Navigation – Plan du site
La parenté de nouveau en question

Kinship as a Relation

La parenté comme relation
Marilyn Strathern
p. 43-61


Les perspectives de recherche actuelles en anthropologie de la parenté peuvent tirer profit des travaux effectués sur la nature récursive des relations, qui apportent un éclairage nouveau sur les notions de maternité et de paternité. Ces deux notions sont indissociables l’une de l’autre, mais le fait que cette interaction soit pensée comme une « relation » qui sera ensuite comparée avec d’autres formes de mises en « relations » est une représentation euro-américaine. D’autant que, même si elle est largement répandue, cette approche de la relation ne peut être envisagée que grâce à un dispositif épistémologique spécifique différent et éloigné d’autres contextes. Le présent article, s’appuyant sur des études anglophones, compare diverses manières de concevoir la maternité et la paternité.

Haut de page

Texte intégral

1Anyone writing about kinship today does so in the middle of a long anthropological conversation. Here I focus on a largely unexamined but much used concept, the relation, and on a practice that veers from being the object of intense scrutiny to being taken for granted, that of description. Some brief evidence suggesting why these are interesting issues to consider together serves as an introduction.

A Question of Choice

2At one point in his splendid account of the Korowai people of West Papua, Rupert Stasch refers to kinship terms as describing, and he quotes a fellow anthropologist’s felicitous phrase, forms of « being-in-relation ». The context is a discussion of kin categories : « the categories say that a persons is his or her relations with certain others » (Stasch 2009 : 132, emphasis in the original). The quotation seemingly spells this out, continuing : « The terms of kinship are inherently linking terms ; [...] they render the self in and through its relation to certain others » (Faubion 2001 : 3). In a way Stasch hardly needs to back up his statement – his description of how Korowai reflect on the world is replete with references to a relationality articulated as kinship. Thus the Korowai term for « relative or kin » can refer to anyone with whom a connection is recognised, and indeed Stasch remarks that it does the functional work of his own terms, « link, relation, and belonging » (2009 : 104-105, emphasis omitted). He also makes it clear that such relations enfold what he calls otherness and strangeness within them, as in the emphasis put on maintaining relations with one’s mother’s brother. Differentiated by clanship and place-ownership, nonetheless, as a nephew said, squeezing his own thigh, « I love my uncle. He’s like the flesh of my leg » (Ibid. : 109).

  • 1 That is, not mediated by the act of naming social relations, relationality is simply assumed.

3The nephew’s gesture echoes a moment in Tony Crook’s account from Bolivip, over the border in Papua New Guinea. His Angkaiyakmin teacher pretended to make an incision in his thigh, then put his hands onto the ethnographer’s skin ; he said « that now his skin has gone onto mine, and that my skin has gone onto his […] we are “one skin” » (2007 : 29). The result of an exchange between them (advice for care), teacher and ethnographer mutually encompassed each other. The shiny membrane covering the thigh muscle, which the teacher’s gesture had opened up, is Crook was told, « another word for love » (Ibid. : 113). Like Stasch he describes the way connections make divisions, and how every phenomenon is a combination of differences. Unlike him, however, Crook claims that Angkaiyakmin have no language for relations : « [p]articipation of persons and things in each other is conceived without an idiom of “relationality” » (Ibid. : 28)1. Crook is here deliberately standing apart from the usage of most fellow anthropologists in his explicit aim « to make “social relationships” disappear from description of Bolivip » (Ibid.).

4Two rather different gestures, but two images – descriptions – of bodily mutuality. However, the point of comparison lies in the anthropologists’ descriptions. The one chooses to emphasise relationality at every turn, the other tries to expunge it from (part of) his vocabulary. Both do a highly successful job, but how come there is choice ?

5Choice was foreshadowed in the opening sentences that brought together two ways in which the idea of relation operates : identification and linking (of terms, of persons). Thus Stasch refers to « an almost seamless » identification of persons with relations, and then goes on, in good English idiom, to talk of relations « between » kin (2009 : 132). Now the English language makes it quite possible either to speak of two kinds of relations, or else to take one kind (the linking of terms), as an exemplar of relations while choosing to deal with the other (identification) as an absence of relation. Where Stasch wants to underline Korowai sensibilities towards relations and their need to make them, Crook wants to ask what it is that Angkaiyakmin make when they take relationality for granted.

6Such linguistic choice is not confined to the concept of relation, but I suggest that this concept is one of many places where the English language – and the anthropology it encourages – teases its users. The problems it at once solves and generates are probably not translatable. If only because of the extent to which « kinship » is in English invariably, and tautologously, expanded into « kinship relations », the concept of relation is worth some interrogation. I focus on certain constructions of motherhood and fatherhood to do so. But first let me be explicit about the approach to kinship entailed in re-rendering this material.

An Approach to Kinship : Making Description a Problematic

7The central problematic for the anthropologist is mundane : what has to be done (understood, analysed, theorised) in order to gain adequacy of description ? Briefly put in relation to kinship, now that we know what kinship is (Sahlins 2011), the question is how to describe it.

8One must be immeasurably grateful to Marshall Sahlins for his at once elegant and extensive elucidation of the idea of kinship as that « mutuality of being » or those « mutual relations of being » that define how people imagine themselves as intrinsic to one another’s existence. It is notable that his account is shot through with telling images drawn from the many ethnographic sources informing his general proposition ; we shall return to this. Now in one sense the identification or definition of kinship constitutes a description. Yet, in another, my question is opened up in his own conclusion, for it hints at further links. He points to Eduardo Viveiros de Castro’s work as a « revelation of a certain cultural order of intersubjectivity in which kinship takes a fundamental place » ; for kinship is to be understood in terms « of its relations to other dimensions of cultural order » (Sahlins 2011 : 239). Kinship’s relations to cultural order ? It would seem to follow that much will depend on how kinship in this or that context is rendered in the first place.

9Their ethnographers make it clear that, in talking about people’s effects upon one another, both Korowai and Angkaiyakmin are also talking about what English speakers would call the world, the cosmos, nature-culture. The manner in which kinship is articulated will indeed bear on how the world is lived (the specific truth of this for English-speakers will become apparent). The corollary for anthropological description is obvious : choice of concept influences the way in which the anthropologist’s subject matter is formulated. Hence Crook’s proposition, how it looks when social relationships are made to disappear.

  • 2 Relations as used between phenomena, social or natural, do not have their basis in the kin usage ; (...)

10Although other matters were in his mind, Louis Dumont’s well known complaint about the difficulties the British have « with their own language when it comes to defining kinship » (2006 [1971] : 3), and how this gets in the way of creating a scientific vocabulary, could have been applied to relations. Like Korowai, the English idiom for kinspersons is « relations » (also « relatives »), a concept used equally for any kind of identification or linking of persons and to the connecting of ideas and observations about the world at large, in short, to relations between phenomena. As we shall see, this does not make the world a kin-based one2. On the contrary, English-speakers have no difficulty in keeping these connotations distinct by context. So, whether compartmentalising the connotations or running them together, we are already in a rather particular world.

  • 3 Kinship does not disappear when it is perceived as a dimension of cultural order ; rather, it is ma (...)
  • 4 Although she sees « worlding » as a proposition about context, in its expanded form it refers to «  (...)
  • 5 With a different focus, this is what Tony Crook’s volume (2007) takes seriously ; its intriguing ou (...)

11Now the topic of kinship does not render the approach taken here any different from an approach to anthropological description at large. At the same time, there is a specificity about anthropological writings on kinship that invites further elucidation. The opening accounts from today’s Melanesia rehearse a substantive issue that has lain behind generations of kinship studies. At once an attribute of kinship systems as such and an attribute of how they might be described, the issue is the significance of relationality. Relations and relationships suffuse anthropological accounts in general, but it is in trying to convey the specificity of kinship that anthropologists dwell on relating as a veritable force in social life. The idea of kinship, we might conclude, evokes a cultural order3. We could as well speak of conceptual world, as Roy Wagner (1977 : 627) argued a generation ago (« [We should] begin our analysis with some particular conceptual world, and with a set of assumptions about conceptual worlds in general, rather than a general orientation regarding “kinship systems” ») ; or of Tsing’s (2009) adoption of the term worlding4. This chapter is about cultural orders or conceptual worlds in which relations are key. In the course of it, practices of description may start looking a little less mundane5.

12A preliminary hint is there in Dumont’s comments. One criticism he had of British kinship analysis in his day was the propensity to distribute « social reality into different sub-systems » (2006 : 11). What conceptual world is implied here ? The English usage of « relations » is illustrative. The contextualising practices by which English-speakers effortlessly keep as separate relations between kin and relations as a narrative of connection relevant to any phenomena, natural or otherwise, support a general propensity to compartmentalisation. So while « domain » may have gone largely out of fashion in anthropological parlance, « context » remains indispensable. I do not have to add that in vernacular social life English kinship emerges as a distinct sub-system or domain of relations (for historical and cultural reasons that make the English view of family life something of an oddity in Europe), so Anglophone anthropologists working in places such as Melanesia will find « kinship » occupying another kind of social space. Yet they may circumscribe its dimensions by implying that in being « related to » other phenomena kinship itself constitutes a separate domain. The English language gets them into this, despite efforts to argue the contrary. So how can one avoid being left with the compartments one begins with ? A preliminary conclusion is that anthropologists probably need to go on doing what they have always done, and supplement what they mean through diverse imaginative devices.

Motherhood and Fatherhood

13The propensity to divide mother’s kin from father’s kin, a general feature of kinship systems in Melanesia, springs as much from the marriage that brought them together as from distinctness of parental origins. The divisions may be depicted as two « sides » (of persons, of relations). The theoretical interest of this lies in the generalised sense in which all those to whom someone is related through mother or father share qualities in which the relevant parent also participates. To refer to these qualities in generic terms, it is helpful to take from English concepts indicating general states of being : motherhood or fatherhood. I have not started with a discretely conceived « mother » or « father ». Rather, thinking of a mother’s role, say, as part of a larger category of motherhood is a concession to those systems of nomenclature where we would not wish to prejudge whether the one through whom ties are traced is taken as the source of the qualities in question or as an exemplar of them.

14There is another reason for starting, in the Melanesian case, with a division between kin. It encapsulates the constant splitting and combining of social identities across the generations. Splitting and combining also accompany the English vocabulary of relations. While, as we have seen, one may differentiate relations of identification from relations of linking or connecting (the connecting of separate elements), taken together these may also be understood as types of – or subsumed under – connection. At the same time, connection can be opposed to separation or differentiation, and frequently is in English idiom, so that a formal definition of a relation as (a combination of) connection and separation goes against the thrust of vernacular usage. The formal usage commends itself to kinship analysis, for it often seems to be echoed in the way in which people arrange their affairs. All this has been said many times (myself included, Strathern 1988 : ch. VIII), but is repeated here since, in approaching motherhood and fatherhood, we shall encounter some very deliberate strategies of differentiation and separation (of persons, of relations). These at once lend credence to the formal position and, it has to be admitted, often seem in excess of it.

  • 6 As « context » can « get in the way » (Tsing 2009 : 47).

15Motherhood and fatherhood are clearly implicated in one another. The existence of each is predicated on them both existing. But what does one need to know to call this a relation ? And what kind of relation might it be ? Briefly put, how is one to understand all the work that people do in formulating their own descriptions of connection and separation ? Might the way anthropologists draw on relations get in the way6 of that understanding ? I stay with Melanesia in this (and longest) section, before coming to Euro-American ideas as exemplified in English kinship. It should be clear that this is an address to languages of description and not an ethnographic exercise.

Describing Relations Through Relations

  • 7 See James Weiner (1988) for a fully worked out account.

16We have already met a form of motherhood in the Korowai « mother’s brother ». More or less since kinship studies began, anthropologists have sprung on English language users (among others) the consequences of different modes of kin reckoning. Think of « male mothers » and « classificatory fathers ». Such descriptives invariably guide the analyst to relations that appear to depend either on a mediating link, such as mother in the case of mother’s brother, or on a relation of identity, as might hold between a father and his brother or son. I shall return to the point that one kin relation is not just categorically distinct from another but frequently comes into being performatively, through active differentiation7. Melanesian motherhood, and the same is true of fatherhood, often implies or contains within it two kinds of relations with siblings (same sex and opposite sex), that show up in terminological or other consequences for their respective offspring. The way mothers relate to their siblings’ children is bound to the way they relate to their own children. In other words, what it means to be a mother (father) is implicated in what it means to be a brother (sister). Each, too, is a condition for the existence of the other. Add generational time, and it becomes evident that it is in their mutual relationship that brother and sister each anticipate their own and the other’s fatherhood and motherhood. In this configuration, it has long been recognised, the separation of sister from brother is key to the connection of mother with father. This is at once a social and a conceptual proposition.

17In the language of relations one could say that relations appear to be described by (other) relations. Of course specific configurations of relations typify specific systems, but it was something like this that caught the anthropological imagination in its concern for many years with marriage exchange. An example from West Papua that was taken up when this debate was current in British and American anthropology comes from the Tor area described by Gottfried Oosterwal (1961) and was put forward as evincing the « simplest » type of system (Rubel & Rosman 1978). In condensed form, one sibling’s marriage is described through that of his or her cross-sex other. Tor practice obviously has its own particularities, however, and I preface the account with a brief comparative observation.

18If separation of maternal from paternal origin is a prerequisite for connection, it has also long been recognised (even if not articulated quite like this) that separation must be reproduced. And to reproduce separation over time, past ties may have to be cut to make possibilities for future ones. Speaking of Mekeo (Papua New Guinea), Mark Mosko (1983) described such cutting as « de-conception » : ancestral unions are un-done so that two sides once bound to each other are no longer bound, and their descendants can intermarry again. The un-doing in Mekeo is achieved through reversing or annulling the original marriage payments that made it possible for the fertility of one set of kin to contribute to that of another.

  • 8 Whether marriage is feasible with parallel cousins varies between the thirteen Tor tribes, though a (...)
  • 9 See James Weiner (1988 : 77) on Foi, where it is men’s responsibility to maintain conceptual distin (...)

19Tor people (as when Oosterwal described them) achieve separations through stratagems that on the face of it seem counter-intuitive : far from cutting relations, they duplicate them. Counter-intuitive also in that these tiny populations do not separate themselves through group ideology or seek spouses from afar – they marry largely within the immediate village (« tribe ») – and there is a marked bilaterality in kin reckoning, with minimal terminological differentiation of mother’s and father’s sides. The ideal marriage, that of a brother-sister pair marrying a brother-sister pair, is from a man’s point of view sister-exchange. (It is ethnographically apposite to refer to men’s need for wives, for nearly half the men fail to find someone to marry.) Until a woman has been returned for his wife, a husband is in debt to his wife’s elder brother ; indeed, an individual man can only marry if he has a younger sister or relative to bestow (Oosterwal 1961 : 103). In the next generation, the children of these two unions may be marriage partners themselves8. Cousin twice over, a woman’s husband may be at once her maternal uncle’s son (MBS) and the son of her paternal aunt (FZS). Through whichever parent the relation is traced, the cousin is encountered as an outcome of the separate marriages of brother and sister. Can we describe this as separation duplicated ? If so, this form of sister exchange detaches sister from brother through the pivotal creation of the brother-in-law (WB/ZH). Separation of motherhood from fatherhood appears contingent on men’s agency9.

20Perhaps this is not so far off what Tor men themselves describe. A ritual distinction between father and mother’s brother creates in another register a division between connections through men and through women, for brothers-in-law treat their respective children differently. A boy’s mother’s brother is responsible for his first initiation. At this the boy is given an initiation flute by his father, in effect a gift to the boy and his MB, who together form a « flute line » (brothers-in-law give to one another’s flute-owning lines). Giving a flute to the son does not identify father and son ; rather, the gift emphatically differentiates them, for the father belongs to the flute-line of his nephew (ZS) (Ibid. : 235). In effect, the father is describing his son as his brother-in-law’s nephew. The boy, thereby seen from two perspectives, reflects back the distinction between motherhood (MB) and fatherhood (F) among men.

  • 10 « [M]arriage by exchange always implies a double exchange » (Oosterwal 1961 : 235). The ideal marri (...)

21Women, absent from the flute-blowing rituals, are crucial to this relation between brothers-in-law. The dependency of men on women, for procreation, for food (also in short supply ; women are the sole providers of sago), is a theme of Oosterwal’s monograph. Through the flutes, men of the Tor repeat all over again the fact that dependency on women entails interdependency between men10. Flutes cause things to grow, and men say that without flutes the sago on which they live would die. The flute is a giver of life, just as a bride is. But the one cannot ordinarily substitute for the other (see footnote 10) : brothers-in-law give different life-giving powers to their offspring because men need both. A boy’s relations with his father (from whom a flute comes) and his mother’s brother (from whom a wife is seen to come) reiterate the prior distinction between wifetaker and wife-giver that foreshadowed his own existence. Divisions between persons are in this sense performative, in that they bring about the conditions for procreation and growth.

What is Concealed in this Description ?

22Within the description just given are elements of many anthropological conversations, including engagement with kin terminology, exchange relations, and attempts at cross-cultural analyses of kinship, both formally and – as in Sahlins’s case – substantively. They are not rehearsed here. I should, however, say that the present focus on motherhood and fatherhood was inspired by the formal elucidation of a cross-cultural model presented by Michael Houseman (1988) a quarter of a century ago. To return to his argument is beyond the scope of this chapter ; I disembed from its theoretical location there his linking (relating) of two concepts, recursiveness and indeterminacy. The link fleshes out what might lie behind the impetus to describe or duplicate relations through relations.

23In a comparison of two West African tales about marriage and procreation, Houseman demonstrates an asymmetry between fatherhood and motherhood in each case, in that one (not both) of the parental figures is shown to have an indeterminate character that is only resolved through recursive reference to other figures. Fatherhood in one case and motherhood in the other, the reference is « to higher-order (and temporally prior) versions of themselves » (1988 : 670). Procreation cannot ensue without the parent’s capacity being thus validated. For men of the Tor perhaps there is something like indeterminacy in sister exchange. It would rest in how a sister should be defined, in the uncertain lien men have on female relatives to give in return for wives, such that the claim (the « ideal ») always has to be predicated on other versions of itself. For it is « sisterexchange » that appears to be repeated, not cousin marriage. These versions encompass, then, both prior enactments of sister exchange and the possibility of future substitutions. A sister’s daughter will do if an unmarried man has no sister to bestow. Pragmatically this may bring a mother’s brother and his nephew into conflict ; formally it is a version of sister exchange that ignores generational difference but keeps the difference between father and maternal uncle (a father can never bestow a daughter in this way [Oosterwal 1961 : 103]). The indeterminacy of sister exchange renders contingent all the benefits men can receive from a fertile union. Asymmetry is evident. The sister/wife remains outside this activity : it is men who have to work to ensure they have sisters and wives.

24We begin to gain an insight into what might be concealed in descriptions of « relations » as the subject of reduplication or recursive referencing. Or, better put, we have to understand what relations under such conditions are. If there is a strongly performative element in iterations of the appropriate alignment of relationships, this needs to be brought out in the anthropologist’s description too. Michael Houseman (1988 : 672) remarks of one aspect of his model that the way indeterminacy is resolved never reaches an interpretation of relations between motherhood and fatherhood that is conclusive or complete : we may say of Tor that diverse interpretive possibilities as to how a man finds a « sister » to bestow always remain. Maybe recursion and the reproduction of indeterminacy serve local ends, likely to slip out of sight when the (anthropologists’) descriptive focus is on relations that evince « kinship ». Maybe people’s overt concerns with propagation are part of the performance. In Melanesia these often appear as an obsession with growth.

  • 11 His FZDD, who is of the « stream of blood » of his FM (Rio 2007 : 90). This is of course a highly t (...)

25If Tor kinship supposedly exemplifies a simple system, that of Ambrym, from Papua New Guinea’s neighbour, Vanuatu, has canonically been taken as the very opposite. Knut Rio’s modern ethnography clears away the confusion. What it shows, repeated in many registers, is the way in which women « circulate in marriage in order to produce […] [an] essential shift of generations between father and son » (2007 : 90). Women who leave their father’s place as sisters come back, after two generations, as wives, a process concealed by the image of paternal kin as « layers of men » in one place. The image is striking : it is a man not a woman who is seen standing in the house door (Ibid. : 118). But how is a man to be replaced by his son ? The indeterminacy of this reproductive ambition, we might say, requires the son enacting a version of what produced the father : in the « exchange of mothers » between them, the father marries the son’s mother and the son marries the father’s « mother »11.

26This « ideology of mother exchange » precipitates a « circle of women », invisible from the house door, but evident from a point that visualises past and future marriages simultaneously. Knut Rio talks of a recycling of relations ; Ambrym men point to yams (Rio links the two as « circles of production » [Ibid. : 102]). Yams grow on their mounds in a circle, though that becomes hard to discern in the mass of mature vegetation. The end in sight is growth (a yam vine « grows into a circle and “comes back” in order to produce growth [underground] » [Ibid. : 119]). This requires putting out of sight (in the ground) and, I extrapolate, rendering uncertain just how the old yam that is planted will have produced new bodies.

  • 12 « The feeding of cooked yams to family members […] recapitulates the conversions of harvested yams (...)

27Discerning human propagation in the yam garden is a familiar theme in the Trobriand Islands, ancient and modern. Mark Mosko refers to present times in elucidating the « fractal recursions » of three points, base, body and tip, in the embodiment of diverse activities. He starts with the reproduction of yams, although he could as well have started with the notion of lives succeeding lives, with the nature of human sexual arousal or the child as a fractal recursion of its parent (2009 : 693-694)12. Parents do not just procreate, they also grow children, right up to the « tip » of human life in old age « when parents’ contributions to adult children emanate chiefly from their minds in the form of magical and other secret knowledge » (Ibid. : 693). Such growth is open to numerous possibilities of diversion ; it has no guarantee. We might conclude that growth is the iteration of conditions for growth.

28Finally, what was also missing, in the sense of not being explicit, from the description of Tor marriage relations was the extent of my re-description. Taking up new issues in attending to old material is propagation of a kind. And there are new purposes : thus I have introduced indeterminacy and recursion in order to lay specific grounds for approaching relations in an English context. Motherhood and fatherhood ? Let me come to the point, and to why this account positively requires « imaginative devices ».

The Tip of the Account13

  • 13 Angkaiyakmin stories return to their beginning – only after they have branched out and reached the (...)

29I reintroduce Angkaiyakmin in order to conclude with an observation made of them. They too have a tripartite image of the transformation (growth) of entities into one another, drawing for inspiration on trees as well as taro gardens. If the way connections are traced through men or women creates asymmetrical « sides », the father’s side with one kind of kin, the mother’s with more than one kind, it is a version of the contrast found in story-telling that goes straight for a while, like a tree trunk, then branches into diverse directions. This feature of narrative supplements the way in which bodies are seen as composed of bodies, for – as on the Trobriands – knowledge is part of what transpires between kin : « knowledge is composed through other people’s bodily resources » (Crook 2007 : 29).

  • 14 On the part of writer hopefully conveyed to reader (of the description).

30Now foregrounding the elucidation of relations serves the anthropologist’s own knowledge-seeking ends, insofar as relations between elements of an exposition have the virtue of showing how phenomena are linked, can be put into context, are functions of other phenomena, and so on. However, « kinship relations » in the Melanesian material considered here are often credited with performative effects or outcomes (reproduction, growth, emergence of [old] life [newly] embodied). The counterpart « performance » for the anthropologist engaged with description as a mode of knowledge-making is, simply, effective description. Exposition is not the only relational form available, however, and turning to the images and analogies by which people describe things does more than simply extend one’s subject matter. Sago-growing flutes from the Tor, an Ambrym man in the doorway, the base and tip of the Trobriand yam : these work as highly condensed moments of recognition, of identification14. So this is motherhood ! – or fatherhood ! They are openings to conceptual worlds. In these particular conceptual worlds, activations of motherhood and fatherhood may be distributed across diverse relations ; if they are not tethered to what in English appear as mothers and fathers, they are not tethered to a relation between themselves either.

31Tony Crook is clear that relations are central to anthropological discourse, and has a deft solution for how to push into the background a vocabulary based on them (see footnote 5). Entailed in this is the role of knowing in Angkaiyakmin kinship. When his Angkaiyakmin teacher in an imagistic gesture opened his thigh, the shiny membrane under the skin was understood to be « advice » or « knowledge ». So perhaps we had already met in Crook’s teacher a form of fatherhood, for a man’s main source of important knowledge is his father. There is an internal asymmetry. A father imagines all a son has to do is listen to what the father says over and again ; by contrast, what sons hear sounds incomplete, and to make things intelligible they branch out by adding what they learn from other sources, especially the mother’s brother (Crook 2007 : 115-116). This duplication gives rise to further uncertainty. Only healthy bodies (person, taro) give consistent evidence of knowledge’s shining effect. Knowledge shapes growing taro plants as particular kinds of talk shape kinship. For effective knowledge has an effective outcome : « [t]he point of knowing in Bolivip is growing » (Crook 1999 : 239).

Knowledge as Kinship

  • 15 For the sake of argument I draw on particular interpretations of « English kinship » presented in M (...)

32It might seem perverse to introduce, as a brief entry into English kinship15, an example of motherhood in the absence of fatherhood, but it has a certain ethnographic verity. In the context of concerns over genetically modified food, Cathrine Degnen interviewed mothers in northern England about what their children ate. Anxieties about the effects of food on children’s bodies were most acute for infants ; mothers would buy organic food (« as nature designed it ») for their babies even when they did not eat it themselves (2009 : 55). Of a recalcitrant child, a mother said : « it were such a worry for me because I wanted him […] to grow and develop and feed his brain […] and I were like “he’s not eating, he’s only eating custard creams [a sweet biscuit] !” » (Ibid. : 54). I have noted elsewhere the extent to which the challenge to a woman’s motherhood includes her responsibility for the knowledge she has of the world. Proper growth depends on that. A second mother : « [I]t used to strike me as a bit revolting when you’d see another young baby eating or chewing […] really sort of greasy bakery sausage roll things and I’d think [apropos the baby’s caretaker] “you don’t know what is in that !” » (Ibid. : 53). The point of knowing is growing, we (anthropologists) might almost say ; in fact we (English-speakers) do not say it like this.

  • 16 Cathrine Degnen (2009 : 55) asks a mother how her « relationship with » organic food developed.

33A mother applies her knowledge to the way she raises her child ; its source, and much of its authority, lies outside her relationship to the child or the child’s father – it is knowledge about child-raising. Childraising is a compartment (Dumont’s « sub-system » !) of knowledge about the world. Describing this16, the ethnographer might say the mother « relates » it to her situation. Indeed the conceptual world in which (as elsewhere) relations receive expression as « kinship » is also (here) to be understood as a world of knowledge-making ; in the hands of English-speakers relating is a significant epistemological device. We should ask what kind of world is thereby opened up to knowledge, that is, how the relating is done.

  • 17 On the contrary, « nature », in this context, is an object of constant recursive referencing : the (...)

34One prominent area in which knowledge would seem intrinsic to kin relations also renders it extrinsic ; this is the relation between motherhood and fatherhood as a matter of procreative capacity. The indeterminate nature of paternity means that how a father is known is crucial to his identity (though not to his fertility). In today’s modes of procreation such uncertainty can also affect women ; thus between commissioning mother and surrogate mother, the paramount question is not who is the surrogate, it is who is the « real » mother (whose role is closest to what we know motherhood really is). In neither case is indeterminacy resolved with recursive reference to other parental figures ; in any event it is impossible for generational time to go back on itself. Reference will be to what is known in other compartments of knowledge, for example in the one case about biology, in the other about the legality of a contract. This is not because recursion is unthinkable17 ; it is perhaps because of the way a (kin) relative must appear as a role-player, a person. We return to this shortly.

35Ignorance of paternity is expressed as an issue of knowledge in many cultural contexts. Everything turns, of course, on how knowledge behaves in the conceptual world in question. Here, ignorance is defined by practices of verification, and what is relevant is the shape given to truth by « relating ». Dividing knowledge (the world) into compartments serves at once as a powerful exemplar of how apparently separate entities may be brought into connection (including those of cause and effect) and, in the constant need to know, reinstates indeterminacy. (It was the genius of the scientific revolution to suffuse « the world » with uncertainty about what it was.) Thus desiring to know about a world conceived of as natural raises questions about nature ; pari passu, social and society. The contingent or indeterminate nature of a cosmos thus conceived is controlled through relating insofar as its particulate entities can be rendered intelligible in relation to one another ; once relations are established, fresh demonstrations of relations can be added. At the same time this conceptual world has a huge capacity for internal expansion, for every domain contains elements that can also be seen as parts of other domains. Switching viewpoints in this way give us what I have called merographic relations (Strathern 1992 : 73).

  • 18 It will be appreciated how abbreviated this account is. There are in the vernacular many kinds of k (...)
  • 19 David Schneider’s comments on the relative as a person, « playing a role in real life », remain per (...)
  • 20 Sandra Bamford (2009 : 170) enlarges on some of the relational consequences of genetic modification (...)

36So how are motherhood and fatherhood « related » in this cosmology ? And what kind of relation is it ? First, a relation of comparison. Insofar as both entail procreation and child-raising, motherhood and fatherhood are each instances of parenthood – in English, fathers are not another kind of mother ; they are another kind of parent. Second, not unconnected, merographic relations. Kinship evokes a role-playing or relational « part » of the individual person (concomitantly, kinship is thought of as a « part » of society). Whatever the effectiveness of knowledge18, in this context it is tied to relationships in a specific role-playing way. So motherhood and fatherhood exist as domains of interaction, which may or may not be compared, in which individual persons participate or through which their identity is defined19. Attributes are attached to the figures of mother and father before they appear as attributes of others : a maternal uncle may or may not evince « motherly » behaviour. Indeed, in general, keeping attributes distinct is a responsibility that not only mothers feeding their children may evince towards the world20. How people behave in relation to the expectations of their roles (Miller 2007) can be a focus of intense moral concern.

37And the outcome ? It is how persons are known as « mothers » and « fathers » that seems so important. The outcome of this knowledge is not procreation as such ; that has an autonomous trajectory. For, although knowledge is crucial to the flow of (moral/legal/social) identity that in this system comes with birth, this kind of relating does not enhance fertility or parturition. Nor, although the same flow of identity is crucial to the keeping of genealogies, does it effect the movement of persons through time and space ; among other things « genetics » determines distance of connection. Above all this kind of relating tethers motherhood and fatherhood to specific persons. Nothing here of the Melanesian recursiveness of kinship positions that leads, often in ways that would appear contradictory in a situation where identities are separated out by context or sub-system, to descriptions of relations through relations.

  • 21 I am trying to convey a folk construction ; it is of course tautologous.

38English motherhood and fatherhood are not divisions of the cosmos (rather, the cosmos is divided by knowledge of it) or even of the world of kin. They relate to each other as particulates through the comparison of particular characteristics and circumstances. These in turn can be related, merographically, to numerous contexts outside « kinship » in which motherhood and fatherhood are also embedded (nurture, education, the law, transmission of disease). There is an epistemological premise of concordance, or homology, between entities and the contexts in which they are found, and, correspondingly, « difference » across contexts21. However many internal differences, comparison always implies an overarching context that provides grounds for similarity – and does so for relations both of identification (« brotherhood is in itself the general form of the relation » [Viveiros de Castro 2004 : 18]) and of linking. In effect, one can only relate what is related. The work of relating includes the work of making explicit – making knowledge out of – relations that are already there. As to recursiveness, then, one is invariably led back to the constant referencing of knowledge to other orders of itself that relations enable.

  • 22 Obviously, insofar as differentiation (the individuality of entities) is taken for granted (Wagner (...)

39Knowledge-making includes gathering together and making distinctions in like measure. But might one hazard that the efficacy sought in descriptive relating – creating connections, telling a story – affects the overridingly positive gloss put on relating in the vernacular ? If so, anthropologists might in turn be alert to the effects on their descriptions of a language derived from a cultural order where relationality is first and foremost connection (of disparate entities). There is an illustration to hand. « Differentiated by clanship and place-ownership, nonetheless, as a nephew said, squeezing his own thigh » ; writing in English I had to insert the qualifying « nonetheless » (that is, despite distance, the two [MB and ZS] are close), or else it would have read oddly. However many times the (Anglophone) anthropologist insists on separation and differentiation as equally intrinsic to the act of relating, the language pulls away. And of all concepts, kinship (« families and relatives ») is somehow connection first and differentiation second22. It takes great effort, as Rupert Stasch attests, to pull the language around for use elsewhere. Idiomatic usage that privileges connection is like the usage that understands sociality as sociability or mutuality as benign before it is malevolent.


40This re-description of mine dilates on what Roy Wagner (1977) expressed so succinctly in the opening pages of « Analogic Kinship », though with a particular interest in knowledge-making as entailed in description. A bit mischievously perhaps, the title for the previous section was a phrase deployed by James Leach (2009) for Reite, Papua New Guinea. Knowledge there, he says, connects people as a form of kinship. So close to what we might say of English kinship thinking, and so far from it. English kinsfolk are preoccupied with the potential enlargement of anything they know about kin relations by anything they know about the world. The substantive contrast between these contexts provides a reason for considering the kinship of motherhood and fatherhood according to the conceptual worlds they evoke, and I have taken ones in which « relations » have distinct work to do. Relating cannot be the same kind of device in the two locations, nor have the same kinds of effect.

  • * This article originated in an invitation from Laurent Barry, Klaus Hamberger and Michael Houseman t (...)

41In the former, the relation between mother and father, that might be imagined as the starting point for thinking about motherhood and fatherhood, is more outcome than premise. It is an ongoing outcome of numerous relations, of which the conjoined separations of brother/sister and husband/wife have been touched upon. Constant referencing of other relations creates the condition for iterations of a performative kind. In the latter, mothers and fathers exist as individual persons, and anything to be asked about the relation between motherhood and fatherhood depends on the context of the question, including how they are implicated in one another. It is specific compartments of human life that are aligned with, related to, one another, just as a mother relates her responsibility for child-raising to how she shops*.

Haut de page


Bamford, Sandra, 2009 « “Family Trees” Among the Kamea of Papua New Guinea : A Non-Genealogical Approach to Imagining Relatedness », in Sandra Bamford & James Leach, eds, Kinship and Beyond. The Genealogical Model Reconsidered. New York, Berghahn Books : 159-174.

Corsín Jiménez, Alberto, [2004] The Form of the Relation, or Anthropology’s Enchantment with the Algebraic Imagination. Unpublished MS, University of Manchester.

Crook,Tony, 1999 « Growing Knowledge in Bolivip, Papua New Guinea », Oceania 69 (4) : 225-242.

Crook,Tony, 2007 Exchanging Skin. Anthropological Knowledge, Secrecy and Bolivip, Papua New Guinea. Oxford-New York, Oxford University Press.

Degnen, Cathrine, 2009 « Eating Genes and Raising People : Kinship Thinking and Genetically Modified Food in the North of England », in Jeanette Edwards & Carles Salazar, eds, European Kinship in the Age of Biotechnology. New York, Berghahn Books : 45-63.

Descombes, Vincent, 2000 « The Philosophy of Collective Representations (transl. by A. C. Pugh) », History of the Human Sciences 13 : 37-49.

Dumont, Louis, 2006 [1971] Introduction to Two Theories of Social Anthropology. Descent Groups and Marriage Alliance. Ed. and transl. by Robert Parkin. New York, Berghahn Books.

Edwards, Jeanette, 2009 « The Matter in Kinship », in Jeanette Edwards & Carles Salazar, eds, European Kinship in the Age of Biotechnology. New York, Berghahn Books : 1-18.

Faubion, James D., ed., 2001 The Ethics of Kinship. Ethnographic Inquiries. Lanham, Rowman & Littlefield.

Houseman, Michael, 1988 « Towards a Complex Model of Parenthood : Two African Tales », American Ethnologist 15 (4) : 658-677.

Leach, James, 2009 « Knowledge as Kinship : Mutable Essence and the Significance of Transmission on the Rai Coast of Papua New Guinea », in Sandra Bamford & James Leach, eds, Kinship and Beyond. The Genealogical Model Reconsidered. New York, Berghahn Books : 175-192.

Miller, Daniel, 2007 « What is a Relationship ? Kinship as Negotiated Experience », Ethnos 72 (4) : 535-554.

Mosko, Mark S., 1983 « Conception, De-Conception and Social Structure in Bush Mekeo Culture », Mankind 14 (1) : 24-32.

Mosko, Mark S., 2009 « The Fractal Yam : Botanical Imagery and Human Agency in the Trobriands », Journal of the Royal Anthropological Institute 15 (4) : 679-700.

Oosterwal, Gottfried, 1961 People of the Tor. A Cultural-Anthropological Study of the Tribes of the Tor Territory (Northern Netherlands New-Guinea). Assen, Van Gorcum.

Rio, Knut M., 2007 The Power of Perspective. Social Ontology and Agency on Ambrym Island, Vanuatu. New York, Berghahn Books.

Rubel, Paula G., & Abraham Rosman, 1978 Your Own Pigs You May Not Eat. A Comparative Study of New Guinea Societies. Chicago, University of Chicago Press.

Sahlins, Marshall, 2011 « What Kinship Is », Journal of the Royal Anthropological Institute 17 (1) : 2-19 (part one) ; 17 (2) : 227-242 (part two).

Schneider, David M., 1968 American Kinship. A Cultural Account. Englewood Cliffs, Prentice-Hall.

Stasch, Rupert, 2009 Society of Others. Kinship and Mourning in a West Papuan Place. Berkeley, University of California Press.

Strathern, Marilyn, 1988 The Gender of the Gift. Problems with Women and Problems with Society in Melanesia. Berkeley, University of California Press.

Strathern, Marilyn, 1992 After Nature. English Kinship in the Late Twentieth Century. Cambridge-New York, Cambridge University Press.

Strathern, Marilyn, 2005 Kinship, Law and the Unexpected. Relatives Are Always a Surprise. New York, Cambridge University Press.

Tsing, Anna Lowenhaupt, 2009 « Adat/Indigenous : Indigeneity in Motion », in Carol Gluck & Anna Lowenhaupt Tsing, eds, Words in Motion. Toward a Global Lexicon. Durham, Duke University Press : 40-66.

Viveiros de Castro, Eduardo, 2004 « Perspectival Anthropology and the Method of Controlled Equivocation », Tipití 2 (1) : 3-22.

Wagner, Roy, 1977 « Analogic Kinship : A Daribi Example », American Ethnologist 4 (4) : 623-642.

Weiner, James F., 1988 The Heart of the Pearl Shell. The Mythological Dimension of Foi Sociality. Berkeley, University of California Press.

Haut de page


1 That is, not mediated by the act of naming social relations, relationality is simply assumed.

2 Relations as used between phenomena, social or natural, do not have their basis in the kin usage ; on the contrary, the kin usage arguably has its historical precedent in the narration (the « relating ») of connections between phenomena. In any case relations between persons do not necessarily mean between kinspersons ; see Vincent Descombes on « concepts and social relations » as « two sides of the same coin » (2000 : 41).

3 Kinship does not disappear when it is perceived as a dimension of cultural order ; rather, it is made to appear, as in Marshall Sahlins’s gift of a definition. To some extent the text that follows tells its own tale ; thus, alerted to the term, the reader will encounter manifold usages of the concept of relation in this chapter.

4 Although she sees « worlding » as a proposition about context, in its expanded form it refers to « world-making dreams and practices » (Tsing 2009 : 62).

5 With a different focus, this is what Tony Crook’s volume (2007) takes seriously ; its intriguing outcome is best encountered in the text itself.

6 As « context » can « get in the way » (Tsing 2009 : 47).

7 See James Weiner (1988) for a fully worked out account.

8 Whether marriage is feasible with parallel cousins varies between the thirteen Tor tribes, though all allow marriage with cross cousins (« bilateral cross cousin marriage »). Where parallel cousin marriage is forbidden, such cousins may find they can trace their connections in other ways (Oosterwal 1961 : 106). Oosterwal was also interested in the variable emphasis that kin terminologies put on bilaterality, distinctions ignored here.

9 See James Weiner (1988 : 77) on Foi, where it is men’s responsibility to maintain conceptual distinction between the sexes.

10 « [M]arriage by exchange always implies a double exchange » (Oosterwal 1961 : 235). The ideal marriage entails both the direct exchange of sisters (within one generation) and the indirect exchange of flutes (across generations). The father’s gift is said to compensate his brother-in-law’s line for the loss of the sister’s life-giving powers, that is, detaches her from it. In rare circumstances a flute may be given, in lieu of a sister, for a bride.

11 His FZDD, who is of the « stream of blood » of his FM (Rio 2007 : 90). This is of course a highly truncated account, adapted for my purposes to make concrete the interweaving of human and vegetative propagation ; it is as « mothers » that men plant a father-yam that will bear a son-yam (Ibid. : 117-118).

12 « The feeding of cooked yams to family members […] recapitulates the conversions of harvested yams into […] garden heaps, storage houses, [and through exchange] male wealth, banana-leaf bundles, and so on. In all these transformations, fractal recursions of body, base, and tip unfold as life-stages culminating in death, fruit, and new life sequences » (Mosko 2009 : 693).

13 Angkaiyakmin stories return to their beginning – only after they have branched out and reached the crown [as in a tree] is the base revealed (Crook 1999 : 229).

14 On the part of writer hopefully conveyed to reader (of the description).

15 For the sake of argument I draw on particular interpretations of « English kinship » presented in Marilyn Strathern (1992, 2005), and elsewhere, to which I only once make separate reference. On « Euro-American » kinship thinking, see Jeanette Edwards’ pertinent discussion (2009).

16 Cathrine Degnen (2009 : 55) asks a mother how her « relationship with » organic food developed.

17 On the contrary, « nature », in this context, is an object of constant recursive referencing : the infant close to nature that must be fed with natural food ; the real (natural) parent referred back to natural (real) behaviour or attributes.

18 It will be appreciated how abbreviated this account is. There are in the vernacular many kinds of knowledge, including « emotional knowledge », that would qualify the depiction here.

19 David Schneider’s comments on the relative as a person, « playing a role in real life », remain pertinent : each of the « different elements which are combined into the definition of a person [comes] from its own domain » (1968 : 57-58).

20 Sandra Bamford (2009 : 170) enlarges on some of the relational consequences of genetic modification. (It will be appreciated that keeping attributes distinct is different from being responsible for bringing distinctions into being ; see footnote 9.)

21 I am trying to convey a folk construction ; it is of course tautologous.

22 Obviously, insofar as differentiation (the individuality of entities) is taken for granted (Wagner 1977 ; Weiner 1988 ; Viveiros de Castro 2004).

* This article originated in an invitation from Laurent Barry, Klaus Hamberger and Michael Houseman to elucidate contemporary perspectives on kinship through the presentation of a specific approach ; they suggested it might be interesting to take up the topic of motherhood/ fatherhood within the framework of the organization of kinship relations in general, in which a Melanesia/Euro-American divide could provide the springboard for an overall cross-cultural hypothesis. I am grateful to them for the impetus. A version of this article was presented to colleagues at the University of Toronto, for whose stimulating comments I am very grateful, as I am for an earlier conversation with Michael Houseman..

Haut de page

Pour citer cet article

Référence papier

Marilyn Strathern, « Kinship as a Relation », L’Homme, 210 | 2014, 43-61.

Référence électronique

Marilyn Strathern, « Kinship as a Relation », L’Homme [En ligne], 210 | 2014, mis en ligne le 16 mai 2016, consulté le 20 janvier 2018. URL : ; DOI : 10.4000/lhomme.23542

Haut de page


Marilyn Strathern

University of Cambridge, Cambridge (England)

Haut de page

Droits d’auteur

© École des hautes études en sciences sociales

Haut de page
  • OpenEdition Journals