Kinship and Evolved Psychological Dispositions - Dan Sperber

723

C u r r e n t A n t h ro p o l o g y Volume 43, Number 5, December 2002� 2002 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved 0011-3204/2002/4305-0002$3.00

Kinship and EvolvedPsychologicalDispositions

The Mother’s BrotherControversy Reconsidered1

by Maurice Bloch and Dan Sperber

This article revisits the old controversy concerning the relationof the mother’s brother and sister’s son in patrilineal societies inthe light both of anthropological criticisms of the very notion ofkinship and of evolutionary and epidemiological approaches toculture. It argues that the ritualized patterns of behavior dis-cussed by Radcliffe-Brown, Goody, and others are to be explainedin terms of the interaction of a variety of factors, some local andhistorical, others pertaining to general human dispositions. Inparticular, an evolved disposition to favor relatives can contrib-ute to the development and stabilization of these behaviors notby directly generating them but by making them particularly“catchy” and resilient. In this way, it is possible to recognizeboth that cultural representations and practices are specific to acommunity at a time in its history (rather than mere tokens of ageneral type) and that they are, in essential respects, grounded inthe common evolved psychology of human beings.

m a u r i c e b l o c h is Professor of Anthropology at the LondonSchool of Economics and Political Science (Houghton St., LondonWC2A 2AE, U.K. [[email protected]]). Born in 1939, he waseducated at the University of London (B.A., 1962) and at Cam-bridge University (Ph.D., 1968). His publications include FromBlessing to Violence: History and Ideology in the CircumcisionRitual of the Merina of Madagascar (Cambridge: Cambridge Uni-versity Press, 1986), Prey into Hunter: The Politics of ReligiousExperience (Cambridge: Cambridge University Press, 1992), andHow We Think They Think: Anthropology and Approaches toCognition, Memory, and Literacy (Boulder: Westview Press,1998).

d a n s p e r b e r is Director of Research at the Centre Nationalde Recherches Scientifiques, Paris. He was born in 1942 and edu-cated at the Sorbonne (Licence es Lettres, 1962) and at OxfordUniversity (B.Litt., 1968). Among his publications are RethinkingSymbolism (Cambridge: Cambridge University Press, 1975), OnAnthropological Knowledge (Cambridge: Cambridge UniversityPress, 1985), and Explaining Culture: A Naturalistic Approach(Oxford: Blackwell, 1996).

The present paper was submitted 31 i 01 and accepted 25 x 01.

1. For Jack Goody.

One of the most discussed topics in the history of an-thropology has been the significance of the relationshipbetween mother’s brother and sister’s son in patrilinealsocieties. However, the subject seems to have entirelyfaded from the hot topics of the discipline since the six-ties. We believe that, in reviewing this academic storyof strange excitement and then total neglect, we can bothunderstand some of the fundamental epistemologicalproblems of anthropology and suggest some of the waysin which new approaches might throw light on questionswhich have tended to be abandoned rather than resolved.

The History of the Mother’s BrotherControversy

The behavior which so intrigued anthropologists in-volved the rights, recognized in many unrelated patri-lineal societies, of a male member of the junior gener-ation over the property and even the persons and wivesof senior male members of his mother’s lineage, typicallythe mother’s brother.2 The example which came to bemost discussed was that of the BaThonga of SouthernAfrica because of the particularly full and surprising de-scription of the customs involved given by the early mis-sionary ethnographer Henri Junod in 1912. There therelation primarily concerned the right of mutual insultbetween the sister’s son and the mother’s brother andhis wives and unclear claims to the property of themother’s brother by the sister’s son. The tolerated vio-lence of the behavior, as well as its sexual overtones,contributed to the fascination with the custom and prob-ably titillated the various scholars who discussed thesubject. But it was not so much this one example whichinterested scholars as the conviction that they were deal-ing with a peculiar relationship which occurred againand again in many totally unrelated societies, somethingwhich was all the more unexpected in that it contra-dicted patrilineal organizational principles—sincemother’s brothers and sister’s sons must usually belongto different lineages—and the respect usually accordedto senior generations.

Examples of this peculiar relationship were thought tohave been found among Australian Aborigines and inAmazonia, southern Europe, Oceania, and India, not tomention other parts of Africa. Even today recent eth-nographers have been struck, again and again, by theprominence accorded to this relationship by the peoplethey have studied in many different places, for example,northern India (Jamous 1991), Amazonia (Viveiros deCastro 1992), and Melanesia (Gillison 1993). But this ap-parent recurrence itself raises a problem. The variousmanifestations which so many anthropologists have rec-ognized as instances of the peculiar mother’s brother/sister’s son relationship are clearly cognate. At the sametime, these cases turn out, on closer examination, to bevery varied—sometimes involving symmetrical joking,

2. In a way that is typical of the time, the focus was almost exclu-sively on male roles.

724 F current anthropology Volume 43, Number 5, December 2002

sometimes asymmetrical joking, sometimes avoidance,sometimes significant economic privileges, sometimessexual rights, sometimes only ritual manifesta-tions—and, furthermore, while in some cases it is actualmother’s brothers and sister’s sons who have the rightsin question, the relation sometimes involves broad clas-sificatory groups. The variation is in fact so great that itbecomes very difficult to say exactly what it is that thevarious examples share, and this inevitably has mademany wonder whether the scholars who have turnedtheir attention to the question have not been dealingwith a nonexistent category.

At first, anthropologists assuming a universal historyfor humankind along a single evolutionary path and, im-plicitly, a universal cognitive representation of filiationand marriage saw in such practices as the aggressiveclaim of the sister’s son to his mother’s brother’s prop-erty a survival of mother right and proof of the existenceof an earlier matrilineal stage (Rivers 1924). This expla-nation was then famously dismissed by Radcliffe-Brown(1924), who, using his refutation to demonstrate the char-acter of structural-functional accounts, supplied a syn-chronic explanation for the practice. Thus the contro-versy over the mother’s brother could not have beenmore central to the short history of social anthropolog-ical theory, and the success of Radcliffe-Brown’s argu-ment was a key element in the gradual marginalizationof notions of evolution from the mainstream of thediscipline.

Radcliffe-Brown’s explanation was, at first, mainly interms of the extension of sentiment. He argued that thesentiments of a child toward its mother were extendedto the mother’s family, making the mother’s brother akind of male mother who acted accordingly in a maternalfashion and so gave gifts to his sister’s son. More im-portant, however, was the argument that such customscould only be understood in terms of their function aspart of the total social structure. Radcliffe-Brown’s ar-gument therefore not only went against evolutionismbut also was to be a dramatic demonstration of the valueof what has come to be known as structural-function-alism. For Radcliffe-Brown, therefore, the idea of an iden-tical and single history of humankind was to be aban-doned, but a universalistic element remained in that heassumed a universal cognitive basis for the representa-tion of kinship; mothers were always mothers, and pat-riliny’s attempt to underplay this caused problems whichhad to be resolved by strange customs. Furthermore, be-cause of the commonality of the fundamental buildingblocks of kinship systems, large-scale comparisons couldbe made between societies, which were to be the foun-dations of the new “natural science of society.”

In turn, Radcliffe-Brown was criticized by Fortes (1953)and then by Goody (1959), who, while retaining the fun-damental principle of a synchronic explanation in termsof a systematic social structure, criticized Radcliffe-Brown’s explanation for being overgeneral, since it wouldpredict a much greater degree of universality and uni-formity than the evidence warranted. Goody’s criticismtook the form of noting that, although the sentiments

of children toward their mothers were everywhere thesame, the specific practice in question was found onlyin certain societies that had patrilineal descent groupswithout the counterbalance of matrilineal inheritanceand that any explanation had to be tied to the occurrenceof this type of group. Furthermore, and here followingthe later Radcliffe-Brown, he specified the character ofthe institution much more narrowly than the earlier ev-olutionist writers, insisting on the element of privilegedaggression in the snatching of property by the sister’sson in ritual contexts. This strange custom he explainedas did Fortes, in terms of the contradiction between whathe argued was a universally bilateral kinship system andthe occasionally occurring unilineal descent system. Heargued that sister’s sons were grandchildren of theirmother’s fathers in the kinship system and thereforetheir heirs, while in the descent system they were in noway their successors, since descent was traced only inthe patrilineal line. The tolerated snatching of meat bythe sister’s son at his mother’s brother’s sacrifices re-solved this contradiction because in this way he recov-ered some of his grandparental inheritance from the sonof his maternal grandparents, who had (abusively interms of the kinship system but legitimately in termsof the descent system) received all of that inheritance.Goody clinched this argument with a comparison of twoclosely related groups with different property systems inwhich the degree of inheritance “deprivation” of the sis-ter’s son was correlated with the importance of meatsnatching.

This piece of work is a particularly fine example ofthe structural-functional analyses of its time. It assumes,with a characteristically confident tone, that comparisonof the social structures of different societies will revealrecurring connections between different features which,it can then be assumed, are related in a synchronic causalway. This sort of comparison also implies a belief thatthe basic institutions of societies are everywhere ofmuch the same kind, that they are represented in muchthe same way, and that we know that all human societieshave men and women, marriage, and filiation. Accordingto this way of thinking, patriliny is a particular perspec-tive imposed on the universally recognized facts of pro-creation. The belief in the universality of the basic rep-resentations of kinship of Radcliffe-Brown is thusmodified but not abandoned, since these representations,when they occur, are about natural, objective facts thatexist independently. Furthermore, the emotional reac-tion to a certain state of affairs, in this case ambiguityover filiation, is assumed to be basically the same for allhumans irrespective of culture and to produce, therefore,similar behaviors in similar circumstances. These dif-ferent but related assumptions of a common ground arewhat made the use of comparison as a discovery proce-dure possible. Variations were significant because itcould be assumed that they occurred within the samenatural field consisting of identifiable elements; thus thegeneral principles of Radcliffe-Brown’s natural compar-ative science of society remained possible.

This identity of the basic building blocks of kinship

bloch and sperber Kinship and Evolved Psychological Dispositions F 725

systems is precisely what came under challenge in thesubsequent development of the discipline. The firstclearly expressed formulation of the coming epistemo-logical shift is to be found in Leach’s 1955 paper on mar-riage, and this shift was emphatically repeated and ex-panded in Needham’s introduction to RethinkingKinship and Marriage (1971). The basis of their argu-ments was that marriage and kinship, as understood bysocial and cultural anthropologists, were not externallyexisting phenomena but merely glosses for loosely sim-ilar notions found in different cultures. As Needham putit, there was no such thing as kinship. Subsequently, ina more empirical mood, Schneider (1984) attempted todemonstrate that Austronesian kinship was a funda-mentally different phenomenon from European kinshipand therefore aiming at understanding the former withthe words appropriate for the latter was a source of con-fusion. Thus, generalizing comparisons of kinship sys-tems were impossible because they did not, as was pre-viously assumed, involve comparisons of like with like.

Similar in inspiration but even more startling—thoughto many less convincing in its extreme forms—was thepoint made by a number of feminists that there were nosuch things as women and men beyond a specific cul-tural context. Explicitly drawing on Schneider’s critiqueof kinship, Collier and Yanagisako (1987) argued that thedifferentiation between female and male that anthro-pologists had incorporated into their analyses was a “cul-tural construction” and of a quite different order fromany sexual difference between organisms that might ex-ist in nature. These anti-naıve-empiricist points had twoconsequences for the kind of argument that Radcliffe-Brown, Fortes, and Goody had presented. First of all, aswas noted above, it could be argued that the grand com-parisons of structural-functionalism involved operationslike adding apples and pears, and, secondly, the socialunits, for example, lineages, were not similar “naturalthings” occurring in different societies but different andunique historical/cultural representations constructed indifferent settings and therefore incommensurable (seeKuper 1982). The only reason, according to these writers,that kinship had seemed so similar among different hu-man groups across the globe was an ethnocentric ten-dency to see similarities and overlook differences. Fi-nally, the last universalistic element in the Goodyargument, the similarity of behavioral response in allhumans to similar situations, also came under attack byanthropologists who claimed that emotions too were cul-turally constructed (Rosaldo 1980) and could thereforenot be intuited from introspective sympathy.

The implication of all this for the type of comparativeenterprise that Goody and others had been engaged inseemed clear: it made it impossible. It led, if not nec-essarily at least quite directly, to the deep relativism ofmuch modern anthropology. The systematic comparisonwhich for the structural-functionalists was to be a firststep toward scientific generalizations became clearly il-legitimate if there could be no assurance that the unitsof analyses were commensurate. Those who studied kin-ship had deluded themselves that they had been dealing

with biological facts, which it would be reasonable toassume would be severely constrained by nature andtherefore comparable, while in reality they had beendealing with representations which, it was implicitly as-sumed, were the product of unique histories and there-fore could take any form at all. In the case of the par-ticular example of the mother’s brother controversy, therecurrence of the institution which had intrigued theearlier writers was a mirage. Every case was different,and the very terms of the relationship—mother, brother,sister, and son—did not indicate the same kind of thingin different cultural contexts. Thus, just as structural-functionalism had dealt the first blow to anthropologyas a natural science, the culturalist attack on structural-functionalism seemed to have destroyed any hope of gen-eralization. We had been left with nothing but anecdotesabout the infinity of specific situations in which humanbeings find themselves.

The theoretical history we have just traced can be seenas unidirectional; it is the history of the gradual aban-donment of belief in the possibility of anthropology asa generalizing science. It assumes that because humanbeings can transmit information between individualsthrough symbolic communication they are entirely freeof any natural constraints and essentially different fromother animals, who transmit most, if not all, informationgenetically. Animals must wait for changes in their ge-nomes to become different; humans, in contrast, changewith their representations. The existence of these rep-resentations is made possible by the learning and com-putational potential of the human brain, but their con-tents, it is implicitly assumed, are not at all constrainedor even influenced by genetically inherited brain “hard-ware.” These contents are determined, rather, by his-torico-cultural processes. Human history is therefore lib-erated from biology, and people may represent the worldand each other as they please. The belief in the need forcross-cultural regularities resisting historical specificitybecomes simply wrong, the product of a category mis-take. The extension of the aims of natural science to thestudy of culture and society would be like studyingsmells with rulers.

The aim of this paper is not to deny the validity of atleast some of the criticisms of earlier anthropologicalapproaches which have just been touched on. Indeed, werecognize the relevance of those arguments, and there isno doubt that the whole enterprise of Radcliffe-Brownianstructural-functional analysis rested in part on the du-bious foundations of misplaced naıve realism. We agreewith Leach, Needham, and Schneider that the phenom-ena described by anthropologists under the label of “kin-ship” are cultural and therefore historical constructionsand that people’s thoughts and actions are about theseconstructions rather than about unmediated facts of bi-ological kinship. The implicit argument which wouldsee representations of kinship, marriage, and gender asmerely the inevitable recognition of “the way things are”will not do. We will argue, however, that this does notmean that the attempt to invoke natural factors or evenbiological factors as explanations of such cultural rep-


resentations must be abandoned as though these repre-sentations and the people who hold them had somehowfloated free from the earth into the immaterial clouds ofhistory. Antirealism too can be utterly naıve.

We choose the example of the mother’s brother/sister’sson relationship in patrilineal societies to demonstrateour argument simply because it has been so critical inthe history of the discipline, and we try to show that itis possible to envisage, in a case such as this, an approachwhich combines the particular with the general, al-though we must recognize that the actual carrying outof such a study lies beyond what we can do here.

The abandonment of overpowerful theories in anthro-pology came, in the first place, from the realization thatthe implicit and explicit cultural “universals” of tradi-tional anthropology were not as uniform as they hadbeen assumed to be. But anthropologists who argue fora radically relativistic constructivism often seem to lackconfidence in their own arguments. Their reasoning hastaken them to a point that negates what all those witha reasonable acquaintance with the ethnographic recordknow—which is that the regularities which have fasci-nated the discipline since its inception are surprisinglyevident. Thus, it is common for younger anthropologists,reared on the diet of relativism which the studies men-tioned above exemplify, to be shocked by discovering theold chestnuts of traditional anthropology in their field-work just when they had been convinced that these weremerely antique illusions.3

The dilemma that this particular history reveals is, infact, typical of the subject matter of anthropology as awhole. What happens is that, first of all, some cross-cultural regularities are recognized: the incest taboo, forexample. These lead to quick explanations in terms ofthe evolution of culture and their “function” for societyas a whole, for individual well-being, or for reproductivesuccess. These explanations are then shown to be basedon a gross exaggeration of the unity of the phenomenato be explained. Then explanation is abandoned alto-gether and declared impossible, leaving anthropologistsand, even more, the wider public with the feeling thatthe original question has been more evaded than faced.In this way the very idea of the possibility of anthro-pology is destroyed.

The Epidemiological Approach toRepresentations

The aim of this paper is to shun such evasion and tosketch a theoretical model applied to a particularcase—in other words, to see how a possible explanationmight be framed in the case of a particular example of

3. Maurice Bloch remembers how, as a student, he was bored withthe mother’s brother controversy and convinced that it was an in-significant aberration in the history of the subject but subsequently,during fieldwork in Madagascar, had to listen all night to a drunkendlessly repeating, “I am your sister’s son, and it is your duty togive me a drink.” He then felt haunted by Radcliffe-Brown.

one of these “obvious” regularities, the varied but similarpeculiar relationships of the mother’s brother and thesister’s son in different societies. We want to do thiswithout either exaggerating the unity of the phenome-non or avoiding the problems discussed above concern-ing misplaced realism, which recent theoretical criticismhas well illuminated.

What is involved in explaining a cultural phenome-non? Here is a way of framing the question. All membersof a human community are linked to one another acrosstime and space by a flow of information. The informationis about themselves, their environment, their past, theirbeliefs, their desires and fears, their skills and practices.The flow has rapid and slow currents, narrow rivuletsand large streams, confluence and divisions. All the in-formation in this flow is subject to distortion and decay.Most of it is about some here-and-now situation and doesnot flow much beyond it. Still, some of it is more stablein content and more widely distributed, being shared bymany or even most members of the community. Whenanthropologists talk of culture, they refer to this widelyshared information.

What explains the existence and contents of culturein the social flow of information? An answer of a sort isprovided by modern interpretive anthropology, whichaims to show that the elements of a culture (or of acultural subsystem) cohere and constitute an integratedworldview (see in particular Geertz 1973). This is notthe approach we favor. Without denying the insightful-ness of such interpretive scholarship and the relative sys-tematicity of culture, we are among those who have ar-gued that this systematicity is often much greater in theanthropologists’ interpretation than in the culture itself(e.g., Leach 1954, Bloch 1977, Sperber 1985a) and there-fore is exaggerated (as is acknowledged by James Boon[1982:3–26], who speaks approvingly of the “exaggerationof cultures”). More important, even if cultures were assystematic as claimed, this would fall far short of ex-plaining the spread and stability of these coherentwholes, unless one were to take as given that there arefactors and mechanisms in the flow of information thatsomehow promote systematicity. Rather than assumingtheir existence, we favor studying the factors and mech-anisms actually at work in the spread and stabilizationof cultural phenomena and leaving as an open questionthe degree and manner in which they may indeed pro-mote systematicity.

Our explanatory approach to this flow of informationin society is that of the “epidemiology of representa-tions” (Sperber 1985b, 1996). It is naturalistic—that is,it aims at describing and explaining cultural phenomenain terms of processes and mechanisms the causal powersof which are wholly grounded in their natural (or “ma-terial”) properties. More specifically, the kind of natu-ralistic explanation of cultural phenomena we favor in-vokes two kinds of small-scale processes: psychologicalprocesses within individuals and physical, biological,and psycho-physical interactions between individualsand their immediate environment (including interac-tions with other individuals) that we call “ecological”


processes. Typically, the scale of the processes invokedis much smaller than that of the cultural phenomenadescribed and explained in terms of them. It is the ar-ticulation of large numbers of these microprocesses thatallows one to redescribe and explain cultural macro-phenomena. This contrasts with more standard socialscience accounts that explain cultural macrophenomenain terms of other social and cultural macrophenomena.4

We view, then, the flow of information as a naturalprocess occurring in the form of causal chains of mi-croevents that take place both in individual mind/brainsand in the shared environment of the individuals in-volved. Inside minds, we are dealing with processes ofperception, inference, remembering, decision, and actionplanning and with the mental representations (memo-ries, beliefs, desires, plans) that these processes deploy.In the environment, we are dealing with a variety ofbehaviors often involving artifacts and in particular withthe production and reception of public representationsthat can take the form of behaviors such as gestures orutterances or of artifacts such as writings. We call theserepresentations “public” because, unlike mental repre-sentations, they occur not within brains but in the sharedenvironment of several persons. Thus not just discourseaddressed to a crowd but also words whispered in some-one’s ear are “public” in the intended sense. Mentalevents cause public events, which in turn cause mentalevents, and these chains of alternating mind-internal andmind-external events carry information from individualto individual. A simple example is provided by a folktale,in which the main mental events are those of compre-hension, remembering, recall, and speech planning andthe main public events are tellings of the tale. Whatmakes a particular story a folktale is the fact that re-peated sequences of these mental and public events suc-ceed in distributing a stable story across a populationover time.

All these events taking place inside and outside in-dividual minds are material events: changes in brainstates, on the one hand, and changes in the immediateenvironment of individuals, on the other. As materialevents, they possess causal powers and can be invokedas causes and effects in naturalistic causal explanations.They differ in this respect from the abstract meaningsinvoked in interpretive explanation (see Sperber 1985a:chap. 1). That meanings can be causes is contentious,and what kind of causal power they might have, if any,is obscure (see Jacob 1997). For instance, attributing toa folktale a meaning that coheres with, say, basic valuesof the culture in which it is told may, in a way, “makesense” of the tale, but it does not come near explainingits distribution and hence its existence as a folktale inthat particular culture.

It could be objected that the microevents invoked inan epidemiological approach are at the level of individual

4. Of course, explaining cultural phenomena in terms of micro-interactions is not new in anthropology. The work of Fredrik Barth(e.g., 1975, 1987), for example, has been a source of inspiration tothe epidemiological approach.

minds and behaviors. How, then, can their study helpexplain cultural macrophenomena that exist not on anindividual but on a societal scale? We have already sug-gested that these macrocultural phenomena are made up,at a microscopic level, of these causally linked micro-events. To this it is sometimes objected that the vastmajority of these microevents cannot be observed: an-thropologists will never witness more than a very smallsample of the public microevents involved, and mentalevents cannot be observed at all. Here, however, the com-parison with medical epidemiology should help disposeof this objection.

Epidemiological phenomena such as epidemics aremacrophenomena occurring at the level of populations,but they are made up of microphenomena of individualpathology and interindividual transmission. In mostcases individual pathological processes are not directlyobservable and are known only through symptoms andtests, while the vast majority of microevents of diseasetransmission go unobserved. This, however, has been achallenge rather than an impediment to the developmentof medical epidemiology. In the epidemiology of repre-sentations the situation is, if anything, better than in theepidemiology of diseases. Our communicative and in-terpretive abilities give us a great amount of fine-grainedinformation about the representations we entertain andabout the process they undergo, whereas pain and otherperceptible symptoms generally provide much coarserand harder-to-interpret information about our patholo-gies. Also, most events of cultural transmission requirethe attention of the participants, whereas pathologicalcontagion is typically stealthy. Hence cultural transmis-sion is much easier to spot and observe than diseasetransmission.

In spite of the limited evidence at its disposal, medicalepidemiology has provided outstanding causal explana-tions of epidemiological phenomena. It has done so onlyoccasionally by following actual causal chains of trans-mission and much more often by helping to identify thecausal factors and mechanisms at work both within andacross individual organisms. Mutatis mutandis, the taskof the epidemiology of representations is not to describein any detail the actual causal chains that stabilize (ordestabilize) a particular cultural representation (althoughin some cases it is of great historical interest to be ableto do so) but to identify factors and processes that helpexplain the existence and effect of these causal chains.For instance, showing that a particular folktale has anoptimal structure for human memory and that there arerecurring social situations in a given society in whichpeople are motivated to tell it or to have it told helpsexplain why the tale is told again and again with littleor no distortion of content in that society.

The central question on which an epidemiological ap-proach focuses is what causes some representations andpractices to become and remain widespread and rela-


tively stable in content in a given society at a given time.5

In so framing the question, we depart from the goal ofgenerally explaining all or even most sociocultural phe-nomena in one and the same way, either as fulfilling afunction (a coarse functionalist approach) or as contrib-uting to reproductive success (a coarse sociobiologicalapproach). True, from an epidemiological point of view,all explanations of sociocultural phenomena will haveto invoke both mind-external ecological factors linkedto the transmission of cultural contents and mind-in-ternal psychological factors linked to the mental repre-sentation and processing of these contents. However, theparticular factors at play and the way they combine varywith each case (just as, in medical epidemiology, a dif-ferent combination of organism-internal physiologicalfactors and of organism-external environmental factorscharacterizes each disease). Because of this multiplicityof co-occurring causes, we aim only at identifying someof the factors that contribute to explaining particularinstances. These factors play a causal role only in specifichistorical and environmental circumstances and there-fore can never be sufficient to explain the local culturalforms. Caused in part by the same factors, these formshave recognizable similarities—which we aim to helpexplain. However, we merely identify a couple of im-portant and recurring factors among many other diverg-ing factors: each cultural form in its full local specificsis therefore unique to its particular historical context.

This, of course, is, first of all, simply to return, thoughmore explicitly and critically, to the general multifac-torial explanations that were typical of anthropology be-fore its recent relativist turn. Two things may be new,though. Rather than accepting implicitly some nonde-script naturalism or objectivism about kinship, we ap-peal quite explicitly to naturalistic considerations aboutevolved, genetically transmitted psychological predis-positions. The result of this explicitly naturalistic ac-count is, however, weaker in its predictive pretensionsthan the type of account found, for example, in Goody’sfunctionalist thesis. There the sister’s son’s privilegeseemed an almost necessary solution to a structuralproblem found in certain patrilineal societies. Similarly,this solution was to account for the particular form ofthe institution, for example, the snatching of significantproperty. According to our more explicitly naturalistic

5. How stable do representations have to be to count as “stable”?From the epidemiological viewpoint, there is no expectation thatthere will be a neat bipartition, among all representations that in-habit a human population, between individual representations thatnever stabilize in the community, on the one hand, and culturalrepresentations that are transmitted over time and social space withrelatively little modification, on the other. We expect, on the con-trary, to have a continuum of cases between the idiosyncratic andthe widely cultural. This viewpoint differs quite radically from thememetic approach to culture of Richard Dawkins and others (e.g.,Dawkins 1976, Blakemore 1999), for which memes are true repli-cators and other mental contents are not. One might wonder, then,when a representation is stable enough to be seen as a culturalrepresentation. We argue, against that very question, that, from ananthropological point of view, representations are best viewed asmore or less cultural depending on the breadth, duration, and sta-bility of their distribution.

but at the same time more modest account, there aresome factors that increase the chances that the sister’sson privilege will stabilize as a cultural form in thesesocieties, and we can expect and not be disturbed by awide range of unexplained variation in practices becausethese will always be combined with many other factorsand many different histories. We avoid, or so we hope,the too-strong explanations of functionalism, old-stylecultural evolutionism, and sociobiology without givingup on causal explanation.

A few simple examples will give an idea of the rangeof factors that an epidemiological approach would con-sider relevant and the complex interrelation betweenmind-internal and mind-external factors. Density of pop-ulation is a mind-external factor in the stabilization ofdrumming as a means of communication. The fact thatpercussion sounds tend to preempt human attention isa mind-internal factor in the culturally stabilized usesof percussion instruments. The relative ease with whichhuman memory retains texts with specific prosodies isa mind-internal factor in the stabilization of variousforms of poetry; familiarity with specific, historicallyevolved poetic forms is a mind-internal factor in the ac-ceptability, learnability, and therefore chances of cul-tural stabilization of new poetic works. The effectivenessof internal combustion engines for moving vehicles is amind-external factor contributing to the stabilization ofthe techniques involved in constructing and maintainingthese engines. Untutored human minds do not, however,spontaneously or even easily acquire these techniques;hence the recognition of the effectiveness of internalcombustion is a mind-internal motivating factor in thesetting up of appropriate institutional teaching withoutwhich the relevant technologies would not stabilize. In-stitutional teaching itself involves a complex articula-tion of mind-internal and mind-external factors.

As these examples illustrate, both mind-external andmind-internal factors explaining cultural phenomenacan pertain just to the natural history of the human spe-cies and its environment or involve also the socioculturalhistory of the populations involved. On the mind-exter-nal side, density of population is a natural factor that isfound in all living species but can be modified by culturalfactors. Demographic density has a wide variety of cul-tural effects, the stabilization of drummed communi-cation in some low-density populations being a marginalbut obvious illustration. On the mind-external sideagain, the presence in the environment of vehicles pow-ered by internal combustion engines is a wholly culturalfactor—which does not mean that it is nonnatural (it is,after all, the product of evolved mental mechanisms ex-ploiting natural laws)—that contributes, among manyother sociocultural effects, to the stabilization of thetechniques necessary for their construction and main-tenance. On the mind-internal side, the tendency of hu-man attention to be preempted by percussion sounds,although it can be culturally modified, is basically a nat-ural trait that humans share with other animals. Theability to organize knowledge in a hierarchy of conceptsis typically human, and although it is likely to have a


strong natural basis it is certainly enhanced by language,writing, and formal teaching. Familiarity with specificpoetic forms is a wholly cultural trait. This illustratesan important difference, among several, between the ep-idemiology of diseases and the epidemiology of repre-sentations: culture occurs both inside and outside ofminds, whereas diseases qua diseases occur only insideorganisms.

The epidemiological model therefore does not deny thecomplexity of the process of human history. It fully rec-ognizes that culture is both in us and outside—that it isnot (even remotely) just a matter of human beings withgenetically determined mind/brains reacting to diverseenvironments according to the dictates of their nature.But the recognition of this complexity and of the uniquefact that humans are beings that, in a strong and im-portant sense, make themselves still leaves room for con-sidering, inter alia, the role of factors such as humanpsychological dispositions resulting from natural evo-lution. However, just as cultural patterns are never sim-ple phenotypic expressions of genes, they are never sim-ple social-scale projections of the individual mind.Culture is not human mentation writ large. It is, rather,the interaction of psychological dispositions with mind-external factors in a population that can best explain thesporadic recurrence of certain types of behaviors andnorms in a whole variety of guises. The inability of othermodels to do this—an inability common in the socialsciences—has left anthropology ill-equipped to explainmany of the cross-cultural regularities which have, inthe past, rightly fascinated it.

A rich example of the relationship between evolvedpsychological dispositions, mind-external factors, andcultural phenomena is afforded by the case of language.A common assumption in cognitive psychology is thathumans come equipped with a language faculty. Thislanguage faculty is neither a language nor a dispositionthat generates a language in the individual ex nihilo buta disposition to acquire a specific language on the basisof external linguistic inputs. The disposition is assumedto work like this: Infants react differently to sound pat-terns typical of human speech: they pay particular at-tention to these sounds, analyze them differently fromother sounds, look for special evidence such as speaker’sgaze in order to associate meaning with sound, structuremeaning in partly preformed ways, test their knowledgeby themselves producing speech, and generally developa competence in the language of their community. Thatthe language acquired by the members of a communitydepends on the public linguistic productions encoun-tered in this community is a truism. However, the lan-guages found in all human communities depend on thepsychological disposition that individuals bring to thetask of language acquisition. Generally, human lan-guages have to be learnable on the basis of this dispo-sition. More specifically, phonetic, syntactic, and se-mantic forms are more likely to stabilize when they aremore easily learnable. All so-called natural human lan-guages—that is, languages the evolution of which is es-sentially the output of spontaneous collective linguistic

activity—will therefore exhibit structural features thatmake them highly learnable as a first language byhumans.

Languages—Chinese, English, Maori, and so forth—differ because they have different histories, with a varietyof factors such as population movements, social strati-fication, and the presence or absence of writing affectingthese histories in subtle ways. However, these mind-ex-ternal, place-and-time-specific factors interact in everygeneration with the language faculty found in every hu-man. It is this interaction that determines the relativestability and the slow transformation of languages andputs limits on their variability. For a variety of socio-historical reasons, topics of conversation, preferredwords, socially valued patterns of speech, and so on, varycontinuously over time in such a way that every gen-eration is presented with a somewhat different samplingof linguistic inputs, to which it reacts, in the acquisitionprocess, by unconsciously bringing about minor changesin the underlying grammar. Generally, whereas day-to-day cultural changes in language use may introduce newidiosyncrasies and difficulties such as hard-to-pronounceborrowed words, the language-learning disposition op-erating at the generational time scale pulls the mentalrepresentations of these inputs toward more regular andmore easily remembered forms. For instance, the moredifficult phonology of borrowed words or the more dif-ficult semantics of meanings stipulated as part of so-phisticated theories are likely to be normalized by lan-guage learners in the direction of easier forms. Thisdetermines a slow evolution of languages that is con-strained both by the necessity of intergenerational com-munication and by the universal constraints of languageacquisition.

The case of language learning, therefore, illustrateshow the existence of a genetically inherited dispositionis a factor in the stabilization of cultural forms not bydirectly generating these forms but by causing learnersto pay special attention to certain types of stimuli andto use—and sometimes distort—the evidence providedby these stimuli in specific ways. This, of course, leavesroom for much cultural variability. Moreover, disposi-tions capable of affecting cultural contents may be moreor less rigidly constraining, the language-acquisition de-vice envisaged by Chomskyans being on the more con-straining side. In general, cultural representations de-parting from those favored by underlying dispositions,though possible, do not stabilize as easily. In the absenceof other stabilizing factors counterbalancing the dispo-sitions (e.g., institutional support), hard-to-learn repre-sentations tend to get transformed in the process of trans-mission in the direction favored by the dispositions.

The epidemiological approach to culture provides away of understanding the relationship between psychol-ogy and culture that neither denies the role of psychologynor reduces culture to mind. In a nutshell, the idea isthat psychological dispositions in general (whetherevolved basic dispositions or culturally developed dis-positions) modify the probability—and only the proba-bility—that representations or practices of some specific


tenor will spread, stabilize, and maintain a cultural levelof distribution.

How might all this help explain the regularities in therelationship between mother’s brother and sister’s sonin patrilineal societies that are the topic of this article?To this we now turn.

Applying the Theory to the Mother’s Brother/Sister’s Son Relation

Underlying the theories of the structural-functionalistsconcerning the mother’s brother/sister’s son relation inpatrilineal societies was the assumption that all humanbeings really reckon kinship bilaterally. This made theoccurrence of unilineal rules to form descent groupssomething which somehow “went against nature.” ThusFortes (1969) contrasted the domestic domain, in whichrelations were governed by biology and natural emotions,with the lineage domain, which was constrained by po-litico-jural considerations in conflict with this biology.For him, therefore, the claims of the sister’s son were akind of reassertion of underlying bilaterality. Goody, al-though distancing himself somewhat from the Fortesianformulation, seemed to imply something similar in thatthe reason the sister’s son was being “cheated” of hisinheritance by the patrilineal rule was that in reality he,like the maternal uncle’s children, was a true descendantof his mother’s parents. The objection to Fortes’s andGoody’s position, however, has been, as we have seen,that they seemed to assume that people acted in termsof genetic relations rather than in terms of a very dif-ferent thing, their representation of socially specified re-lations. But what if there were some indirect causal linkbetween social representations and genetic relations?Then the accusation of naıve empiricism might fall awayand the Fortes/Goody argument might be partly rein-stated. How this might be possible is what much of therest of this paper is about.

We begin by noting that support for the structural-functionalists’ assumption of the universal bilateralityof kinship seems to come from an unexpected source.This is Hamilton’s (1964) neo-Darwinian explanation ofkin altruism and its development in sociobiological the-ory. However, this kind of theory has been rejected outof hand by most social and cultural anthropologists (e.g.,Sahlins 1976). It is necessary to outline the theory of kinaltruism and why it has been rejected to see if, after all,it might not be used legitimately in favor of the kind ofargument implicit in the writings of Goody and Fortes.

The by-now familiar kin-altruism argument can besummarized as follows: Genealogical relationships in thestrict biological sense exist among all organisms, in-cluding humans. The transmission of heritable biologicaltraits through genealogical relationships is what makesnatural selection possible. Natural selection favors geneswhich have the effect, given the environment, of ren-dering more probable more replications of themselves infuture generations. This includes genes that promote the

reproduction of the organism in which they are located,genes that promote behaviors favorable to the survivaland reproduction of descendants of the organism inwhich they are located, and also—and this is fundamen-tal to Hamilton’s thesis—genes that promote survivaland reproduction in yet other organisms which, beinggenealogically related, are likely to carry copies of thesame genes. A gene causing an organism to pay a costor even to sacrifice itself for the benefit of its lateral kinmay thereby increase the number of copies of itself inthe next generation not through the descendants of thecost-paying or self-sacrificing organism (which maythereby lose its chance of reproducing at all) but throughthe descendants of the “altruistic” organism’s kin, whoare likely to carry the very same gene.

The potential contribution of kin altruism to what isknown as “inclusive fitness” favors the emergence of adisposition to helpful behavior adjusted to the genealog-ical distance between the altruist and the beneficiary.For such a disposition to exert itself, the organism musthave the possibility of discriminating kin from nonkinand, among kin, degrees of relatedness. This does notmean, of course, that the organism must have the con-ceptual resources to represent genealogical relatednessand its degrees precisely and as such. What it means isthat, if the ecology is such that degree of relatedness can,at least roughly, be discriminated thanks to some simplecriterion such as smell, appearance, or habitat, then adisposition exploiting this possibility may be selectedfor.

The importance of the theory of kin altruism for ev-olutionary biology and for the sociobiological study ofanimal behavior is not in dispute, but what are its con-sequences, if any, for the study of human behavior? Atfirst sight this theory, transposed directly to humans,would predict that the requirements of this altruismshould, in humans, favor an instinctually based universalbilateral recognition of kinship. This would be a priorisupport for the structural-functionalists’ assumption.Here, however, is where the objections of most anthro-pologists come in.

These objections are fundamentally two. First, thegreat variability in kinship systems throughout the globeseems unaccountable in terms of panhuman character-istics. Second, humans live in the world via their rep-resentations, and how one gets from genes to represen-tations or norms has simply not been thought throughin the sociobiological literature (which has been criti-cized precisely on this ground by evolutionary psychol-ogists [see Tooby and Cosmides 1992]).

The first objection means that the explanation interms of genes is far too direct. One should note, how-ever, that the sociobiological position not only is com-patible with the recognition of some degree of variabilitybut also purports to explain it. The expression of genesis always contingent on environmental factors, and itmay be part of the contribution of a gene to the fitnessof the organism that it has different phenotypic expres-sions in different environments. For instance, the sex ofmany reptiles is determined not directly by their genes


but by the temperature at which eggs are incubated, fe-males developing better, it seems, and being more oftenborn in a warmer environment and males in a colder one(Shine, Elphick, and Harlow 1995).

Closer to our present concern, Alexander (1979) offersan explanation of both matrilineal inheritance and sis-ter’s-son rights in patrilineal societies in terms of un-certainty of paternity. An evolved disposition to favorkin should be sensitive to degrees of doubt or certaintyof relatedness. In particular, a man’s investment in hisputative children should be sensitive to his degree ofconfidence that he is actually their biological father. Ifthere are reasons that this degree of confidence shouldbe low, then a man’s closest relatives in the next gen-eration may well be his sister’s children. On this basis,Alexander predicts “that a general society-wide loweringof confidence of paternity will lead to a society-wideprominence, or institutionalization, of mother’s brotheras an appropriate male dispenser of parental benefits”(1979:172). One may accept the premise that there is anevolved disposition to favor kin that is sensitive to con-fidence in relatedness and yet doubt Alexander’s conclu-sion, in particular regarding the institutionalization ofmatrilineal inheritance. True, there is ethnographic ev-idence that confidence in paternity tends, with excep-tions, to be lower in matrilineal than in patrilineal so-ciety, as the case of the 19th-century Nayars illustrates(Gough 1959), but it is most probably even lower in so-cieties which have neither matrilineal nor patrilineal de-scent groups (Gibson 1986, Stack 1983). Furthermore, acorrelation is not sufficient to determine that there is adirect causal relationship, let alone what the directionof such a causal relationship might be.

The ethnographic and historical record shows thatmatrilineality and patrilineality and related patterns ofinheritance are fairly stable systems, with very rare doc-umented examples (such as Barnes 1951) of a society’sshifting from one to the other. In contrast, changes insexual mores toward or away from greater permissive-ness and associated lower confidence in paternity arevery common and may be caused by rapidly shifting ec-onomic, demographic, or ideological factors. It cannot bethe case, then, that a lowering of confidence in paternitysystematically or even frequently leads to the institu-tionalization of matrilineality. Alexander’s claim, there-fore, is at best unconvincing. One could, for that matter,argue that the lower confidence in paternity in matrilin-eal society is an effect rather than (or as much as) a causeof the descent system. When the inheritance system ismatrilineal, then a man knows that his heirs will be hissister’s children rather than those of his wife. Hischances of investing in his wife’s children’s welfare maybe further reduced by rules of separate residence of thespouses such as are often found in matrilineal societies.To the extent that the opportunities for a man to investresources in his wife’s children are limited, it may matterrelatively less whether these children are biologically hisown, especially if the counterpart of greater paternitydoubts is a greater chance of having children with othermen’s wives. This fits well with the common ethno-

graphic observation that in most matrilineal societiesthere is less control over the sexual fidelity of women.

Extending Alexander’s line of reasoning to the casewith which we are concerned here, one would predictthat the chances of having institutionalized privilegesfor the sister’s son in an otherwise truly patrilineal sys-tem will be greater when paternity doubts are greater(but not great enough to tip the system over towardmatrilineality). In this case, however, we know of noevidence of a correlation between institutionalized priv-ileges of sister’s son and paternity doubts, let alone acausal link in the hypothesized direction.

The second standard anthropological objection to a bi-ological account implies that, even if we accept that adisposition to Hamiltonian kin altruism is biologicallyadvantageous and therefore likely to have somehowevolved (something which is clearly plausible), it is notclear at all what would follow regarding cultural normsof human behavior. The answer is probably nothing di-rectly and unconditionally, since dispositions to behav-ior need not actually lead to behavior, let alone to cul-turally codified behavior; they may be offset or inhibitedin many ways. Moreover, assuming that a disposition isnot inhibited, it still need not be reflected in a culturalnorm. In most human society, for instance, the dispo-sition to use, under certain conditions, an eyebrow flashas a sign of recognition is both uninhibited and culturallyuncodified (see Eibl-Eibesfeldt 1975). Should we, then,as do most cultural and social anthropologists, simplyforget about all this biological stuff and, along with thetheologians and philosophers of old, recognize that thecategorical uniqueness of human beings frees them com-pletely from animality?

The epidemiological approach offers a way of avoidingthis type of dismissal while taking into account what isvaluable in the objections. Let us accept, as a hypothesis,that there is an evolved disposition to try to differentiatepeople in a way sensitive to their degree of genealogicalrelatedness to self. It is most unlikely that such a dis-position would be such as to cause the individual to seekactual genealogical information. It would be rather a dis-position merely to seek whatever available informationmight indicate relatedness to self.6 Now, such a dispo-sition would favor the cultural stabilization of systemsof representation providing for such ego-centered differ-entiation without determining their exact nature. Thedisposition would not be the source of these represen-tations; these would arise as part of the process of dis-tribution of ideas and practices—the historical dialecticof thought and communications, so to speak—and itsinteraction with the individual cognitive developmentof the members of every new generation. The epidemi-ological approach seeks factors explaining the transfor-mation and stabilization of representations in the pro-

6. Hirschfeld (1984) can be read as suggesting a similar approachand insisting, quite rightly, that an essential relatedness and notjust any kind of relatedness is aimed at, but his description of thiskind of relatedness in terms of a “natural resemblance” seems tous inadequate.


cess of their transmission, including biological factors.It does not pretend, as might a classical sociobiologicalapproach, that these biological factors somehow generatethe representations or that culturally sanctioned behav-iors are phenotypic expressions of genes.

One prediction that would follow from the hypothesiswe are considering is that individuals would tend toshow interest in evidence of relatedness, whether or notculturally codified. For instance, if a single kinship cat-egory included full sibling, half-sibling, and more distantrelatives, with the same cultural norms of behavior vis-a-vis all, individuals would nevertheless tend to differ-entiate both cognitively and behaviorally between thedifferent types of individual falling into this category(see, e.g., Bloch 1998). This further interest could be car-ried out individually without being particularly cultur-ally condoned, as we have just envisaged, or it couldcontribute to the stabilization of further cultural repre-sentations (e.g., folk theories, tales, alternative or com-plementary terminologies for kin) drawing finer-graineddistinctions than the basic kinship-terms system. Inother words, whenever representations involving clas-sifications and norms which distinguish kin in terms ofcloseness appeared amidst the babble and multiplicity ofother representations caused either by individual imag-inations and circumstances or by more general socio-historical circumstances, these particular representa-tions would seem strangely “right,” “attractive,”“natural,” or “obvious” to people. This would be the casewithout individuals’ being at all sure why these repre-sentations had these qualities, and even if they gave rea-sons these reasons would often be merely post hocrationalizations.

Assuming this general framework, we would make thefollowing predictions: In unilineal systems where trans-mission of rights and goods and generally helpful behav-ior creates an inequality of treatment among individualsthat are equally closely related to ego and therefore goesagainst the predisposition in question, there should be ageneral, nondeterministic tendency to compensate forthis imbalance. Norms or institutions capable of playing,in such a system, a compensatory role would simplystand a greater chance of stabilizing than in systemswhere the imbalance did not exist in the first place. Thespecial rights of the sister’s son found in some patrilinealcultures could well be a case in point.

The relationship between biological disposition andcultural norm that we are envisaging in this case is onebetween a biological causal factor that is obviously notsufficient and maybe not necessary but such as to rendermore probable the emergence and stabilization of normsof the type in question. We emphasize that this moresophisticated naturalism makes, in this case, weakerclaims than the common-sense naturalism of anthro-pologists such as the 19th-century cultural evolutionistsand Malinowski, Radcliffe-Brown, Fortes, and Goody.According to their common-sense naturalism, there arenatural kinship facts that people are somehow aware ofand that guide their sentiments and behaviors. Thismakes a strong universalistic claim about human cog-

nition, emotion, and behavior, which are taken to beneatly attuned to natural facts. If these classical claimsappear misleadingly weaker and more acceptable thanthose we are tentatively considering here, it is only be-cause they are made, for the most part, implicitly,whereas we have tried to spell out a possible naturalisticapproach.

According to the approach we are considering, thereare indeed biological facts and, in particular, genealogicalrelationships. These, however, need not be cognized assuch by people. A predisposition to attend to reliablecorrelates of these relationships cognitively, emotion-ally, or behaviorally in one or several of a multiplicityof possible ways is likely to have evolved in many spe-cies, including the human species. In humans, this at-tention to relatedness encounters a wealth of relevantcultural inputs. More specifically, developing children,searching their environment for evidence of relatednessto others, find kinship terms (“kinship” now in the cul-tural rather than the biological sense), people identifiedas related to them by means of these terms, do’s anddon’ts relating to kinship categories, folk theories, etc.Because of their evolved disposition, they attend to thisinformation or even seek it, retain it, use it to guide theirbehavior, and become, in turn, transmitters of suchinformation.

At this stage we seem to be just defending a weakened,updated and explicit version of the implicit or less ex-plicit naturalistic claims of Fortes and Goody regardingthe mother’s brother/sister’s son relation in certain pat-rilineal societies. In fact, given the sweeping and carelessway in which these claims have been dismissed, this isworth doing in any case. We are defending them, how-ever, in a way that is not contradicted by the very realuniqueness of each case. Furthermore, in contrast to so-ciobiologists assuming a fairly direct connection be-tween genes and culture, we claim only an indirect re-lationship of genetically favored receptivity to specificinformation, favoring in turn the stabilization of culturalrepresentations of a more or less specific tenor.

Why Ritualized Transgression?

From Junod to Goody, ethnographers have stressed thetransgressive style in which the sister’s son’s rights areexerted. This may take many forms, from ritualized in-sults among the BaThonga to ritualized snatching ofmeat among the Lo Dagaba. Why should it be so? Thegeneral approach we are proposing might help us under-stand not just the recurrence of the recognition of thesubsidiary rights of the sister’s son in his mother’sbrother’s property but also the ritualized transgressionsso often involved in exerting those rights.

From a cultural-epidemiological point of view, culturalnorms (such as the norm that authorizes a Lo Dagabaman to snatch meat from his mother’s brother) are justa kind of representation that is widely distributed in apopulation through various processes of transmission.What makes them norms is the fact that they represent


the way things are required or allowed to be. In the socialscience literature, norms are mostly envisaged as causesof behaviors conforming to them. However, norms playother causal roles which may be no less important. Inparticular, they serve to confer approval or blame on be-haviors attributed to oneself or to others or just on be-haviors that occur very rarely, if at all, but the very pos-sibility of which captures the imagination and definesthe limits of what is acceptable. In most societies, forinstance, norms against cannibalism are much more im-portant as a topic of narrative and conversation than asa guide for behavior. It would be interesting to knowhow much the norm permitting a sister’s son to take hismother’s brother’s goods in one or another ritualized wayresults in actual taking of goods with significant eco-nomic effects as opposed to being a topic of conversationwith occasional symbolic enactments, serving to definesocial roles more than to reallocate economic resources.Alas, the literature does not seem to offer the kind ofdata that would answer this question. Moreover, thingsare likely to differ in this respect across different soci-eties and times.

Norms are not just causes of behaviors but also effectsof behaviors. Their spreading is caused by the differenttypes of behaviors that they promote. In other words,norms are cultural to the extent that they are distributedby causal chains in which mental representations of thenorms and public behaviors (including public statementsof the norm) alternate. Again, it would be interesting toknow how much a norm such as that permitting goodssnatching is maintained by actual acts of snatching andhow much by statements of and about the norm.

Both universal and culture-specific factors may con-tribute to the acceptability and attractiveness of a normand therefore to its chance of reaching, in a given socio-historical situation, a cultural level of distribution.Whatever the extent to which a norm permitting ritu-alized transgression causes behaviors that conform to it,the cultural stability of the norm is a sign of its psycho-logical acceptability and attractiveness—which have tobe explained. Here we propose some considerations rel-evant to such an explanation.

Suppose that there is a type of behavior that, for dif-ferent reasons, is simultaneously attractive and unat-tractive in the same society. As a result, there are, inthat society, factors that would favor the stabilization ofa norm approving this behavior and other factors thatwould favor the stabilization of a norm prohibiting it.Under such conditions, the stabilization of one of thetwo types of norm is an obvious obstacle to the stabi-lization of the other, opposite type.

In such a case, things can go in one of three ways. Thefirst possibility is that indeed the stabilization of onenorm effectively counteracts factors that would have fa-vored the stabilization of the other. For instance, reli-gious iconoclastic movements have, in different socie-ties, effectively suppressed any type of image eventhough receptivity to iconic representations, we assume,was still psychologically present and would otherwisehave favored the cultural approval of image production.

Here a psychological disposition, although present, failsto favor any direct cultural expression. The second pos-sibility is that the factors favoring opposite norms endup stabilizing some compromise norm, as when imagesare accepted and even encouraged but only with religiousthemes. Then there is a third possibility, in which thestabilization of one norm contributes to the stabilizationof a well-contained, ritualized form of the opposite norm.One norm dominates, but the other norm applies inclearly insulated circumstances. This state of affairs mayactually contribute to the stability of the dominant normby highlighting the exceptional character of its occa-sional violation. Thus Bloch (1987) has argued that thesexual chaos expected at certain stages of Malagasy royalrituals must be seen as “scene setting” for the extremedomestic order dramatized in the next stage.

The behavior studied by Goody might well be such acase of a potential conflict of norms that results in thestabilization of two sharply contrasting cultural normscaused by very different factors. One, patrilineal descentand inheritance, is wholly dominant, while the other,the rights of the sister’s son, takes the form of an au-thorized transgression with ritual aspects the very trans-gressive character of which contributes to the stabili-zation of the dominant patrilineal norm. This suggestionis, of course, reminiscent of a line of argument famouslyinitiated by Gluckman (1954) and developed by the Man-chester school, in particular in the work of Victor Turner(1969). What the epidemiological approach does and theGluckman-type explanation does not, however, is seekto explain the macrocultural fact of the asymmetricalequilibrium between a dominant norm and its author-ized or even prescribed transgression in terms of factorsaffecting the microprocesses of cultural transmission.

Given the stabilization of a patrilineal norm (the ex-planation of which is not the topic of this article) andthe persistence of evolved psychological factors favoringinvestment of resources in all close kin, whether patri-lineally or matrilineally related, we may expect individ-uals to welcome expressions of these psychological fac-tors provided that they are not incompatible with thepatrilineal norm they have internalized. These psycho-logical factors may find an expression through the in-formal helping by the mother’s brother of his sister’schildren. Here, however, we are talking of individual at-titudes rather than of a culturally sanctioned practice. Acultural practice that acknowledges the rights of one’ssister’s children would normally go against the patrilin-eal norm and would be unlikely to stabilize (unless thepatrilineal norm itself was in the process of destabili-zation). Expressing interest in the sister’s son/mother’sbrother relationship while highlighting the fact that thisrelationship does not ground normal, regular rights ofsharing or inheritance is a way of reasserting by contrastthat very patrilineal norm. More specifically, ritualizedtransgression practices of the type we are discussing hereunderscore the out-of-the-ordinary character of a sister’sson’s rights over his mother’s brother’s goods and therebycontribute to highlighting the normal character of pat-rilineal transmission of goods. Thus the combination of


the dominant patrilineal norm internalized by all mem-bers of the society and the psychological factors favoringall close kin renders people receptive and welcoming toa norm of ritualized expression of sister’s-son rights.

The norms and practices of ritualized transgressionthat are likely thus to stabilize are “catchy” because oftheir psychological rather than because of their eco-nomic effects. These are first and foremost “symbolic”practices that need not have any significant—let aloneany major—effect regarding the actual allocation of re-sources between direct and lateral descendants. This isa further contrast between the epidemiological accountwe are sketching here and any sociobiological accountthat would explain such practices in terms of their pu-tative effects, through reallocation of economic re-sources, on social stability or biological fitness.

All that we have said, of course, does not amount toa comprehensive explanation of the particular forms ofthe sister’s son’s privileges in any one of the societiesdiscussed by so many ethnographers, and it is importantto understand why. There are two reasons for this—besides the very sketchy character of our attempt. First,we have relied on the hypothesis that there is an evolvedhuman disposition that is aimed at modulating behaviorin a way sensitive to degrees of biological relatedness,but this hypothesis is based on speculation, howeverwell-motivated, more than on conclusive hard evidence.Secondly, we are not offering an explanation for why, forexample, Lo Dagaba sister’s sons behave in precisely theway they do. Indeed, we think a unifactorial or bifactorialexplanation of such an ethnographic datum would in-evitably be insufficient. Actual cultural practices, as per-formed by specific individuals at a given time, are em-bedded in the sociohistorical processes that havedistributed, stabilized, and transformed cultural repre-sentations and practices in the population to which theseindividuals belong. Each of these historical flows isunique. These processes are influenced by many typesof factors, evolved psychological predispositions beingonly one of them. Mostly, cultural processes are influ-enced by other cultural processes. People’s behavior, inparticular their conformity or nonconformity to norms,is guided by the representations they have of the worldrather than by the way the world is. People’s represen-tations are influenced in several ways by the phenomenathey are about, but they are influenced also—and to agreater extent in most cases of interest to anthropolo-gists—by other representations, in particular culturallytransmitted ones.

All these difficulties and caveats do not mean that weneed to abandon generalizing explanations of the kindwe have attempted here. In other words, the recognitionof the value of the objections to kinship studies of suchas Needham and Schneider need not lead to a denial ofthe relevance of general unifying causes, amongst whichare some universal human dispositions likely to havebeen naturally selected in the course of evolution. Sucha method, precisely because it sets nonabsolute condi-tions for the expression of general factors, can overcomethe difficulty which we highlighted at the beginning of

this paper and which seems to have overwhelmed an-thropology. Reasoning in terms of such things as evolvedhuman dispositions has all too often produced too pow-erful explanations, while the refusal to try to explainobvious though partial recurrences across cultures in theend seems perverse and inevitably leaves anthropologicalquestions to be naıvely answered by others.

Comments

monique borgerhoff mulderDepartment of Anthropology and Graduate Group inEcology, University of California, Davis, Calif. 95616,U.S.A. ([email protected]). 4 v1 02

Bloch and Sperber’s paper is remarkable for its honestrecognition of general patterns seen across human so-cieties and their need for an explanation. For anyonedeeply committed to an evolutionary and comparativeapproach, it is a delight to read. I hope that its eloquentportrayal of the complex and indirect links betweendominant institutions, specific circumstances, andevolved psychological dispositions will help dispel theantagonism that most anthropologists still feel towardcomparative studies, evolutionary frameworks, or both.My commentary elaborates on the complementaritiesand differences between the approach of evolutionarysocial scientists and Bloch and Sperber’s so-called epi-demiological model (an unfortunate term because of itslinks with memetics [Laland and Brown 2002], an ap-proach that they would undoubtedly reject).

I strongly agree with them on the need to return tothe multifactorial explanations that were typical of an-thropology prior to its engagement with relativism andto dissolve the nature/culture distinction. We agree alsoon a naturalist approach that includes, as one element,the existence of evolved, genetically transmitted psy-chological predispositions that cause people to pay at-tention to certain types of stimuli; these mechanismsinteract with a host of historical factors and current cir-cumstances to produce behavioral outcomes that willthemselves shape norms and institutions in the future.Most of us also recognize that a functional model cannotdeny the full force of human history. Human behavioralecologists have recently begun to borrow methods frombiology to tackle this issue. With the introduction ofcomparative methods designed to control for shared his-tory, it is now possible to incorporate quantitatively intoan evolutionary approach the extent to which historyinfluences current social practice and norms (BorgerhoffMulder et al. 2001).

Bloch and Sperber seek to restore legitimacy to com-parative studies without overlooking previous criti-cisms, particularly with regard to the universality andontological status of elements to be compared. Theircommitment to what they call “non-naıve” comparisonappeals strongly to evolutionary anthropologists inter-


ested in patterns of cultural variability, though they donot take their empirical analysis as far as a human be-havioral ecologist would. With respect to the issue oflineality, they are right in arguing that simple biologicalrelatedness fails to explain why, when r (genealogicalrelatedness) is equal, paternal kin are stronger targets ofaffection, altruism, and reliance than maternal kin orvice versa, but they fail to emphasize (though theyclearly appreciate) that r is simply a weighting term inthe predictive model. As has long been understood byanthropologists, individuals will choose to associatewith kin related through males or females depending onconsiderations of property, defense, labor cooperation,etc. These ideas and their implications for an evolution-ary approach to sociocultural diversity in kinship organ-ization were carefully analyzed by Irons (1979) over 20years ago. Since then human behavioral ecologists andothers have conducted extensive comparative analysesof the extraordinary variation in kinship institutionsworldwide, focusing not just on linearity but on marriagepayments, polygyny, polyandry, residence patterns, in-heritance, etc. Each of these studies could easily be re-formulated within Bloch and Sperber’s framework of rep-resentations, and doing so would promote more carefulmechanistic thinking.

One more complementarity between the approacheslies in Bloch and Sperber’s focus on institutions. As inother areas of anthropology guided by methodologicalindividualism (Ensminger and Knight 1997), evolution-ary social scientists grapple poorly with the ontogenyand persistence of institutions. As Bloch and Sperbernote, how you get from genes to representations or normsis not well thought-through. Their focus on institutionsrather than individual behavior nicely complements be-havioral ecology’s more micro-scale analyses.

Finally, the differences: Bloch and Sperber contrasttheir epidemiological approach with a “coarser socio-biological account” that would explain practices interms of putative effects on fitness. While evolutionarysocial scientists differ radically with respect to their con-cern with measuring fitness differentials (Smith, Bor-gerhoff Mulder, and Hill 2001), human behavioral ecol-ogists would indeed look to see how sister’s son privilegevaried in relation to the stability and appropriateness ofthe dominant patrilineal ideology in a given context.Therefore they might predict greater elaboration of thesister’s son’s rights (ritual or other) in certain circum-stances. Low paternity certainty might be one such fac-tor, though, as Bloch and Sperber and others (Flinn 1981,Gaulin and Schlegel 1980) note, it is not necessarily acausal variable in these scenarios; other factors thatmight promote a sister’s son’s privileges include changesin the economic patterns of production, security consid-erations, and migratory options, each of which can im-pinge on the adaptive value of purely patrilineal inher-itance. Absence of support for such predictions wouldnot falsify Bloch and Sperber’s broader hypothesis re-garding the role of evolved psychological factors in sta-bilizing norms of ritualized expression. But the test itselfenables human behavioral ecologists to move away from

a troubling tendency among evolutionary psychologiststo universalize (and essentialize) evolved human predis-positions, particularly with respect to the emergence ofthese dispositions in hypothetical “Pleistocene” condi-tions. We like to keep open the conceptual possibilitythat human predispositions are an accretion of ancientand more modern influences. We do this by designingmethodologies that allow us to evaluate the adaptivevalue of revisions, modifications, and variations of be-havior. My hope is that Bloch and Sperber’s path-break-ing and integrative approach to cross-cultural regularitieswill retain the same open-mindedness.

james s . bosterDepartment of Anthropology, University ofConnecticut, Storrs, Conn. 06269-2176, U.S.A.([email protected]). 11 vii 02

The principal proposal of this paper, that humans have“an evolved disposition to favor relatives,” should notbe controversial. After all, such a disposition is well doc-umented among other animals, as when free-tailed batsreturning to a colony find and nourish their own off-spring among thousands of pups (Balcombe and Mc-Cracken 1992) or when tiger salamanders preferentiallycannibalize nonkin over kin (Pfennig 2002). It would bestingy to deny the capacity for kin recognition and kin-based altruism to humans as well, if it were not also welldocumented (e.g., Burnstein, Crandall, and Kitayama1997, Cialdini et al. 1997, Essock-Vitale and McGuire1985, Kendrick 1991, Kruger 2001, Matthews et al. 1981).The key question is how the evolved disposition givesrise to different behaviors in different ecological con-texts. Bloch and Sperber do not address this question,and it is hard to see how their account is an improvementover Goody’s (1959) treatment of the tolerated theft frommother’s brother by sister’s son in patrilineal societies,as they themselves astutely state.

As did Goody, they explain the tolerated theft in termsof a universal human attention to bilateral kinship, dif-fering in explicitly stating that the disposition is anevolved one. However, Goody’s account is superior inthat it more explicitly explains variation: his model pro-poses and demonstrates a correlation between theamount of disparity between the flow of inheritance todescendants of daughters versus sons and the amount oftolerated theft. In contrast, Bloch and Sperber’s modelposits a universal disposition to favor kin which predis-poses humans to accept some ideas more readily thanothers. What ideas are available is a matter of historicalaccident. In other words, they propose a regression modelwith a constant (the universal disposition) and an errorterm (the historical accidents) but no variables.

A related problem is that they present neither new datanor a complete reanalysis of the existing data. Instead,they offer a style guide for just-so stories—a “visuali-zation” or sketch of the structure of an analysis but notthe analysis itself. This makes it difficult to resolve someof the apparent contradictions in the argument (e.g., their


floating between extreme cultural relativism and a skep-tical realism). Their failure to present any new evidenceis particularly troubling when they use an absence ofevidence to dismiss Alexander’s (1979) argument. It istrue that, as they state, there is “no evidence showing acorrelation between institutionalized privileges of sis-ter’s son and paternity doubts,” but this is because thestudy has not been done. If it were to be done, it verylikely would show the relation Alexander expected, be-cause other studies have shown diminishing paternal in-vestment with decreasing paternity certainty (e.g., thepapers in Beckerman and Valentine 2002). An absence ofevidence is not evidence of absence.

Although theorizing about the social distribution ofcultural beliefs is an interesting project, I believe thatthe metaphor of an epidemiology of representations isnot the most useful for the enterprise. It puts too muchagency in the ideas themselves rather than the individualhumans who believe them. A more useful metaphor isan economic one (Boster 1991), for it encourages one toview representations as constructed, distributed, andused in ways that serve the interests of the human be-lievers rather than the ideas believed. Ironically, eventhough Sperber has critiqued Dawkins’s (1976) conceptof the meme, his concept shares with Dawkins’s a pre-sumption of the agency of ideas and the passivity of theirhuman hosts. This contributes to the problem men-tioned above: beyond the universal human dispositionto favor kin, all the rest is chance because the differentialacceptance of certain ideas (e.g., the tolerated theft) isnot viewed as serving the interests of the individual hu-man actors.

A final problem is the use of “sociobiology” as a foil.Who are the authors talking about when they refer tosociobiologists? The only sociobiologist cited is Alex-ander (1979), the date of whose work is, coincidentally,the median publication date of their references. Al-though the literature on kinship from a structural-func-tionalist point of view may have been quiescent sincethe late seventies, research in human behavioral ecologyand evolutionary psychology has been booming. I do notknow of any contemporary advocate of an evolutionaryperspective on human beings who believes that genesdirectly encode cultural representations. The dominantcontemporary Darwinian perspective on humans is thatof behavioral ecology, which asserts that humans are ge-netically similar but phenotypically (and culturally) di-verse: The same genotype is facultatively expressed asdifferent behavioral phenotypes depending on signalsfrom the environment. The position Bloch and Sperbertake is actually more similar to the one that they stig-matize as “sociobiology” than any taken by behavioralecologists in that it is more committed to a biologicallybased, invariant human nature. This is puzzling becauseSperber, in his other work, has shown a deep appreciationof how evolution has crafted human minds to think dif-ferently about different things (e.g., Sperber and Hirsch-feld 1999).

In sum, I agree with the proposition that humans havean evolved disposition to favor kin. I agree that that dis-

position is part of an explanation of the question at hand.And I exhort the authors to follow their “visualization”with an actual research project that more directly en-gages the theories, issues, and results of contemporarybehavioral ecology and evolutionary psychology.

michael f . brownDepartment of Anthropology and Sociology, WilliamsCollege, Williamstown, Mass. 01267, U.S.A.([email protected]). 4 iv 02

Bloch and Sperber’s reanalysis of one of the discipline’sfoundational riddles may be a sign that anthropology isslouching toward a Bethlehem in which the practice ofcultural comparison will be reborn. When we abandonedthe comparative project two decades ago, we ceded it toreductionists (think evolutionary psychology) who arepleased to have this particular sandbox to themselves.By offering a nuanced, nonreductionistic approach tocomparison, Bloch and Sperber make a case for renewedattention to cross-cultural similarities and differences.

One only wishes that the results of their effort weremore arresting. An uncharitable gloss goes somethinglike this: “In societies with unilineal descent, practicesinvolving special relations with close matrilateral kinwill arise naturally unless other factors intervene to pre-vent this from happening.” Such a modest conclusion isunlikely to bring comparativists back from the academicElbas to which they have been exiled. A more sympa-thetic rendering of Bloch and Sperber’s thesis is that byidentifying strong panhuman tendencies, which can bemuted or redirected by the complex web of ecological orculture-historical forces in which every social systemoperates, anthropologists are better able to discover newand illuminating comparative questions. This is still notelectrifying news but worthy of serious consideration.

What I find mystifying is Bloch and Sperber’s reluc-tance to acknowledge affinities to the growing body ofwork on complexity, emergent structures, and self-or-ganizing systems. They use some of the idioms of thisliterature—when, for instance, they refer to culture as a“flow of information”—but resist others. How is an ex-pression like “epidemiology of representations” differentfrom or preferable to “memes” (Dawkins 1976)? It is easyto understand why Bloch and Sperber want to distancethemselves from Dawkins’s genetic determinism, butmore recent work (see, e.g., Balkin 1998) has moved be-yond simplistic models of selfish genes and maximiza-tion of inclusive fitness. An attraction of the latest it-eration of complexity theory is that it promises toreconcile or transcend the exhausted dichotomiesagainst which Bloch and Sperber struggle in this essay:mind/brain, constructed/real, software/hardware, cul-ture/nature (Taylor 2001:230).

Complexity theory has gained a following among ar-chaeologists and ecological anthropologists. In sociocul-tural anthropology, however, it hasn’t proved “catchy”(to use another of Bloch and Sperber’s metaphors), per-haps because few scholars have succeeded in translating


the resolutely abstract theories of Luhmann (1986) andothers into ethnographically satisfying accounts. DoesBloch and Sperber’s essay lay the groundwork for a waveof complexity-focused comparative analysis? It is toosoon to tell, but I found myself engaged by the attempt.

rosario calderonDepartamento de Biologıa Animal y Genetica,Apartado 644, Universidad del Pais Vasco, E-48080Bilbao, Spain ([email protected]). 30 v 02

“All members of a human community are linked one toanother across time and space by a flow of information”that is not just cultural but genetic. Bloch and Sperber’scentral question, “what causes some practices to becomeand remain widespread and stable,” remains unan-swered. The statement that “psychological dispositions. . . modify the probability—and only the probabil-ity—that representations or practices . . . will spread,stabilize, and maintain a cultural level of distribution”seems a rather vague and unhelpful response.

I do not agree that “the phenomena described by an-thropologists under the label of ‘kinship’ are cultural andtherefore historical constructions . . . rather than . . .facts of biological kinship.” Both standpoints shouldhave been considered. Animals and humans (another an-imal species) change their genomes to become geneti-cally different. Whether humans can change their ge-nomes, consciously or not, through cultural rules andtherefore differ significantly from the other animal spe-cies is another question. We must distinguish among (a)kinship as a biological fact, (b) kinship as a personal orsocial perception, and (c) patrilineal inheritance as amode of property acquisition.

Biological kinship between two individuals is a quan-titative value that may be close or distant. This valuewas unknown in primitive societies and could only beevaluated after the discovery and development of Men-del’s laws. Biological kinship affects the biology of hu-man populations through their genotypic distributionsand genetic loads.

Social kinship is a qualitative rule relating personswith or without biological kinship. It usually involvesclose biological kinship (two generations above and twobelow ego) as well as affinity. Even though social andbiological kinship are not related, people identify a cer-tain (usually erroneous) association between the two thatvary over time and from one society to another. Familyrelationship has been related to transmission of blood(representing the spirit) and of flesh. This idea is quitedifferent from genetic transmission as we know it today.Social kinship causes marriage preferences and avoid-ances among some members of society and affects bio-logical kinship. Consequently, the two concepts are notequivalent, and historical societies have misunderstoodwhat biological kinship means.

Patrilineal inheritance is basically a discontinuous,qualitative character in many societies. For example, oneperson may inherit all the patrimony while his relatives

are dispossessed. The context in which these cases occurmust be taken in account. A certain degree of social de-velopment is required to introduce the norm of inheri-tance; there must be property to be transmitted and thepower to ensure conformity to the norm. The utility ofthe transmission of undivided property must be recog-nized, and there must be life span sufficient to permitthe heir to manage the inheritance properly and a certainfamily size. Patrilineal inheritance would not have af-fected biology, offering the heir increased life expectancyor higher fertility in relation to the dispossessed. Fur-thermore, it is quite unlikely that the heir’s advantagecould have been sustained for several generations. More-over, the relative importance of other modes of propertyacquisition such as trade, warfare, and raiding must beconsidered.

The transgressive and tolerated violence of the behav-iour analysed does no harm to inheritance rules and thesystem of property in the community, which remain un-changed through time. It affects neither biological norsocial kinship. Thus this kind of behaviour should beseen as a ritual that does not threaten but consolidatesthe property rules of the societies in which it is practised.

jack goodySt. John’s College, Cambridge University, Cambridge,U.K. ([email protected]). 31 v 02

Bloch and Sperber’s article is important for much morethan the mother’s brother. It attempts to apply the ar-guments of a branch of cognitive science to the study ofkinship in rather the same way that their colleague Pas-cal Boyer has tried to do for religion. It would need anarticle at least to deal with all the careful philosophicaland scientific points that it raises. The project is ap-pealing for several reasons. In the first place, it takes usbeyond the uniqueness argument (to which they them-selves resort at the end of the article) that has boggeddown creativity in much contemporary social science;of course each blade of grass is different, but that doesnot stop us from looking for common features. Secondly,it is prepared to deal with “universals.” I myself do notsee the problem as an opposition of “cultural particu-larity” to “universals”; comparison can deal with wide-spread but universal phenomena such as the character-istics of redheads, which may have a biologicalcomponent but equally may be (as with the footballerFreddie Lindberg and his imitators) a matter of dying thehair. The problem can only be resolved by empirical in-vestigation (or possibly by conceptual reformulation).Lastly, it is important to try and make contact with thefindings of related disciplines. Sociocultural anthropol-ogists are too prone to think that they alone hold thekey to human behaviour and that others should followtheir unilineal shifts from one approach to another.

Bloch and Sperber are right historically about the moveaway from comparative studies and towards a vaguelypostmodern cultural particularism. They are, I think,wrong in assuming that the comparative students of kin-


ship “deluded themselves they had been dealing withbiological facts.” I think Levi-Strauss, Leach, Fortes, andothers would have strongly resisted this suggestion morestrongly than Malinowski. In my view, these anthro-pologists did not perhaps allow enough room for biolog-ical facts, though in behavioural terms these are alwaysmodified by cultural factors. Even if it is agreed that theywere dealing with representations (and is not all socio-cultural life?) and all had “unique histories,” it is diffi-cult to see why cross-cousin marriage or incest, whichare of course human constructs, do not also have com-mon elements. And, indeed, were the differences purelyrandom affairs as the uniqueness argument assumes?

Who, except perhaps Levi-Strauss in the atom of kin-ship and Fortes at a similar level, was ever looking for“universals”? If there was a tendency to set aside biologyat this level (as most strongly with Leach and Schneider),perhaps not “universals” but at least widespread simi-larities were often taken into account without biology,although all are constrained by the human mind and byits psychology. The question we have to ask all the timeis how far constrained, and that seems to me an empiricalquestion.

I am not at all sure that the dichotomy of mind-ex-ternal and mind-internal factors is very satisfactory.Have not analysts and social scientists long talked aboutthe process of internalization? Is my reading of a text(book) or indeed my interpretation of an utterance aninternal or an external event? Regarding the problem ofthe mother’s brother, it is an error to think that Fortescontrasted a domestic domain, governed by biology andnatural emotions, with a politico-jural domain. The do-mestic domain was also characterized by jural rules, andthe family was by no means purely biological; indeed,as I have said, my own view is that perhaps not enoughweight was given to the latter by him and by many othersocial anthropologists. It was not that all “human beingsreally reckon kinship bilaterally” or that the sister’s sonis “a true descendent of his mother’s parents” or thatthere is “a truly patrilineal system.” These are analyticconcepts, representations, whose truth or reality is pro-visional, related to notions that are useful rather than“real.” The world is a more complex matter but one inwhich “biology” is involved.

I am intrigued at the attempt to make a link withHamilton’s influential notion of kin altruism, which isnot simplistically sociobiological. If Hamilton is right,there has to be some predisposition in favour of recog-nizing genealogical relationships, but that seems such along way from cultural “kinship” even at the level ofthe elementary unit of reproduction—which biologicallyhas to involve both man and woman and therefore bebilateral for descendants—that I wonder how much ithelps with the mother’s brother. Bloch and Sperber have,however, made a most enterprising attempt to seek analternative explanation of a widespread but far from uni-versal phenomenon.

I wonder too about the frame of the argument con-cerning iconophilic and iconophobic behaviour. It ap-pears to assume that we find a psychological disposition

in favour of iconic representations that is suppressed bycertain religious movements. Is there also a psycholog-ical disposition towards iconoclasm? That would seema weak supposition. I am also worried about the notionof a dominant (patrilineal) norm and its transgression(especially since this is “a double-descent system”) anddo not see the snatching by the sister’s son (of which Ihave seen recurrent examples) in this light. I myself haveargued that the switch between the two attitudes to im-ages needs to be examined above all at the level of rep-resentations, a necessary feature of language-using ani-mals facing the world; these, because they are never thething itself but always a facsimile, give rise to ambiva-lence, doubts, and contradictions. One result is not sim-ply religiously inspired iconoclasm but also a moregeneral secular “disposition” to reject images (as para-digmatically in Plato). Bloch and Sperber take the expla-nation of the phenomenon to a more “naturalistic” level;I pitch it at a more “cultural” one, though I do not rejecta psychological underpinning as effect as well as possiblepredisposition of a less immediate kind.

edward d. loweCenter for Culture and Health, Department ofPsychiatry and Biobehavioral Science, University ofCalifornia, Los Angeles, 760 Westwood Plaza, C9-752NPI, Box 951759, Los Angeles, Calif. 90095-1759,U.S.A. ([email protected]). 28 vi 02

How do we study human similarities without dismissinghuman differences or human differences without dis-missing obvious similarities? Bloch and Sperber remindus, first, that the direction of History has always beentoward greater diversification and complexity, but theway in which the products of History constrain and allowthe development of other products over time depends onthe level of the history with which one is concerned.This is because, rather than being independent of oneanother, the various levels are “fused” in such a waythat the systemic organization of the products of historyat one level constrains the development of products ofhistory at other levels (Ford and Lerner 1992). The levelsof History that develop at a slower pace tend to placegreater constraint on the development of faster-devel-oping levels than the other way around. For example, theradiation of species always proceeds much more slowlythan the production and diversification of culture, andthe development of culture proceeds much more slowlythan that of any individual human. As a result, the prod-ucts of genetic history should place powerful constraintson the products of human social histories, and bothshould place powerful constraints on the developmentof individuals. The production of culture—the wide-spread adoption and reproduction of particular forms ofrepresentation—does have some degree of historicallycontingent arbitrariness associated with it—partly be-cause of a genetic capacity for individual agency. But thearbitrary character of cultural production is limited bythe sedimentary residua of both genetic-historical and


social-historical processes. Anthropologists may inves-tigate human similarities while giving equal considera-tion to human differences by investigating these twinprocesses of diversification and constraint in the flowsof various levels of History.

Given the tendency toward diversification in the pro-duction of culture, that the norms that govern the re-lationship of mother’s brother and sister’s son are highlyvariable among the societies that have “hypercognized”(Levy 1984) it is unremarkable. The observation that thisrelationship, among all others possible, is recognized andritually elaborated in terms of a degree of informalityfrom younger to older and nurturance from older toyounger in so many historically unrelated patrilineal so-cieties is remarkable. Some constraint must be operat-ing, but what is it?

There may indeed be some genetically determined do-main-specific competence in the human brain/mind thatgives special recognition to people based on their degreeof relatedness to ego. A psychological disposition mayindeed emerge in the development of individuals wholive in patrilineal societies that combines the kin-selec-tion competence with principles of patriliny in such away that social solutions for the obvious disjuncturesbetween the two attitudes will seem particularly attrac-tive to them, motivating them to adopt those solutionsand reproduce them over time. But to my knowledge noevidence of such a domain-specific competence exists,as Bloch and Sperber admit.

Alternatively, we could dialogue with current neuro-science and developmental psychology to identify can-didates for constraint. Two possibilities come to mind:a domain-specific competence for social categorization(Hirschfeld 1994) and the capacity for extending the pri-mary emotional attachments that typically develop be-tween mother and infant to other exchange partners inone’s immediate social environment, typically close kin,and sustaining these attachments throughout life (Lowe2002). The combination of the two competences couldeasily produce the bilateral kin recognition tendenciesthat seem central to the case under discussion notthrough genetic determination but through the emer-gence of particular interpersonal relations made moreprobable by the interaction of each genetic competencewith the typical social experiences of individuals livingin these societies. If a woman’s attachments to her sib-lings, particularly her brothers, were sustained when shemarried and had children (perhaps through forms of cul-turally sanctioned occasional exchange), they could eas-ily be the basis for her directing her children’s attentionto her brother as a source of support, in opposition topatrilineal principles. Indeed, she might encourage therecognition of the relationships between her son and herbrother as an alternative source of support to the child’sown lineage mates. Her brother might also feel partic-ularly sympathetic with his sister’s (likely) precariousposition in her husband’s household, where she mightnever gain full acceptance. Sibling attachment and theinsecurity associated with living among affines (partlyresulting from being a different kind of social person)

could easily produce the kind of psychological disposi-tion that would make ritualized snatching and relatedinformal joking behaviors in the mother’s brother/sis-ter’s son relationship particularly “catchy.”

joseph poulshockLanguage Evolution and Computation Unit,Department of Theoretical and Applied Linguistics,University of Edinburgh, Edinburgh, Scotland, U.K.([email protected]). 3 vii 02

Bloch and Sperber’s article is a welcome attempt at anapproach to an understanding of cultural phenomena bymeans and mechanisms grounded in natural or materialproperties. It is welcome because it attempts to trace ourunderstanding of cultural experience to empirical andreal causes as opposed to logically coherent argumen-tation and speculation based on diverse biological andcultural facts which relate to explicanda but cannot beconfidently traced to putative causes in our explicans.Moreover, although Bloch and Sperber disassociate theirmethod from a memetic approach to culture, in manyways it is like memetics—or an improvement on me-metics. They differ from the main memeticists by claim-ing to expect a continuum of cultural representationsfrom idiosyncratic to broadly cultural—as opposed toDawkins’s (1976) and Blackmore’s (1999) memetic di-chotomy between true replicators and other mentalevents. However, their epidemiology of representationsis still memetic in that it fits into the meme-as-germschool of thought (Lynch 1996) rather than the meme-as-gene school (Blackmore 1999). Moreover, their ap-proach, like memetics, is a naturalistic and Darwinianway of explaining and comprehending culture that, whileavoiding the meme-gene analogy and the focus on re-plicators, still attempts to explain culture as an epide-miology of representations. Memetics is rife with em-pirical and conceptual difficulties and has yet to provideany tangible results (Aunger 2001). Bloch and Sperber’sapproach, however, with its emphasis on a mechanismfor representational transmission, holds out the promiseof putting tread on the tires of memetic-like research.

Their modest claim that a biological predispositionsuch as an engendered sensitivity toward kinship infor-mation stabilizes cultural representations of a similarkind is suggestive and interesting. Moreover, the ideathat a psychological predilection rebels ritualisticallyagainst an internalized norm and yet affirms that normby contrasting with it is an intriguing and counterin-tuitive conclusion. Nevertheless, though evocative, suchan assertion does not provide us with any real empiricalinsight into this cultural trait. Moreover, and more se-rious, this is probably because of problems with the ap-proach. Bloch and Sperber claim that their method “aimsat describing and explaining cultural phenomena interms of processes and mechanisms the causal powersof which are wholly grounded in their natural (or ‘ma-terial’) properties.” However, their mechanism is seri-ously underdefined. Human agents who make up cul-


tures possess bio-psychological predispositions thatsupport cultural norms. But what exactly is the mech-anism that curses some cultural representations to sta-bilize? Can we see this mechanism in action? Moreover,are we absolutely sure that the putative predispositionsin question exist? For example, are there ways to test fora human psychological tendency to look for informationrelevant to kinship ties? In addition, does our currentunderstanding of the way in which we gather kinshipinformation belong to the realm of fact, truism, or as-sumption? Lastly, can we trace cultural representationsto causes empirically, or are we going to accept infer-ences based on assorted facts and assumptions combinedwith careful speculation?

Besides these problems, there appears to be an evenmore basic issue at the root of Bloch and Sperber’s ap-proach. The informational norms at the center of theirresearch question do not seem not to fit easily into mech-anistic categories. Human agents create cultural repre-sentations, and while there may be biological predispo-sitions for much of culture, I wonder whether we canreduce such creative acts to mechanisms. Perhaps thereis a rigorous way to approach culture in the terms thatBloch and Sperber suggest that will produce knowledgeabout the stabilization of cultural representations, butat this point, though their effort is a worthy one, theirresults appear still somewhat immaterial.

peter j . r icherson and robert boydDepartment of Environmental Science and Policy,University of California, Davis, Calif. 95616, U.S.A.([email protected]). 9 iv 02

Bloch and Sperber’s portrayal of the interaction betweencultural and genetic evolution is mainly correct. Herewe briefly situate their proposal in the history of the fieldof gene-culture coevolution.

Darwin’s (1874) theory of human evolution was re-markably modern (Richerson and Boyd 2001a). Darwinwas not a racist. He believed that culture and environ-ment explained virtually all differences between livinghuman populations except those caused by sexual selec-tion. However, his ideas were incomplete in two crucial,related respects. He thought that culture and organic in-heritance were integrated via the concept of inheritedhabits and, as a consequence, did not clearly articulatea relationship between the evolution of psychologicalrules and the subsequent evolution of cultural traditions,heritable habits, and other forms of inheritance of ac-quired variation. Despite these problems, he gave ex-amples that accord closely with Bloch and Sperber’s pro-posal. For example, he thought that selection betweentribes had given rise in “primeval” times to specific, nowuniversal ethical intuitions such as sympathy and loy-alty and that these “instincts” have guided the evolutionof ethical systems ever since (Richards 1987). Like Blochand Sperber, he was quite clear that such instincts donot rigidly determine what culture traditions evolve orbehaviors occur. For example, he knew that the moral

crimes of Europeans, such as slavery and colonial gen-ocide, existed despite the instinct of sympathy that mo-tivated his own passionate opposition to such practices.

Darwin’s ideas significantly influenced the develop-ment of psychology until the end of the 19th century(Richards 1987). James Baldwin’s (1895) views were par-ticularly sophisticated. As early as 1895, five years beforethe rediscovery of Mendel, Baldwin drew a sharp dis-tinction between the “machinery of heredity” and imi-tation and hence drew the distinction we now make be-tween culture and genes. Nevertheless, Darwin’s and hisfollowers’ influence on the emerging social and behav-ioral sciences was narrow (Bowler 1988) and eventuallydisappeared.

Donald Campbell’s (e.g., 1965) work began a late-20th-century revival of interest in the relationship betweengenetic and cultural processes. Campbell was inspirednot directly by Darwin or Baldwin but by what he calledthe “blind variation and selective retention” analogy be-tween natural selection on genes and mental selectionof variant behaviors essayed by individuals during learn-ing. He argued that the analogy applied very broadlyacross knowledge systems, including cultural systems.He also reasoned that selection would favor what hecalled “internal selective criteria that are vicarious rep-resentatives of external selectors,” especially natural se-lection. Specifically, evolution would have built humanbrains with vicarious selectors to manage culture. Theparallel with Bloch and Sperber’s argument is close, ifnot exact.

In 1973, Luca Cavalli-Sforza and Marcus Feldman in-troduced the idea of using the formal analogies betweengenes and culture to motivate the borrowing of modelsfrom population genetics for the study of cultural evo-lution. They also noted the analogy of culture to epi-demic contagions (Cavalli-Sforza and Feldman 1981:46–53). These models are powerful tools for deducing themacro-level consequences of micro-scale processes, aproblem that Bloch and Sperber correctly note is centralto understanding evolution.

Our own efforts along these lines (e.g., Boyd and Rich-erson 1985, Henrich and Boyd 1998) have been mainlydirected at elucidating the mental mechanisms that in-dividuals use to manage the flow of cultural informationand deducing the macro-level consequences of suchmechanisms. For example, we have shown that the useof a conformist rule in converting public to mental rep-resentations is adaptive under a surprisingly wide varietyof circumstances. Selection very likely built such rulesinto the innate structure of our minds. One macro-levelconsequence of conformity is that it makes group selec-tion on cultural variation easy compared with group se-lection on genes. Darwin’s proposal that between-tribeselection resulted in the evolution of sympathy and loy-alty is plausible if we suppose that first cultural insti-tutions evolved enjoining sympathy and loyalty and thenthe innate instinct coevolved in response. In this sce-nario, violators of cultural rules experienced fitness-re-ducing punishment by those innately better prepared tofollow them. This potential inversion of the evolutionary


causal order generally assumed by sociobiologists (e.g.,Lumsden and Wilson 1981) was noted by Baldwin and isoften called the “Baldwin effect.” The hypothesizedmental products of cultural group selection and relatedprocesses do seem to exist and to account well for theunique patterns of human sociality (e.g., Richerson andBoyd 2001b). The models also suggest several mecha-nisms by which historical contingency arises in culturalevolution (Boyd and Richerson 1992), underlining theexplanatory modesty noted by Bloch and Sperber.

We believe that Darwinian methods can be applied tothe analysis of culture and its psychological foundationsalong the lines indicated by Bloch and Sperber. We alsobelieve that there already has been enough progress inthis enterprise for it to make serious claims to anthro-pologists’ attention.

christ ina torenDepartment of Human Sciences, Brunel University,Uxbridge, Middlesex UB8 3PH, U.K. ([email protected]). 4 vii 02

Perhaps the most profound observation to be derivedfrom the corpus of ethnography is that the peopled worldprovides for all our historically constituted descriptionsof it, rendering these always partial descriptions objec-tive in different ways. Difficult though it is to credit thatother people’s ideas are as objectively warranted by theworld as one’s own, it is only to the extent that one doesso that, as an anthropologist, one recognizes the neces-sity for a theory of human being that is capable of ex-plaining how this comes to be so. It is important there-fore to attempt anthropologically valid explanations,explanations that seek to render analytical the conceptspeople use to describe and/or account for their own ideasand practices, especially where, as here, we are asked toaccept that a series of conjectures derived from socio-biology and evolutionary psychology illuminates “thesignificance of the relationship between the mother’sbrother and sister’s son in patrilineal societies.”

Elsewhere I have argued for a model of human beingthat recognizes that mind is a function of the wholeperson in intersubjective relations with others (see Toren1999a). I have proposed that each of us makes sense ofthe world by making meaning out of meanings that oth-ers have made and are making—a process in whichknowledge is transformed in the very process of beingmaintained. The development of mind over time is thusan aspect of physical development in which mentalstructures are continually differentiated through func-tioning to assimilate additional experience. But becausewe are biologically social creatures, our understandingof the world is always mediated by relations with others.Thus language acquisition is crucial to the form our un-derstanding takes. Even so, we do not, in learning tospeak, simply take on other people’s meanings; rather,we constitute anew the concepts and grammar of thelanguage by which we are surrounded and, in so doing,at once maintain and subtly transform that language.

Our descriptions of the world and ourselves are the ar-tefact of the embodied microhistory that makes each oneof us what we are. The effectiveness of scientific expla-nations and the technologies to which they give rise isa function of their validity, but neither the explanationsnor the technologies are immune from history.

This model shows how we come to be certain that theworld conforms to our particular ideas of it even whilewe may know that others are just as certain of their own.It follows that to understand human ontogeny is to un-derstand how mind as the fundamental historical phe-nomenon imagines the world that warrants its imagi-nation. Because the model shows how ideas aretransformed in the self-same process in which they aremaintained, it is capable of explaining how we come tobe so similar in the ways we differ and so different inthe ways we are the same. And because it has no recourseto distinctions between biology and culture, evolutionand history, it enables us to avoid the sterile debate con-cerning what is “in the genes” and what is “transmittedthrough the environment.”

My holding that anthropological explanations shouldbe consistent with what we know about biology does notentail that I accept the entirety of Dawkins’s model, forI am more convinced by Gould’s (see Sterelny 2001). Iam also concerned that anthropological practice be sci-entific, but this does not entail that I accept the idea of“evolved psychological predispositions” that Bloch andSperber take for granted. The implication of my argu-ment is that the idea of psychological predispositions asformulated in evolutionary psychology (see Tooby andCosmides 1992:92–93) is unnecessary, that an epidemi-ology of representations has no explanatory value, andthat, from an anthropological perspective, both ideas areunscientific (see Toren 2002).

The strength of anthropological analysis is its abilityto embrace complexity. Thus Levi-Strauss’s (1978[1973])analysis of “the atom of kinship” that is constituted inthe relations between brother, sister, father, and sonshows how it allows for a relation of consanguinity, af-finity, and filiation and how, further, relations betweenmen and women in the senior generation stand in cor-relative opposition to relations between men across gen-erations (1977[1963]:46). In Fiji (where I do my fieldwork)this should mean that reserve between brother and sisteris opposed to familiarity between husband and wife asreserve between father and son is opposed to familiaritybetween mother’s brother and sister’s son, but this setof oppositions does not, in fact, exist. Nevertheless,when we take the point of view of a male ego, at anygiven point in his life Levi-Strauss’s thesis holds; thefather’s sternness to the child contrasts with the playfulfamiliarity of the mother’s brother, while a progressiverelaxation in the relation between father and son overtime contrasts with increasing avoidance betweenmother’s brother and sister’s son. The relation betweena male ego and his father’s sisters undergoes a somewhatless pronounced change in contradistinction to the re-lation between him and his mother (Toren 1999b; cf.Levi-Strauss’s discussion of how, for the Lele, the system


of relations in the atom of kinship changes as a functionof “succeeding phases of individual life” [1978(1973):100].) “The avunculate” emerges as an artifact of theanalyst’s focus on just one of the variety of structurallylogical possibilities presented by the atom of kinship(Levi-Strauss 1978[1973]:100), and it entails an idea ofunilineal descent that is arguably an anthropological il-lusion (see Kuper 1988:190–209).

Selection may indeed work primarily by selecting lin-eages of genes according to their differing capacities forreplication (Dawkins 1982), but the matter is by nomeans decided even among evolutionary biologists. Evenif, however, we accept this thesis and, further, accept kinaltruism and its entailment of “an instinctually baseduniversal bilateral recognition of kinship,” this servesonly to lead us back to Levi-Strauss’s atom of kinship.

Reply

maurice bloch and dan sperberParis, France. 27 viii 02

The central purpose of our paper was to outline the kindof theory which took account of and understood the prob-lems older comparative and generalizing projects in so-cial and cultural anthropology have run into. Thirtyyears ago or so, most social and cultural anthropologistswere still involved in systematically documenting thevariety of human cultures and institutions and trying toexplain them. However, the comparative and general-izing projects of old ran into serious problems. Today,most anthropologists are critical if not downright con-temptuous of these goals, as if they were unattainableand in any case unattractive. They are even more hostileto the various biologically inspired, theoretically ambi-tious approaches to culture that have recently flourished.Often, it is true, proponents of these approaches—sociobiology and memetics in particular—have shownlittle or no interest in standard anthropological evidenceand issues, as if the knowledge and competence that hadaccumulated were so defective that the field had betterbe reinvented from scratch. They have rarely understoodthe problems of the comparative enterprise, especiallyin the study of kinship. Our approach, by contrast, takesinto account recent criticisms of classical theorizingwhile also drawing on evolutionary thinking in waysthat we hoped might be unobjectionable and even ap-pealing to social/cultural anthropologists, at least tothose who have not altogether given up on the originaltheoretical goals of the discipline.

The commentators well understand and in most casesapprove of our goal of reestablishing fruitful communi-cation between those on the “biological side” and thoseon the “culturalist side.” However, with a couple of ex-ceptions, only the former seemed to have been suffi-ciently interested in our argument to offer comments.

This silence from the culturalist side is not to be inter-preted as tacit approval.

Our aims were several and went beyond just improvingscientific communication. We wanted to develop a the-ory which could account for cultural recurrences suchas the special relation of mother’s brother to sister’s sonin certain societies that had fascinated earlier anthro-pologists, and we wanted to do so without exaggeratingtheir commonness or their uniformity. To this end wewanted to explain, defend, and illustrate an “epidemi-ology of representations,” a theory that differs from clas-sical anthropological approaches and recent biologicalapproaches to culture in various ways, in particular inthe central role it assigns to cognition. We feel that thespecificity of our approach is not always fully recognizedor discussed by the commentators, with the exceptionof Borgerhoff Mulder (who in some ways explains ourenterprise better than we did ourselves).

We discuss the general theoretical issues raised by thecommentators in the first part of this response. In thesecond part, we discuss kinship issues.

Anthropological theory. Richerson and Boyd usefullyplace our attempt in the context of earlier cognate work.Indeed, they, as well as Boster and Borgerhoff Mulder,justly hint that we should have drawn more attentionto attempts similar to ours. Had we done so, however,we would not have told quite the same story as our com-mentators. The idea of viewing culture as an epidemi-ological phenomenon of course did not originate with usor even with Cavalli-Sforza and Feldman (1981). It ismuch older: it was, for instance, developed more than acentury ago by Gabriel Tarde in sociology (Tarde 1985)and by diffusionists in anthropology. This idea is quitebroad in any case. An epidemiological phenomenon is adistribution of conditions (such as diseases, accidents, oraddictions) in a population. An epidemiological approachto a phenomenon such as culture consists in viewing itas a historically evolving distribution of habits, practices,beliefs, and so on, and trying to identify and explain pat-terns in this distribution. In contrast to holism, epide-miological approaches look for explanations of macro-phenomena in microprocesses, and in contrast tomethodological individualism, they do not privilege in-dividual rational choices and actions as explanatory mi-croprocesses (Boster sees this last point as a weaknessand prefers a more individualistic approach). In a broadsense, all recent evolutionary approaches to culture are“epidemiological.”

When, however, Cavalli-Sforza and Feldman (1981),Dawkins (1993), Lynch (1996), and others use the epi-demiological idiom, they have something much morespecific in mind, namely, the epidemiology of infectious,typically bacterial or viral, diseases. They see culturalitems as spreading, as do viruses or bacteria, through acopying process, with novelty resulting from mutations(which, in the cultural case, may sometimes be intendedrather than random). Infectious diseases, however, areonly one kind of epidemiological phenomenon. Addic-tions are another common kind and arguably one morerelevant to the study of culture than infectious diseases.


The epidemiology of representations that we are de-fending was first put forward in Sperber (1985b) and hassince been elaborated and illustrated in a number ofworks (e.g., Atran 1998; Atran and Sperber 1991; Boyer1994, 2001; Hirschfeld 1996; Norenzayan and Atran n.d.,Sperber 1996). It differs from the works we have justmentioned in that it takes the idea of epidemiology inits broad sense and makes the processes of human cog-nition central. Therefore, contrary to what Poulshocksuggests, our approach does not fit into the meme-as-germ school of thought. It is not a memetic approach atall. We hope that this clarifies our use of the term “ep-idemiology” (to which Borgerhoff Mulder objects be-cause of the link she sees with memetics) and answersBrown’s question “How is an expression like ‘epidemi-ology of representations’ different from or preferable to‘memes’?”

The epidemiology of representations denies that mostor even much of culture is to be explained by processesof selection among competing genelike or germlike re-plicators (or “memes”). It insists that different kind ofprocesses and factors are involved in the distribution ofdifferent types of cultural phenomena. It stresses, amongmany other things, the causal role of biologically evolveddomain-specific psychological dispositions and suscep-tibilities in the stabilization of cultural representationsand practices. In this sense, it fits better with the programof evolutionary psychology as formulated by Tooby andCosmides (1992) than with less psychologically rich ev-olutionary approaches, be they memetic or sociobio-logical (but see below for some differences with Toobyand Cosmides).

One of the best ways of identifying what we are notsaying and how our approach differs from memetics isreturning to Boster’s charge that we attribute “agency”to representations and “passivity” to humans. This maybe a valid criticism of the meme idea, but it certainly isnot of ours. For us, representations do not have agency.They are more or less successfully transmitted from oneperson’s mind to another’s because of the way mindswork. The receptivity of individual minds to differentcontents is determined by a variety of factors, in partic-ular evolved psychological dispositions. When, say, in-dividuals come to believe what they have been told, itis not a case of their minds’ being passively invaded byan active idea. Rather, what happens in them is a com-plex inferential process of comprehension and accep-tance. This process results in the construction of a men-tal representation influenced by that of another personthrough the production and perception of a mind-exter-nal public representation. This public representationdoes not have intrinsic meaning, let alone agency; it hasonly the meaning attributed to it by its producer andcomprehenders. Where we may differ from Boster is thathe seems to favor a Homo economicus model of thehuman agent whereas we appeal to a model which takesinto account the richness of human cognition andemotion.

Goody is not sure that the dichotomy of mind-internaland mind-external factors is very satisfactory: “Have not

analysts and social scientists long talked about the pro-cess of internalization? Is my reading of a text (book) orindeed my interpretation of an utterance an internal oran external event?” We would argue that the metaphorof “internalization” has done more harm than good. Tosee this, think of a case in which internalization occursin a literal sense, for instance, when you insert a diskettein a computer. Public representations, utterances, andbooks do not literally enter our minds. Rather, they stim-ulate our auditory or visual nerve endings. These nervesdo not serve to “internalize” anything. They provide in-put to complex cognitive processes that construct ratherthan take in mental representations. The internalizationmetaphor has been successful in the social sciences, wesuggest, because it helps keep the psychological mech-anisms and processes involved in cultural transmissionvague.

One of our problems with meme theory (see also Bloch2000 and Sperber 2000) is that it seems to reproduce,within a naturalist framework, something like the oldontological contrast between nature and culture whichso encumbers Calderon’s comments. The elimination ofthis contrast, both Borgerhoff Mulder and Brown recog-nize, was one of our goals. We want to insist that cultureis always natural. Meme theory suggests that, with thecoming of the “mind invaders” (Dennett 1995), a dis-continuous and radical shift occurs in primate evolu-tion—so radical a shift, indeed, that it can look verymuch like the folk notion of the “passage from natureto culture.”

Unlike memeticists, for whom the evolution of me-mes is radically distinct from that of genes, we are amongthose who assume that biologically evolved psycholog-ical dispositions not only make culture possible but alsoplay some causal role in the stabilization of specific cul-tural contents. Richerson and Boyd go a step farther andsee our proposal as based on the idea of gene-culturecoevolution (to the development of which they have somuch contributed). We recognize the importance of thisidea and of Baldwin’s discovery of the effect that is nownamed for him. We are willing to pay homage to DonaldCampbell in the matter (although we fail to see the par-allel between his and our arguments). However, wehardly made use of the idea of coevolution in this paper.We assumed an evolved disposition to look for evidenceof relatedness and to favor related individuals. We re-flected on the causal role that such a disposition couldplay in the stabilization of specific cultural practices.Given the different rates of biological and cultural evo-lution stressed by Lowe in his commentary, it is gen-erally much easier to see how biologically evolved dis-positions might have influenced cultural practices thanthe other way around. Still, it is possible that the dis-position in question might itself have evolved in reactionto culture, yielding a true case of coevolution. We wouldwelcome suggestions to that effect, but we ourselves didnot have any to make. Therefore, when Richerson andBoyd praise us for, in substance, following a well-troddenpath, we deserve neither the compliment nor the gentlyimplied criticism.


The main theoretical point of our paper was to sketchand illustrate an original way of articulating evolved dis-positions and cultural features. To explain its originalityin a different way than we did in the article, let us dis-tinguish four current positions on the nature of cultureand explain how ours differs from all four. A first positionis shared by traditional anthropologists and memeticists.They converge in assuming that the role of evolveddispositions in culture is hardly more than to make cul-ture possible. A second, diametrically opposite position,which today no one holds, would be to assume that genesrigidly determine cultural representations. Rather, athird position is dominant among sociobiologists and be-havioral ecologists, and it is that (to quote Boster) “hu-mans are genetically similar but phenotypically (and cul-turally) diverse: The same genotype is facultativelyexpressed as different behavioral phenotypes dependingon signals from the environment.” (A “facultative phe-notypic expression of genes”—degree of sun tanning, forinstance—is one of several possible expressions that iscoded by the genes and depends for its occurrence onlocal factors.) Tooby and Cosmides (1992) hold an orig-inal fourth position that combines elements from thefirst and third ones. They distinguish “evoked culture,”that is, locally occurring behaviors that are facultativephenotypic expressions of genes, and “transmitted cul-ture,” that is, behaviors that are caused by social and, inparticular, intergenerational transmission, and assumethat both sorts of culture exist.

We disagree with all these positions. In contrast totraditional anthropologists and memeticists, we holdthat evolved psychological dispositions play a role inshaping cultural contents. In contrast to sociobiologistsand behavioral ecologists, we do not believe that culturalrepresentations and practices are phenotypic expressionsof genes, even facultative ones. In contrast to Tooby andCosmides, we reject the distinction between evoked andtransmitted culture. Rather, we hold that there is a con-tinuum of cases between cultural items that are stronglyshaped by evolved dispositions and others that areweakly shaped. All of culture is partly evoked and partlytransmitted, in different proportions.

Now, how might evolved dispositions and transmis-sion processes interact in generating culture? It may beeasier, in answering this question, to start with the ex-ample of a cultural artifact. It would make little senseto say that a hamburger is a phenotypic expression ofgenes. It is very much a historical collective product andnot a response of individual organisms to the environ-ment. At the same time, it is obvious that a hamburgerprovides a combination of glucids, protids, and lipids anda texture that meet evolved human food preferences par-ticularly well. A culturalist would point out that ham-burgers are not popular in all cultures and that food pref-erences are themselves culturally shaped, which is truebut only to a certain extent (no culture, for instance, hasgiven up on any of the three basic kinds of ingredients).However, the cultural shaping of taste is not fullyachieved until adulthood. In modern societies, in partic-ular the U.S.A., children can to a large extent impose

their food preferences, and industrial interests influencethe acceptability of foods more than, say, religious con-cerns. These factors combine to give relatively untutoredtastes a strong influence on the success of a foodstuff.In such societies, if a foodstuff like the hamburger is onoffer at all, it stands a good chance of reaching a veryhigh level of distribution. Cultural representations andpractices are like cultural artifacts. They are not coded,even as facultative expressions, by the genes. They arenot phenotypes. They are historical products of collect-ivities.

Some of the commentators express theoretical pref-erences or concerns different from ours but offer opinionsrather than arguments to which we could try to respond.Brown criticizes us for not acknowledging “affinities tothe growing bodies of work on complexity, emergentstructures, and self-organizing systems.” The body ofwork he is referring to and the more standard cognitiveapproach that we adopt have common origins in the the-ory of automata and that of information of the ’30s and’40s, but otherwise the affinities are rather weak. UnlikeBrown, we are not convinced that complexity theory isthe way to go (nor are we hostile, for that matter). Poul-shock worries that our informational-representationalperspective on culture does not fit well with the mech-anistic view that we also adopt. Here, however, we arejust in agreement again with the standard cognitive ap-proach, which aims to be both representational andmechanistic. Toren (with whose very general remarkswe would otherwise have little difficulty in findingpoints of agreement) flatly asserts that an epidemiologyof representations has “no explanatory value” and is “un-scientific.” She praises Levi-Strauss’s work based on theassumption of a universal human drive to structure cul-tural information in a particular way, a drive which heno doubt considered the product of human evolution,and then tells us that she does not believe in evolvedpsychological predispositions—a remark which, takenliterally, would make her a rare creationist contributorto current anthropology.

Several contributors complain about the vague andtentative character of our proposals. Calderon, for in-stance, finds our theoretical claims “vague and unhelp-ful.” Poulshock asks, “Are we absolutely sure that theputative predispositions in question exist? For example,are there ways to test for a human psychological ten-dency to look for information relevant to kinship ties?”The answer to his first question is of course no. Thereare no absolute certainties in the empirical sciences. Theanswer to his second question is that this is an ordinarypsychological issue quite similar to others that have beenexperimentally studied. We can understand the frustra-tion with our enterprise of scientists seeking relativelysimple causal explanations in strictly controlled exper-imental situations. Although the epidemiological ap-proach to culture has benefited from experimental work,it also needs the vaguer but much richer evidence pro-vided by anthropological fieldwork. We are, after all, try-ing to contribute to anthropology, not to displace it.

Kinship. In this article, we used the example of the


mother’s brother/sister’s son relationship merely as anillustration of our theoretical project. Apart from Boster,who complains that we “present neither new data nor acomplete reanalysis of the existing data,” the commen-tators understand this. We argued that “classical” an-thropologists, Goody in particular, had been generallyright in their analysis. Brown and Boster are clearly dis-appointed by this lack of radical novelty, but in fact itprecisely suited our purpose of showing that classicalanthropological concerns, contrary to what most an-thropologists of today think, could be fruitfully ad-dressed by adopting a naturalistic perspective.

Boster, whose commentary combines very interestingsuggestions of his own with repeated misunderstandingsof ours, asserts that the “principal proposal of this paper[is] that humans have ‘an evolved disposition to favorrelatives.’ ” Not only was this claim not our principalproposal but we even refrained from making it. Theclaim is indeed uncontroversial on the biological sideand at the same time unacceptable in principle on thecultural side. We, for our part, see it as highly plausible(and Boster explains well why it is so), but rather thantry to persuade our more culturalist readers to accept itwe wanted them just to entertain it and see that it hadplausible and interesting consequences. Lowe, thoughwilling to entertain the claim, prefers an original com-bination of Hirschfeld’s domain-specific competence forsocial categorization with a disposition to extend sen-timents toward the mother to other relatives. A similarextension-of-sentiments thesis has been refuted byNeedham (1962), who pointed out that things that wedo willingly do not normally need to be backed up bysocially organized sanctions. The problem that we triedto address and that Lowe does not deal with is how in-dividual dispositions might play a causal role in theemergence and persistence of institutions, given that weare dealing with institutionalized rules of behavior.

More generally, the area of kinship illustrates partic-ularly well both the necessity and the difficulty of ar-ticulating the biological and the cultural points of view(a concern at the center of Calderon’s comments). Thebiological notion of kinship is a relatively clear one, andit applies to humans as to other animals. The questionof how biological kinship is being managed by humansocieties obviously arises, and it may have seemed un-problematic that this is done by social institutions of“kinship” in a now anthropological sense of the term.Classical anthropologists did not mistake social insti-tutions for biological phenomena (and if we gave theimpression that we thought they did, Goody is right tocorrect us), but they took for granted a relatively clearrelationship between the two. Such writers as Goodyhimself, Fortes, and Levi-Strauss (who, according to To-ren, said the last word on explaining the recurrence ofthe special relationship of sister’s son to mother’s brotherin patrilineal societies) did not think that kinship sys-tems simply reflected biological closeness but assumedthat “kinship” was always the same isolable kind of phe-nomenon, although which links were given social sig-nificance varied.

We believe that the critiques of Leach, Needham, andSchneider have shown how fundamental the various phe-nomena labeled “kinship” really are. What in the pasthave been taken to be straightforwardly comparable in-stitutions have revealed themselves not to be so. Kinshipitself is not a type of social relation but a very loosefamily (see Carsten 2000). “Kinship types” which havebeen labeled by terms such as “patriliny” turn out toidentify very little, and it is this realization which, aswe said in the paper, has been the major cause of theabandonment of the comparative programme in socialanthropology. Just how radical this problem has been hasnot always fully sunk in, and we would argue that thisis the case for such authors as Irons, although praised byBorgerhoff Mulder. Given how muddled the issue re-mained for a long time in social/cultural anthropology,it is not surprising that many writers on the biologicalside have not grasped the problem.

Boster’s confident assertion that a correlation could bedemonstrated between paternity doubts and institution-alized privileges of sister’s son illustrates the issue (andis precisely the kind of assertion which, for good or badreasons, makes most social/cultural anthropologistscringe). Quite apart from the difficulty of measuring “pa-ternity doubts,” the problem is that “institutionalizedprivileges” are, as we point out, of totally different kindsin the different examples and the term “sister’s son” canrefer to quite different persons. For example, “sister’sson” would in some cases be own sister’s son, in othersany child of a woman of ego’s descent group. Given thisstate of affairs, it is not clear what is supposed to cor-relate with what and what would be the significance ofsuch a correlation were it to be established.

The history of anthropology is littered with such quix-otic attempts. This is why we propose an approach thatneither takes for granted nor needs a taxonomy of so-ciocultural macrophenomena as its starting point. Cul-tural forms are extremely varied and labile, and, in spiteof the illusion created by traditional anthropological ter-minology, they do not fall into discrete classes. They arejust ever-changing patterns in the history of social in-teraction. To consider the cultural variability revealedby the ethnographic record the product of history is notto state that it is random. There is, for us, an explanatorylevel, and it is that of the microprocesses that shape thesechanging sociocultural patterns and of the diverse factorsthat favor or impede specific microprocesses. Like med-ical epidemiologists with diseases, we do not expect, ingeneral, to be able to fulfill Poulshock’s request to “seein action” the mechanism by means of which some cul-tural representation or practice stabilizes. Micro-mechanisms can be anecdotally documented and, insome cases, experimentally studied. To understand theircausal role in a formally appropriate way would involvemodeling their cumulative effect in specified conditionsand matching the model to anthropological observationsof macrophenomena. At present, formal tools may beavailable (for instance, in the work of Boyd and Rich-erson), but what is sorely lacking is sufficient knowledgeof micromechanisms and macroanthropological descrip-


tions that are hostage neither to obsolete ontologies norto antiscientific ideologies.

In the particular case of kinship, to say that there isno distinct and universal type of social institution whichhas as its function to manage biological kinship does notat all imply that biological kinship and sociocultural in-stitutions are wholly unrelated. As we suggested withthe minor but, we hoped, illustrative case of the mother’sbrother/sister’s son, there may be evolved psychologicaldispositions to attend to evidence of biological related-ness and take it into account in one’s thoughts, emo-tions, and behavior. These (together with other factorswhich may or may not be linked to biological kinshipand may be more or less specific to time and place) favor,in the long term, the emergence and stabilization of rep-resentations, terminologies, practices, and institutionswith particular contents. They do not, however, deter-mine a discrete type of cultural form.

Conclusion. With the exception of Goody, whose workon this topic antedates the attitude we criticize, and To-ren, all the social and cultural anthropologists who wereapproached for comments, even those who are expertsin kinship, declined to do so. We are thus left with onlythe possibility of commenting on their silence. Thereseems to be a kind of holy horror among most culturaland social anthropologists concerning any suggestionthat a genetic element might contribute to explaininganything of anthropological significance. This is espe-cially striking in the area of kinship, given its importanceboth from a cultural and from a biological point of view.This horror is explained in part by the connection which“biological” approaches have too often had with varioustypes of racism and also by the frequent and fundamentalmisunderstanding of the complexity of cultural data andthe significance of history in human cultural systems onthe part of many of those who have taken such ap-proaches. However, it is not as if all biological ap-proaches were so guilty. Richerson and Boyd justly pointout the modernity of Darwin’s own positions in this re-spect. Therefore, the antibiological stance of much cur-rent anthropology is not fully explained by a mere excessof moral and intellectual concern. Plain disciplinary pa-rochialism and conservatism obviously play their part,perhaps made all the worse by the bizarre institutionalstructures of anthropology departments. In any case,there is nothing in our article or in our work which couldeven remotely justify a suspicion of ideological misuseof biology. Moreover, far from ignoring the complexityof historical-cultural data, one of our main purposes wasto explore how biological considerations could play apart in explaining it. The silence of social and culturalanthropologists in this debate illustrates the fact that somany of them no longer want to engage with the fun-damental questions concerned with the explanation ofvariability and uniformity in human populations exceptto express occasional supercilious annoyance at the naıv-ety of those who do. Yet, it is surely this attitude thatmakes ordinary people—who naturally, as do all of us,want the best possible scientific understanding of humannature and history—turn to scholars who share this goal

but may not have benefited from the accumulated ex-perience and disappointments of the discipline. No one,whether anthropologist, biologist, or interested reader,stands to benefit from this refusal of anthropologists toengage in the debate.

References Cited

a l e x a n d e r , r . d . 1979. Darwinism and human affairs. Seat-tle: University of Washington Press.

a t r a n , s . 1998. Folkbiology and the anthropology of science:Cognitive universals and cultural particulars. Behavioral andBrain Sciences 21:547–609.

a t r a n , s . a n d d . s p e r b e r . 1991. “Learning without teach-ing: Its place in culture,” in Culture, schooling, and psycholog-ical development. Edited by L. Tolchinsky-Landsmann. Nor-wood: Ablex.

a u n g e r , r . 2001. Editor. Darwinizing culture: The status ofmemetics as a science. Oxford: Oxford University Press. [jp]

b a l c o m b e , j . p . , a n d g . f . m c c r a c k e n . 1992. Vocalrecognition in Mexican free-tailed bats: Do pups recognizemothers? Animal Behavior 43:79–87. [jsb]

b a l d w i n , j . m . 1895. Mental development in the child andthe race: Methods and processes. New York: Macmillan. [pjr,rb]

b a l k i n , j . m . 1998. Cultural software: A theory of ideology.New Haven: Yale University Press. [mfb]

b a r n e s , j . 1951. Marriage in a changing society. Rhodes-Liv-ingstone Paper 20.

b a r t h , f . 1975. Ritual and knowledge among the Baktamanof New Guinea. New Haven: Yale University Press.

———. 1987. Cosmologies in the making: A generative approachto cultural variation in New Guinea. Cambridge: CambridgeUniversity Press.

b e c k e r m a n , s t e p h e n , a n d p a u l v a l e n t i n e . Editors.2002. Partible paternity: The theory and practice of multiplefatherhood in South America. Gainesville: University of Flor-ida Press. [jsb]

b l a k e m o r e , s . 1999. The meme machine. Oxford: OxfordUniversity Press.

b l o c h , m . 1977. The past and the present in the present. Man,n.s., 12:279–92.

———. 1987. “The ritual of the royal bath in Madagascar: Thedissolution of death, birth, and fertility into authority,” in Rit-uals of royalty: Power and ceremonial in traditional societies.Edited by D. Cannadine and S. Price. Cambridge: CambridgeUniversity Press.

———. 1998. “Commensality and poisoning,” in Food: Natureand culture. Edited by A. Mack, pp. 133–51. Social Research 66(1).

———. 2000. “A well-disposed social anthropologist’s problemswith memes,” in Darwinizing culture. Edited by R. Aunger,pp. 188–203. Oxford: Oxford University Press.

b o o n , j . a . 1982. Other tribes, other scribes: Symbolic anthro-pology in the comparative study of cultures, histories, religionand texts. Cambridge: Cambridge University Press.

b o r g e r h o f f m u l d e r , m . , m . g e o r g e - c r a m e r , j .e s h l e m a n , a n d a . o r t o l a n i . 2001. A study of East Af-rican kinship and marriage using phylogenetically-based com-parative methods. American Anthropologist 103:1059–82. [mb]

b o s t e r , j a m e s s . 1991. “The information economy model ap-plied to biological similarity judgment,” in Socially sharedcognition. Edited by Lauren Resnick, John Levine, and Ste-phanie Teasley, pp. 203–25. Washington, D.C.: American Psy-chological Association. [jsb]

b o w l e r , p . j . 1988. The non-Darwinian revolution: Reinter-


preting a historical myth. Baltimore: Johns Hopkins UniversityPress. [pjr, rb]

b o y d , r . , a n d p . j . r i c h e r s o n . 1985. Culture and the ev-olutionary process. Chicago: University of Chicago Press. [pjr,rb]

———. 1992. “How microevolutionary processes give rise to his-tory,” in History and evolution. Edited by M. H. Nitecki andD. V. Nitecki, pp. 179–209. Albany: State University of NewYork Press. [pjr, rb]

b o y e r , p . 1994. The naturalness of religious ideas: A cognitivetheory of religion. Berkeley and Los Angeles: University of Cal-ifornia Press.

———. 2001. “Cultural inheritance tracks and cognitive predis-positions: The example of religious concepts,” in The debatedmind: Evolutionary psychology and ethnography. Edited by H.Whitehouse, pp. 57–89. Oxford: Berg.

b u r n s t e i n , e . , c . c r a n d a l l , a n d s . k i t a y a m a . 1994.Some neo-Darwinian decision rules for altruism: Weighingcues for inclusive fitness as a function of the biological impor-tance of the decision. Journal of Personality and Social Psy-chology 67:773–89. [jsb]

c a m p b e l l , d . t . 1965. “Variation and selective retention insociocultural evolution,” in Social change in developing areas:A reinterpretation of evolutionary theory. Edited by H. R. Bar-ringer, G. I. Blanksten, and R. W. Mack, pp. 19–49. Cambridge:Schenkman. [pjr, rb]

c a r s t e n , j . Editor. 2000. Cultures of relatedness: New ap-proaches to the study of kinship. Cambridge: Cambridge Uni-versity Press.

c a v a l l i - s f o r z a , l . l . , a n d m . w. f e l d m a n . 1973. Cul-tural versus biological inheritance: Phenotypic transmissionfrom parents to children (a theory of the effect of parental phe-notypes on children’s phenotypes). American Journal of Hu-man Genetics 25:618–37. [pjr, rb]

———. 1981. Cultural transmission and evolution: A quantita-tive approach. Princeton: Princeton University Press. [pjr, rb]

c i a l d i n i , r . , s . b ro w n , b . l e w i s , c . l u c e , a n d s .n e u b e r g . 1997. Reinterpreting the empathy-altruism rela-tionship: When one into one equals oneness. Journal of Person-ality and Social Psychology 73:481–94.

c o l l i e r , j . , a n d s . y a n a g i s a k o . 1987. “Towards a uni-fied analysis of gender and kinship,” in Gender and kinship.Stanford: Stanford University Press.

d a r w i n , c . 1874. 2d edition. The descent of man and selec-tion in relation to sex. New York: American Home Library.[pjr, rb]

d a w k i n s , r . 1976. The selfish gene. Oxford: Oxford UniversityPress.

———. 1982. The extended phenotype. Oxford: Oxford Univer-sity Press. [ct]

———. 1993. “Viruses of the mind,” in Dennett and his critics.Edited by B. Dahlbom, pp. 13–27. Oxford: Blackwell.

d e n n e t t , d . 1995. Darwin’s dangerous idea: Evolution andthe meaning of life. New York: Simon and Schuster.

e i b l - e i b e s f e l d t , i . 1975. Ethology: The biology of behavior.New York: Holt, Rinehart and Winston.

e n s m i n g e r , j . a n d j . k n i g h t . 1997. Changing socialnorms: Common property, bridewealth, and clan exogamy.current anthropology 38:1–24. [mb]

e s s o c k - v i t a l e , s . , a n d m . m c g u i r e . 1985. Women’slives viewed from an evolutionary perspective. 2. Patterns ofhelping. Ethology and Sociobiology 6:155–73. [jsb]

f l i n n , m . v. 1981. “Uterine vs. agnatic kinship variability andassociated cousin marriage preferences: An evolutionary bio-logical analysis,” in Natural selection and social behavior. Ed-ited by R. D. Alexander and D. W. Tinkle. New York: ChironPress. [mb]

f o r d , d . h . , a n d r . l e r n e r . 1992. Developmental systemstheory: An integrative approach. Newbury Park: Sage. [edl]

f o r t e s , m . 1953. The structure of unilineal descent groups.American Anthropologist 55:17–41.

———. 1969. Kinship and the social order. London: Routledge.g a u l i n , s . j . c . , a n d a . s c h l e g e l . 1980. Paternal confi-

dence and paternal investment: A cross-cultural test of a socio-biological hypothesis. Ethology and Sociobiology 1:301–9. [mb]

g e e r t z , c . 1973. The interpretation of cultures. New York: Ba-sic Books.

g i b s o n , t . 1986. Sacrifice and sharing in the Philippine high-lands. London: Athlone.

g i l l i s o n , g . 1993. Between culture and fantasy. Chicago:University of Chicago Press.

g o o d y, j . 1959. The mother’s brother and sister’s son in WestAfrica. Journal of the Royal Anthropological Institute 89:61–88.

g l u c k m a n , m . 1954. Rituals of rebellion in South-East Af-rica. Manchester: Manchester University Press.

g o u g h , k . 1959. The Nayars and the definition of marriage.Journal of the Royal Anthropological Institute 89:23–34.

h a m i l t o n , w. d . 1964. The genetical theory of social behav-ior. Journal of Theoretical Biology 7:1–52.

h e n r i c h , j . , a n d r . b o y d . 1998. The evolution of conform-ist transmission and the emergence of between-group differ-ences. Evolution and Human Behavior 19:215–41. [pjr, rb]

h i r s c h f e l d , l . 1984. Kinship and cognition. current an-thropology 27:217–42.

———. 1994. “Is the acquisition of social categories based on do-main-specific competence or on knowledge transfer?” in Map-ping the mind: Domain-specificity in cognition and culture.Edited by L. A. Hirschfeld and S. A. Gelman, pp. 201–33. Cam-bridge: Cambridge University Press. [edl]

———. 1996. Race in the making: Cognition, culture, and thechild’s construction of human kinds. Cambridge: MIT Press.

i ro n s , w. 1979. “Investment in primary social dyads,” in Evo-lutionary biology and human social behavior: An anthropolog-ical perspective. Edited by N. A. Chagnon and W. Irons, pp.181–213. North Scituate, Mass.: Duxbury. [mb]

j a c o b , p . 1997. What minds can do. Cambridge: CambridgeUniversity Press.

j a m o u s , r . 1991. La relation frere-soeur. Paris: EHESS.j u n o d , h . 1912. Life of a South African tribe. Neuchatel,

Switzerland: Attinger Bros.k e n d r i c k , d o u g l a s . 1991. Proximate altruism and ultimate

selfishness. Psychological Inquiry 2:135–37. [jsb]k ru g e r , d a n i e l j . 2001. Psychological aspects of adaptations

for kin directed altruistic helping behaviors. Social Behaviorand Personality 29:323–30. [jsb]

k u p e r , a . 1982. Lineage theory: A critical retrospect. AnnualReview of Anthropology 11:71–95.

———. 1988. The invention of primitive society: Transforma-tions of an illusion. London: Routledge. [ct]

l a l a n d , k . n . , a n d g . r . b ro w n . 2002. Sense and non-sense: Evolutionary perspectives on human behaviour. Oxford:Oxford University Press. [mb]

l e a c h , e . 1954. Political systems of highland Burma. London:Bell.

———. 1955. Polyandry, inheritance, and the definition of mar-riage. Man 55:182–86.

l e v i - s t r a u s s , c l a u d e . 1977 (1963). “Structural analysis inlinguistics and anthropology,” in Structural anthropology. Har-mondsworth: Penguin (Peregrine) Books. [ct]

———. 1978 (1973). “Reflections on the atom of kinship,” inStructural anthropology II. Harmondsworth: Penguin (Pere-grine) Books. [ct]

l e v y, ro b e r t i . 1984. “Emotion, knowing, and culture,” inCulture theory: Essays on mind, self, and emotion. Edited byR. Shweder and R. LeVine, pp. 214–37. Cambridge: CambridgeUniversity Press. [edl]

l o w e , e d w a r d d . 2002. A widow, a child, and two lineages:Exploring kinship and attachment in Chuuk. American An-thropologist 104:123–37. [edl]

l u h m a n n , n i k l a s . 1986. Ecological communication. Trans-lated by John Bednarz Jr. Chicago: University of Chicago Press.[mfb]

l u m s d e n , c . j . , a n d e . o . w i l s o n . 1981. Genes, mind,and culture: The coevolutionary process. Cambridge: HarvardUniversity Press. [pjr, rb]


l y n c h , a . 1996. Thought contagion: How belief spreadsthrough society. New York: Basic Books. [jp]

m a t t h e w s , k . , c . b a t s o n , j . h o r n , a n d r . ro s e n -m a n . 1981. The heritability of empathic concern for others.Journal of Personality 49:237–47. [jsb]

n e e d h a m , r . 1962. Structure and sentiment: A test case insocial anthropology. Chicago: University of Chicago Press.

———. 1971. “Remarks on the analysis of kinship and mar-riage,” in Rethinking kinship and marriage. Edited by R.Needham. London: Tavistock.

n o r e n z a y a n , a . , a n d s . a t r a n . n.d. “Cognitive and emo-tional processes in the cultural transmission of natural andnonnatural beliefs,” in The psychological foundations of cul-ture. Edited by M. Schaller and C. Crandall. Hillsdale, N.J.:Lawrence Erlbaum Associates.

p f e n n i g , d a v i d w. 2002. “Kin recognition,” in Encyclopediaof evolution. Edited by M. Pagel. Oxford: Oxford UniversityPress. [jsb]

r a d c l i f f e - b ro w n , a . 1924. The mother’s brother in SouthAfrica. South African Journal of Science 21:542–55.

r i c h a r d s , r . j . 1987. Darwin and the emergence of evolu-tionary theories of mind and behavior. Chicago: University ofChicago Press.

r i c h e r s o n , p . j . , a n d r . b o y d . 2001a. Built for speed notfor comfort: Darwinian theory and human culture. History andPhilosophy of the Life Sciences 23.

———. 2001b. “The evolution of subjective commitment togroups: A tribal instincts hypothesis,” in Evolution and the ca-pacity for commitment. Edited by R. M. Nesse, pp. 186–220.New York: Russell Sage Foundation.

r i v e r s , w. h . r . 1924. Social organisation. London: KeganPaul.

ro s a l d o , m . 1980. Knowledge and passion. Cambridge: Cam-bridge University Press.

s a h l i n s , m . 1976. The use and abuse of biology. London:Tavistock.

s c h n e i d e r , d . 1984. A critique of the study of kinship. AnnArbor: University of Michigan Press.

s h i n e , r . , m . j . e l p h i c k , a n d p . s . h a r l o w. 1995. Sis-ters like it hot. Nature 378:451–52.

s m i t h , e . a . , m . b o r g e r h o f f m u l d e r , a n d k . h i l l .

2001. Controversies in the evolutionary social sciences: Aguide for the perplexed. Trends in Ecology and Evolution 16:128–35. [mb]

s p e r b e r , d . 1985a. On anthropological knowledge. Cam-bridge: Cambridge University Press.

———. 1985b. Anthropology and psychology: Towards an epide-miology of representations (The Malinowski Memorial Lec-ture, 1984). Man, n.s., 20:73–89.

———. 1996. Explaining culture: A naturalistic approach. Ox-ford: Blackwell.

———. 2000. “An objection to the memetic approach to cul-ture,” in Darwinizing culture: The status of memetics as a sci-ence. Edited by Robert Aunger, pp. 163–73. Oxford: OxfordUniversity Press. [jsb]

s p e r b e r , d a n , a n d l a w r e n c e h i r s c h f e l d . 1999. “Cul-ture, cognition, and evolution,” in MIT encyclopedia of thecognitive sciences. Edited by Robert Wilson and Frank Keil, pp.cxi–cxxx. Cambridge: MIT Press.

s t a c k , c . 1983. All our kin: Strategies for survival in a blackAmerican community. New York: Harper Collins.

s t e r e l n y, k i m . 2001. Dawkins vs. Gould: Survival of the fit-test. Cambridge: Icon Books/New York: Totem Books. [ct]

t a r d e , g . 1895. Les lois de l’imitation. Paris: Felix Alcan.t a y l o r , m a r k c . 2001. The moment of complexity: Emerg-

ing network culture. Chicago: University of Chicago Press.[mfb]

t o o b y, j . , a n d l . c o s m i d e s . 1992. “The psychologicalfoundations of culture,” in The adapted mind: Evolutionarypsychology and the generation of culture. Edited by J. Barkow,L. Cosmides, and J. Tooby. New York: Oxford University Press.

t o r e n , c h r i s t i n a . 1999a. Mind, materiality, and history:Explorations in Fijian ethnography. London: Routledge. [ct]

———. 1999b. Compassion for one another: Constituting kinshipas intentionality in Fiji. Journal of the Royal AnthropologicalInstitute 5:265–80. [ct]

———. 2002. “Anthropology as the whole science of what it isto be human,” in Anthropology beyond culture. Edited byRichard G. Fox and Barbara King. Oxford: Berg. [ct]

t u r n e r , v. 1969. The ritual process. London: Routledge.v i v e i ro s d e c a s t ro , e . 1992. From the enemy’s point of

view. Chicago: University of Chicago Press.

Documents

Kinship and Evolved Psychological Dispositions - Dan Sperber