Navigation – Plan du site
Kinship, gender, and social relatedness

A quality of being: embodied kinship among the Tiwi of Northern Australia

Andrée Grau

Texte intégral

Setting the scene

1Thirty years ago, almost exactly, I argued in my PhD thesis that, among the Tiwi, dance was kinship-in-action, not so much because it reflected kinship practices in other social domains - which it did - but because I observed on the dance floor kinship relationships being explored and new ones being generated. As I put it then ‘the kinship dances are not models of, but rather models for kinship practices’ (Grau 1983: 333).

2The anthropologist Jane Goodale coined the term ‘kinship dances’. In her book Tiwi wives, she described them as ‘special types of dance performed by individuals or groups of kin who have a particular relationship to the deceased’ (Goodale 1971: 300). It is important to note here that among the Tiwi, almost all dance performances are dedicated to someone, dead in the context of mortuary rituals, alive in the modern contexts of celebrations for weddings, birthdays, or graduations. The bereavement term and status given to an individual becomes by extension the label given to the dance they are entitled to perform and the modern contexts follows the structure of the mortuary rituals.

3In my work I have contrasted this part of the dance repertoire with another, which I have labeled the Dreaming dances.

4In 2007 the artist Pedro Wonaeamirri, Vice President of the Jilamira Arts and Craft Association commented, in the context of an exhibition of the works of late Kitty Kantilla at the National Galery of Victoria.

  • 1 The original punctuation of colon rather than full stop renders the Statement non sensical. Here I (...)

5I perform my brolga dance and my tide dance in ceremony when I dance with my cousin brothers [.W]e1 share the same songs when we dance for my uncle and aunty. But when I’m dancing my totem, it’s separate, because when I am dancing brolga, it’s a dance that belongs to me and also a special song to go with the brolga. When dancing in a ceremony, our dances are based on the relationship with the deceased person. Only sometimes I dance brolga at ceremonies. (Wonaeamirri 2007).

6One can therefore see a contrast between the two groups: the Dreaming dances belonging to specific individuals and lineages, the kinship dances belonging to all, the former marking the relationships between dancers and environment, the latter cementing social relationships.

7Although I knew from my reading prior to fieldwork that understanding kinship was essential to working in an Aboriginal community - that indeed kinship was then still the hard core of anthropology - I had not quite anticipated that I would end up compiling vast amount of genealogies in order to understand dance. On my very first day on Melville Island, however, I was made aware of the very close connection the Tiwi made between dance and kinship. When I mentioned my interest in dance, I was told: ‘you dance like this (with your fists clenched opposite your mouth) for your mother’, ‘like this (with the hands underneath your breasts) for your daughter’, and ‘like this (holding your leg) for your brother’. It was quite clear therefore that kinship was going to have a prominent place in my analysis of dance and I learned quickly that the kinship dances were at the core of the mortuary rituals.

8Kinship, as Alan Barnard and Anthony Good would have it, ‘appears to be the one area of anthropological discourse where the ground rules are clearly laid down’ (Barnard and Good 1984:2). I could therefore be quite systematic in going deeper into the kinship dances. Many scholars had worked among the Tiwi and most had listed kinship terms and discussed kinship organisation. Drawing on this work, I started to ask questions of the type ‘how do I dance if such and such a relative dies?’, slowly building up the kinship dance system. Alongside I compiled genealogies so that I could correlate what was happening in practice, noting who danced what, with whom, and for whom and compared it with the model, being especially attentive to the instances where practice did not fit the system, since I was very much aware that one has to be careful with the very notion of prescriptive models. I was therefore contrasting what Laurent Dousset has referred to as the ‘rather formal grid of rules and expected behaviour tied to marriage prescription, kinship terminologies and kin categories’ with what he describes as ‘the more flexible form of relatedness’ (Dousset 2005: 21).

9My first observation was that many of the kinship dances were focusing on specific body parts. The connection between body parts and kinship, or body part and land is fairly common in Aboriginal Australia. Indeed according to the linguist Michael Walsh ‘Australian languages are pervasive in the use of body parts metaphors’ (1996: 328). The anthropologist W. H. Stanner, for example, commented: ‘Our word “land” is too spare and meager […] the Aboriginal would speak of “earth” and use the word in a richly symbolic way to mean his shoulder or his side’ (Stanner 1979: 30).

10What does not seem to be so common, however, is to have a whole body of kinship knowledge explicitly articulated through the dancing body, many body parts being linked to specific kin: the legs representing a certain kind of sibling, the cheek: another, the abdomen: a woman’s children or a man’s sisters’ children, the shoulder: the in-laws, especially a man’s mother-in-law, the groin: a man’s children or a woman’s brothers’ children, the big toe: the husband.

11Many of the kinship dances are simple choreographically in contrast to the other dances of the repertoire, consisting often – but not always - of holding the specific body part with one or both hands while performing generic feet movements, shared with most other dances. Whilst this relationship between body parts and specific individuals was used in everyday life, for example when a man touched his shoulder to allude to his mother-in-law, nowhere else was it as clearly articulated as in dance.

12Furthermore I would argue that the literal incorporation of kin within the dancing body is not metaphorical or symbolic, but that it is “real” in the sense that it is about creating a special quality of being and of relatedness between individuals. To use Michael Houseman’s expression, I contend that a special ‘quality of action’ (Houseman, 2006, p. 413) is created in the dance. I have discussed elsewhere (Grau 1993) how being a mother or father did not depend on sex or gender, for example: Tiwi men danced that they were pregnant, giving birth and breastfeeding, whilst women danced about finding spirit children. Being a mother or a father was therefore primarily a quality of relationship rather than being a biological one and both women and men experienced both roles.

13The dance floor was also a space where the agendas of competing constituencies – often kinship based but not always - could be articulated and confronted with one another and where the heterogeneous aspects of the society could be sorted out (cf. Grau 1995). Through dance some Tiwi manipulated their social reality. Depending on how one looks at a genealogy, looking at the relationship between two individuals via the mothers, via the fathers, or via the spouses one can have different results about what role an individual should or could take in the dancing. Sometimes, in such cases, the dancers performed all the dances associated with their roles. Sometimes they chose to perform one role only, thus emphasising a specific relationship. And occasionally they manipulated the system to avoid performing the dance that genealogical connections prescribed them, and performed another dance related to the message they wanted to send. I have described elsewhere, for example, a woman who used dance in such a way as to protect her daughters from men she did not want them to marry (cf. Grau 1994).

14I would therefore argue that for the Tiwi, and for many Aborigines, their relatives are inside their bodies. When individuals felt a pain, for example, they usually interpreted it as a warning of illness or even death of specific relatives and more often then not this was indeed the case. To me this embodiment of kinship is no different to the embodiment of land that is often discussed, as for example, when Aaron Stuart, the chair of the Arabana Aboriginal Corporation RNTBC (Registered Native Title Bodies Corporate) commented last year ‘Our land is our identity; It’s who we are’ (Stuart, in Hull 2012: 5). In neither case, I would argue, are people speaking metaphorically or symbolically.

15Let me now turn to the kinship dances.

The kinship dances of the Tiwi

16The Tiwi recognised both patrilineal and matrilineal principles of descent and belonged to descent units of both types. It was well illustrated in the kinship dance system, which can be divided into categories of patrilineal, matrilineal, and affinal dances. Goodale too recognised this broad division. Her analysis, however, described six kinship dances, whilst I observed ten. It is worth noting that some of these dances were physically almost indistinguishable one from another, unless one was constantly aware of them in movement terms and kept checking what the movements meant. Eric Venbrux confirmed my model. In his book A Death in the Tiwi Islands, he also refers to ten dances and, drawing on my and Maria Brandl’s work, he provides descriptions very similar to my own, given twelve years earlier (cf. Venbrux 1995: 235-237).

17Time does not allow me to present the whole system here and in any case I have described it elsewhere (Grau 1998). I will first make some general remarks and then focus on one dance, the mutuni, the dance for the “cousin brothers” that Wonaemirri referred to earlier, as illustration. Throughout the following account I will use the past tense as it is based on observations made in the early 1980s and the late 1990s, rather than now.

18Looking at the dances overall one sees that, although the kin terms stressed a speaker’s own and related matriclans (there were twelve terms used for people in this category as distinct from eight terms for patrilineal relatives) the dances stressed the patrilineal side of the family. Altogether there were five patrilineal dances, three matrilineal dances, and two dances associated with relatives acquired through marriage.

19Four dances were reciprocal, i.e. the person for whom the dance was being performed would have danced in the same way if the roles had been reversed as they were of a similar, though not necessarily always equal, status. In these dances one sees that the mutuni, which I will discuss later, emphasised the patriline of the deceased, the paputawi stressed his/her matrilineal affiliations, and that the impala and the amparuwi recognised his/her affinal affiliations. Six dances were non-reciprocal dances, i.e. the person for whom the dance was being performed would have performed a different dance if the roles had been reversed. The units mamurapi turaa/wunantawi pulanga and krimirika/kiakiae emphasized the patriline and the meeting between Spirit Children and their fathers and fathers’ fathers. When men and children found each other the children received part of their identity in receiving their Dreamings. The unit mamurapi pularti/wunantawi pularti on the other hand emphasised the matriline and the link to the ancestresses, the daughters of Pukwi, the Sun-woman, who created the Tiwi world and from whom every Tiwi is descended.

20Looking at kinship through the kinship dances gives a dimension different to those previously given by other anthropologists working among the Tiwi. It has been said, for example, that in Tiwi society individuals looked towards their matrilineal kin in the field of physical welfare but that in spiritual matters and in ritual obligations they looked towards their patriline (Brandl 1971:229). From the perspective of the dance, however, this observation needs to be qualified: it is true that the patrilineal dances were mainly concerned with Spirit Children, but the movement of the respective dances also showed the physical making of the child, as well as the physically caring relationship between fathers and children. Similarly although the matrilineal dances were about the physical welfare of individuals, in that they dealt mainly with pregnancy and motherhood, they were reaffirming the links with Pukwi and the Dreaming. In addition, the accompanying songs often related to specific Dreamings and the spiritual associations between individuals and their land.

21The organisation of the final mortuary ritual followed a specific order of performances. The system, however, did allow a certain amount of flexibility in that not all categories of personnel necessarily had available members at the time, either because people were not around or because they simply did not exist. Usually the order was: first the patrilineal relatives, then the in-laws, then the matrilineal relatives, and last the amparuwi, people who were in the category of widows/widowers.

22Let’s now turn to the mutuni dance.

The mutuni dance

  • 2 It is interesting that kinship terms for the Tiwi are always preceded by a possessive pronoun.

23The patrilineally organised sibling group, the aminiatuwi performed the mutuni dance. One was mutuni if one called the deceased ngiaingkalapini (paternal half-brother belonging to a matrilineal «clan» other than one’s own), ngia ingkalapa (paternal half-sister belonging to a matrilineal «clan» other than one’s own)2 or if one was aminiati (pl. aminiatuwi) with the deceased, that is if one shared one amo (FM, FFZ, MFZ, FZD) or one amini (FF, MF, FMB, FZS). Aminiati was the only kin term, which implied equality between the people who used it. All the individuals who comprised the mutuni group were considered the «same». During the mutuni dance the performers held one or both cheeks with one or both hands. The accompanying songs often referred to the cheeks looking like rocks, such as those found on Pungalo.

  • 3 Goodale was based in Milikapiti, whilst I was in Pirlangimpi, about 30km away, two villages of some (...)

24It is interesting to compare the accounts and discussions about the aminiatuwi given by different anthropologists. Charles William Hart argued in the 1930s that members of this unit maintained exclusive rights to a geographical area validated by the presence of the grave of the common fathers’ father in the area, and that only those individuals descending from an important man (a man who had enough wives to produce enough children and grandchildren to enforce their claim and demonstrate their existence as a localised social group) belonged to such a group (1930:173-5). Goodale, on the other hand, who did her first Tiwi fieldwork in 1954, followed by a short one in 1964, and again a long one in 1980, when I started my own,3 disagreed with this view of the aminiatuwi being a non-universal social group. For her, it was a patrilineal sibling set, ‘conceptually parallel to the matrilineal sibling set so that members of such a set were considered as being «one person» by kinsmen outside the set’ (1971:97). However she saw it as being ‘overshadowed by the matrilineal system of affiliation’ (1971:98)

25My own analysis brought another emphasis. Documenting the mutuni dance I was told that it ‘marked’ a relationship between equal individuals related through someone important, amini or amo (it is worth noting too that linguistically amini and amo are also masculine and feminine adjectival suffixes meaning «big»). Looking at who performed the mutuni dance together, my data also showed unequivocally that the sex of the linking relative was not always male, as had been argued by both Hart and Goodale, though it usually was.

26The importance of this group was reflected throughout the Tiwi ritual life: during the annual Kulama yam ceremony men had to compose and perform over three days a song cycle; at one stage the songs had to be about people who had died during the year, at another they were about being specific women. In all the songs I studied over my different periods of fieldwork, covering six Kulama, the individual sung about, belonged to the same aminiatuwi set as the performer. I was told that the performers were ‘making them important’. During mortuary rituals members of the same aminiatuwi set as the deceased were also the decision-makers as well as the choreographers and composers.

27What is to me particularly interesting in the kinship dances is that, through them, every Tiwi experienced all the social roles of the society, even those that were impossible to execute or socially unacceptable in everyday life. In dance both men and women found Spirit Children even though in everyday life women were not even supposed to know what they looked like; in dance one could be a spouse or potential spouse to both men and women regardless of one’s gender; and both men and women were pregnant, gave birth and nurtured the baby. Thus not only did the Tiwi have a system where men and women generally had valued complementary roles but they also had the opportunity through performance somehow to «become» the other, therefore experiencing the whole world rather than just the one limited by their gender or age. Within the kinship dance system they could somehow imagine and create social reality as well as reflect it. Through the kinship dances, one could argue, everyday life was transcended and the performers danced through the three parallel worlds of the Unborn, the Living and the Dead.

28It was through my meticulous compilation of genealogies and movement observations that I was able to go into such details and add to the analysis of earlier anthropologists, and especially Jane Goodale who had been such a source of support in my early days in the field and with whom I shared my Kulama notes, as circumstances did not allow her to attend them during her stay (cf. Grau 1999 for a discussion of this). In her obituary, Laura Zimmer-Tamakoshi recalled Jane’s commitment to ‘grounded ethnography based on engaged, experiential, dialogic, ethical, reflexive, holistic, and culturally sensitive fieldwork recognizing individual agency and variation’ (Zimmer-Tamakoshi 2010: 346). Whilst I do not pretend to have achieve all this, in a small way this is what I attempted in my fieldwork and analysis.


Barnard, Alan and Anthony Good (1984) Research Practices in the Study of Kinship: Research Methods in Social Anthropology Waltham, Massachusetts: Academic Press.

Brandl, Maria (1971) Pukumani: The Social Context of Bereavement in a North Australian Tribe Unpublished PhD thesis, University of Western Australia.

Dousset, Laurent (2005) ‘Structure and Substance: Combining ‘Classic’ and ‘Modern’ Kinship Studies in the Australian Western Desert’ The Australian Journal of Anthropology 16 (1): 18-30.

Grau, Andrée (1993) ‘Gender interchangeability among the Tiwi’ in Dance, Gender, and Culture, Helen Thomas ed., London: MacMillan.

Grau, Andrée (1994) ‘Dance as politics: the use of dance for the manipulation of the social order among the Tiwi of Northern Australia’ Proceedings, 17th Symposium of the International Council for Traditional Music, Study group on Ethnochoreology, Nafplion: Pelopponesian Folklore foundation.

Grau, Andrée (1995) ‘Ritual Dance and «Modernisation»: the Tiwi Example’. In Dabrowska Grazyna and Bielawski Ludwik eds Proceedings of the 18th Symposium of the ICTM Study group on Ethnochoreology, Warsaw: Polish Academy of Sciences.

Grau, Andrée (1998) ‘On the Acquisition of Knowledge: Teaching Kinship through the Body among the Tiwi of Northern Australia’, in Verena Keck ed. Knowing Oceania: Constituting Knowledge and Identities, Oxford: Berg Publishers, 71-94

Grau, Andrée (1999) ‘Fieldwork, Politics and Power’ in Buckland Theresa ed. Dance in the Field: Theory, methods and issues in dance ethnography, London: Macmillan Press, 163-174

Goodale, Jane (1971) Tiwi Wives: A Study of the Women of Melville Island, North Australia seattke: University of Washington Press.

Hart, Charles William Merton (1930) ‘The Tiwi of Melville and Bathurst Islands’ Oceania 1(2):167-180

Houseman, Michael (2006) ‘Relationality’ In J. Kreinath, J. Snoek and M. Stausberg (Eds.) Theorizing Rituals. Classical Topics, Theoretical Approaches, Analytical Concepts, Annotated Bibliography (pp.413-428). Leiden: Brill.

Hull, David (2012) ‘This is Arabana Country: An interview with David Hull, director of Arabana Aboriginal Corp RNTBC’ Native Title Newsletter, December, 4-5.

Sutton, David and Renate Fernandez (1998) ‘Introduction’ Anthropology and Humanism 23(2): 111-117.

Venbrux, Eric (1995) A Death in the Tiwi Islands: Conflict, Ritual, and Social Life in an Australian Aboriginal Community Cambridge: Cambridge University Press

Walsh, Michael (1996) ‘Body parts in Murrinh-Patha: incorporation, grammar and metaphor’ in Chappell, Hilary and William Mc Gregory eds The Grammar of Inalienability: A Typological Perspective on Body Part Terms and the Part-Whole Relation Berlin : Walter de Gruyter, 327-382

Wonaeamirri, Pedro 2007 ‘Tiwi Art and Culture and the First Old Lady’ in Judith Ryan (ed) Kitty Kantilla Melbourne: National Gallery of Victoria (available online at (Consulted 20th January 2011).

Zimmer-Tamakoshi, Laura (2010) ‘Obituary: Jane Carter Goodale (1926–2008) American Anthropologist 112 (2): 344–347.


1 The original punctuation of colon rather than full stop renders the Statement non sensical. Here I assume that there is a typo

2 It is interesting that kinship terms for the Tiwi are always preceded by a possessive pronoun.

3 Goodale was based in Milikapiti, whilst I was in Pirlangimpi, about 30km away, two villages of some 400 inhabitants on Melville Island.

Pour citer cet article

Référence électronique

Andrée Grau, « A quality of being: embodied kinship among the Tiwi of Northern Australia », in Australian Aboriginal Anthropology Today: Critical Perspectives from Europe (« Les actes »), 2014, [En ligne], mis en ligne le 13 juin 2014, Consulté le 23 mars 2017. URL :


Andrée Grau

Roehampton London University

Droits d’auteur

© Tous droits réservés