What is kinship?

This work is licensed under the Creative Commons | © Stephan Feuchtwang. Attribution-NonCommercial-NoDerivs 3.0 Unported. ISSN 2049-1115 (Online). DOI: http://dx.doi.org/10.14318/hau3.2.016

Book Symposium

What is kinship?

Comment on SAHLINS, Marshall. 2013. What kinship is—and is not. Chicago: University of Chicago Press.

Stephan FEUCHTWANG, London School of Economics


True to magnificent form, Marshall Sahlins’ short book is a polemic against biological reductionism and economic individualism, but also a substantial argument in its own right. I cannot hope to respond in kind, having neither the wit nor the compendious authority of its ethnographic evidence. Instead, I will tell something of the struggle I have experienced in making clear to myself what his argument is. After one reading I still wanted to know what kinship actually is for Sahlins. Writing this commentary is an attempt to work out an answer, with occasional reference to kinship in the ethnographic region with which I am most familiar: China.

I have read and re-read the book because, while easy and a pleasure to read, I still had to sort out what kind of mutual being kinship is. I think the following facts from the ethnography and history of China (and of course other ethnographic regions, especially those with elaborate states) may well prompt the same question.

Chinese cosmology is not a cosmology of kinship, though patrilineal ancestors and affinal relationships play a big part in Chinese rituals, particularly in the always crucial death rituals. Relations with gods, who are cultural heroes and legends, historical and mythical at once, are equally (if not more) important than relations of kinship and affinity in China. They can, as named figures, be taken and treated as ancestors of those with the same name and from the same place of putative origin. But for the great majority of those who address them, they are gods and not ancestors. Further, relations establishing friendship and interpersonal trust, including affinity but also beyond matrilateral and same-generation conjugal kinship, are hugely important and are understood in some fundamental way to differ from “kinship.” Yet all these are consolidated by actions creating mutual being, including those of great intimacy, the sharing of suffering, of food, of touch, and of blood in blood brotherhoods. The language of kinship is used to describe the closest and best maintained of these relations of “brotherhood” and “sisterhood”— what anthropologists as outsiders used to call “pseudo-kinship.” The emperor was designated the Son of Heaven (Tianzi), and loyalty (zhong) was paired with filial duty (xiao) in the ethics of propriety learned by the emperor’s subjects. But filial duty to kin sometimes superseded loyalty to the imperial state and was, in any case, distinguished from loyalty as an analogue, just as friendship was knowingly treated separately or as an analogue with kinship. Without states, kinship is about what you do with strangers. With states, strangers are co-citizens, and many people exist who are neither strangers nor kin—friends, business associates, neighbors.1

So, I still ask: what is kinship?

It cannot just be mutuality of being. Yet, Sahlins extends mutuality of being to great scales of inclusion. Mutual being, at its most extensive, is for Sahlins either cosmological or human. Let me take the cosmological first. Through Eduardo Viveiros de Castro’s triadic synthesis of kinship, magic, and gift exchange, mutuality of being is bound into one ontology that is a universe. Through Philippe Descola’s ethnography of the Achuar, mutuality of being extends to other intentional things that the Achuar silently invoke with their desires (2013: 58–60). Chinese personifications of the forms of the universe with its encompassing Tian (Heaven) are also invoked for the communication of desires, not by silent song but by pledges and promises accompanied by food and other offerings, particularly incense and a text, both burned as an act of transmission. As I have already said, these personifications are not ancestors. The Chinese equivalent to what Sahlins calls animism is a conception of all things, celestial and terrestrial, as forms of the circulation, constructive or destructive, of vital energies and the secular as well as cyclical changes they bring about. The whole cosmos is animated, including human bodies and minds, internal selves and other internal selves, and beings differentiated from the human. By centering activities, which include disciplines of eating, sex and other exercises, and of ritual—always including the offering of food, drink, flame, and incense—humans mediate the energy circuits inside and outside their selves, celestial and terrestrial (Feuchtwang, forthcoming). So, where does mutual being cease being kinship? Or, put positively, what kind of mutual being is kinship?

Let us turn now to the human by way of Sahlins’ “Parenthesis on human nature” (Sahlins 2013: 37–44). Shared intentionality has been demonstrated by neuroand cognitive psychology to distinguish the human being from other species’ being. Maurice Bloch, for example, has been writing about this from his chosen scientific experiments; Marshall Sahlins has chosen his own. As Bloch (2007; 2012, see especially ch. 8) puts it, our human capacity to enter each others’ minds and accept third-party authoritative, shared knowledge—through speech or other media—is the most general way in which we go into and out of each others’ bodies. Again, I must ask where does the interpenetration of selves, or mutuality of being, stop being kinship?

Marshall Sahlins’ answer is, where a cultural transmission says it ceases, literally

by certain conventions of the use of the possessive first person pronoun (though, again, other uses go beyond kinship); kinship is the formation of differentiations of this mutual identification. On the one hand there are differentiations of kind and, on the other, differentiations or gradations of distance (87). All kinship systems, including Chinese patriliny, ascribe what counts as close, but they also accept selection and to some extent creation—through fostering and adoption, for example—of close kinship by actions of intimate care and mutual obligation, and the making of this “kind” by affinal, conjugal, and matrilateral relations. They also distinguish the substances of kind and closeness from other substances of mutual being, or at least the intensity of the sharing of the same substances.

So, kinship is mutuality of being but mutuality of being is not co-extensive with kinship. There are differentiations that mark out kinship as a distinct mutuality of being. For instance, in Chinese death rituals and other rites in which “ghosts” are invoked and fed, there is a sharp differentiation between the ghosts of forgotten ancestors and the ghosts of strangers, though food is shared with both. At the same time, there is a sharp ritual differentiation of ghosts from gods; ghosts are low, gods are high. But hospitality to gods and to the ghosts of strangers is hospitality to those who are outside, where the inside is kin—including affinal kin, who are differentiated and particularly honored in death rituals—and the surnamed but forgotten ghosts who are ancestors.

Affines are kin made by the intimate bodily interpenetration of selves, that is sex and procreation, even when its other side is that wife-takers are also raiders. Nothing controversial there. But it does lead to a question beyond Sahlins’ and Bloch’s undoubtedly different answers to what kinship is. Is there a peculiar intimacy that differentiates kinship from other relations?

The answer could be (though not exactly or explicitly offered by Marshall Sahlins) that kinship is mutual possession as prescribed and created through naming—to distinguish it from other forms of ‘I’dentification and ‘you’dentification. That is one way to narrow down Sahlins’ formulation: “for understanding kinship, much is gained by privileging intersubjective being over the singular person as the composite site of multiple others . . . a culturally relative hypostasis of common being” (28). Kinship is mutual possession that creates an inside, as expressed in the Chinese formulation zijiren (“our people”) when it is also the sharing of a family name, plus the kin made by marriage (in Chinese, qinqi). But that is not enough. On its own it would over-stress the purely formal and ascribed designation. In addition, following Sahlins but maybe again forcing him to be more exact than he wants, the activity of mutual being that is kinship could be specified as a special quality of creating an inside, namely an intimacy of participation of being. It is one of care but also of intense rivalry and, where witchcraft exists, of harm. It is an intensity of ambivalence in mutual possession, which is acted out in death rituals, during which affines separate the living kin from the dead first by their participation (as among the Gawa), and then by their own actual ritual departure, which is followed by the mourners being separated by descent from their possessed and possessing dead. In China, both agnates and affines are inscribed as an inside because they are named and thus distinguished from the anonymous dead— who are pitiable ghosts that must be fed and then expelled again—while another rite releases the living from their dead so that they can return to life and turn their dead into ancestors.

If I may allow Chinese ethnography to further help me provide an answer: Kinship is that particularly vivid mutuality and moral answerability created in the care of children and the dependency between children and their nurturers, reversed when the children are adult and the nurturers are old. I am combining the work of Maurice Freedman (1979, see especially part four) on asymmetrical lineage segmentation as a creation of closeness in shared inheritance of a name, on family as a fraught unity, and on marriage as the creation of matrilateral kinship with Charles Stafford (2000) on the cycle of nurturance.

So, kinship is not just mutuality of being. It is a particularly intimate sharing of name and dependency, and a ritual definition of an inside in which the mutuality of self is both prescribed and performed. I would add that kinship creates the most formative of personal and learned memories because it is recreated in the most frequent physical and sensual interpenetration of selves. And, one last thing, kinship mutual being is distinguished by its own temporality: the temporality of generational reproduction. It is not necessarily a genealogy, but a reproduction of name as common existence in which the spatial unit of reproduction is domestic. States and hierarchical religions with central organizations of transmission introduce other kinds of temporality: historiographies and liturgical repetitions, calendars and almanacs. Kinship has its own temporality among these others.

One might remark: Isn’t all this obvious? But it does need to be stated without recourse to any bio-determinism and assumed ego-definition . . . and without cultural relativism, following Marshall Sahlins’ good example.


Bloch, Maurice. 2007. “Durkheimian anthropology and religion: Going in and out of each other’s bodies.” In Religion, anthropology, and cognitive science, edited by Harvey Whitehouse and James Laidlaw, 63–80. Durham, NC: Carolina Academic Press.

———. 2012. Anthropology and the cognitive challenge. Cambridge: Cambridge University Press.

Feuchtwang, Stephan. forthcoming. “Coordinates of body and place: Chinese practices of centring.” In Cosmologies: Making contemporary worlds, edited by Allen Abramson and Martin Holbraad. Manchester: Manchester University Press.

Freedman, Maurice. 1979. The study of Chinese society: Essays by Maurice Freedman. Selected and edited by G. William Skinner. Stanford: Stanford University Press.

Sahlins, Marshall. 2013. What kinship is—and is not. Chicago: University of Chicago Press.

Stafford, Charles. 2000. “Chinese patriliny and the cycles of yang and laiwang.” In Cultures of relatedness: New approaches to the study of kinship, edited by Janet Carsten, 35–54. Cambridge: Cambridge University Press.



Stephan Feuchtwang
Department of Anthropology
London School of Economics and Political Science
6th Floor, Old Building, Houghton Street
London WC2A 2AE
United Kingdom


1. Many thanks to Tom Boylston for this formulation and for many other suggestions.