CHAPTER 4
In the jilimi: mobility

The high residential mobility of Warlpiri people through the camps of Yuendumu, can be approached from two directions. One is to follow individuals on their trajectories through camps; the other is to take a camp-centric view and examine the flow of people through it. I here undertake the latter, and analyse mobility in the jilimi in which Celeste, Joy, Nora, Polly and I lived.

The flow of people is a reality in all kinds of camps at Yuendumu and is by no means exclusive to jilimi; however, its volume is particularly high in these. During the period I lived in the jilimi, for example, more than 160 other individuals stayed there, at different times and for different lengths of time. For a four-bedroom house, this is a staggering number of residents, even if not all were ever present at the one time.¹

Moreover, none of this mobility was random. Each person arrived in and left the jilimi for their own reasons, which included choosing to stay at this particular jilimi because others they knew and related to were staying there.

In order to understand the patterns underlying mobility, we also need to examine the relationships between people. Our jilimi thus serves well to illuminate the entanglements of physical structures (a four-bedroom house), values underpinning particular ways of being in the world (here, mobility), and social practice (the relationships of Joy, Nora, Celeste and Polly to the house and the jilimi on the one hand, and their roles in facilitating the mobility of others through the jilimi on the other).

The magnitude and the nature of Warlpiri people’s mobility through the camps of Yuendumu made it impossible for me to reconcile the reality I was observing (and participating in) with the analytical terms anthropologists commonly apply in similar contexts. For example, how can it be possible to find a social unit at Yuendumu that can be defined as ‘a household’ when people move so much? The constant flux in the social composition of people who sleep in any one camp does not at all lend itself to trying to draw boundaries around units. Yet many people who work in Aboriginal Australia use the term. It is habitually applied within the post-contact, on-going colonial and post-colonial contexts (rather than in reference to pre-contact camps where anthropologists commonly speak of bands and the domestic cycle, see amongst other Peterson 1978).

Much of the discussion about Aboriginal households is underwritten by an implicit and often explicit understanding of money and access to it as being the triggering force behind the dynamics of the formation of households as social units. Such studies stress the mutual interdependence of mobility and financial factors, and portray Aboriginal households as prone to cyclical changes induced by the vagaries of ‘boom and bust’ or ‘prosperity and poverty’ (see especially Finlayson 1991; Finlayson et al. 2000; Finlayson and Auld 1999). Generally, the practice of ‘demand sharing’ is seen to engender fluctuations in the social composition of Aboriginal residential arrangements.² Focussing on the mobility side of the equation, Sansom (1982) introduced the term ‘concertina household’ into the literature, used by Perth welfare workers to describe the high fluctuations common in dwellings occupied by Aboriginal people there. This fluctuating social composition, Sansom suggests, is a crucial feature of the contemporary Aboriginal commonality.³ I agree, and maintain that in the face of this, the term ‘household’, with its suggestions of boundedness, is rendered meaningless.⁴ Similarly, and almost three decades ago, Yanagisako, in her review of the extensive anthropological literature on households (1979: 200), concluded that terms such as household are ‘merely “odd-job” words, which are useful in descriptive statements but unproductive tools for analysis and comparison’.⁵ Outlining the analytical dilemmas associated with this term is important however, reaffirming as it does my reasons to maintain a methodological distinction between domestic space on the one hand and social practice relating to this space on the other. This distinction allows for an examination of mobility as process and value in its own right, rather than viewing mobility as effect on bounded categories (such as the household). Accordingly, the questions I address are: where do people come from, where do they go, why do they stay, why do they go, and what are the reasons and meanings underlying their mobility? To answer them I first outline the quantitative realities of mobility through the jilimi, and then analyse these statistical findings by contextualising them ethnographically.

Extent and volume of mobility in the jilimi

During my fieldwork, I took a census of who slept where and next to whom each morning when I woke up. I slept in this particular jilimi and recorded census data there for 221 nights, over a period of some 467 nights. This means that for 246 days of that period I have no data about who stayed at the jilimi because I stayed elsewhere. This is significant in itself, because while I considered the jilimi my home for this period of my fieldwork, the data shows that I spent fewer than half of those nights there. As my own mobility was largely determined by the mobility of the people I was living and working with, this may well be quite representative of the amount of time people generally spend ‘at home’ and ‘elsewhere’.

As we have seen, more than 160 individuals slept in the jilimi during those 221 nights. Compared to an average four-bedroom house in non-Indigenous Australia, this is an astonishing number. Moreover, when I say ‘at least 160’ persons stayed in the jilimi over the census period, this is a conservative estimate. As I was by no means the first person up every morning, the chances are that I have failed to count people who left early in the morning. Further undercounting has, I suspect, occurred during the initial stages of fieldwork in relation to children, not all of whom I knew. More importantly, one difficulty arose out of the fact that initially I did not know the names of many of the people who lived in the jilimi.⁶ The census data encompasses named individuals (105 in total) and three other categories: adults whose names I did not know (60), children whose names I did not know (146),⁷ and sorry mobs. The latter term describes groups of people who travelled to Yuendumu from other settlements to participate in mortuary rituals, and stayed in or adjacent to the jilimi and accessed its facilities during the time mortuary rituals were performed. Sorry mobs comprised between ten and twenty people at any one time.

Including named and un-named individuals (but excluding sorry mobs), as indicated in Table 1, on average 17 people stayed in the jilimi every night, on average 12 adults and 5 children. Over the census period, the minimum number for adults was 6 and the highest 19. For children, the highest number was 11 and the minimum was one child present.

Table 1: Average numbers of adults and children sleeping in the jilimi over 221 nights

Such averages have to be treated carefully, as they conceal another dimension. These average numbers encompass different people at different times, which distinguishes them from comparable non-Indigenous statistics. In a case study about another Yuendumu jilimi I presented elsewhere, there was an average of 22 people, but in actual fact:

Over the fortnight there were a total of 27 different adults and 15 different children sleeping at the house; that is, a total of 42 different persons. Moreover, a ‘core’ of 11 persons (seven adults and four children) slept at the house for the whole two-week period. (Musharbash 2000: 59)

The point is that average numbers of residents conceal the actual flow of people through camps. The figures in Table 1 should thus be treated with caution and be read as indicating statistical realities rather than actual practice. The fact that more than 160 people stayed in the jilimi over the census period is at least as important as the fact that on average 17 of them were present on any one night. Further analysis of the figures draws attention to other features of the jilimi population. The graph in Figure 10 presents the number of nights the 105 named individuals spent in the jilimi on those 221 nights I slept there (the graph excludes both un-named individuals and sorry mobs).

Each column of the graph represents the number of nights a named individual slept at the jilimi. Because it is based on the nights I myself spent in the jilimi, the tallest column represents myself and the 221 nights I spent there. The next column in line represents the person who spent the next highest amount of nights in the jilimi, while I was there. It is important to keep in mind, however, that this person, like most others covered by the graph, would have been present on many of the nights I was not. The graph is skewed by my own mobility; it does not present data for a continuous period of time, and it may well under-represent some people and over-represent others (especially those like Celeste, for example, whose patterns of mobility I often followed). Nonetheless the data does reflect well the way people relate to the jilimi. In analysing the data in the graph it is helpful to divide the columns into four sections, splitting the 105 individuals into categories according to how many nights they spent in the jilimi.

Figure 10: Nights/people in the jilimi

The first section comprises eleven individuals who during the census period all slept in the jilimi more than 100 nights (133–221 nights), and distinctly more than everybody else. I call these the core residents and discuss them under this heading below. Then the curve takes a deep plunge, and the next section is comprised of those individuals who spent distinctly fewer nights in the jilimi than the core residents but distinctly more than the remaining individuals — these I call regular residents (44–76 nights/12 individuals). Then the curve peters out slowly and I divide it into two more sections, one comprising those individuals who slept in the jilimi on an on-and-off basis (8–36 nights/36 individuals) and the other those who slept there sporadically (1–6 nights/48 individuals).

Table 2: Types of residents

	Number of nights	Number of individuals
Core residents	100+	11
Regular residents	76–44	12
On-and-off residents	36–8	36
Sporadic residents	6–1	48

Core residents

These are the eleven individuals who stayed in the jilimi for distinctly more nights than all other residents over the same period, and who can be described as relating emotionally to the jilimi as ‘home’ for most or all of the census period. They would have thought of and talked about the jilimi as ‘home’ when they were sleeping there as well as on the many nights they slept elsewhere. In contrast to Bell’s description of jilimi residents as senior, ritually active women (1980a, 1993), these eleven individuals belong to four generations and include children as well as elderly women beyond the state of social seniority (see Musharbash forthcoming on the loss of social seniority). Next to Polly, Joy, Celeste, Nora and I, the others are (see also Figure 11):

1) Neil (aged thirteen), who is the adopted son of Celeste, the biological son of Celeste’s sister and Polly’s grandson. Polly and Celeste, while co-residing, shared the responsibility of bringing up Neil.

2) Nangala (in her eighties), Joy’s frail and blind mother, cared for by Joy, assisted by Polly and Celeste.

3) Kiara (aged ten), who lives with her adoptive grandmother Joy, and who is is the daughter of Polly’s son whom Joy adopted, and thus is, in actual fact, Polly’s grand-daughter and Celeste’s niece.

4) Toby (aged twelve), one of Nora’s grandsons, whom she looks after.

5) Pearl (in her seventies), a close sister of Nora’s. These two, Pearl and Nora, co-resided and shared resources and time over long periods of their lives.

6) Annie (in her fifties), who is the daughter of a close brother of Polly’s deceased and Joy’s former husbands. Annie is not a co-dependant of any of the four women and in fact moved into Nora’s room when she left, taking up a similar focal position in the jilimi as Nora held before.⁸

Figure 11: Genealogy of core residents

Who comes to stay at the jilimi, when, why and how long for, is largely, but not exclusively, determined through the relationships people have to any of these core residents. Based on age, social status and life history, Polly, Joy, Celeste, Nora and Annie in this regard certainly had a greater gravitational pull than the other core residents, and I describe them as focal in this regard (however, people also came to stay with Kiara, Toby, Neil, Nangala and myself, as well as with some of the less regular residents). In the following case studies, I provide some examples of why people came to stay in the jilimi, who they stayed with, how long for and so forth, by paying particular attention to the nature of the relationships between these residents and Polly, Joy, Celeste, and Nora respectively. In these case studies, I also discuss these residents’ stays in regards to the classification by amount of nights slept in the jilimi during the census period.

Staying with Polly: Amy

Just counting close kin, Polly has many descendants: next to her own eight children, twenty-two grandchildren and nineteen great-grandchildren, she has also ‘brought up’ a number of other children. All of these and their respective children and grandchildren make up the substantial pool of her descendants. However, not all of them spend equal amounts of time with Polly, nor do they all have the same access to her resources. A few never stay with Polly, and the others are distributed across the four categories of residents. This fact hints at a crucial difference between types of residents: although people in one category may be equally closely related to a focal woman as those in another, they stay with her for different amounts of time and for different reasons. Residence may thus acquire different qualities depending upon the factors underlying it.

Polly is Amy’s paternal grandmother; the reciprocal kin term for this emotionally often close, caring and comfortable relationship (between father’s mother and son’s children) is yaparla. Amy is in her late twenties, and she has been married twice. Her first marriage was to a man from Nyirrpi with whom she has a teenage daughter who spends most of her time with her paternal grandmother in Nyirrpi. Amy’s second husband is from Kintore and she has a five-year-old son with him, who usually stays with Amy and sometimes with his paternal grandmother in Kintore. Amy herself mainly lives in one of the Alice Springs town camps.

Then Amy became gravely ill. She told us that after several checks at Alice Springs hospital, the doctors there decided they could not tell what was wrong with her. Since her illness could not be determined with certainty, everybody suspected sorcery, and Amy had to find a safe place to stay and be cared for (this case is study is further discussed in Musharbash 2008a). Neither the Alice Springs town camp, nor Kintore or Nyirrpi seemed good places, since in all of them lived affines (in-laws) of Amy — and affines, especially if marital relations are not too good and there are fights over children, are the first suspects in cases of sorcery. Since her parents are no longer alive, the place for her to go, then, was to her yaparla, Polly. During her sickness, Amy stayed physically close to Polly, sleeping next to her every night, and also shared her money and food with Polly. Polly in turn looked after Amy and organised a number of trips to go with her to other settlements to visit traditional healers to find out the causes of and cures for her grand-daughter’s illness. Amy stayed in the jilimi for 36 nights (while I was there), while her son (Frederico) stayed with her for 28 nights.

In this instance, Amy and Frederico can be classified as on-and-off residents who have a particular reason for staying in the jilimi for a substantial period of time (Amy’s illness). Once recovered, Amy went back to Alice Springs, and when she came to Yuendumu after that she generally stayed in other camps. During Amy’s illness, her daughter Cassandra came from Nyirrpi to visit her, but while in Yuendumu Cassandra often preferred to stay with some of her paternal relatives and only spent a total of six nights in the jilimi. Cassandra thus falls into the category of sporadic residents, and her example serves well to illustrate some core differences. Polly lived in the jilimi, Amy came to stay with Polly during the time of her illness and Polly, Amy and Frederico shared space and resources equally. Cassandra, who also is Polly’s great-grand-daughter, sometimes came to stay with her mother, and although Amy looked after her while she was in the jilimi, Polly did not get involved much. Cassandra’s access to the jilimi was conveyed through Amy and did not come from Polly. The fact that Cassandra spent fewer nights in the jilimi and more with her paternal relatives points to the reasons why she is here classified as a sporadic resident, while Frederico is classified as an on-and-off resident. Both Cassandra and Frederico are related to Polly in exactly the same way (both are children of Polly’s grand-daughter Amy); however, Frederico is emotionally closer to Polly than Cassandra is (and there is more shared history between them) and Cassandra thus stays elsewhere more often.

Staying with Joy: Charity, Jenna and Megan

Joy’s adopted son (Polly’s actual son) had four children, all of whom were brought up by Joy, as was one of their sisters, Charity, daughter of the same mother but a different father (Charity is thus not a grandchild of Polly’s). The youngest of these grandchildren, Kiara, stayed with Joy at all times (and was one of the jilimi’s core residents). Her siblings are in their teens and early twenties, and most of them are involved in tempestuous marriages. Kiara’s next eldest sister, Charity, was highly mobile and oscillated between her father’s place, her young husband’s parents’ place, her close grandfather’s place, her close sister’s place and the jilimi. There, she spent 34 nights (during the census period) and is thus a typical example of an on-and-off resident. This kind of ‘restless’ or ‘unsettled’ residency behaviour is fairly common among young Warlpiri girls, who until they settle down with a husband and children are extremely mobile within a fairly stable and limited number of residences. Where Charity would sleep each night depended upon what happened during the day, with whom she spent time and where she ate dinner. If she had dinner in the jilimi, as she sometimes did, she might simply stay there. On other nights she would have spent time with Kiara at the disco and then came home with her, and so forth. The point is that in all her usual residences there would always be a place for her to stay, and if she felt like it she would come ‘home’ to the jilimi and simply crawl under the blankets with Joy and Kiara.

Kiara’s and Charity’s sister Jenna stayed in the jilimi for 18 nights, and like Charity, Jenna oscillated between a number of places: her young husband’s parents’ place, Mt Theo outstation and the jilimi.⁹ Being in a somewhat more stable marriage than Charity, Jenna would only stay in the jilimi when fighting with her husband, or to be with her siblings if all of them were at the jilimi at the same time.

Megan, the eldest sister, stayed 20 nights in the jilimi. She falls into the category of on-and-off resident as well, but her story is somewhat different to the others. Megan had been married to a man from the south and lived there until their marriage deteriorated. When she returned to Yuendumu she moved in with Joy, and lived in the jilimi as a single woman until she got married again and moved with her new husband into her close grandfather’s camp. Megan’s and her sisters’ patterns of and reasons for staying were quite different; however, they all can be classified as on-and-off residents and they all came to stay with Joy.

Staying with Celeste: Adrian and Stella, Jemima and Angelina

Many years ago, Celeste separated from her husband (who lives in Willowra) and has been living in a number of jilimi since. Celeste’s son Adrian lived with her until his initiation, after which he began living in jangkayi (men’s camps). When Adrian married Stella, a girl from Hermannsburg, they had to face the problem of where to live. Normally today, a young, newly married couple sets up their yupukarra in the camp of either partner’s parents.¹⁰ However, since they wanted to live in Yuendumu, where Adrian had employment, this was not possible: Stella’s family lived in Hermannsburg, and all of Adrian’s paternal relatives were in Willowra, while the older members of his maternal family at Yuendumu were women living in jilimi. At first, Adrian and Stella stayed in a derelict house close to the jilimi for a while, using the facilities of the jilimi and joined by a number of other young couples. Winter approached, and the derelict house provided almost no shelter and no warmth, so all but Adrian and Stella deserted it. Finally, without protection from the weather and without marlpa (company), they used Celeste’s room in the jilimi as a yupukarra (married people’s camp). Due to the substantial period they spent in the jilimi while looking for a new place (Adrian stayed for 48 nights and Stella for 54), they can be classified as regular residents. When Adrian was elsewhere, their yupukarra ceased to exist and Stella shared a yunta with other women in the jilimi. While the arrangement of having a yupukarra in the jilimi did not please anybody (neither the couple nor the other residents), everybody agreed that for the time being there was no other option. During the day, Adrian was at work at the Mining Company or away with his brothers and cousins, and he had previously, while unmarried, often been present in the jilimi at mealtimes anyway. The only difference during their stay in the jilimi was that the door to Celeste’s room was closed at night and access to the room was restricted, symbolically marking the separation between jilimi and yupukarra.

Jemima and Angelina are the six-year-old twin daughters of Camilla, who is Celeste’s deceased sister’s daughter, making the twins Celeste’s close grand-daughters and Polly’s great-grand-daughters. Celeste and Camilla spend much time together and also work together at Yuendumu’s Childcare Centre. When Camilla is away shopping in Alice Springs or taking courses at Batchelor College, she asks Celeste to look after her daughters. Angelina and Jemima spent 20 and 15 nights in the jilimi respectively. The difference between the two girls’ stays is due to the fact that while Camilla looks after Angelina, her twin Jemima is being brought up by her paternal grandmother who spends substantial amounts of time in Murray Bridge. Thus Angelina sometimes comes to stay with Celeste on her own, while her twin Jemima is only ever in the jilimi when Angelina is there too. When they stay in the jilimi, they always stay with Celeste, not with Polly.

Staying with Nora: Sharon, Leah, Eva and Ray

Nora’s daughter Sharon often came to visit the jilimi during the day, to gamble, gossip, and pass time. She also spent a substantial number of nights there because of her rather stormy marriage. Her husband often left Yuendumu for business trips that frequently turned into long absences. Whenever he left Yuendumu, or when they had a fight, Sharon would, according to Warlpiri practice, move into the jilimi. However, Sharon’s relationship to her mother Nora was almost as stormy as that with her husband. She would thus move into the jilimi in which her mother was staying, but only stay in the same yunta as her mother when their relations were smooth. More often, she would set up her mattress next to some other people staying at the jilimi at the same time, for example with Polly’s daughter Marion when she was staying in the jilimi, or with Joy’s grandchildren. Sharon spent 29 nights in the jilimi, and is thus yet another on-and-off resident. But due to her age (she is in her early fifties) and her familiarity with the jilimi and its residents, she moved into the jilimi as a free agent as much as she did as Nora’s daughter.

Nora’s sister Leah (in her late fifties) and Joy’s half-sister Eva (in her fifties), who have co-resided for many years, came to stay in the jilimi with Nora twice for different reasons. The first time they came to care for Nora, who had returned home from Alice Springs after having been hospitalised with pneumonia. The second time they came because of trouble at their former residence. Leah and Eva for a while had been living with Leah’s daughter and her husband. However, that camp was known to be a locus of violence, and after things got out of control one too many times, they moved into the jilimi. With them came Ray, Nora’s grandson and Toby’s half-brother. Previously, Ray had moved between the two camps, and stayed alternately with Nora or with Leah, as a regular resident in either camp. Now, these women, together with their half-sister Pearl, who had been staying with Nora already, and Nora’s grandsons, formed a tight-knit group, sharing resources between them and living in close proximity. In fact, when Nora received ‘her own house’, they all moved into that house together (except for Pearl who joined them there much later). In early 2005, they still lived in that camp joined by a continuous flow of people coming to stay with them in turn, and later that year moved en masse into another house because their previous one had come under a death-related taboo (yarrkujuju, see also Musharbash 2008b) when one of Nora’s grandsons passed away.

Two things are interesting here. First, Eva, who is more closely related to Joy, stayed with Nora because Leah did so. This is a case were friendship ties, those between Leah and Eva, were more significant than kinship ties, those between Joy and Eva. Second, while Leah and Eva were on-and-off residents in the jilimi where they did not stay long and for necessity only, once they moved into the new house with Nora, all three of them, as well as their grandsons, became core residents there.

Residential categories and personal networks

As these case studies illustrate, many of the other jilimi residents are some of the relatives of the focal resident women, but by no means all. They constitute a relatively small number of the substantial pool of relatives and close associates that each of these women can draw upon. In relation to the focal women, the twelve regular residents (44–76 nights) in this jilimi and during the census period, were comprised of sisters (four), daughters (three), sons (two) and one son’s son, father’s sister, son’s wife and daughter-in-law respectively. It needs to be kept in mind that each focal woman has a much larger number of people in each of these kinship relationships, and that it is only a few who regularly stay with them. This point is important as it underlines the fact that other factors are at work besides the kinship status of the people involved.¹¹

There were more on-and-off and sporadic residents (36 and 48 respectively) than core and regular residents (11 and 12 respectively), underscoring the fact that people regularly stay in camps other than their home ones for short periods. While staying less often or for shorter periods in the jilimi than core and regular residents, on-and-off residents are nonetheless recruited from the same pool of kin surrounding the five focal women. In the main, on-and-off residents relate to these women as children, grandchildren, great-grandchildren, sisters and cousins. Further, there are some other individuals in this category whose relationships can be traced genealogically to the focal women but who are more accurately described by their relationship to the jilimi. Take Greta, for example. She used to be the ‘owner’ of one of the rooms; in her own words, ‘the whole building’ used to be hers.¹² While I was staying at the jilimi, she had two other main residences that she oscillated between, but she also frequently came to stay at the jilimi. Having once been a focal woman in this particular jilimi, she stayed in the jilimi because she felt it was her right to do so, and also to express her closeness to some of the more recent focal women through residing with them occasionally. Often when she came to stay, her grand-daughter and her sister joined her, and these two never came to stay in the jilimi without her.

Another woman in this category, Celia, moves backwards and forwards between Willowra and Yuendumu, and when in Yuendumu she stays for substantial periods of time in the jilimi. Although she often put up her yunta on the spatial margins of the jilimi, in many respects she paralleled the focal women in terms of access to rights and space within it, as well as through the gravitational position she was in. When living in the jilimi, she was often joined there by her daughter-in-law and a number of grandchildren.

A final woman in this category was a close sister of a former focal woman who had passed away some time ago. Since this woman was also a close classificatory relative of some of the current focal women, she continued to stay at the jilimi for a period before moving elsewhere for good.

The 48 sporadic residents (1–6 nights) are similar to the on-and-off residents in that they are made up of actual and close classificatory relatives of the focal women. Some differences between the two categories are that the former includes ex-focal women and that persons in the latter stayed in the jilimi less frequently and for shorter periods, often for one night only (during the census period, that is). It can safely be assumed that these patterns of occasional short stays were repeated at other times.

Those latter, most infrequent short-term stayers fall into two different kinds. Firstly, there were those who would have a number of other options to explore before staying at the jilimi, whereas for many individuals described as regular and on-and-off residents the jilimi would be the first choice — after their own ‘home’. And secondly, such short-term stayers are also made up of those who came from other settlements and stayed in Yuendumu for brief periods only. While only a few individuals in the above categories have their usual place of residence elsewhere than Yuendumu, a striking difference about sporadic residents is that almost half of them are individuals usually based elsewhere. Many of these are grandchildren of the focal residential women visiting from other settlements, as well as sisters and cousins. And while no adult core resident is male (some of the children are, though), and the categories of regular and on-and-off residents contain one and two men respectively, four of the sporadic residents were men. Men do stay overnight in the jilimi, but not many and rarely for very long. These men were sons and grandsons of the focal women, and two were the husbands of the daughter and grand-daughter respectively of one of the focal women.

It is important to note that all these (named) residents can and do trace their relationships to people already staying in the jilimi, in many cases (but not exclusively) to the focal women. Apart from myself and possibly some people in the sorry mobs, there were no ‘strangers’ who came to stay in the jilimi. However, a genealogical link alone is not enough cause to come and stay in the jilimi; what matters is the actuality of such relationships. Such relationships need to be lived, sustained and continually affirmed — practices which create the formation of personal networks.

I discovered these personal networks through participation rather than by being told about them. Warlpiri people do not generally teach the anthropologist by answering questions; they insist on one doing things (see also Harris 1987; Morphy 1983; Myers 1986a: 294). ‘You did this and now you know’ were words I often heard. In respect to mobility and residence choices it is only in retrospect that I realise what I have learned, and created. When I now return to Yuendumu I have choices as to where and with whom to stay. There are those people I am closest to, but should they be elsewhere I would not be homeless. There are a number of others whose camp I could join with equal ease. I cannot approach somebody and tell them, ‘I’ll stay with you for a while’ just because they are my classificatory mother, sister or daughter; however, I can do exactly that with somebody who is part of my personal network. Personal network relationships are based on shared experiences, shared residency in the past, and continued practice of reciprocal exchanges based on demand sharing. To be able to walk into a camp with one’s swag and put it down next to a person there, to stay there for an unspecified period, is possible only once one knows from whom one can demand hospitality.

This perspective on the personal trajectories of people’s residentiality alerts us to the fact that the classification into four types of residents according to the length of stay is arbitrary in some crucial regards. It only makes sense from a camp-centric perspective, during a specific period of time. Had I taken the same census a year later, people who were in one category might have been in another, or in none at all; others that did not stay there when I actually did take the census, stayed at the jilimi before or after, and so forth. Most importantly, if one examines the residential trajectories of individuals, one would find that each person falls into each category in different camps at different times. For example, while Polly, Joy, Celeste and Nora were core residents of the jilimi during the census period, they were not so a year before or after, and, what is more, during the census period they often stayed in other camps — where in turn they fell into one of the other residential categories.¹³

Both the camp-centric and the personal network perspectives provide crucial insights into contemporary mobility. The quantitative data presented here express the high rate of movement of people through the jilimi and are characteristic of social life throughout the settlement. The qualitative data suggest that people’s mobility patterns reflect their need for sanctuary when ill or involved in marital disputes, their involvement in mortuary rituals, their visiting of relatives, their need for or provision of help, their arrangements for childcare, and simply socialising. Above and beyond the particular reasons for any move stands the common practice of creating and the need to maintain personal networks through face-to-face interaction, for Warlpiri social relations depend on such networks. Mobility is a taken-for-granted aspect of life; in order to create and maintain the conditions for one’s own mobility one needs to accommodate the mobility of others.¹⁴

Thus, in order to truly understand mobility at Yuendumu, we must also take a step back and look at mobility not only as a practice but also as the value that it so obviously is. It shapes people’s everyday experiences and their lives. Why should this be the case? A clue to this lies in the Aboriginal English term for ‘staying at’ or ‘living with’, which at Yuendumu is stopping. As Warlpiri residential patterns are processual, to stop is an apt term underscoring the halt that the flow of people through the camps comes to every night. In this vein, when I return to Yuendumu these days, my friends and I reminisce about ‘that time we stopped in that jilimi’, or, when giving somebody a lift home, a sensible question to ask first is ‘where are you stopping?’. Stopping aptly characterises residential patterns in a life where they change on a regular basis.

In the olden days people moved from one place to the next, where they stopped and set up camp before moving on to the next place, where they stopped again, and so on. The contemporary settlement of Yuendumu is fixed in place, as are its houses. Yapa, however, continue moving — not across their country as in the olden days — but in such a way that each night, when putting up yunta, they stop, bring the events of the day to a halt and arrange themselves in ever-changing camps in and around the houses of the settlement.¹⁵

Conclusion

At Yuendumu, the flow of people is halted each night when people stop. Warlpiri people move; nobody lives in the same place with the same people permanently. Rather, people follow their own paths, which continually crisscross and occasionally run parallel to those of others, forming flows of people through camps. These flows of people through the camps cannot be captured within the standard terminology. Applying terms such as ‘the household’ in this context would entail the creation of analytical boundedness in a situation where there is none.

Nonetheless, I would like to examine one last definition of the household, or rather in this case, the ‘residential group’¹⁶ as it illustrates its own non-applicability in the Warlpiri context in particularly pertinent ways. This definition is the result of Verdon’s (1979) effort to criticise the original concept of the ‘residential group’. He argues that it is futile to define it by activities such as cooking, eating, pooling of resources and labour, as they are commonly engaged in with others from other dwellings. Instead, he proposes that the ‘residential group’ should be defined by the only one activity shared exclusively by all people in one dwelling, namely sleeping. Elaborating on the possible range of relationships between sleepers, he says that

in every society with residential groups, one thus observes a certain limit of internal complexity in their composition, some kind of ‘breaking point’ which is only exceeded in uncommon demographic, economic or physical circumstances (Verdon 1979: 420).

His proviso that this is true only for societies with residential groups begs the question who these are. Maybe Warlpiri people do not fall into this category. Or maybe the ‘uncommon demographic, economic or physical circumstances’ of settlement life cause them to exceed ‘certain limits of internal complexity’. Or, to make a third suggestion, maybe our obsession with searching for bounded categories blinds us to the possibility that reality (in some circumstances) is better understood without them. The extensive internal complexity of residence in Yuendumu camps, I hope to have shown, is the result of focussed agency, rather than random activity. There are as many reasons to stop in the jilimi as there are people who do so on any one night. The flow of people through the camps of Yuendumu is an expression of social practice being lived out, it is the result of negotiated relationships, and we should interpret it as such. The volume and rapidity of residential mobility is a direct consequence of the way Warlpiri people organise their everyday life around relationships, while the particular shapes residential mobility takes today are formulated in dialogue with the permanent structures of houses and the settlement.