddedness? The iconography reproduced below has been chosen to illustrate the Indigenous answers to each of these three questions.

The minimal social unit

The illustration in Figure 1 was drawn by a woman representing herself as the lonely vertical line in the centre (1). Above her, she draws four lines (2). These are her actual children. They are underlined as they were all born in the same community: (9) they stand on the same ground, affiliate to the same community. Below her are three children she either fostered for long periods or adopted (3). They are clearly distinguished from her actual children, but the line below them is the continuation of the line below her actual children. They are recognised as having identical ties to the community or locale to which her own children affiliate, for adoption is considered to create similar ties to birth ties. Indeed, a person's prohibited spouses, because of geographical or genealogical closeness, are also prohibited for that person's adoptive siblings, and vice versa.


Interesting in relation to my question is the word nyuyurlpa, used to name an adoptive mother or adoption in general. The semantics of this word contribute towards considering the minimal social unit as being distinct from what is usually considered the nuclear family. Indeed, nyuyurlpa is polysemic and besides meaning the 'adoptive mother' also means 'the one who adds wood to the fire', 'the one who entertains the fire or hearth', adding to a unit based on social ties, formulated around procreation and adoption, the character of locality and shared activity:
Nyuyurlpa kutjupankatja
The one who entertains another one-from
the fire

mantjinu kanyinu purlkanu.
took kept [and] made
 big [grew].

The circle (in Figure 1) that circumscribes these eight people, among whom, one may have noted, the father is absent, with the mother or hearth in the centre, testifies to this unity. This circle was explained as representing ngurra. This word, as is well known (for example, Hamilton 1979:15), is polysemic and, in fact, metonymical as well. It can be used in relation to the camp or house, the local community, the sites to which one affiliates, the area that contains these sites, the entire dialectal territory or area, and, in some cases, the entire known, or heard of, world. It is also used in relation to birth and conception sites, and to territorial affiliation in general. To be fully understood, the metonymical character of the notion must be linked to its social counterpart, walytja (Figure 2).


The woman's drawing in Figure 1 illustrates both the most elementary level at which ngurra has a semantic value, and what could be called the minimal social unit, the smallest quantum at which ngurra and walytja have social significance. It is based on the metaphor of mother-hearth, and is reflected by, or obtains meaning in, the lowest or most interior level of space structuration.

Among Ngaatjatjarra-speaking persons, the understanding of genealogy as a concept is inseparable from that of spatial structure and does not systematically exclude people who would have been omitted when applying the genealogical concept. Both the iconography used in the woman's drawing, and the parallelism between ngurra and walytja, illustrate these principles.

Ascent and social distinctions

The next example (Figure 3) was chosen to illustrate the answer to the question regarding the conception of ascent as a mechanism for establishing social distinctions among contemporary people.


Every vertical line is a person having a strong link to an important area southwest of Kintore, in the Gibson Desert. The individuals are organised in what was described as three sets of siblings living in three distinct communities. From a purely genealogical point of view, however, they are not actual siblings, as some have only their mother in common, some only their father, and some are offspring of same-sex siblings. The three sets are themselves considered actual siblings. Such wide sets of 'social' siblings are sometimes attributed a subsection so that they are qualified not only as being all Tjupurula, for example, but also the Tjupurula from such-and-such an area. The subsection name has, in these contexts, a rather specific meaning. Tjupurula, of course, is one category of the social organisation in the area and there are probably hundreds of Tjupurula people across Central Australia. With reference to the persons in this drawing and the site to which they affiliate or have been affiliated, Tjupurula becomes a common denominator reflecting closeness or identity through linkage to the same area: they are siblings with respect to their territorial affiliation and their classificatory position. Actual and precise genealogical position matters little in this context. (10)

Within this specific Tjupurula set, people are distinguished with respect to their knowledge of songs relevant to the area, and with respect to their seniority, especially among those who are still living. In the set on the left, such distinctions are correlated to the size of the lines. The person on the far left is the most 'knowledgeable', the person on the right the least, for this area. The drawer, a man, is the third line starting from the left.

The global sibling set has been distinguished with respect to its members' residence, creating the three subsets supposedly living in three distinct communities. However, the represented persons were all born in the bush and walked later, mostly in the 1940s, to stations or to missions. The double horizontal lines in the central and right sets indicate that these people have, in fact, all passed away. Interestingly, some died even before they arrived at the communities for which they are described as residents, and for some even before these communities were actually established. A legitimate question is, therefore, to ask how these people have come to be distinguished according to a place of residence, when most of them have never actually lived there.

The answer is that the people in these sets have not been distinguished with respect to their own residence, but with reference to the residence of their descendants and, therefore, this illustrates one mode of employing ascent to establish contemporary social distinctions within and between actual residential communities.

The drawing exemplifies a conflation of actual genealogical positions into general categories, horizontally speaking, and also conflates generations socially by referring to former generations through current people's residence and affiliation. The dead are differentiated with regard to the living's residential community.

In this sense, social distinctions and identities are reflected and represented through the landscape and actual people's practice within this space. They are not conceived upon actual genealogical relationships in the strict sense but, rather, reflect social distinctions established through (1) global kin category identity, (2) affiliation to sites and areas, and (3) residence of cognatic descendants.

Alliance and embeddedness in social networks

The last example (Figure 4) refers to my third question: how does marriage reflect networks and how does the representation of alliance mirror embeddedness?


Before discussing the drawing, it is necessary to outline the principles operating in relation to marriage among Ngaatjatjarra-speaking people. Marriage takes place between classificatory cross-cousins or cross-relatives two generations removed. Spouses should also be spatially distant; that is, ideally, they should not have identical birthplaces, not have a conception totem (tjuma) linked to the same site or area, not express land affiliation to the same area, and preferably live in distinct communities before marriage. Still, today, the large majority of marriages follow these principles.

A man has at least two promised wives though initiation, one from the initiator, who is a person from the MB (kamuru) category, and henceforth becomes a waputju (WF, initiator). The latter will promise his daughter or close 'daughter' to the young man, who will refer to this woman as his pikarta ('the one obtained through pain'). At the same time, he will also be promised a second wife from a kurntili, that is, a person of the FZ category. This FZ then becomes his yumari (WM). In the case of the wife promised through initiation, however, the WM is called pampurlpa. This promised wife is called pampurlpa-ku yurntalpa (the daughter of pampurlpa) and reflects a mother-in-law bestowal, as the daughter is usually unborn when promised.

It is not compulsory for a man to actually marry his promised wives. He may be able to marry persons promised to him in a non-ceremonial context, which is called karlkurnu (i.e. making presents to the parents-in-law), or through elopement, warngirnu. In many ways, it is beneficial for both the man and his relatives if he does not marry his ceremonially promised wives. Indeed, bestowal through initiation and marriage both create affinal links that involve similar obligations and rights and, therefore, duplicate the number of persons and families with whom one stands in emotionally close and, economically speaking, reciprocal relationships (Dousset 1999).

The drawer of the diagram in Figure 4 is an older man who represents himself as a dot at the bottom. The middle dot in the three above him is his initiator. The dot on the far left is the initiator's daughter, whom the drawer actually married in this case. This woman is his kurri (wife) and pikarta ('the one obtained through pain'). On the right hand stands the drawer's pampurlpa. That is, she is a wife's mother who promised him a daughter during initiation. He did not actually marry that daughter, nor is pampurlpa the wife of waputju (WF), although they obviously stand in a spouse category to each other. The lines between the dots symbolise the links between these people, their mutual relationships.

What is particularly interesting is that the dots and lines also represent sites and tracks in space. This 'genealogical' drawing, which illustrates social relationships, is also a spatial map. Ego is plotted at his birthplace (ngurrara), his wife is located at her conception site (tjuma ngurra), and the father-in-law and the mother-in-law in their respective 'countries' (ngurra; no further details recorded for the origin or reason of affiliation). The lines between the sites/people are the representation of three relationships simultaneously: (1) the inalienable links established between the people though betrothal, initiation and marriage; (2) the inalienable rights for mutual access and obligations of responsibility established between the actors relative to these sites; and (3) the tracks that traditionally would have been travelled to move between these sites.

With respect to the question asked on the relation between alliance and embeddedness, my conclusion is rather simple and certainly not revelatory: the social network created through alliance, initiation included, is understood or conceptualised, at least in the Western Desert, as a network in space.


The selective construction and interpretation of a genealogical grid may lead the researcher to define a kinship system, or to apply a predefined system to a group or even a culture. However, a genealogy in itself is not a kinship system. The most immediate and straightforward confirmation for this rests in the history of the anthropological discipline itself, where theorists, although sometimes building their analysis on identical ethnographic material, may disagree on the definition and typology of systems.

An 'objective' or 'biological' genealogy, including all relationships and individuals without omission and selection, is a practical impossibility. Therefore, each genealogy constructed is the product of three interpretative processes: first, the consequence of the Indigenous capacity, will and strategy in the exertion of genealogical memory; second, the inherent Western background of the genealogical concept and method, that is, the legitimisation of the present condition through selective demonstration of ties to the 'origins'; and, third, the difficulty in apprehending genealogies without an abstraction in which general genealogical patterns are interpreted as automatically engendering formal descriptions of systems. While these three processes are, in my opinion, productive for anthropology as a discipline, as they allow for the discussion and refinement of tools and concepts as well as the formulation of new working hypotheses, they only take half of the available data into account, missing out the possible existence of Indigenous modes of representing social relationships, be they iconographic or discursive.

I have tried to show that, as an alternative to the genealogical concept, there are Indigenous modes of representing relationships between people, at least in the Western Desert, and that, more importantly, they simultaneously include affiliations to, and structuration of, space. The conclusions that can be drawn from the illustrations reproduced here are twofold. Indigenous modes and iconography for representing 'genealogies' reflect a cultural schema, just as the anthropological ways of drawing genealogies reflect the schema of the 'genealogical concept', where direct biological ties to origins legitimate contemporary claims and status. This Indigenous schema can be summarised as an unalienable link (or identity) between people and locales on the one hand, and relationships and routes/tracks on the other. Agenealogy in the Ngaatjatjarra understanding is a representation of both social and spatial affiliation and structure. It is a synchronic network approach with nodes and links that conform to the classificatory--but not automatically to the genealogical--principles inherent in a kinship system and its cognition.

Far from insinuating that the existence of such Indigenous iconographies must be pan-Australian or even universal, or that the 'classic' anthropological genealogies have little to tell about social structure and practice, I am nevertheless suggesting that investigating such phenomena is a worthwhile lead to follow up and, at least in my opinion, a more convincing argument in demonstrating what, in the last instance, is implicitly depicted as cultural authenticity within the current Australian legal and political framework.


My profound gratitude goes to Lizzie Markilyi Ellis, a Ngaatjatjarra-speaking person, without whose collaboration and work the collection of genealogies would not have been possible. My gratitude also goes to Peter Sutton, Patrick McConvell and the anonymous reviewers who commented on an earlier version of this article, all of whom expressed strong, albeit not always unanimous, criticisms of it. I believe I have addressed most of them, but remain responsible for any shortcomings and misinterpretations. A shorter version of this paper was presented at the Workshop on Genealogies, AIATSIS, Canberra, 5-6 October 2002.


(1.) Of course, the construction of genealogies is not limited to the Western world but is found in many societies, such as among the Arabs or among the so-called high cultures of Central and South America. However, extrapolating it to a so-called scientific context is probably a process rather specific to Judeo-Christian culture.

(2.) Many facts concerning the history of the notion have their source in Durye (1988 and 1999).

(3.) One must note, however, that there are currently no appropriate electronic tools available for accurately processing genealogies in the field of applied anthropology.

(4.) See Ives (1998) for good examples.

(5.) See Hirschfeld's (1986) discussion of this problem and the comments following his article by Scheffler, Schneider and others. In another article, Parkin (1996) explicitly discusses the necessary distinction between category and genealogy.

(6.) I do not think Sutton's (1998:31-2) critique of Layton's description of the Pitjantjatjara system as being ambilineal is convincing. Layton (1983:24; see also Peterson 1977 for an interesting comparison between the Pitjantjatjara and the !Kung) describes how 'people speak of choosing' in relation to the land-owning group, from which he correctly deduces an ambilineal mode of affiliation. Northwest of the Pitjantjatjara, among Ngaatjatjarra-speaking people, another mode of affiliation is operating. It could be described as gendered duo-lineal (or ambilineal) indirect affiliation, that is, a man preferably but not exclusively refers to his father's father's country, and a woman preferably but not exclusively to her mother's mother's country. The secondary affiliations to other members of preceding generations are cognatic but do not constitute cognatic descent groups based on genealogical links. In European culture, the cognatic recognition of descent or ascent does not constitute a group as well, and, if otherwise, then one specific line, usually the male's line, is dominant and the other eliminated or complementary.

(7.) See Holy (1996:115ff) for further discussions on cognatic descent.

(8.) See, for example, Turner (1980:ii).

(9.) Birthplace is an important affiliative criterion. However, since most children are born in the Alice Springs hospital nowadays, the parent's community of residence is often chosen as the child's birthplace.

(10.) It would equally be erroneous, of course, to consider a subsection as a landholding unit or a social corporation.


Barnes, J.A. 1967 Genealogies. In A.L. Epstein (ed.), The Craft of Social Anthropology, Tavistock, London, 101-27.

Bernd, L. 1966 Ambilineal Descent Groups in the Northern Gilbert Islands, American Anthropologist 68(3), 641-64.

Bouquet, M. 1996 Family Trees and Their Affinities: The Visual Imperative of the Genealogical Diagram, Journal of the Royal Anthropological Institute 2(1), 43-66.

-- 1997 Reply to Robert Parkin, Journal of the Royal Anthropological Institute 3(2), 375-6.

-- 2000 Figures of Relations: Reconnecting Kinship Studies and Museum Collections. In J. Carsten (ed.), Cultures of Relatedness: New Approaches to the Study of Kinship, Cambridge University Press, 167-90.

Coult, A.D. 1964 Role Allocation, Position Structuring, and Ambilineal Descent, American Anthropologist 66(1), 29-40.

Dousset, L. 1999 L'alliance de mariage et la promesse d'epouses chez les Ngaatjatjarra du Desert de l'Ouest australien, Journal de la Societe des Oceanistes 108, 3-17.

-- 2002 Accounting for Context and Substance: The Australian Western Desert Kinship System, Anthropological Forum 12(2), 193-204.

Durye, P. 1988 Genealogie, PUF, coll. Que sais-je?, Paris.

-- 1999 Genealogie, Encyclopaedia Universalis, CD-ROM.

Finlayson, J.D., B. Rigsby and H.J. Bek 1999 Introduction. In J.D.

Finlayson, B. Rigsby and H.J. Bek (eds), Connections in Native Title: Genealogies, Kinship and Groups, Research Monograph no. 13, Centre for Aboriginal Economic Policy Research, The Australian National University, Canberra, 1-12.

Firth, R. 1957 A Note on Descent Groups in Polynesia, Man 57, 4-8.

Fortes, M. 1959 Descent, Filiation and Affinity, Man 59(309 & 331), 193-7; 206-12.

Fox, R. 1996 (1967) Kinship and Marriage, Cambridge University Press.

Hamilton, A. 1979 Timeless Transformation: Women, Men and History in the Australian Western Desert, PhD thesis, University of Sydney.

Heritier, F. 1981 L'Exercice de la Parente, Hautes Etudes, Gallimard, and Le Seuil, Paris.

Hiatt, L.R. 1968 Ownership and Use of Land among the Australian Aborigines, In R.B. Lee and I. De Vore (eds), Man the Hunter, Aldine, Chicago, 99-102.

Hirschfeld, L.A. 1986 Kinship and Cognition: Genealogy and the Meaning of Kinship Terms, Current Anthropology 27(3), 217-42.

Holy, L. 1996 Anthropological Perspectives on Kinship, Pluto Press, London and Chicago.

Ives, J.W. 1998 Developmental Processes in the Pre-contact History of Athapaskan, Algonquian, and Numic Kin Systems. In M. Godelier, T.R. Trautmann and F.E. Tjon Sie Fat (eds), Transformations of Kinship, Smithsonian Institution, Washington and London, 94-139.

Layton, R. 1983 Ambilineal Descent and Traditional Pitjantjatjara Rights to Land. In N. Peterson and M. Langton (eds), Aborigines, Land and Land Rights, Australian Institute of Aboriginal Studies, Canberra, 15-32.

Leach, E.R. 1962 On Certain Unconsidered Aspects of Double Descent Systems, Man 62(214), 130-4.

Levi-Strauss, C. 1967 (1947) Les structures elementaires de la parente, Mouton, Paris.

-- 1968 Gidjingali Marriage Arrangements: Comments and Rejoinder, Discussions Part IV, 22b. In R.B. Lee and I. De Vore (eds), Man the Hunter, Aldine, Chicago, 210-11.

Malinowski, B. 1963 (1913) The Family among the Australian Aborigines, Shocken Book, New York.

Murdock, G.P. 1960 Cognatic Forms of Social Organization. In G.P.

Murdock (ed.), Social Structure in Southeast Asia, Quadrangle Books, Chicago, 1-14.

Parkin, R.J. 1996 Genealogy and Category: An Operational View, L'Homme 139, 87-108.

-- 1997 Kinship with Trees [comments on 'Family Trees and Their Affinities: The Visual Imperative of the Genealogical Diagram' by M. Bouquet in Journal of the Royal Anthropological Institute 1996 (2:1) 43-66; with reply by Bouquet, see this issue, 375-6], Journal of the Royal Anthropological Institute 3(2), 374-5.

Peterson, N. 1977 Territorial Adaptations among Desert Hunter-Gatherers: The !Kung and Australians Compared. In P.C. Burnham and R.F. Eller (eds), Social and Ecological Systems, ASA Monograph in Social Anthropology, Academic Press, London, 111-31.

Rivers, W.H.R. 1910 The Genealogical Method of Anthropological Inquiry, Sociological Review 3, 1-10.

Sackett, L. 1976 Indirect Exchange in Symmetrical System: Marriage Alliance in the Western Desert of Australia, Ethnology 15(2), 135-49. Sutton, P. 1998 Native Title and the Descent of Rights, Native Title Tribunal, Perth.

Turner, D.H. 1980 Australian Aboriginal Social Organisation, Australian Institute of Aboriginal Studies, Canberra, and Humanities Press, Atlantic Highlands.

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

Laurent Dousset. After fieldwork of more than two years between 1994 and 1997 in the Western Desert, the author completed his doctoral studies at the Ecole des Hautes Etudes en Sciences Sociales in Paris in 1999. He was then appointed as post-doctoral research fellow at the University of Western Australia, where he obtained an ARC fellowship. In 2003 he returned to EHESS (Marseille), teaching as associate professor. His fields of research include kinship and social organisation, socio-cultural transformations, contact history in the Western Desert, indigenous rights within international agreements and conventions and, more generally, hunter-gatherer societies and their transformations.

EHESS-CREDO (Centre de Recherche et de Documentation sur l'Oceanie) Maison Asie Pacifique, Universite de Provence. 3, place Victor Hugo, 13003 Marseille, France. <> <>
COPYRIGHT 2003 Australian Institute of Aboriginal and Torres Strait Islander Studies
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2003 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Dousset, Laurent
Publication:Australian Aboriginal Studies
Geographic Code:8AUST
Date:Mar 22, 2003
Previous Article:Euphemism, banality, propaganda: anthropology, public debate and Indigenous communities.
Next Article:The creation of Wilkinkarra (Lake Mackay) in Pintupi/Kukatja Dreamings.

Terms of use | Privacy policy | Copyright © 2022 Farlex, Inc. | Feedback | For webmasters |