Strathern: What is a parent?

This work is licensed under the Creative Commons | © Marilyn Strathern.

What is a parent?*

Marilyn Strathern, University of Cambridge

Drafted in 1991, this paper lies at an intersection of interests in debates following the UK Human Fertilization and Embryology Act (1990), academic discussions about interpretation, first sightings of fractal imagery, then recent modelings of Melanesian sociality, and in the role of knowledge in English kinship thinking. Inspired by questions about parenthood and kinship posed of West African materials by Houseman, it explores certain comparisons—and the possibility of comparison—between motherhood and fatherhood.

Keywords: origins, English kinship, New Reproductive Technologies, parenthood, metaphor, deconstruction, Melanesia, West Africa (Beti and Samo)

“Everything begins with reproduction.”
Derrida, Writing and difference, 1967 (quoted in Smith 1988: 45).

“Perhaps I can make the significance of this a little more clear by giving your lordships an analogy.”
Hapgood, Archbishop of York, House of Lords, December 1989 (quoted in Morgan and Lee 1991: 72).

I wish to bring together certain issues raised by anthropologists working in the French tradition as well as the British one, though that is not the reason for juxtaposing these remarks. They are offered as a commentary upon one grounding to current debates in the field of new reproductive technology. The ground is, not to be paradoxical, the realization that fundamental grounding facts of existence no longer appear the limiting conditions that they once did: “there never was an original ground or image in relation to which [a subject’s] boundaries could be established” (Hayles 1990: 218).

One could go supplementing such remarks: for each is posing a question about the absence of limits, and there seem no limiting conditions to the question itself. Rather, we simply confront the question in different forms, at particular moments (as we shall see: Derrida’s moment is the act of interpretation, the Archbishop’s the decision to which the House is to come, and Hayles, the recognition of what the person as subject has done to his or her own image). I therefore decide on the form of the question: what is a parent?

Obviously this is a question about reproduction. What is being reproduced when a person recognizes his or her parenthood in a child?

Let me start with some general Euro-American understandings of the terms themselves. Now whereas “child” notionally exists independently of its origins—one can be both a child and a child of someone—“parent” is a relative term (a “kinship” term) that is defined only with respect to another—one is parent to one’s offspring. Its orientation is prospective: to be a parent is to have a child. At the same time, parent actually signifies a retrospective state; to have a parent is to have had an origin in another person. What is signified is the prior (“independent”) existence of that person as producer or originator. This slippage suggests to me that, as a kin category, parent is quite different from any other category in the kinship repertoire. Aunt/nephew, sibling, even grandparent/children all function as reciprocal terms that recognize the mutuality of each kinperson recognizing the other. The same is true of mother/father in relation to son/daughter.

“Parent” and “child,” however, turn on an irreducible asymmetry. As we shall also see, the asymmetry recurs within the definition of parent itself when two aspects of parenthood are distinguished by sex. There is a difference between “mother” and “father” that is not replicated in that between daughter and son. That is one of the reasons for my interest. Moreover, the asymmetry is of an order comparable to that between (say) a person and a set of relationships, or individual and society. As I shall try to show, what is at stake are two different kinds of knowledge about connections between persons. To think of mothers and fathers and sons and daughters as “parents” and “children” summons certain specific ways of knowing.

Now it was the “recognition” of the uniqueness of parentage with respect to children which established the distinctiveness of descriptive kinship nomenclatures so-called (as Morgan discovered 130 years ago). I want to go further and argue that Euro-American conceptualizations of the kind of relationships that flow from parenthood do indeed mark off such kinship systems from certain others found in (say) Africa or Melanesia. And that it is not a question of terminology, though terminology indicates that question: what is at issue, though, is the kind of recognition that the terminology presupposes.

One of the questions that Euro-Americans, including late twentieth century people in Britain, pose to themselves is to whom the term should apply. How (on what grounds) should one recognize who is a parent? The New Reproductive Technologies (NRT) have introduced a need to be explicit about the grounds of knowledge, and thus “construct” the basis for the “recognition.” I shall have reason to return to the difference between these terms. In fact, the question belongs very much to the late twentieth century as far as Euro-Americans are concerned. More than a century ago, Morgan and others did indeed make explicit a fundamental grounding to our attention to parenthood in the recognition of natural processes of procreation. I suspect it is continuing to search for the validation of this ground that has taken the ground away (validation: trying to ground the grounds in further realities). In a very real sense, the cultural issue is how we make knowledge for ourselves.

The question about the grounds on which one should recognize a parent is fascinating as a question about itself. A parent is an origin point; to establish grounds is to explore the origins of one’s definition of parent, and to come on the knowledge as recognition implies that what is known has a prior existence that makes it the origin of the act of knowledge. If spelling this out introduces a sense of vertigo, I shall try to show how real that sense is to current Euro-American perceptions of what is at stake in kinship.

The Archbishop is not immune. He, too, is contextualized by the cultural practices of the late twentieth century, as is our kinship thinking. Derrida (deconstructive critical practice) and Hayles (who is discussing Baudrillard, critic of consumer culture) serve to triangulate the Archbishop’s remarks. I start with their context.

Receding mirrors

When Derrida says that everything begins with reproduction, he means that there is no ultimate origin that corresponds to the present thing, no originary text. The observation occurs in the context of his discussion of Freud (“the substitution of signifiers seems to be the essential activity of psychoanalytic interpretation,” cited in Smith 1988: 45).

Indeed, there can be no originary text. In this view, signifying events depend on differences, but these differences are themselves the products of events (Culler’s phrasing, 1979: 164). One forever shifts between perspectives: “When one focuses on events one is led to affirm the priority of differences, but when one focuses on differences, one sees their dependence on prior events.” Priority, or origin, the very signification of an autonomous point of reference, thus turns out to be relative to, or dependent on, its outcome. In the act of interpretation this means there is no stopping place (origin) in the concepts that present themselves—every signifier leads to (is substituted) by others. When what they will become is part of the understanding of them, they are constituted by the (as yet) absent others-to-be. For literary criticism, and historical analysis (Kramer 1989), this means that there is no context (reference point) that is independent of the text;1 or rather, that every context is another text.

However, this is also a limit of sorts. In commenting on the limitation Freud set himself (there is no original text to find that could be “metaphorically re-written in the preconsciousness or in consciousness itself”), Derrida in fact comments on his own discovery of limitations.2 The impossibility of locating origins means that interpretation stops when, in Smith’s words, a critical decision is taken. In other words, a limit is established in the recognition of a particular moment of decisiveness. The subject “recognizes” that he or she has (is at) a point of critique. In Smith’s view, the impossibility of finding origins is an enabling limitation to the act of interpretation, precisely in that it forces that sense of participation on the interpreter. It is he/she who constructs the critical moment, or recognizes it (“The interpretations select themselves,” cf. Smith 1988: 47).

The conscious text is thus a transcription, because there is no text present elsewhere as an unconscious one to be transposed… The unconscious text is already a weave of pure traces … a text nowhere present, consisting of archives which are always already transcriptions. Originary prints. Everything begins with reproduction. Always ready: repositories of a meaning which was never present, whose signified presence is always reconstituted by deferral. (Derrida [trans.] 1978: 211; emphasis removed)

The difference between the conscious and unconscious need not detain us here as anything more than a re-rendering of text and context: context can only be retrieved (so to speak) as text. Nature is fabricated in the perception of it. In short, “recognition” symbolizes the possibility of an original while rendering it relative to the fact and mechanics of symbolization.3

Hayles’ comment on original ground is made in the context of reflections on Wiener, the coiner of the term cybernetics (the science of communication and control). Insofar as cybernetics is concerned with the flow of information “through and around systems,” in Hayles’ words, it changes the way boundaries are perceived. Consider, she asks, “how a human being looks from a cybernetic point of view” (1990: 215). The operative unit is determined by the flow of information, such that the mechanisms through which we communicate become part of ourselves—are intrinsic to the activity of communication. Conversely, “the body is revealed as a construct” (218), is made through the communicative (symbolizing) acts. It is in this context that she introduces Baudrillard via Lacan.

Lacan is concerned with the moment at which the body’s boundaries are recognized. (The child enters the symbolic realm of language in order to compensate for the loss [absence] that its perception of its own separateness brings: its critical realization of its own image as an autonomous and bounded entity.) But imagine, says Hayles, what happens when the child that has seen itself in a mirror sees itself again. It can never again see “itself,” its original, for what it sees the second time is the image that it saw the first time in the mirror. The mirror now reflects “a cybernetic amalgam” (217). And why Baudrillard? Because if the “first” recognition of itself launches the child into the Lacanian project of the Symbolic order, then the “second” recognition launches the subject into what Baudrillard has called hyper-reality. “Already initiated into language and the linguistics of absence, the subject in this second moment of recognition realizes that there never was an original ground or image in relation to which her boundaries could be established. At that moment all boundaries become negotiable, at best no more than permeable membranes through which information flows in and out.” (217-8).

This galaxy of twentieth century names—Derrida, Lacan, Baudrillard—that evoke other names—Freud, Wiener—are here in my account simply to remind ourselves that we always live in a world already inhabited. There is nothing original to the problem of origins, and everything to suggest it is inherent in knowledge practices revealed by philosophy, psychoanalysis or whatever. The interesting phenomenon, of course, is the extent to which in the 1970s/1980s both critical theory and culture theory were swept off their feet by deconstructionism, despite its name a kind of synthesis of these possibilities. We “recognize” that there is no context that is not itself a text; no body that is not itself a construct; no organism that is not a bit of information; no origin that is not constituted by the backward glance. In fact, one could as well have looked to the new science (Best 1991) as to the new criticism.

Such figures have retrospectively become part of a flow of ideas that seem to eddy everywhere. It is the same question repeated: what is a parent? That is, out of what does one (thing) appear? The mirror makes apparent what is already there, but which could not “be” there until reflected back. Reality is shaped by the textual process that apprehends it. There are no points of origin that are not recreated in the present. And so on. At the moment at which representationalism “dies,” a kind of (re)creationism is reborn.

But there is a liberty enjoyed by the cultural critics that is not everywhere replicable. This lies in the essential atemporality of the text or vision or consciousness. Recognition implies a priority, but recognition constructed as a self-conscious act has the character of simultaneous realization, as the body and its image are simultaneously realized in the mirror reflection. Such reproduction does not require time. The dynamic is non-linear.4

Time, however, is held to be crucial to the perception of human reproduction as twentieth century Euro-Americans have it. What is true of their perception of biological process (“growth”) is also true of their perception of the developmental progression of generations. For all that twentieth century Euro-Americans fantasize about non-linear time, that belongs to an unlived world—a feature of science fiction or an attribute of the universe one can leave to physicists and mathematicians to model. For they have adhered to a reproductive model of time in their depiction of human relations—the manner in which one generation produces another follows in linear sequence. One can think of time going back on itself, but not relationships. And the sequencing of generations is fixed in the immutable direction of procreation: parents beget children.

Yet that grounding in natural process is already at odds with one cultural reality: that the parent does not exist before the child. In many parts of the world, the child is prefigured in the parent, the bride already pregnant so to speak, the initiate already a father. But in the kinds of kinship systems that have characterized twentieth century Euro-American culture, it is only in “having” the child that the parent is made.

Perhaps it is no accident, then, that in trying to come to grips with some of the implications of reproductive technology, the Archbishop of York should appeal to a non-linear dynamic (cf. Best 1991: 200). And evokes another popular hero, one who was already there, Mandelbrot.

Metaphors within metaphors

The statement that the Archbishop of York made to the Rouse of Lords in the course of debate, and that is reproduced by Morgan and Lee “as a metaphor [which could serve] not just for embryonic existence and human life but for the whole of the moral debates which these issues engendered” (1991: 72, my emphasis) appeals to analogy.5 That is, it appeals to a semantic domain—an aspect of chaos theory—which will act both as a reference point and as a vehicle or mechanism through which to make the subject of his discourse—embryo development—appear. In this construction, chaos theory is not under dispute and can thus be taken in and of itself. We might say that he stops the flow of analogy (interpreting one thing through another) by the very act of creating the particular analogy (producing the decisive parallel to what he wants to say). The act is deliberate, a “construction.” Here is the text.

By and large a biological approach to life is rooted in gradualism…. The same is true in the development of individual lives. They begin with chemistry and they reach their fulfillment in mystery…. Biologically speaking we are looking at a continuous process. Perhaps I can make the significance of this a little more clear by giving your lordships an analogy. Exactly ten years ago a mathematician called Mandelbrot first discovered what is now called the Mandelbrot set. It is a set of points which can be mapped out as a computer graphic to form the most amazing, beautiful and complex structure that it is possible to imagine. It is a picture of literally infinite depth. If one magnifies the details of any part of the picture, one finds that in them are whole worlds of further detail which are always beautiful, which never repeat themselves and which always reveal more and more detail, on and on, ad infinitum. How is the Mandelbrot set made? It is made by the use of an absurdly simple equation with only three terms. The secret lies in the process. It is a process whereby the answer to one use of the equation becomes the starting point for the next. In other words, it is a cumulative process, just like evolution in which one life form builds on another and just like embryology in which the development of one cell provides the context for the development of its neighbors and its successors. (House of Lords, Official Report, 7 December 1989, col. 1020).

On fastening on the point of process, the speaker is able to use a non-linear model to model what his listeners ordinarily think of as a linear progression. One life form building on another is simultaneously a depiction of progressive growth and a non-temporal displacement of one moment by the next. We might say that in building on another form, the new form is not seen simply to “extend” or “reproduce” it, but substitutes its own complexity for the form that preceded it. The new form, or signification, is both there and not there in its predecessor. This was the nub of the parliamentary debate.

It was a debate concerned at this point only indirectly with parenthood. its principal concern was with the disconcerting manner in which it was both possible and impossible to imagine a single process by which (pre-) embryonic cells “become” a human being and individual person. The interpretive impasse is one with which Derrida and others have now made us familiar. But it is glossed over in the apparently innocent question about when one can recognize in the natural form the beginning of the social one.

For the very notion that the one “grows” into the other obscures the displacement or switch between terminologies. In his description, the speaker moves between semantic domains—from the language of natural, biologically defined material to the language of cultural definitions of social entities. “What is happening embryologically,” the Archbishop said (quoted in Morgan and Lee 1991: 71), “is the creation of persons through a process, which although it begins with genetic union, is not simply about a union of genes but also depends on a certain cellular identity.” That identity is thus prior to and an origin of the “person” who eventually appears. But this construction of ideas about stages of growth is not at all constructed like Mandelbrot’s set.

I want to suggest that there is indeed something very interesting about the metaphor on which the Archbishop draws. But its interest for me lies in the manner in which the analogy was drawn. In fact, I want to suggest that the mathematical set is a very poor analogy for the processual growth under debate when that growth is perceived of as from cells to embryo to baby to person (all terms used in the legislative debates). It is the way he uses it (the kind of knowledge he offers) that provides the apt analogy. I would see an analogy for that Euro-American model of growth lying in the very manner in which the Archbishop brings together two semantic domains—chaos theory and embryology—that are of a quite different order. It is the relationship between the two domains that in my eyes models the relationship between cell and person. And, as in the way one grounds the other (as cellular identity in the statement above grounds the individual person as its origin), this is a construction whose pattern is repeated over and again, and is at the heart of Euro-American middle class models of the relationship between parent and child.

Although this need not detain us here, I would even turn the proposition around and say that the “partial” manner in which Euro-Americans deploy analogies—one which allows every parallel also to be situated asymmetrically as context and text, grounding proposition and dependent one—inheres in their kinship thinking. But let me give some substance to my earlier suggestion.

“What is happening” is not Mandelbrot’s set at all. That set emerges and re-emerges from the fractal realization of the “same” elements. In the same way as one text becomes a text (context) for another, or the construction of the body is seen through what constructs it, or interpretation works on what is already transcribed, the collection of points that constitutes a Mandelbrot set remains a collection of points. There is a continuity of “substance.” In these languages, one signifier substitutes for another signifier, one concept carries over or supplements another concept (as in Derrida’s notion of supplementation), in the same way as points reproduce points. The criticism and the mathematics both turn on revealing the homologies (displacement effects) between similarly placed parts that substitute for one another. Now in verbal or conceptual continuum between biological cell and social person, “cell” and “person” are similar signifiers or points and homologously similar to one another. Yet the purpose of Hapgood’s analogy was to reveal not the homology but the transformation of one into the other.

The very use of metaphor is regarded as transformative in this way. When the Archbishop drew his analogy with chaos theory, he was not, I think, intending to displace one signifier by another signifier—to substitute one analogy for another analogy and thus bring chaos theory and embryology into a homologous relation as well. Rather, his was the (representational, foundational) strategy of supposing that by his analogy he was illuminating something altogether different. Evolution and embryology have a real existence. They are simply made to appear to the imagination—are realized or recognized—in the analogy with the Mandelbrot set. They are transformed from their natural status to being “known” through an exercise of the imagination.

Now it is the “image” that the person has of him or herself in the mirror and the “image” that he/she has of him/herself that simultaneously displace one another in a cybernetic circuit. But the notion that cells grow into persons combines concepts from different domains, the difference being constituted in a temporal process that could be interpreted either as continuous or discontinuous. Explicitly under debate was the fact that one started off with one order of phenomena and ended up with another. As “orders” of phenomena, one can say there was a parallel or analogy between them. Yet they occupy quite different places in this discourse, were not homologies, and were not intended to substitute for each other. One was regarded as the ground or origin for the other, but discontinuously so. Moreover, the entire parliamentary process, in Parliament’s eyes, was premised on the fact that laws had to be made in an area where laws were not always already in existence. In the same way, they knew that the biology had to be made known by being transformed into objects of the imagination which could then be grasped for legislative processes.

Here is my parallel.

The kinds of critical replications mentioned in the previous section, chaos theory included, are poor substitutes for conceiving what Euro-Americans conceive of as reproduction and as the “process” that the Archbishop is trying to capture. They are helpful as a vehicle for conceptualizing the singularity of Euro-American kin constructs in this context, the parent as origin, while offering a critique that shows the manner in which “recognition” is constituted by making apparent what is already there but making it there only in effecting its appearance.6 They thus allow a way for talking about origins that are not origins—entities brought into being, created, only at the instant that their outcome also is. But they do not afford a way of grasping the central element of the reproductive model, namely the asymmetry between parent and child. This has both an atemporal and a temporal dimension.

The asymmetry is constituted in the fact that like persons and relations or biology and society (cell/person), “parent” and “child” are phenomena of different orders. We might even speak of a formal hierarchy: either term may be encompassing of the other but neither exists as the simple analogue or reciprocal of the other. Rather the analogy is between this asymmetrical construction and others like it. For example, one might say that the formal relationship between the terms parent and child is analogous to the natural/social distinction that differentiates cell from person. In this Euro-American formulation, then, parent and child are not analogous to each other; but the relationship of parent and child is analogous to that between nature and society (cell/person). The terms within each analogy are, as we have seen, made up of different orders of knowledge. They are thus irreducible to each other. Here we may take account of a temporal differentiation. They are not substitutable insofar as they are transformations of each other.7 Or, rather, one is held to grow out of the other.

This irreducible relationship is repeated over and again. For instance, the same is true of language and imagination. Euro-Americans hold that the parent produces the child in reality; the child “produces” the parent in metaphor. The formulation describes itself, for language is felt to be a substitute for reality (bridging Lacan’s gap), that is, a displacement of a non-equivalent order, one that is also a decisive transformation, or “merely” a representation, depending on how one will have it.8

The very process of debate in which the parliamentarians were engaged spoke to a sense that language provided a description, no more nor less, in the same way as the legislative decisions would put the stamp of human resolve on what was and what was not to count as relevant. In fact, there is a double process going on in the terminologies, that is, in the way the participants were interpreting their own language use. The parliamentarians were assisting themselves to understand complex processes, and there was a self-consciousness about their deliberations. The Archbishop’s analogy brought the metaphoric process to everyone’s attention: but he did not intend everything he said to be taken metaphorically. I have already suggested that he probably imagined he was illuminating a non-metaphoric order of phenomena by his metaphors.

It is in the constant replication (reproduction) of an irreducible relationship between different orders of knowledge that I find a parallel to the Mandelbrot set. No one of pair of terms can stand as an originary text. For a summarizing formula I shall use the distinction between recognition and construction, and for reasons that will become apparent.

If we take “construction” to be the encompassing indigenous form that recognition is understood as taking,9 then there are (at least) two kinds of constructions involved that the actors operate. One we can call “recognition,” the other “construction.” By recognition I mean that the activity of the imagination is regarded as being exercised in order to bring to the imagination facts that are not regarded as dependent upon it. Thus Mandelbrot’s set is brought to bear on the natural facts of embryology growth as a way of forcing people to a certain realization. They are thereby enabled to recognize, in the sense of perceive or acknowledge, the nature of the process being described. We might say that what is recognized is always already there in the language of factuality and information. The process of recognition simply involves the imaginative assessing of information. In construction, however, in this worldview, it is appreciated that human decisions have to intervene in the description of events in order to create categories that can then (in this case) be legislated upon. Thus the care with which the concept of person was, for many speakers, reserved for a late “stage” in the growth of the human being reflected a self-consciousness about the jural implications of using such a term. A concept is created. Such constructions, it was openly recognized, were social conventions.

These two forms of construction can exist in a transformative relation to each other. Thus the perception of the natural facts (recognition) acts as a ground or reference point for decisions about what social conventions (constructions) should apply. Conventions can be seen as transformations of the facts—facts being put to human use. They can also be seen as flowing from the facts, and as having their “origin” in the facts.

The distinction between recognition and construction (precept and concept; perceptual image and referential code; the factual and conventional)10 is asymmetrical. We may regard either as encompassing the other, and the distinction itself is familiar to the way in which Euro-Americans think about what their knowledge represents. It is repeated in one of its instantiations. Between parent and child, the child is recognized; the parent, by contrast, is constructed.

By this I mean that the child is regarded as autonomously produced by biological processes in the same way as it is intended to grow into an autonomous adult (person). It does not even need to “know” its parents in order to exist as a child, though to be the child ofsomeone depends on parentage. Yet that parentage may be presumed as much as known. To see a child is thus to “recognize” a natural fact. But parents exist only in (always) already having been recognized: that is, they exist only insofar as their children not only themselves exist but are known to exist. A person is not presumed to be a parent unless there is some way of knowing they have “had” children. Whether socially or legally, by contrast with the child, parenthood is thus always “constructed” as an object of knowledge.

This introduces the issues that will be the focus of the rest of the paper. At least as far as mid-twentieth century Euro-American kinship thinking is concerned, the category parent contains the same distinction within it. The mother is recognized; the father, by contrast, is constructed.

Parenting knowledge

I do not need to adduce examples of the extent to which Euro-Americans have regarded motherhood as a “natural” phenomenon in a way that fatherhood must always be a “social” one. It was long thought there was a certainty about the former—childbirth was the evidence—and a corresponding and intrinsic uncertainty about the latter—paternity could only ever be presumed. This was instantiated in law. “In English law husband is presumed to be the father of any child born to his wife unless it can be proved that he is not its natural father” (Wolfram 1987: 121). Indeed, the anthropology of kinship began, so to speak, in debates that arose from exactly this point. Whereas maternity was known through the obviousness of the birth process, paternity had to be inferred from the conjugal union. In short, the mother is constituted in her connection with the child, where fatherhood is constituted in his relationship to the mother.

We can now add a modification to the Euro-American view that a “parent” is created by its offspring. The difference between parent and child in this regard is repeated in the difference between mother and father.11 As Cannell (1990: 672) puts it, “biological motherhood and biological fatherhood are constructed in asymmetrical ways.” In conventional Euro-American thinking, the mother is regarded as created by her offspring in the act of her (visibly) giving birth. But the father is at a further remove. He is created through the mother-child connection, that is, the mother’s demonstrated connection to the child is necessarily prior to his claims with respect to the child. The certainty of the connection between mother and child is thus necessary to the certainty of the father’s claim (cf. Cannell 1990: 673). Insofar as these claims turn on his relationship with the mother, they are regarded—certainly in middle class English kinship practice—as dependent on a social relationship. We should not be confused here by the issues of legality. The so-called “natural” father has to demonstrate a “social” relationship to the mother quite as much as the jural father whose paternity is established through marriage. In short, to see a mother is thus to “recognize” a natural connection, whereas the father exists insofar as his connection with the mother is known to exist. And a father is not presumed to be a father unless there is some way of knowing about that connection. Fatherhood is thus always “constructed” as a (conventional) object of knowledge.

To put it another way. A mother does not have to make her motherhood known—it is constituted in the act of carrying and giving birth. But the fatherhood has to be declared, whether through the public relationship of marriage or the private acknowledgement of the mother. Thus it is assumed that a woman always knows if she has given birth, whereas a man may be in doubt as to whether or not he is a father, since knowledge rests not in the facts of his own actions but in their relevance to the connection between mother and child. I labor the obvious point to make an observation about how these relationships are construed.

This kinship system places special emphasis on the act of procreation. The social relationship between mother and child is regarded as inseparable from the process of gestation and childbirth (post-birth nurture is another issue). Fatherhood, on the other hand, separates social relationship from procreative process: the procreative act that establishes his relationship with the child only exists as a sexual act with the child’s mother. In the case of the mother, the sexual act is subsumed under the subsequent gestation and birth of the child who results from it, and indeed is inferred to have taken place by virtue of the woman becoming pregnant. In the case of the father, the sexual act itself has to be acknowledged, the point being that it cannot be acknowledged in the person of the child—it can only be acknowledged via his demonstrated connection to the child’s mother.

Alongside this inevitable “uncertainty” goes a theory about origins: the procreative act constituted an endowment of substance so that in genetic terms the father is as much “in” the child as the mother. In this sense, the father can be seen in the person of the child. But the effect of that is, of course, to naturalize the sexual act. That is, if the father has an evident genetic connection with the child, then that implies an axiomatic recognition of his sexual relations with the child’s mother. The asymmetry is again reproduced: between the genetic link which is constructed—since it cannot be directly seen, in this view, it is made known through various means—and the recognition of the prior act. That cannot be seen either, but is taken as self-evident in the pregnancy that follows.

The genetic link that each parent has with the child presumes their joining together in a sexual act, and being the equal originators of the substance of which the child is made. Mother and father appear analogous to each other in this regard. Yet if we consider the entire procreative process, we can also say that of all it takes to “make” a child, in fact it is this component that makes the father most like the mother. What for the mother is only part of her natural contribution to the child’s formation, for the father is the whole. I repeat that this is an artefact of the kinship system. That system supposes, I suggest, that natural fatherhood is modeled after motherhood.

There is something much more interesting in these constructs than simply the representation of maternity as a natural and paternity as a social fact. That summarizing metaphor is itself constituted of myriad elements that repeat themselves at different points in the way in which parenthood is defined—Mandelbrot’s constellation of points, each validating the patterning of prior ones. It is because of this capacity of the symbolic system to replicate itself over and over that it at once attracts and absorbs innovation it changes while appearing not to change.

One does not “change” categories that seem foundational—such as “nature” or “culture,” for instance, as witness the constant appeals to nature that are made on every side of the present NRT debates. Rather, one seeks to validate the foundational categories through the numerous particular instances where they are felt to apply. Thus the increasing sophistication of genetic knowledge only appears to validate what we always, already know about the transmission of inheritance. It “adds” to our knowledge.

Now such supplementation also changes knowledge, introduces new measures and scales of information even where it keeps the scale of (symbolic) complexity (cf. Strathern 1991a). This is the phenomenon that many commentators in the NRT field have commented on—each validation produces new uncertainties that remain to be tackled. We may imagine the relationship between certainty and uncertainty as a version of that between recognition and construction.

For example, the fertilization of the egg has to be made known (constructed), whereas the penetration of egg by sperm is recognized in that knowledge. But if one’s starting point is the interaction of genetic and epigenetic factors in the environment of the egg/maternal body, then new lines of uncertainty are drawn between what one can infer about heredity and the role that other factors play in the emergent phenotype. The process is replicated again in the further realization that hereditary substance does not exist independently of the context of its formation. Embryo growth does not unfold according to a program: “embryos begin by actively creating structures” (Birke et al. 1990: 69, emphasis removed). And so forth.

But this is not the road I wish to go down. The proliferation of knowledge proceeds like Derrida’s supplementation or the fractal replication of patterns to an infinite horizon/point. What is of interest to the anthropologist is that the proliferation of knowledge is (like) the proliferation of symbolic constructs. Each appears to allow a fresh entry point, in this Euro-American worldview, but each reproduces what has gone before—not as a replica but as a growth from it. Thus knowledge and symbolic possibilities appear to “grow.” But what looks like the embryo turns out to be like the mathematical set (a process of replication). So we seem to be back to the position we left in discussing the Archbishop’s application of Mandelbrot’s model. Let me, accordingly, repeat the conclusion.

What is of cultural interest here is the fact that the replication of (homologous) possibilities for knowing more or in greater scale rests on the irreducibility of the elements that make up the relationships being proliferated—what I have called their internal asymmetry. As it has been explained in popular terms (Gleick 1987), the points in Mandelbrot’s set are composed of two irreducible elements, a number and a rule. (For example, take a number, multiply it by itself, add the original number; take the new result multiply it by itself….)12 However high the magnitude of one’s calculations the original complexity remains—one may imagine it as an irreducible relationship between two orders or dimensions (number/rule) that do not map on to each other. The same irreducibility is found with respect to certain mid-twentieth century ideas of person and relation or individual and society. I see the same irreducibility in the Euro-American depiction of the difference between mother and father. That difference (complexity) is replicated throughout “the system.” So what sustains the difference? Will other instances of it help answer the question?

In the case of the irreducible relationship between parent and child, one source of difference lies in the temporal “direction” of the relationship, insofar as the parent is thought to be prior and the parent’s genetic material to be the origin of the child’s.13 Now “mother” and “father” are not temporally separated in the same way at all. But I suggest that there is in these Euro-American constructs an originating directionality nonetheless. I return to the concept of ground or context with which I started out earlier: one order of phenomena provides, by contrast, a model for another.

The modeling may or may not indigenously involve a sense of metaphor. A metaphor is drawn upon by the Archbishop, reiterated by Morgan and Lee, in order to provide a grounding understanding for the processual issues at stake in embryo growth but not to suggest an intrinsic relationship between fractal geometry and reproduction. Or what is held to be a non-metaphorical fact, such as the certainty of motherhood, provides a grounding for the social value given to the mother-child tie as a natural bond. As an anthropologist, I might observe that the natural tie thus provides a metaphor for, or symbolizes, the immutability of the social one, but this would not be a viewpoint shared by many indigenous Euro-Americans. They depict certainty through the ability to say something without having to have resort to metaphor—to recognize rather than construct. It was the context of debate and dispute—the necessity to construct—that inspired Hapgood’s resort to analogy.14

Now as a point of reference or illumination, a metaphor is formally equivalent to a fact instituted as a ground or reason. Substantively, the natives hold that metaphors do not have the same reality as the grounds for knowledge. So to say that motherhood is a metaphor for fatherhood would not be an inference the natives would recognize. The metaphor is regarded as the less real element. That is the crux: motherhood stands for the reality of natural process. We can, though, reasonably infer that in certain ways of thinking about fatherhood, the natives do model their ideas about fatherhood after those that already exist about motherhood. This is true insofar as motherhood represents nature in its natural, originary form, while fatherhood supports its social claims with such natural justification as it can borrow from the idea of motherhood.

My statements refer to various cultural categorizations already in existence (and are not meant to carry ontological weight). This particular “representation” is to be seen in and through various refractions of the ideas about maternity and paternity that have received attention, at least in English debates, in the context of the new reproductive technologies. What lies in wait, though, may well be a redirectioning. Some of the new reproductive interventions make it seem as though it is not fatherhood that is modeled on motherhood, but motherhood that is being modeled on fatherhood.

Taking analogies for granted

But let me first insist on the particularity of these prior, modernist, Euro-American assumptions. They constitute a representational theory of parenthood that finds representation (the difference between construction and recognition) at the heart of what is being represented.

I do not simply want to show this through adducing a contrary case—raiding my Melanesian knowledge, as I have done on many occasions, in order to find a point of critique. However, it is of interest that one recent understanding of kinship at least has discovered Mandelbrot in Papua New Guinea. Wagner (1991) draws on fractal geometry as a way of restating how people there visualize reproduction. It is worth dwelling on for a moment. The interest is in what is described as its motivating irreducibility.

In the Melanesian case, such motivation is not to be found in the difference between mother and father, or parent and child, each of which in the Euro-American system we might imagine as of a different order of knowledge (recognition and construction) to the other. Rather, it rests in an identity between elements, described by Wagner as between person and relation. A person’s relations are at once integral to him/her and carried by him/her. It is as though number and rule were identical: the rule being perceived as an instantiation of number. It is “irreducible” because that identity is repeated whenever Melanesians encounter either persons or relations. A “fractal person” is thus a person with a relationship implied, as one might imagine a number being implied by the rule that expresses it. The one is not replicated without the other. As a consequence, a person in a relationship (as Euro-Americans would express it), e.g. a “mother” or “father,” is analogous to any other person in a relationship.

Kinship becomes a flow of analogies.15 Thus Wagner characterizes a range of non Euro-American kinship “systems” (his type case is from Papua New Guinea) as being founded on an analogical approach: having obviated the distinction between “natural” kin type and “cultural” kin relationship by subsuming terminology and relationship within a single entity, an analogical approach does not incorporate the contrast between “mental” symbolization and “physical” fact. Its constructions are intended as simultaneously conceptual and phenomenal; they belong to a single universe of apprehended cultural construction (and culturally constructed apprehension) that is contiguous with other realms of conceptualization (Wagner 1977: 626-627).

Now insofar as analogy is taken for granted, then it does (can) not have the constructionist force that it has carried in systems that rest on a difference between recognition and construction.

But how can irreducibility that takes the form of identity be motivating; how can it reproduce itself? What the figure of a fractal person contains, so to speak, are other persons: he or she is a person with other persons implied in his/her constitution. Reproduction in this regime turns out to be an act of differentiation, one that divides persons from persons in such a way as to create differences between domains or dimensions. We could say that in an analogic system irreducibility between particulars has to be created (by convention).

In approaching them through irreducibly different orders of knowledge, Euro-Americans find mother and father to be irreducibly different. Each parent appears to belong to a different domain or dimension (for example, nature/culture). What is thus recreated each time is this difference, discovered afresh in the particulate identity of each figure. It is the knowledge practices themselves that afford analogies: these differentiations are endlessly repeated, as I have tried to indicate, and are recreated every time one tries to “know” something. For the difference between whole orders of knowing (recognition, construction) is a relation integral to knowledge. That is why I have insisted that the appropriate analogy for the Mandelbrot set is, for Euro-Americans, in the symbolic order itself. It appears contrived, as analogies and metaphors always do in this system.

But here in Papua New Guinea we have people thinking like Mandelbrot (always and already). They take the analogical process for granted, and see it manifest in the recursive repetition of persons. In fact, all that has to be reproduced are persons. That being so, social life consists in creating differentiations (dividing person from person), and Melanesians accomplish this as though the act of differentiation brought about a difference between orders of knowledge. It is the act that differentiates. That is, reproduction is the act that produces irreducibility between different types of knowledge. Hence the “transformative” effect of actions constantly being framed off from one another—whether through protocols in the way Weiner (e.g. 1987) has described, or all the events anthropologists identify as childbirth, bridewealth, ceremonial exchange, war compensation, sister-exchange, male initiation, female seclusion, salt preparation, and the rest. Differences between types of acts are established such that movement from one point to another is inevitably transformative.16 Indeed, people may mark off specific sequences of events as transformative (ritual), or particular figures for that matter (big men, great men).

When a person is conceived fractally, then, relationships are already implied, one parent is already a version of, analogous to, another parent, and the work of convention is that of differentiation. What have to be established are the particularities of identity, and individual, particulate facts that Euro-American constructions take for granted (cf. Weiner 1988). One Melanesian mode of differentiation turns on establishing origins: the fact of transformation (from one state to another) is reproduced as cause/effect or origin/outcome.

The Euro-American propositions with which this paper began suppose that the linear fact of an origin being prior to its product or outcome can be theoretically subverted in the cultural realization that an origin is also created by its outcome and acquires its significance from it. Thus an origin is recognized retrospectively, but constructed prospectively. The late twentieth century Euro-American idea that there are no origins any more is the realization that recognition, too, is a kind of construction. So we could characterize deconstructionist talk and the culture theory of consumerism as the discovery of analogic reasoning! Similarity is taken for granted; differentiation seems to rest only in the activated preferences of the consumer/the decisive moment of the interpreter. But this is not quite the analogic reasoning of (say) the Melanesian Daribi or Foi. Perhaps the prescriptiveness of consumerism, the duty to make discriminating choices, is proceeding towards Daribi or Foi conventionality: the absolute social necessity to make those differentiations in order to constitute particular entities. But insofar as interpreter and text or consumer and consumer goods remain already differentiated, then analogic realization is not complete. Indeed, I would suggest that it will require a more thorough-going relativization of persons than simply thinking, writing, imagining and advertising will accomplish. (That relativization is the promise of the new reproductive technologies.)

Where analogies are taken for granted, by contrast, we should be alert to differentiation between different types or versions of the same thing. We should expect an interest less in the difference between an origin and its outcome—as between parent and child—than between types of origins. Why? There is no why: but we might understand this as a how. That is, it is the empirical case for Melanesia that establishing origins is also one way of establishing the difference between types of parent, or types of children. And vice versa: one differentiating act enables another to be done. The difference between motherhood and fatherhood is established in the further difference between types of origin: one parent may be seen as originating from ancestral parents whereas the second parent’s ancestry may cease to be carried forward. That in turn is repeated in the differentiation of children: those who do and who do not replicate the parent.

Fatherhood and motherhood remain analogous, in the same way as pointing to either continuity or discontinuity with prior examples of oneself constitute analogous reference points (origins), or brothers’ sons are analogous to sisters’ sons; differentiations have to be imagined as significant moments or events (acts) that dramatically sever one type of relationship from another. (Thus their rehearsing is often done through folktales or mythic narrative that recount the decisive moment; in myth, of course, the decisive moment is re-lived over and again, and must be re-enacted in real life, over and again.)

But while it would be intriguing to explore the Melanesian material with this in mind, I want to borrow the insights without borrowing the substance. And that is because I want to uncover what I think may be apprehended as an example of analogic kinship from a most subtle and complex account that nonetheless rests itself, in the end, on Euro-American representationalism. It is a rich, intelligent and multi-layered analysis. I wish it to demonstrate, simultaneously, what I “recognize” as indigenous analogic kin constructs and the Euro-American kinship thinking that the author has “constructed” out of them. In constructing such a metamessage (Werbner 1989: 224-5), I acknowledge the challenge of the original.

Parenthood and kinship

Houseman (1988) analyses two African folktales (from his own work among the Beti of Cameroons and Heritier-Auge’s among the Samo from Burkina Faso). In each case, one parental role is taken for granted, while the other has to be established through their origins—through “reference to higher order (and temporally prior) versions of themselves” (1988: 670). But in one case lineality seems to flow from that active intervention, while in the other it does not. For the “patrilineal” Beti, the identity of fatherhood is established through recursive reference to prior fathers. For the Samo, the same is true of motherhood; yet that does not make the Samo system “matrilineal.” Perhaps it is this which gives Houseman pause. Motherhood appears the problematic in his account.

Analogically speaking, we could say that the difference between motherhood and fatherhood is established in both cases (though to contrasting effect) in an originary difference between the one parent who already, so to speak, “contains” his or her child and whose parenthood is taken for granted, and the other parent whose parenthood is established only through the explicit instantiation of him or herself as a child of his or her own parent. The senior parent to whom reference is made is of the same sex in both stories; the analogy that both stories seem to want to block is that between opposite-sex parent and opposite-sex partner. That is what makes the systems lineal. But although it is fatherhood that is doubly defined in the one case and motherhood in the other, the reproductive effect, as far as I can see, is similar: reproducing a patriline that is dependent on maternal sources of fertility. Houseman, however, wishes the one case to be the inverse of the other. And that is because his comparison rests on what he sees as an irreducible difference in ways of knowing: there has to be a radical difference between the parent taken for granted and the parent that is doubly defined. Moreover, his insistence takes the form of suggesting that the double definition reduces uncertainty. The necessity for the double definition arises, in his analysis, from an “indeterminacy” in the role in question.

Indeterminacy he defines as uncertainty inherent in the structure of relations. One of the two terms for the relations between mother and father has an autonomous identity where the second (indeterminate) term requires further specification or contextualisation (grounding). Since he discovers that the specification is done in each case through appeal to a reduplicated version of the term (e.g. the father’s father validates the father), he argues that a hierarchical relation is inevitable between the two terms.17

As a consequence, parenthood and kinship (person and relation?) themselves “are founded on irreducibly different premises” (1988: 660), since hierarchy is intrinsic to the first and not the second domain. A hierarchical encompassment of one term by the other is endorsed in an inevitable asymmetry between the terms themselves. The paper turns on explicating the asymmetry between parenthood and kinship; between mother and father; and, as we shall see, within the figure of the mother between elements of her role. Thus in comparing the two cases, he argues that what makes a difference between Beti and Samo is the extent to which “physical” and “jural” parenthood is or is not differentiated in the mother. An absolute identity between these elements, he argues, is only possible in the case of the mother. Therefore it is in her case that one may ask whether or not physical and jural parenthood are taken as distinct or identical. In the father’s case, such identity is not possible (they may be conjoined but that can never be taken for granted) and therefore the question of whether or not there is identity or difference cannot be posed.

He writes:

these constraints remain firmly grounded in ineluctable biological realities. Thus, the very orientation of these two asymmetrical figures, as well as the fact that it is the distinction between physical and jural motherhood and not that between physical and jural fatherhood, which is the determinate variable underlying the pertinence of one or the other of them, derives from the self-evident sexual dissymmetry mentioned at the beginning of this paper. As a result of gestation and childbirth, the relationship between the female sexual (genital-linked) identity and maternal procreative function is a potentially continuous one, whereas the relationship between male sexual (genital-linked) identity and paternal function is necessarily discontinuous…. In other words, an absolute identity of physical and jural parenthood is possible only in the case of motherhood (ibid.: 673, references omitted).

Note that the difference between continuity and discontinuity are held to point not to two types of origins but to a radical disjunction between two orders of knowledge.

Houseman makes it very clear that he is not debating substantives, how one might characterize motherhood or fatherhood (their figurative realization) but what he calls the formal (literal) relationship between terms. Accordingly, I do not quarrel with his discovery of an ineluctable biology, but I would comment on the implicit asymmetry in his differentiation between physical and jural parenthood (the difference between “a child’s recognized social parents” and “a child’s supposed [constructed] biological parents,” 1988: 660). In fact, either the jural or the physical status, I think, may be in doubt, as his examples appear to show. There is no substantive parallel between jural and constructed status, or biological status and recognition. Indeed, this very quotation reverses that potential equation. The parallel lies, of course, in the fact that in each case one term is seen to depend on the other. In that sense, the asymmetry repeats that between parenthood and kinship and between mother and father in his analysis. Indeed, he suggests an analogy in his own reiteration of irreducibility. In what I have so far quoted from his article, it must be apparent that this follows the now familiar lines of Euro-American reasoning. Moreover, it is a reasoning which ultimately turns on an asymmetry between different orders of knowledge.

Héritier-Augé (1989: 164) provides a clue in the Samo saying that “words make descent and words can take it back.” Descent traced patrilineally relies on speech insofar as “it is based on the common will and public acknowledgement of the social link” (ibid.). In order to bring out the significance of this, an ethnographic digression is necessary.

By contrast with acts of public acknowledgement, Samo hold that each generation is replenished by an infusion of maternal blood. That blood in turn is constituted through semen (semen turns into blood inside a woman’s body, so that the blood with which the child is endowed is the husband’s semen transformed; [before marriage?] a woman’s blood is regarded as inherited from her father). But while the transmission of paternal substance (blood/semen) can be thought of as continuous over the generations, the maternal transformation that is necessary to procreation creates discontinuities. (A woman cannot make semen; a man transmits semen by “making” blood via his wife.) Each infusion of maternal blood comes from a different source and indeed the origin of the mother’s blood must be obliterated at each generation, for she does not pass on her mother’s blood, only what she inherited from her father, and her children inherit her blood as a transformed endowment from her own husband. Indeed, the importance of discontinuity in the “female line” is brought out in the Samo story that Houseman quotes.

Briefly, a woman imagines that her husband has two penises (and she also has a double in a cowife). But she is only able to become a proper mother when she realizes that in fact he has one (the absent one is left for her “cowife”). And that occurs after her own mother dies.

I would read this as a statement that the woman has to differentiate between her father/ husband (she cannot reproduce via both of them; be both herself and her mother/co-wife), and that differentiation is done through the event of the mother’s death. She is separated from her mother. More than that, her mother then ceases to be distinct from her father—her blood ceases to transform his—and the woman (daughter) comes to her proper husband properly constituted through paternal blood alone.18 This is, of course, thoroughly consistent with patrilineal logic of a kind.

The woman’s referencing back to her mother is what intrigues Houseman—why the wife’s mother comes into the story at all. An origin is invoked only to be denied. He finds “somewhat paradoxical” (1988: 671) the opposition (as he puts it) of motherhood to itself. So he takes this referencing as evidence of, and the very reduplication of, a role that is indeterminate or “uncertain” (1988: 666). Motherhood in Samo, he argues, has to be achieved. By this he means that motherhood is only achieved in the successful act of giving birth, for there is no possibility of separating jural from physical status. Fatherhood is, by contrast, unproblematic: a Samo man is virtually a father by the fact of his own birth—provided he is married, all the children borne by his wife are credited to his (patri)lineage, and he does not have to prove his own physical paternity. There will always be lineage descendants to sacrifice for him. Yet while a childless man will have someone to perform sacrifice after death, a childless woman in Samo is divested of personhood. She can only be a jural mother, in Houseman’s terminology, through also physically giving birth.

In his account, the asymmetry between Samo motherhood and fatherhood rests on the indeterminacy of motherhood (it has to be established through the actual parturition of the woman); but the need to overcome this indeterminacy (he talks of the role of motherhood being fulfilled or achieved) formally subordinates the father to the mother. To the degree that motherhood is established by reference to other mothering figures, the mother is the superior figure in a hierarchical relationship that includes both mother and father. Now because a Samo mother is established only in being separated from her own mother, he also adduces that in formal terms the hierarchy is of a special (itself “asymmetrical”) type: where a category or class is encompassed by an expression of itself and its opposite. Motherhood and its negation encompass both motherhood and fatherhood.

I wish to pose a question: how did Houseman arrive at the indeterminacy of the Samo mother figure? After all, on first reading, Heritier-Auge’s observation would seem to suggest that the descent created by words that can be taken away by words refers to the less certain figure. Social descent, she points out, cannot be traced through blood—it can only be traced through speech, the common will voiced and affirmed by the social group. Indeed, as we have seen, nothing can be traced through blood, for (in my own words) unlike representations of the transmission of the “same” semen over the generations, blood is constantly divided from itself, as daughter is divided from mother. Maternal blood is constituted in its discontinuity.

Now the statement that when jural and physical motherhood are combined then the motherhood figure will be superordinate, as far as the formal hierarchy of terms is concerned, is not meant to refer to Samo alone. Houseman presents it as a universal proposition: i.e. this type of hierarchy will prevail wherever that combination is found. As we have seen, he links this to a human absolute: the relationship between sexual and parental identity is potentially continuous in the female case where it never can be in the male case. There is always discontinuity for the male, whereas either continuity or discontinuity is possible, and varies, in the female case. Samo motherhood offers a case where continuity must be demonstrated (being evidenced in the axiomatic linking of jural and physical motherhood). And while he analyses the woman’s relationship to her own mother as a break, an antithesis, an opposition, what interests him is the “continuity” or identity that has to be established between juraland physical status. This seems all back to front.

I suggest three reasons for the direction that Houseman’s analysis takes, which partly explain why Houseman differentiates persons as evidence of a difference between certainty and uncertainty. All of them turn out to be commentaries on substantives in Euro-American kinship thinking.

First, the article is premised on there being two orders of knowledge. There is the figure who can be taken for granted, and there is the indeterminate figure who requires activation or definition: the formal contrast we might say is between recognition and construction. Thus a Samo father is recognized by virtue of his own lineage membership and his wife’s having been promised in marriage to him as a girl. The Samo mother, on the other hand, must prove she is a mother (construct her motherhood) by her procreative capacities. A distinction between different ways of knowing is thus held to reveal two quite distinct figures. These figures appear as particulate entities, rather than analogues to each other. Indeed, he says as much in the distinction between kinship and parenthood. He imagines the domain of “kinship” as proceeding from parts to whole, a relational exercise in which particulate building blocks can be put together (constructed). By contrast with kinship, “parenthood” is recognized, a domain that proceeds from wholes to parts, at issue being how the necessary and universal duality of the biological procreative pair is differentiated in different cultures. This proposition in turn contains two orders of knowledge within: in the sexual asymmetry of the reproductive process which means that motherhood is inevitably indeterminate as to whether or not jural and physical parenthood is separate, fatherhood is not since separation is assumed, etc.

Second, I suspect the argument conflates the idea of continuity between jural and physical parenthood with continuity over time. Or to put it another way: the separation of the woman from her mother is regarded as another asymmetry or opposition or antithesis rather than as the description of a definitive temporal, epoch-dividing act. In fact, Houseman’s account downplays the discontinuity (he refers to the daughter’s “individualization” with respect to her mother, and to her having to obtain her mother’s “consent” to conceive even though it requires a “disassociation of the present generation from the preceding one” [1988: 668]); he is left at the end simply with a notion of “opposition.” He defines the self-referencing of the daughter via her mother as the equivalent to a jural validation that is also a physical validation. (I would read this per contra as a description of a radical temporal dislocation: one that makes continuity irrelevant to female potency. Mother and daughter are the same: it is the act, and thus the difference between the penises, which separate them in time.) And that is because, I think, he is modeling the Samo mother on his analysis of the Beti father.

Third, then, the irreducible asymmetry in which he is interested comes from a motivated modeling.19 It looks as though the difference between parenthood and kinship (recognition, construction) are to be found in, and thus modeled after, differences in the maternal figure (the possibility of conflating or distinguishing physical and jural parenthood). Thus motherhood is the figure that is the principal focus of interest in Houseman’s argument. But I suggest that motherhood is anthropologically (theoretically) indeterminate in his description because the father is, in fact, the reference point for the argument, that is, for the anthropological disquisition. The very differences in which he is analytically interested (jural, physical, etc.) are those differences that underlie Euro-American representations of the father. That the constructed figure is the more interesting figure is also what reveals anthropological interests (the discipline being devoted to uncovering social constructions).

I suggested that the figure of the Euro-American mother is a model for those aspects of the father that refer to natural or physical process. It is not, for Euro-Americans, the combination of jural and physical elements that makes motherhood such a model but taken-for-granted certainties about natural process.

But natural processes only ever defined an aspect of the father’s role, and were themselves subject to uncertainty in his case. The question for Euro-American kinship is whether the father can combine jural and physical attributes. To ask that question of themother is to model the mother on the father!

It might be thought that I have skewed my commentary by focusing on Houseman’s rendering of Samo rather than Beti. Beti after all, he says, regard the father rather than the mother as the indeterminate figure—while the Beti father transmits his identity in the patriline, he cannot do so unless he has sons born to him. For that he is dependent on other men’s daughters to bear his children; but he has the possibility of persuading his son-in-law to give up his own jural fatherhood and remain attached to his household as a physical father only. In a Beti story about two penises it is a senior man without sons who tries to turn his daughter into an heir by demanding that only a man with two penises could be his son-in-law. However, the son-in-law designate in turn becomes a proper (jural) father himself when he sheds the (now absent) penis of his father-in-law and takes the woman to his own place. (This seems to me the “same” story as the Samo one, not its inverse as Houseman argues.) He is able to shed the extra penis (restore his bifurcate organ to singularity) by invoking (an admonition from) his own father—in the same way as the Samo woman sheds her claim to the second penis by invoking her mother. Houseman makes much of the fact that the former stresses a connection with the senior parent (father-son continuity), the latter opposition (mother-daughter discontinuity).

It would seem to me that what is at issue in both cases is the assertion of patrilineal continuity, though Samo do this through emphasizing the separation of daughter from mother, Beti through the separation of son-in-law from father-in-law. The Beti son separates himself by making a difference between two types of “fathers”—his father-in-law and his own father. Once that is done, he can then be a father himself. Houseman’s preference is to analyze this as the invocation of jural status: the son-in-law becomes a father in his own right by virtue of his jural identity as a patriline member. Motherhood is not in question and does not require reduplication through reference to a third term. A woman simply becomes a mother by virtue of marriage itself—she may be mother to her own or other children, so that a Beti woman may achieve jural motherhood even if she is not physically the mother. Now a Beti father can achieve physical fatherhood even if he not jurally the father, and vice versa. Why, then, should Beti fathers have any problem at all?

Houseman’s material only makes sense if we accept that Beti prioritize “jural” over “physical” fatherhood: I wonder if the former is the more complete state, not because jural and physical fatherhood are combined, but because the father situates himself in a line of fathers. The question would then be what to “do” with physical paternity outside this context—where to “put” the non-lineage child. I would thus turn Houseman’s argument around. I would hazard that in both his examples what Houseman calls “indeterminacy” rests on the parent who has problems with regard to physical parenthood. I would see the difference between Samo and Beti not in terms of whether or not physical and jural parenthood combines in the mother, but whether or not lineal continuity—connections with origins—is held to depend on the transmission of male substance. Fathers may or may not be required to demonstrate physical connection with their children. Where they are not, the “fact” of physical connection becomes problematic. The reason why the problematic of Houseman’s argument seems to be about the mother is because it turns on the role of physical connection in definitions of parenthood. For Euro-Americans, that is a problem represented by motherhood. But insofar as he raises a question mark at all against the place of physical connection, then that question, in fact, turns on the Euro-American representation of fatherhood.

It is in this sense that I have suggested Houseman’s account of Samo models motherhood on fatherhood. I think he does the same for Beti women. Thus he imagines that women have men’s problems with continuity: referring to the various contradictions both Samo and Beti stories entertain, he suggests both deal with “the inherently problematic transference of female-procreative power” (1988: 668). I would have thought his (patrilineal) examples showed rather clearly that the transference of procreative power is a specifically male problem.

If we reintroduce the logic of analogic kinship, then we might argue that continuity and discontinuity exist as simple versions of each other. The Beti man’s continuity with his father is the same as, an alternative to, his discontinuity from his father-in-law. The Samo woman’s continuity with her child is the same as, an alternative to, her discontinuity from her other self (mother, cowife). Continuity and discontinuity are the “same.” They are made disjunct only in the differences established between persons, for it is they who differentiate the effects they have on one another: father must be divided from father-in-law, daughter from mother. The penis is bifurcate: a double penis is also a divided one. But this appears an inappropriate division, as far as that organ is concerned, for which the stories then proceed to substitute the appropriate one: a division of other persons in its stead.

The new knowledge

Although I have emphasized the anthropologically familiar idea of lineality, I do so to suggest that the “same” patrilineal succession is being constructed in two different ways. Lineality is as well understood as a discontinuous series of substitutions (the son replaces his father) as a continuous flow of substance. In fact, such lineality may not appear “linear” at all. Certainly there seems no simple temporal linearity in the Beti and Samo cases. Persons are either duplicated or divided. Temporality inheres as much in the discontinuity that creates two “kinds” of time as in the continuity that presupposes one. In fact, one could say that Beti and Samo are both telling stories about the fact that “having a child” is just like “having a parent”; whether the parent/child is regarded as the same as or different from oneself will depend on the claims of one’s spouse.

Parent and child are thus analogies of one another, as origin and outcome must also be. Perhaps we do not need a difference between jural and physical parenthood in order to describe a situation where parenthood may be instantiated by an act of either continuity or discontinuity in relation to either the parent’s parent or the parent’s child. But we do in understanding Euro-American middle class kinship.

What gives this latter system its cast of temporal linearity is the fact that only continuity of connection is felt to demonstrate parenthood. It would be a paradox in the Euro-American case to imagine parenthood founded on discontinuity.20 On the contrary, the absence either of a physical connection or a jural one may compromise what had been regarded as a parent-child relationship. Since continuity is all, temporality seems of one kind only, downward flowing, directed towards the future and requiring an origin.

In fact, continuity is ideally established twice over—a matter of recognizing both physical connection and jural connection. The necessary reduplication is precisely of these elements: not persons by persons but type of knowledge by type of knowledge. Insofar as motherhood stands for the specificity of the natural (physical) connection, then motherhood encompasses both motherhood (an instance of the fact) and fatherhood (the social recognition of the fact). Insofar as fatherhood stands for the social (jural) recognition of the fact, then fatherhood encompasses both fatherhood (the necessity to recognize) and motherhood (an instance of such recognition). In this schema, I suggest, “recognition” is reduplicated in the indeterminate term as “construction.” (A construction is a recognition that recognition has taken place.)

Euro-Americans do not regard persons as recursively related—but it lies within their philosophical, critical and psychoanalytic traditions to regard the imagination (knowledge) as constructed that way. What Houseman writes of the recursive trajectories of parenthood is a perfect description of the recursive trajectories of knowledge for some Euro-Americans:21

substantive interpretations of mother- and fatherhood are invariably caught up in recursive trajectories for example, fatherhood being distinguished from motherhood as an aspect of fatherhood, itself distinguished from motherhood as an aspect of fatherhood, and so on—which reproduce the hierarchical form of interrelation of these terms. In other words, particular conceptions of mother- and fatherhood, whether couched in behavioral, physiological, jural or cosmological terms, will inevitably reiterate either the logical subordination of motherhood to fatherhood according to the first figure … or the logical subordination of fatherhood to motherhood according to the second … depending upon which of the two schemes applies (1988: 673).

What is of interest is that such insights do not remain, so to speak, with the professionals of insight. The new reproductive technologies have forced similar realizations into relatively public debate—into areas where the grounds for knowledge have to be established, facts validated, concepts created. Not only is the whole NRT debate a kind of duplication of the necessity in Euro-American culture to recognize parenthood: in the parliamentary debates, that is duplicated over and again in terms of the necessity to provide a legislative stamp to what is constructed as the recognized facts.

In the concluding section I point to two aspects of some of the changes people envisage consequent upon the NRT. Both of them illuminate—and transform—received notions about parenthood.

Now, when Warnock (1985: 37) says apropos egg donation,

Egg donation produces for the first time circumstances in which the genetic mother (the woman who donates the egg) is a different person from the woman who gives birth to the child, the carrying mother. The law has never, till now, had to face this problem. There are inevitably going to be instances where the stark issue arises of who is the mother.

She means that what has to be weighed up between the two women is the weight to be given to their claims of continuity with the child. The way the question is phrased already tells us what mother is: a carrying mother or a genetic mother. But there is a perceived division or bifurcation of natural process here which raises questions about which component suggests the superior claim to continuity. I want to suggest, first, that the very question is raised at all takes away from the idea of motherhood an axiomatic exemplification of what Houseman would call physical parenthood; and, second, this is the same revelation that rests in the necessity to find the grounds on which to recognize and thus construct (jural) parenthood in the new situation—for both motherhood and fatherhood must now be subject to the kind of practice of verification for which fatherhood once stood.

The new kinship

Wolfram (1987) puts the issue starkly. English law previously recognized “two possibilities as to who is the father of the child” whereas there was ever “only one possible ‘real’ mother” (200). The father but not the mother was traditionally bifurcate. But this was never an analogic bifurcation (the division of a single entity). As we have seen, the distinction between possibilities is an asymmetric one between two orders of knowledge.

Wolfram is commenting on Warnock’s recommendations concerning surrogacy: that although the carrying mother must be regarded as “the mother,” the law should allow a genetic mother to adopt the child if necessary. Wolfram compares the right thus accorded the genetic mother with those of the genetic father who is not the mother’s husband. Here we see one of the components of fatherhood providing a model for which to think about motherhood.22

One of the arguments put forward on behalf of mothers who wish to establish their claims to a child through genetic continuity is precisely that it would be an injustice not to treat their claims on equal terms with those of genetic fathers. One such argument is cited by Morgan and Lee (1991: 153) as influencing legislative outcome with respect to surrogacy arrangements. (Human Fertilization and Embryology Act 1990, Section 30; rather than having to go through lengthy and alienating adoption procedures, it was decided that a couple may apply for a court order to establish their legal parentage so long as at least one of the couple was the source of the gametes.) The argument they cite draws an explicit parallel between motherhood and fatherhood. The injustice is that a genetic mother whose genetic offspring are borne by a surrogate mother (the terms continue to bifurcate) would have had no legal rights over “her own” children, whereas the genetic father could draw on existing conventions with regard to illegitimacy. A genetic father already had the right to apply for guardianship. They quote a letter to The Times:“Surely genetic mothers, at the very least, should be accorded the same rights and privileges as genetic fathers?” (ibid.: 154).

Although the argument turns on the recognition of natural facts—both parties contribute genetic material, and on this ground then mother and father—“in these days of the equality of the sexes” (quoted Morgan and Lee 1991: 153)—should be regarded as equal, the argument also turns on the idea that it is the law’s business to recognize the facts in an equal way. That is, what is at issue is an equality of recognition. The argument is that because the genetic father of an illegitimate child can already have his paternity recognized in law, then the mother should be given parity.

Now the genetic claims of the illegitimate father (so to speak) were traditionally established in the same way as the claims of the legitimate one—through an act of sexual intercourse with the wife/mother. The claim to continuity depended on proof that intercourse had taken place. Perhaps this lay behind the Warnock Committee’s strong recommendation that when a man donates semen through means other than intercourse, then he should have no claim over either mother or child.

The giving of semen by other than sexual means has long been regarded as “donation,” an alternative to “giving” by intercourse. We might say that, because the act is potentially modeled on the sexual one, special effort has to be made to distinguish it for legal reasons.23 The concept of donation has been extended to ova. The idea of egg donation thus appears to be modeled on that of semen donation. This is a point that has attracted much critique, among other things because the procedures involved in extracting the gametes are so different in the two cases. Here I simply note that “donating” an egg is not an alternative to any other act that would have occurred between the women. There is no parallel to egg donation in any existing relationships and certainly not in sexual intercourse itself. The only parallel is in donation per se—both the donation of body parts (organs) in general and semen donation in particular. The mother’s act is modeled on the kind of donation that fathers already seem to be doing.

Now as it happens, another parallel was to hand: outside the parliamentary arena, such “giving” is also assimilated not to the presentation of items (gametes, organs) but to the rendering of services.24 Women have been known to “help” one another, or been used to complete the father’s desire for parentage and thus “helped” him, on an informal basis. However, the law did not recognize (so to speak) an illegitimate mother since motherhood was never mediated by marriage: the natural mother was the mother. There was thus no mechanism for adjudicating between competing claims in her case. (See above: there could only be one mother.) Such services, in not being recognized, could not be drawn upon for a model of how to constitute the claims of genetic donors.

Yet semen donation was also a culturally awkward analogy, and its application introduced an asymmetry. After all, in Euro-American kinship, it is not that donation is a single act shared by each parent—that the couple each “give” to the other—as one might say for Melanesia where the gift relationship is divided by the persons who give. Rather, there was always an asymmetry: the father “gave” semen to the mother, but the mother did not “give” eggs to the father—one could only think of the mother’s donation as towards her genetic offspring. And in that context, both parents gave their genetic endowment to the child. Imagining that eggs can be donated then, requires thinking about the rather different act of another person (the father), and implies its formally indeterminate status.

It is of some interest that Section thirty is entitled: “Parental orders in favour of gamete donors.” That the recognition of genetic claims is also modeled after the kinds of claims that the traditional father could make is captured in the terminology. Whereas the phrase “genetic father” was used freely in the debates, and in the commentary on the Act, in the Warnock Report and so forth, at the point of definition in the Act itself the term “father” is withheld. He is simply the gamete donor—in the same way as a woman is a gamete donor—and parentage has to be legally claimed.25

I have suggested that the idea of male semen donation provides a model for thinking about egg donation in a way that women’s services for one another could not. While there are good legislative reasons for the latter (there was no equivalence to the illegitimate father), and good historical reasons for the former (not just semen donation but artificial insemination has a long history in Euro-American practices), I want to point now to a cultural shift. My emphasis so far has been to show the manner in which the recognition (construction) of parenthood has had to rest on what is already recognized (recognizing the genetic mother in the same way as the genetic father was, for instance). This models the process on that which always, already defined the Euro-American father: fatherhood required validation. “We are used to assuming that between a mother and the fruit of her womb, there is bond of blood and body. In contrast, a father is linked to his legitimate offspring essentially on the basis of an assumption” (Sissa 1989: 133). Two different orders of knowledge are at issue. I now want to underline the further fact that the traditional bifurcation of the father’s double relationship to his child has provided a model for thinking about the bifurcation of the mother’s role created by the technical possibilities of In Vitro Fertilization (IVF) with embryo transfer, gamete donation and embryo implantation.

But asymmetry is preserved. Whereas the double element in the father’s case is in common parlance distinguished as “biological” and “social” fatherhood, what is being divided in the mother’s case are two aspects of her “biological” parenthood. And one aspect of fatherhood comes into its own as a single reference point. Both egg donation and surrogacy arrangements have been compared to semen donation by men. Here fatherhood emerges as a single, biological, unitary reference point for what is in the mother seen to be a split between two aspects of a formerly conjoint condition. The attempt to draw an analogy between mother and father thus finds itself reproducing the asymmetry between what is recognized as a fact (semen donation) and what is constructed, modeled after the fact. The substantive analogy is not between the persons, between mother and father as entire entities, but between aspects of their roles. The biological fact of paternal endowment becomes a natural reference point for a new double maternity. I take each kind of motherhood in turn.

We find that analogies are drawn doubly between egg donation/maternal surrogacy and semen donation. In both cases a further asymmetry is discovered between what is appropriate male or female behavior. Neither turns out to be parallel after all!

In investigating the claim that egg donation is the “female equivalent” to semen donation, Haimes (1991) found that those who assert the similarity of the acts on one level or another seem to regard them very differently. She consulted members of the Warnock Committee, among others, and offers an intriguing analysis of the way in which the assumption about the way men and women ought to behave affected views about the propriety of donation.

She points out, for example, that the Warnock Report is straightforward on the issues of preserving anonymity in the case of semen donation, but equivocates in the case of egg donation, since close relatives may be involved. The reasons for insisting on anonymity for semen donors include, Haimes points out, the desire to protect the recipient couple from “the invasion of the third party into the family” (Warnock 1985: 25), whereas in the case of egg donation, reference was made to the potential closeness of the donors to each other. A sense of danger thus surrounded the idea of semen donation. Haimes notes that previous statements on semen donation raised questions about the man’s motives. While in the eyes of some either the donating man or receiving woman might be suspect, their “pathology” was differently constructed: “male concerns in reproduction were presumed to revolve around ideas of virility, genetic continuity and generally being assertive and in control. Female concerns were presumed to revolve around the need to become a mother which led to a form of pathological assertiveness when otherwise the woman’s role was characterized by passivity” (Haimes 1991: 8) (the mother’s assertiveness here being modeled on the father’s).26

The instances of difference are replicated over again—as in the findings of other Committees in Europe that while semen donors may have questionable motives, what in the late 1980s was at issue for egg donors was not a corresponding pathology but a liability to exploitation. Potential female donors were regarded as under risk from pressure being put on them to donate in a way that male donors were not; the possibility of money being involved does not carry the same meaning; and so on.

In her discussions with Warnock Committee members, Haimes found that many disliked the idea of semen donation (both as a “sexual” practice and from the point of view of the donor as a “third” man), but that the idea of egg donation was, by contrast, more palatable. It was more palatable on two grounds. First, since semen donation was already tolerated, there could be no objection to egg donation insofar as it was seen “as a parallel activity” (motherhood modeled on fatherhood). Second, however, the idea of egg donation was assimilated to presumptions about other aspects of women’s mothering roles—altruism, concern for others, their passivity as women and being liable to exploitation (motherhood modeled on motherhood). The presumed similarity (two types of donation) thus encompassed a difference between a similarity (the female act was like the male one) and a difference (the female act was compatible with other female acts and different from the male act).

Lack of regulation seemed to worry people in relation to semen donation in particular—there was a sense of excess, fears of a man siring too many children, and so forth. By contrast, egg donation was regarded as more benign, passive, a domesticated intervention. Haimes relates the difference she uncovers to a familial ideology concerned to protect the family’s boundaries. I would add a further comment. In the discussions, the already recognized semen donation was a discursive reference point, a “natural” origin, for thinking about the new, complex and artificially engineered processes of ovum extraction and implantation. But the male act also carried resonances of an unregulated unpredictable “nature,” by contrast with which the female act evoked a potential that could be domesticated to social ends.

On balance, Haimes found that if semen donation had negative (assertive) overtones, egg donation had positive (passive) ones. But if one turns to the second set of analogies, between semen donation and maternal surrogacy, the balance of values is not the same at all. Referring to European views in general, Sissa (1989: 133) asserts that semen donation “is considered perfectly acceptable social practice, whereas the notion of a surrogate mother is often found distressing and shocking.” In other words, between the two analogies there is a difference that turns on the internal difference of each analogy in turn. Thus while semen donation remains a single reference point for the two aspects of motherhood, the process of pursuing the analogies duplicates an internal split between positive and negative aspects of the male act, the bad and the good donor.

We might say this merely repeats the original relationship between certainty and uncertainty as far as fatherhood is concerned. If so, then we might follow Cannell (1990: 673) in seeing that the presumed indissolubility of the mother-child bond was the principal sticking point in the Warnock Committee’s treatment of surrogacy. As far as the common law presumption of fatherhood on the part of the woman’s husband is concerned, continuity between mother and child has to be taken for granted. Cannell points out that of all the reproductive procedures that the new technologies have made available, using them to create surrogate maternity was rejected outright by the Committee. That is, they advised that it be excluded from legislation. The reason was that surrogacy distorts the very relationship between mother and child. They did not see how arrangements for surrogacy could be properly “recognized” (constructed/legislated upon). The HFE Act did find itself able to make a decision on the rights of gamete-donors in the case of a baby being carried by another woman, as we have seen (Section 30); but it was silent on the practice of surrogacy itself, bar making its commercialization illegal. That is because, for all other intents and purposes, the mother is specified as the carrying mother.

What blocks a full equation between semen donation and surrogacy (the giving of services) is the fact that donation was already defined with respect to the recipient—that is, it was tolerated as bringing relief to the recipients: the caring father and the carrying mother. To then contemplate how to recognize a difference between the kinds of carrying mother—whether what was being carried was one’s own child, another child—would have undone the certainty of the conclusions that rested on the fact of donation.

However, public opinion raced ahead. As Cannell demonstrates, the press satisfied itself fairly quickly that there was a difference between good and bad surrogates. Surrogates were positively valued in the context of creating a child as an act of love or altruism for childless women; negatively when they were seen as prostituting their maternity for money.

Where we find what I would call an analogy, as analogies work for Europeans, then, the reproduction of a difference that can be reproduced only as a difference, is in the irreducibility of the kinds of reasons that go to make up a good/bad donor or a good/bad surrogate. The bad surrogate is like the bad donor in exploiting her or his own capacities for personal ends of gain (however obscure), whereas the good surrogate/donor acts out of compassion for, and in the interests of, others. The problem that legislation has is that while it can “recognize” the first—acts of selfishness—its business is to “constitute” the grounds on which the second may or may not be legally recognized as a basis upon which persons can claim rights with respect to one another.

After the event

I have suggested that the search for origins that preoccupies much Euro-American thinking implies a search for foundations that are simultaneously prior in time and prior in logical status. In fact what for Euro-Americans distinguishes an idea as a ground for a statement (a fact) from an idea that aids the imagination (a metaphor) is that a fact is like an originary parent, a metaphor like a child at once produce by and independent from what it illuminates. As cultural critique would have it, the origin comes to seem the indeterminate term.

The search for origins is the search for parenthood. Euro-Americans find the creation of parenthood a problematic demanded by their ideas of continuity and directional time, that is, those that lead them to think about origins in the first place. It is also the case that by contrast with the child, the parent is the indeterminate term. I said at the beginning that a child is doubly defined as an autonomous person and someone’s offspring. Only in the latter, not the former sense, is it like its parent.

The new kinship? The HFE legislation—and other debates—are an interesting mix of deliberate retention of old ideas while trying to find grounds for new ones. Is it that the reproduction of persons is suddenly seen to depend on the reproduction of ideas?


Best, Steven. 1991. “Chaos and entropy: Metaphors in postmodern science and social theory.” Science as Culture 11: 188-226.

Birke, Linda, Susan Himmelweit, and Gail Vines. 1990. Tomorrow’s child: Reproductive technologies in the 90s. London: Virago Press.

Cannell, Fenella. 1990. “Concepts of parenthood: The Warnock report, the Gullick debate, and modern Mmyths.” American Ethnologist 17: 667-686.

Culler, Jonathan. 1979. “Jacques Derrida.” In Structuralism and since, edited by J. Sturrock, 154-180. Oxford: Oxford University Press.

Derrida, Jacques. 1978. Writing and difference. Translated by Alan Bass. Chicago: University of Chicago Press.

Gleick, James. 1987. Chaos: making a new science. London: Heinemann.

Haimes, Erica. 1991. “Gender, gametes and the new reproductive technologies.” Paper given to the British Sociological Association Annual Conference, Manchester.

Hayles, N. Katherine. 1990. “Designs on the body: Norbert Wiener, cybernetics, and the play of metaphor.” History of the Human Sciences 3: 211-28.

Héritier-Augé, Françoise. 1989. “Semen and blood: Some ancient theories concerning their genesis and relationship.” In Fragments for a history of the human body, Part III, edited by M. Feher, R. Naddaff, and N. Tazi, 158-175. New York: Urzone Inc.

Houseman, Michael. 1988. “Towards a complex model of parenthood: Two African tales.” American Ethnologist 15: 658-77.

Kramer, Lloyd. 1989. “Literature, criticism, and historical imagination: The literary challenge of Hayden White and Dominick LaCapra.” In The new cultural history, edited by L. Hunt, 197-128. Berkeley and Los Angeles: University of California Press.

Gleick, James. 1987. Chaos: making a new science. London: Heinemann.

Morgan, Derek and Robert G. Lee. 1991. Human fertilisation and embryology act 1990: Abortion and embryo research, the new law. London: Blackstone Press Ltd.

Sissa, Giulia. 1989. “Subtle bodies.” In Fragments for a history of the human body, Part III, edited by M. Feher, R. Naddaff, and N. Tazi, 133-156. New York: Urzone Inc.

Smith, Paul. 1988. Discerning the subject. Minneapolis: University of Minnesota Press.

Strathern, Marilyn. 1991a. Partial connections. ASAO Special Publication 3. Savage, Maryland: Rowman and Littlefield.

———. 1991b. “Partners and consumers: making relations visible.” The New Literary History 22: 581-601.

Wagner, Roy. 1977. “Analogic kinship: A Daribi example.” American Ethnologist 4: 623-642.

———. 1986. Symbols that stand for themselves. Chicago: University of Chicago Press.

———. 1991. “The fractal person.” In Big men and great men: The personifications of power in Melanesia, edited by M. Godelier and M. Strathern, 159-173. Cambridge: Cambridge University Press.

Warnock, Mary. 1985. A question of life: The Warnock report on human fertilisation and embryology. Oxford: Basil Blackwell.

Weiner, James F. 1987. “Diseases of the soul: sickness, agency and the men’s cult among the Foi of New Guinea.” In Dealing with inequality. Analysing gender relations in Melanesia and beyond, edited by M. Strathern, 255-277. Cambridge: Cambridge University Press.

———. 1988. The heart of the pearlshell: the mythological dimension of Foi sociality, Los Angeles and Berkeley: University of California Press.

Werbner, Richard P. 1989. Ritual passage, sacred journey: The process and organisation of religious movement. Washington: Smithsonian Institution Press.

———. (n.d.) “Trickster and the eternal return: Self-reference in West Sepik world renewal.” Symposium The mother’s brother is the breast: ritual and meaning in the West Sepik, edited by B. Juillerat. Washington: The Smithsonian Institution.

Wolfram, Sybil. 1987. In-laws and outlaws. Kinship and marriage in England. London: Croom Helm.

Qu’est-ce qu’un parent ?

Résumé : Ébauché en 1991, cet article se situe à l’intersection de questions débattues à la suite de la Loi sur la Fécondation et l’Embryologie promulguée au Royaume Uni (1990), de discussions académiques sur l’interprétation, des premières observations d’imagerie fractale, des modélisations à l’époque très récentes de la sociabilité mélanésienne, et du rôle de la connaissance dans le mode de pensée de la parenté anglaise. Inspirée du questionnement sur la parentalité et la parenté posé par Houseman à partir de matériaux sur l’Afrique de l’Ouest, l’auteure explore certaines comparaisons entre — et la possibilité même de comparer — paternité et maternité.

Marilyn Strathern is Emeritus Professor of Social Anthropology, University of Cambridge. Her interests have long been divided between Melanesian and British ethnography. Projects over the last twenty years are reflected in publications on reproductive technologies and intellectual and cultural property rights, while ‘critique of good practice’ has been the umbrella under which she has written about audit and accountability. Some of these themes are brought together in her last book, Kinship, law and the unexpected (2005).


* This paper was presented in 1991 at Embodiment and sociality, a Melanesian Colloquium convened by Maurice Godelier and myself and held at Manchester University. It was in turn based on notes prepared for discussion by the Research Group on The Representation of Kinship in the Context of the New Reproductive Technologies, ESRC Project ROO 23 2537, and the support of the ESRC and of those colleagues is warmly acknowledged. Michael Houseman kindly made a response to the paper, and I remain grateful to him for his original stimulus, as I do to participants at the Colloquium. Parts of the paper have appeared in other contexts, but it needed more work and was never published in this form. It might, nonetheless, have some historical interest. Marilyn Strathern, October 2011.

1. Social scientists (especially anthropologists) are, of course, well versed in the proposition that no text is independent of context. It is the collapse of the distinction that is of interest here. A context subject to interpretation becomes “text,” in the same way as the investigation of grounding assumptions removes their status as grounds.

2. One of the pair of limitations (see above). The other is the irreducible materiality of the signifier: however provisional, the “matter” of the signifier will not go away. (An example is the signification of parent as originator or producer of the child.) The two limitations can be understood as Wagner’s literal (microcosmic) and figurative (macrocosmic) poles in the activity of symbolization.

3. Cf. Rabinow (to be completed). [I leave this as an open-ended acknowledgement of his influence at the time. MS]

4. Or rather, a direction taken is reversible, as in the perception of the dialectic of differance, or the negative/positive feedback loop that modifies while it sustains oscillation. (“One of the first triumphs of cybernetics was to use this model to explain intention tremors and Parkinson’s disease by creating machines that could mimic the behavior characteristics of these disorders” [Hayles 1990: 219].)

5. Hayles (1950: 1990: 212) quotes Wiener on mathematics—what “most of us see as the most factual of all sciences, constitutes the most colossal metaphor imaginable, and must be judged, aesthetically as well as intellectually, in terms of the success of this metaphor.”

6. If this sounds rather “Melanesian” it is because current modes of interpreting Melanesian ethnography have themselves drawn on or fabricated “alternative” discourse.

7. They may be substitutable in certain contexts, but with respect to the definition of parenthood I am pursuing here are not.

8. If I have returned to the metaphysics of presence, it is not to take one philosophical view against another, but because the management of time and process (of which more below) that we find in the reproductive model speaks to a modernist (pluralist) world that is reproduced over and again in these debates. But it is reproduced in a wider context (is part of another text) that introduces all sorts of possibilities for apparently liberating thought—as Hapgood evidently found. In other words, its modernity is already compromised. Postmodernism does not, of course, come after modernism: it is found within it (was already there). I make no apology for either giving it its own name, nor for moving between apparently modern and postmodern vocabularies. This is exactly what the Archbishop, Hapgood, was doing. It is how one describes what is prior that matters. I have my own vocabulary for signifying to myself what I think is/was distinctively modernist in the reproductive model of mid-twentieth century middle class Euro-Americans. The model is founded on an asymmetrical relationship between nature and society, or parent and child, whose joining or connecting, I call merographic.

9. From the standpoint of the critical commentator. In using the term “construction,” I am trying to keep ethnographically close to modernist perceptions of the role of culture in social life.

10. These are, of course, mutually implicated in anyone “construction”, the terms are Wagner’s (1986). Hence the replication of symbolic moments to which I refer later.

11. Each change of scale—the magnitude of phenomena under scrutiny—reintroduces both the same complexity and specifically the same substantive asymmetries. if one is looking for homologies—like the infinite reproduction of the Mandelbrot set—then they rest in the process of “representation” itself. One reproduces over and over again the “same” constellation of symbolic elements.

12. The number starts with a complex number, that is, one itself composed of elements in different planes, as one might imagine a point on a grid reference mapped by two directions. Here is also the resolution of a Zeno-type paradox, helpful to think about in this context.

If you imagine someone at mid-point along a path, and then imagine him or her at the mid-point along the next half of that, and at the mid-point again along the next half, the traveler would never reach his or her destination. The paradox is that we know of course he/she will, even though he/she passes each mid-point exactly as described. The solution is to see that the imagining of this process involves a fractal extension. That is, at each juncture, two different orders or dimensions are being brought together, and it is that conjunction that is repeated over and again in the posing of the paradox. When we are invited to imagine something halfway along its path, and then halfway beyond what is left, we are not proceeding along a straight line at all. Because the first half is not the same as the second half. The first half is half the whole length; the second half is a half of a different length altogether, that is, the length that was left. A quite “different” length is thus being added to the first (the difference between a divided and undivided length). At each juncture a “new” length is added to an “old” length. What is irreducible—and what means that this line of thinking will never get the walker to his/her destination—is the imagined replication of that relationship between “old” and “new” lengths.

13. In the same way as the parent is also only known by, and is thus created by, the child, so too the genetic material it passes on is only known by its manifestation in the child. Indeed, there is a sense in which it is only “passed on” if it is so manifested.

14. But I need to think therefore about other grounding references of the kind Sarah Franklin has reported, e.g. appeals to experience. A sharp difference is drawn between hypothetical (metaphorical) and personal experience.

15. “Consider, then, a situation in which all kin relations and all kinds of relatives are basically alike, and it is a human responsibility to differentiate them. The responsibility of doing so will be our task in understanding kin relationships, as it is man’s role in perhaps the majority of human societies. A mother is another kind of a father, fathering is another kind of mothering” (Wagner 1977: 623).

16. Hence the Melanesian “transformation” between domains of activity such as political and domestic action. The “switch” to which I refer can be understood as between whole orders of knowledge (ways of apprehending the world), and may thus appear as a transformation between types of sociality. Each transformation works, however, to reveal what is already there explicitly given shape, for instance, in the dramatics of performance in which every “act” or scene recapitulates the previous one (see Werbner n.d.).

17. Hierarchical because: “In both stories the motherhood/fatherhood interrelation is represented as being founded upon the necessary articulation of two generational levels, in which either fatherhood (Beti) or motherhood (Samo) occupies a position on both of these levels. A logical hierarchy is thereby established between these two parenthood roles…. Thus, in the Beti example motherhood may be said to be logically subordinated to fatherhood; in the Samo example the reverse holds true.” (Houseman 1988: 671, original emphasis).

18. The story thus collapses into a single generational succession, what Héritier-Augé (1989: 113) describes as a theory of gradual diminishment that lies on a “threshold of conscious recognition” in marriage rules. The process of obliterating maternal endowments is a gradual one measured by the span of legitimate spouse; this supposes that maternal connections are expunged after three generations.

19. Thus “mother” and “father” are both parenthood and kinship terms, but the relationship between the terms differs in the two cases. It is the asymmetry involved in the definition of parenthood that engages his argument.

20. That the Human Fertilization and Embryology Act preserved the English common law presumption that a woman’s husband is the father is interpreted as securing a “continuing link” between child and presumed father (Morgan and Lee 1991: 155).

21. For example, Best (1991: 197) cites one of the ways in which quantum mechanics understands the world: “the scientist cannot identify both the position and momentum of a sub-atomic particle, but must choose which aspect to measure in the most accurate manner. In a strict inverse relation, the more scientists focus on one characteristic, the less they can discern about the other. Hence, as theorized by Werner Heisenberg, there is a strong element of ambiguity or “uncertainty” involved in quantum physics that cannot be eliminated.”

22. Wolfram (1987: 209) is puzzled that Warnock should have emphasized an asymmetry in the closeness of the relationship between mother and father to their child. (Warnock’s model of closeness is motherhood.) Wolfram’s own view of the English kinship system is that it rests on the belief that parents are equally related to their children—at once a rendering of genetic beliefs (genetic endowment from each side is paired) and traditional (anthropological) approaches that have depicted Euro-American kinship reckoning as bilateral. perhaps we see here a precursor to the modeling of parenthood on father- rather than motherhood.

23. Thus Artificial Insemination by Donor (AID) (Donor Insemination - DI) was initially likened to adultery, that is, as though the sexual act had taken place.

24. On ideas of gifts and giving, see Strathern (1991b).

25. However, when it comes to disclosure of information (Sections 33-35), the term parent is used to refer to persons who might be construed as parents if they were not barred by the law itself (i.e. whose only claim is that of donation).

26. Haimes is here referring to the reports of commissions that preceded the Warnock Committee. She also notes that “the whole context of sperm donation is, in these two reports, associated with ideas of inappropriate sexuality: adultery, masturbation and illegitimacy” (1991: 9). By contrast, the woman’s receipt of the donation is regarded as a passive act, essentially asexual in that no “activity” is required of the woman. I adduce the example as one of asymmetry; public attitudes towards the propriety of semen donation as such have undergone considerable relaxation since then.