Learning A Culture The Way Informants Do:

Observing, Imitating, and Participating

Alan Page Fiske


Most fieldwork in anthropology—indeed most social science research—relies primarily on informants’ verbal descriptions or explanations. Yet research on children around the world shows that adults hardly ever tell children how to do anything or explain anything to them. Children typically learn their cultures by observation, imitation, and participation. Ethnographers should do likewise if they aim to understand the inarticulable practical competence that constitutes much of culture. Recent advances in psychology demonstrate that explicit declarative knowledge is only one of several distinct kinds of competence, each of which is learned in a different way.

The Inexplicability of Action

A society is a group of people who exhibit many resemblances among themselves produced by imitation or by counter-imitation.(Tarde 1900:xii;emphasis in original)

Children, even the older ones, are rarely offered straightforward explanations on social matters, beliefs, ideas, values, or rituals. They must use their eyes and ears and reason a great deal on their own. They are not encouraged to ask questions or to seek explanations on why things are the way they are. When they do so, they will usually be cut short with a remark like ‘that is how it is’, or ‘that is customary’....

I may add perhaps, that I felt this absence of formal teaching quite trying myself, and not very helpful to my endeavors to familiarize myself with the culture. But children have an amazing gift of participation, and they learn to behave though they are given limited instruction.(Nicholaisen 1988:205-206, on the Punan Bah of Sarawak.) 

There is every reason to think that as soon as he reflects on his practice, adopting a quasi-theoretical posture, the agent loses any chance of expressing the truth of his practice, and especially the truth of the practical relation to the practice.Academic interrogation inclines him to take up a point of view on his own practice that is no longer that of action, without being that of science.... Simply because he is questioned, and questions himself, about the reasons and the raison d’être of his practice, he cannot communicate the essential point, which is that the very nature of practice is that it excludes this question.(Bourdieu1990:91)

The anthropologist must relinquish his comfortable position in the long chair on the veranda of the missionary compound, Government station, or planter’s bungalow, where, armed with pencil and notebook and at times with a whisky and soda, he has been accustomed to collect statements from informants, write down stories, and fill out sheets of paper with savage texts.He must go out into the villages, and see the natives at work in gardens, on the beach, in the jungle; he must sail with them to distant sandbanks and to foreign tribes, and observe them in fishing, trading, and ceremonial overseas expeditions.{sic} Information must come to him full-flavored from his own observations of native life, and not be squeezed out of reluctant informants as a trickle of talk.... Open-air anthropology, as opposed to hearsay note-taking, is hard work, but it is also great fun(Malinowski 1954/1926:146–147)

The thesis of this article is that people acquire most of their culture by observing and participating.This participation is often based primarily on imitation of observed practices that people acquire and know motorically, as bodily skills. These kinds of competence can rarely be translated into articulate verbal concepts. Informants pressed to explain practices that they themselves learned by observation, imitation, and participation generally have to make up concepts that have very tenuous, often imaginary relations with the manner in which the informants themselves actually acquired or generate the actions in question. Informants’ translations of such savoir faire into conceptual language tend to be highly problematic, distorted, and confabulated because informants are simply unaware of and quite unable to explain how or why they actually do perform most of their practices. Consequently, interviewing, questionnaires, life history narratives, descriptions of events, explanations of motives or norms, and other verbal reports are not valid primary methods for learning about most of any culture.Fieldworkers have to rely on true participant observation to learn a culture, because that is the only medium in which people can acquire, reproduce, or transmit most of their culture.

Let me give you an example. I am doing fieldwork in a small village among the Moose (pronounced MOH-say; formerly spelled “Mossi”) of Burkina Faso, in West Africa. After eighteen months of intensive instruction and immersion, I have become quite fluent in Moore. Some people are beginning to trust me. I’ve learned that tomorrow there is going to be a major ritual that comes only once a year. I go to talk to some friendly, cooperative informants.

“What’s going to happen tomorrow?” I ask.

“It’s Kiuugu,” they say.

“What’s Kiuugu?” I ask.

“A sacrifice,” they answer, with a combination of mild amusement and perplexity.

“What will happen?” I probe.

They pause, trying to find an answer. “Well, you’ll be there; you’ll see” they finally respond.

“Well, can you at least tell me why you do Kiuugu?” I ask.

They reply with a stock phrase: “It’s what we found when we were born—and we’ll leave behind when we die.”

I try to push, gently, politely, but firmly, for more description, explanation, anything!The more I push my friends, the more irritated and perplexed they become. They don’t understand what my questions mean, or how to answer them. Finally I give up. 

I go to the ritual. It begins with the phrase, “This is what we found when we were born” [identified possessively: our tradition].Afterwards, I try again to get some kind of exegesis. It’s impossible; we all get exasperated at each other. It’s not a matter of secrecy: there is no hidden or privileged meaning. It’s just a tradition. 

It’s as if I grilled you about why you carve faces in pumpkins—and why pumpkins, rather than watermelons? Why on October 31?“That’s just what we do!”That’s what Halloween is!” Similarly, most Americans would probably be at a loss if I asked them why they eat cake and ice cream at birthday parties, why they light candles and then blow them out—or why they celebrate birth anniversaries at all.Could you—or most American informants—give any answer that reflects an articulated understanding of birthday rituals that was in your mind before I asked the question? 

Americans learn about the significance of birthday candles primarily by observing and participating in birthday parties. We rarely, if ever, discuss their meaning with anyone, beyond a few simple ideas such as matching the number of candles to the number of years or asserting that you get your wish if you blow out the candles. You probably do not carve jack-o’-lanterns and put candles on cakes as a result of anyone’s explanation of their significance. You construct Halloween or a birthday party ritual primarily by reenacting memories of past practices in which you have participated. These reenactments are imitations of observed actions, not deductions from propositional rules or conformity to linguistically formulated norms. 

Semiotically, the practice of placing candles on birthday cakes and blowing them out is transmitted by bodily mimesis. It is encoded in the mind almost kinesthetically, as a set of motoric enactments, like mime. As a result, if I inquire about birthday candles, you are likely to be at a loss to provide a verbal articulation that captures the basis for this practice. You did not learn this practice in a linguistic medium, and it is difficult to explain it verbally. Imagine learning to dance, to pitch a baseball, or to flirt. You learn by imitatively attempting to perform the actions you have observed. Conversely, it would be virtually impossible to get it right without ever seeing it done. One demonstration that you can mimic is worth a number of words. 

Of course, there are limits to what we can learn by observation. I’ve watched Michael Jordon on a number of occasions, and I still can’t quite manage some of those moves. Unfortunately, interviewing won’t solve that problem. Think about it: if Michael Jordan could explain to me the somatosensory and motoric processes that enable him to hit those baskets and make those passes, and if I could just translate his explanations back into somatosensory and motoric competence, I could be a short, aging, feeble Michael Jordan! (I’d settle for that!)

But for the Moose, it goes beyond the problem of exegesis. They seem unable even to give me a verbal description of the ritual. This Kiuugu ritual is extremely important; I deduce that it is the single most important enactment and constitutive marker of village solidarity. Nevertheless, even my most motivated and intelligent informants can not give me a verbal script—let alone an exegesis—for a ritual they have performed every year of their lives. 

In some cultures, people do describe and discuss their rituals with each other. However, likemany other peoples in Africa, the Moose have no indigenous tradition of reflective analysis of their own practices. They have a rich, elaborate religion, but no theology. They have a complex society, but no ethnosociology. Like many other African peoples, they have virtually no mythology or cosmology. They have a sophisticated political system, but no political science. They live their lives in practice, but without any great interest in reflecting on it, analyzing it, or trying to explain it. 

Did you ever dance? Can you describe to me, in words alone, how to dance? Have you ever analyzed the meaning of dance steps? (Explain the mambo or jitterbug, if you can.)Did you play basketball or field hockey? Did you devote your energies to doing these things, or to accounting for them? Did you ever analyze the reasons for the having precisely five players on each side, or for the rule against kicking the basketball? Clearly, practice, even the most refined practical competence, need not necessarily give rise to reflective analysis. 

Moose learn their many rituals by observing them, then participating in minor roles, and eventually carrying them out with others.Moose evidently encode, think about, and reproduce their rituals in a kinesthetic or sensori-motor mode that resembles the way a dancer, a gymnast, a magician, a surgeon, a carpenter, a weaver, or a fly-caster encodes, thinks about, and reproduces the relevant skilled practices.As an ethnographer, I had to do likewise. Eventually, Moose carried out many rituals with me, often for the health and welfare of me and my family. But they never described them or verbally prescribed how to do them. They just performed them, and then left me to carry on performing them mimetically. Later, three different diviners independently transmitted to me what they asserted was the capacity to see the moral meanings of misfortune in the patterns of cowry shells tossed on the ground. They passed on to me the magical implements and legitimated my personal powers, anointing me and my implements in special rituals. But none of the diviners ever thought to explain or even demonstrate divination to me pedagogically. Nor did they recognize the point in doing so when I asked them to teach me. 

That leaves me, as an ethnographer, with the responsibility for translating these practices into a written text, oral talk, diagrams, charts, or figures for my own academic audiences. And it certainly leaves me to explain these practices; the Moose have very little interest in doing so.These are difficult semiotic and analytic problems, but they are properly my problems: there is no reason to try to force the Moose to do something they are not accustomed to doing, and do not see the point in doing.Attempting a verbal representation of a ritual is unnatural and infelicitous—never mind a discursive exegetical analysis. 

The Moose are not unique in this respect.Victor Turner found that Ndembu have hardly any mythology or explicit cosmology, and he found it rather difficult to make sense of their rituals. Then he ran into Muchona, a wandering, marginal man who loved to talk about ritual and about his own activities as a healer. Muchona’s interpretations of ritual symbols were uniquely detailed, clear, consistent, and cogent. Turner was enthralled by these elaborate exegetical discourses, paying Muchona handsomely for them and using them as the basis for most of his analysis of Ndembu ritual. Turner discounts as mere jealousy the skepticism of his Ndembu research assistant, who ridiculed Muchona’s accounts and said he was lying. Turner acknowledges that many Ndembu scoffed at Muchona, points out that Muchona “delighted in making explicit what he had known subliminally about his religion” (p. 138), and observes poignantly that when Turner left, Muchona “could no longer communicate his ideas to anyone who would understand them” (p. 150).So we have to ask whether Muchona’s singular verbal explications have anything much to do with the way most other Ndembu understand, remember, reproduce, and use rituals or find them compelling. The same problem arises with respect to the famous cosmology generated by the Dogon philosopher Ogotemmeli in his conversations with the French ethnographer, Griaule. Subsequent research among the Dogon has completely failed to uncover any corroborative evidence or resonance of this cosmology among other Dogon (van Beek 1991).It seems as if the religious practices of the Moose, Ndembu, and Dogon are fundamentally sensori-motor enactments, motorically represented and transmitted. There may be little or no linguistically explicit conceptual foundation for them at all. 

This lack of articulable knowledge poses big methodological problems. In a few cultures, such as many of those in Europe and South Asia, reflective exegetical analysis is a widespread cultural practice. In these cultures, widely-known and discussed indigenous accounts of cultural practices may sometimes feed back to transform these practices. But even where they are readily forthcoming, informants’ explanations may be far removed from the generative mechanisms that actually produce the actions in question.Ethnosociology and ethnopsychology are appropriate topics of research in their own right, but they are not valid substitutes for scientific sociology, psychology, or anthropology. 

This is important, of course, because most anthropological and psychological investigations have relied primarily on verbal data. Interviewing is the core of most fieldwork. Language has also been the focus in many or most studies of the child’s constructive acquisition ofculture. Indeed, some researchers have even focused on meta-language, utilizing interviews, narratives, or other linguistically-formulated representations of language (e.g., Miller and Hoogstra 1992). This work has suggested that language learning is closely associated with certain aspects of the acquisition of social competence (Ochs and Schieffelin 1984; Ochs 1988; Watson-Gegeo and Gegeo 1977). However, this discourse-oriented research tradition generally fails to consider the overall question of how children or adults use different semiotic media to acquire or construct their culture. Indeed, many developmental and psychological anthropologists effectively ignore the existence of any other mode of communication or learning aside from language. Anthropological fieldwork often consists primarily of interviewing, supplemented by recording of other verbal communications, without regard for any of these epistemological, cognitive, or semiotic issues.(In two informal samples representing hundreds of recent Ph.D. dissertations in anthropology, an enormous majority relied almost exclusively on interviews—typically using translators;S. Ferzacca, personal communication.) Social and clinical psychologists rely even more exclusively on verbal data collection. Rarely do anthropologists or psychologists even stop to consider whether the competence, knowledge, practice, or action they are studying is verbally articulable.If they recognize that they are studying something that is non-verbal, they usually do not go beyond stating that it is implicit or embodied—lumping together all that is inarticulate without attempting to characterize or differentiate the manner in which it is learned, remembered, reformulated, or produced.

Non-Verbal, Non-Conceptual Skills

The capacity to do something does not entail discursive knowledge of how or why it is done—or even awareness that one is doing it. Developing a long tradition in philosophy and psychology, Merleau-Ponty (1962/1945) contrasted the explicit, verbally formulated, objectifying symbolic understanding of the conscious mind with the praktognosia (practical knowledge) contained in bodily action. Ryle (1949) labeled this distinction “knowing how” versus “knowing that”. To introduce an example, a person may know how to ride a bicycle without being able to describe or explain how to do so, without being able to control the necessary movements reflectively and self-consciously, and without remembering the occasions or the manner in which she first learned the skill. (Did you know that to start a left turn you momentarily pull the handle bars a little to the right?) This distinction has been developed, revised, and elaborated by many philosophers and psychologists. Schutz (1977/1951) contrasted social interactions involving communication though concepts (whose meaning can be grasped at a given moment in time) with communication based on meaning that is inherently temporal, based on a joint experience of the flux of activities that are articulated temporally in a step-by-step sequence, such as music (see Lindsay’s 1996 phenomenological account of making music together, Feld’s 1982 ethnography of music, sound, and emotion among the Kaluli, and Highwater’s 1985 discussion of the meaning of dance). One consequence of this awareness in philosophy and social science has been the development of phenomenological approaches (see Jackson 1996 for an excellent review of the philosophical and anthropological contributions to phenomenology). 

Psychologists have collected a considerable body of experimental and clinical evidence demonstrating the importance of this disjunction between what people can verbally articulate and what they can do—or between the cognitive and affective processes that actually shape their behavior and their conscious (albeit private) representations of their motives and minds.Nisbett and Wilson (1977) reviewed the early evidence about “Telling More than We Can Know,” showing that people are often unaware of the stimuli that influence their own actions, unaware of the influences of dispositional and situational factors that affect their actions, or unaware of their own actions. Because they are unaware of the actual causes, people often mistakenly explain even their own actions by referring to their a priori, implicit theories of behavior. Subsequent research has shown that people’s verbal reports of their own thought processes—or of their attitudes and behaviors—may be accurate in some circumstances, but they may also misrepresent many aspects of cognition (Ericsson and Simon 1993). Clearly, competence does not imply cognizance. We can adequately explainhow we do only a very few of the things we are able to do (see Borofksy 1994, for some implications of this).

Donald Fiske (1978, 1981, 1986) pointed out that the ambiguity and shifting denotation of words (in such instruments as rating scales and written records of behavior) often make the validity of verbal reports of one’s own or others’ actions problematic as the basis for a scientific psychology. Major problems in the use of verbal reports stem from the sensitivity of subjects’ responses to the precise manner in which questions are formulated and also from the variation in responses that depend on the subject’s perceptions of the interviewer. Furthermore, subjects may construct invalid responses to questions that request knowledge they do not have. Consequently, verbal reports are often invalid, unreliable, and misleading. (See Susan Fiske 1995 for strategies for dealing with these problems in social psychological experiments, and Rowe 1997 for an overview focusing on issues regarding self-reports.) 

In particular, social inferences exhibit this distinction between processes of memory/learning/competence that are accessible to conscious reflection or verbal articulationand processes that are not. Bargh has demonstrated the pervasiveness of unintended, involuntary, effortless, autonomous, nonconscious processes of social attention, perception, categorization, attribution of meaning, evaluation, affective response, motivation, and goal-establishment (e.g., Bargh 1990, 1992, 1994 1997; Spielman, Pratto and Bargh 1988; Uleman and Bargh 1988). Since people are unaware of these “automatic” social processes, they cannot report them or explain why they make the resultant categorizations, interpretations, and evaluations, or why they pursue the resultant goals. Many of the most important kinds of social competence are procedural, in this sense: people can make skilled social inferences or utilize stereotypes and other categories and relational schemas without knowing how they do so, or even that they do so (e.g., Smith 1989, 1994, 1997;Smith and Branscombe 1987). For example, in a score of studies we found that people think about others in terms of a set of implicit relational models they could not label or characterize; memory for events and persons, errors of action and naming, judgments of similarity and classification of social relationships are all based on relational models that are not represented in the surface lexicon and are not ordinarily articulated as such (summarized in Fiske & Haslam 1996). People were not even aware that they were thinking about others in relational terms rather than focusing on individual attributes. 

Some anthropologists have always been aware of this issue. In 1887 when E. B. Tylor prepared a guide for the collection of ethnographic data in northwest Canada, he cautioned against reliance on asking preset questions, recommending instead the observation of religious rites and the transcription and translation of myths (Stocking 1983:72–73).In a famous aphorism, Marett (1929/1909:xxxi) criticized intellectualist theories of religion for being 

too prone to identify religion with this or that doctrine or system of ideas. My own view is that savage religion is something not so much thought out as danced out; that, in other words, it develops under conditions, psychological and sociological, which favor emotional and motor processes, whereas ideation remains relatively in abeyance.

Marett’s student James (1917) joined him in his critique of Frazer and Durkheim’s ideational theories, arguing that “In the first place savages live out rather than think out their cult. To them, ‘religion’ is not a matter of theory but of practice” (p. 5). Myths, James argued, were ex post facto explanations and justifications of how and why they conducted their rites (p. 217).What is required in Australian aboriginal rites, for example, is the performance of the ceremony in a prescribed manner—beliefs, theories, theology and dogma may grow out of ritual, but are not the original source of it (p. 224).While the thesis of Marett and James suffers from the attribution of concrete enactive thought to “primitives” or “savages,” leaving abstract theories to more “civilized” peoples, their contribution is in recognizing that daily life, including religion, is often embodied in action—not abstraction. 

Later Malinowski raised the issue again. 

A man who submits to various customary obligations, who follows a traditional course of action, does it impelled by certain motives, to the accompaniment of certain feelings, guided by certain ideas. These ideas, feelings, and impulses are moulded and conditioned by the culture in which we find them, and are therefore an ethnic peculiarity of the given society. An attempt must be made therefore, to study and record them.

But is this possible? Are these subjective states not too elusive and shapeless? And, even granted that people usually do feel or think or experience certain psychological states in association with the performance of customary acts, the majority of them surely are not able to formulate these states, to put them into words.(Malinowski 1922:22). 

Mauss (1973/1936) made very concrete contributions to the demonstration that culture consists of more than ideas, values, and institutions; it also consists of bodily techniques or “habitus”.People sit, walk, swim, eat, sleep, and gaze according to their cultures. Bourdieu’s (1990) development of Mauss’s concept of habitus is expressly intended to capture the idea that the core of culture consists of generative dispositions, principles, and kinds of competence that are ordinarily outside of and incompatible with consciousness. (You can demonstrate this for yourself, at some peril, if you attempt to walk down stairs or ride a bicycle by consciously and reflectively deciding upon each of the necessary movements). Bourdieu contrasts habitus with rule following or intentional conformity to explicit norms. “Practice excludes attention to itself (that is, to the past). It is unaware of the principles that govern it and the possibilities they contain; it can only discover them by enacting them, unfolding them in time” (Bourdieu 1990:92; cf. Connerton 1989:101–102). Agents generally cannot take their practices out of their practical temporal context; consequently, abstract “theoretical replications transform the logic of practice simply by making it explicit” (Bourdieu 1990: 93). Most cultural practices are taken for granted; they ‘go without saying.’ As Connerton (1989: 102) observes, the performativity and formalization of many collective rituals makes them especially immune to discursive questioning or critical scrutiny.

Bourdieu (1990:56) contrasts habitus with consciously formulated intentions. He characterizes habitus as “embodied history, internalized as a second nature and so forgotten as history.... a spontaneity without consciousness or will.”Practical sense, Bourdieu (1990:69) writes, is “social necessity turned into nature, converted into motor schemes and body automatisms” without agents being fully aware of what they are doing or how they do it.Habitus, Bourdieu (1990:73) writes, is acquired by practical mimesis based on identification, not by conscious effort to imitate something explicitly taken as a model per se. Similarly, the reproduction of habitus takes place “below the level of consciousness” without memory or reflexive, articulated knowledge. Especially in societies without schools, “the essential part of the modus operandi that defines practical mastery is transmitted through practice, in the practical state, without rising to the level of discourse.... Schemes are able to pass directly from practice to practice without moving through discourse and consciousness” (pp. 73–74). As Connerton (1989:101–102) points out, many performative bodily memories operate at the societal level, as skilled habitual cultural performances that may take place and be transmitted without conscious reflective attention to them.

Sapir (1949/1927) got much more specific and concrete. He described speakers’ unconsciousness of the conceptual and sound systems that form the basis of their language, along with people’s lack of awareness of the premises of their systems of exchange and accumulation of wealth. Sapir demonstrated that people operate with reference to a myriad of historically transmitted cultural patterns that they take for granted as given in the nature of things and which they cannot understand in explicit terms. Whorf (1956a/1937) built on Sapir’s analyses to describe many covert categories or cryptotypes of language (such as intransitive verbs) that are not overtly marked by any surface morpheme. Whorf (1956b/1940) also wrote about “background phenomena.”He pointed out that speakers cannot readily reflect on the linguistically-relative grammatical distinctions that their language requires them to make; moreover, speakers are unaware of the effects of these distinctions on their everyday action and thought.Using these ideas and Linton’s characterization of covert culture, Kluckhohn (1943) analyzed covert patterns of culture that he called cultural configurations—unstated premises that informants use to organize their behavior without being aware of doing so. Even within the broad category of overt culture that is visible to an observer, Kluckhohn distinguished between explicitly stated, normatively sanctioned patterns of culture such as formal ideals and actual behavioral patterns.What people say they do may be far from their real practices. 

From the start, psychoanalytic anthropologists have assumed that basic dynamic processes, by their very nature, inevitably ensure the inaccessibility of the most fundamental collective and personal meanings of cultural practices and symbols. (See Anna Freud’s 1973/1936 classic account of defense mechanisms; on repression in relation to cultural practices, see Johnson, 1997). One of the longest debates in anthropology concerns the variability of the unconscious infantile object cathexes toward each parent and the resolution of this Oedipus complex (Malinowski 1927, Anne Parsons 1969, Spiro 1982, Obeyesekere 1990, Kurtz 1992).In this view, most cultural institutions function to provide acceptably transformed outlets for unconscious drives that people are inherently unable to satisfy directly or even acknowledge. People adopt and sustain religion, mythology, the arts and many other aspects of culture because they provide socially supported mechanisms of defense against these libidinal drives that cannot be directly expressed (Kardiner 1939,Róheim 1943).Conversely, defense mechanisms such as repression, projection, or sublimation are culturally constituted to a considerable degree (Spiro 1965).What this means is that resistance against recognizing (much less communicating) what underlies everyday cultural symbols, activities, and institutionsis by no means adventitious.The axiom of psychoanalytic anthropology is that personality and culture are largely the result of the fact that humans cannot normally admit their fundamental motives to consciousness. 

From another point of view, the founder of anthropological structuralism, Lévi-Strauss (1953:526–527), stressed the importance of unconscious structural models that generate people’s kinship system, social structure, mythology, and other communicative action. Often, he argued, people have no conscious models of their communicative structures and the conscious collective models—norms—that people do construct for themselves are generally unsatisfactory from an explanatory point of view.

In his interpretations of Ndembu ritual symbols, Turner (1967a) distinguished between three kinds of meaning. The exegetical meaning consists of the explanations given by informants in response to the ethnographer’s questioning; these responses may be uniquely personal, common lay knowledge, or esoteric knowledge of specialists. The operational meaning of a ritual symbol is evident in how Ndembu use it: who are the users, what are their emotions when using it, and who is absent when the symbol is employed.“For the observer must consider not only the symbol but the structure and composition of the group that handles it or performs mimetic acts with direct reference to it” (1967a: 51). The third type of meaning is positional, deriving from its relationship with other symbols in the total Gestalt. The operational and positional meaning of ritual symbols is largely inaccessible to most ritual performers, while their exegetical knowledge may be quite circumscribed—and, we might add, typically peripheral to their performance as practice. 

It is clear at this juncture that there is much more to culture than just talk: much is habitus, practical knowledge, procedural, automatic, unconscious, covert, embodied, experiential, sensate. Some anthropologists such as Stoller (1989), Desjarlis (1992), Devisch (1993), Csordas (1994), Jackson (1995) and have recently experimented effectively with a more embodied or phenomenological methodology. But we have only just begun to address the basic question: What are the basic modes of cultural construction? How do people acquire, remember, reformulate, constitutively perform, and transmit culture?

Artifacts, architecture, socially transformed landscapes and ecologies, domesticated plants and animalsall play major roles in the reproduction and transformation of culture. The use of space is important—for example, the distribution of bodies when people sleep (Shweder, Jensen, and Goldstein 1995). An especially important medium for the constitution and conveyance of culture is the entry into the body of substances such as food, drink, tobacco, kola, betel, psychotropic drugs, medicines, and other persons. For example, Rabain (1979, Zempleni-Rabain 1973) points out the importance of body contact, posture, proximics, and the giving and sharing of food—especially nursing and egalitarian exchange of food among siblings—in the molding of social relations among Wolof children in Senegal.Elias (1993) has detailed the historical processes through which the manner of eating came to mark social status in renaissance Europe.In virtually every culture, but especially traditional ones, the sharing of food is the most important marker of inclusion and solidarity: leaving a co-present person out of a meal or a smoke or a drink or a share of any other comestible is a sign of hostile exclusion. Conversely, partaking in ceremonial food together—sacrificial meat and libation beer, or Communion wafers and wine—is the basic performative marker of unity or participatory equivalence that I call Communal Sharing (see Fiske 1991).There are certainly other essential modalities as well. But one medium stands out above all others as a core channel for the reproduction of culture: motoric imitation. 


Tarde (1900/1890) and Baldwin (1897) regarded imitation as the basis of society, socialization, and the formation of the self.Tarde (1900:73) wrote that a society is a group of people in so far as they are imitating each other or, if not currently imitating each other, who resemble each other because their common traits are previously copied from a common model.[2]In one of the first books on developmental psychology, Baldwin (1895) criticized and reformulated Tarde’s ideas, analyzing imitation as the principal source of mental development and showing how it gives rise to volition and the self. Describing his own children, Baldwin (1895:362–365) also illustrated how they learned adult roles by imitative play (see also Baldwin 1910).Baldwin (1897, 1910) developed his theory of imitation into a sophisticated theory of cultural transmission (“social heredity”), the constitution of social norms, and what we now call agencyBaldwin emphasized that children imitate selectively and generalize from their imitations, learning how to learn and to invent. Baldwin described how sociality results from imitation: children learn and incorporate the subjective perspectives of their social partners (the “socius”) by imitating them. Then when children observe others acting as they themselves have acted, they attribute to them similar volition and emotions, developing empathy and intersubjectivity as they develop a social self. Thus imitation is the basis of social organization but also of social change (“progress”). 

In one of the first books on social psychology, McDougal (1908; as a member of the Torres Straits expedition, one of the first fieldworkers) followed Tarde and Baldwin in attributing collective mental life and shared ways of doing to imitation. McDougal agreed with Tylor that many practices persist by imitation, long after their original meanings have been lost—sometimes resulting in the subsequent creating of new meanings for such survivals. Much later Miller and Dollard (1941) offered a Hullian behaviorist theory of imitation that they tested in various experiments: they confirmed that children (and adults) are more likely to be rewarded when their behavior matches that of higher status persons. Bandura and Walters (1963) extended and revised these learning theory accounts. They conducted further experimental investigations of social imitation, focusing on the acquisition, inhibition, disinhibition, or eliciting of “prosocial” and “deviant” (particularly aggressive) responses.

More recently Carol Eckerman has been studying imitation naturalistically by analyzing videotapes of everyday interaction among American children in the first two and a half years of life. Eckerman has shown that imitation is the principal medium for social coordination and the matrix in which verbal conversation takes form. She found that 12-month-old children observe adults and then tend to select corresponding objects to manipulate—while smiling, vocalizing, gesturing to, and approaching the adult (Eckerman, Whately, and McGhee 1979). At 24 months, when an adult imitates their actions, toddlers respond by seeking eye contact, prolonging their own actions, and generating imitative games (Eckerman and Stein 1990).From 16 to 32 months, children increasingly coordinate their actions with peers, primarily through mutual imitation (Eckerman, Davis, and Didow 1989).During this developmental period, toddlers tend to initiate conversation after first coordinating with each other through non-verbal imitation (Eckerman and Didow 1996).

Research with apes and other animals indicates that humans have evolved a specialized capacity to imitate (Nagel, Olguin, and Tomasello 1993; Tomasello, Kruger, and Ratner 1993; Heyes and Galef 1996). The exceptional capacity for social learning through imitation is a human adaptation which appears to be an essential foundation for the development of culture (Donald 1991, 1993, 1995).And in fact ethnographies of childhood show that imitation is a core medium for acquiring culture in virtually every society that has been investigated. Moreover, these ethnographies consistently reveal that there is much less child-rearing than there is culture-seeking. Adults do little training but children learn a lot on their own initiative. In the first ethnography of socialization, F. C. Spencer (1899) emphasized imitation as the principal mechanism of socialization for Pueblo children: children imitate adult work and by age five or six begin to participate in subsistence activities and child care. Zuni children learn complex rituals though participation that initially is passive but becomes increasingly active and responsible. Mead’s (1975/1930) famous ethnography of growing up in Manus (an island near New Guinea) describes children learning to dance, to drum, to make war, and to shoot fish by imitating their elders; young children also play drawn-out word-repetition games with adults (pp. 36–45, 132, 153–154). As Mead (1975:120–129) points out, this imitation is highly selective, and Manus children show little interest—indeed disdain—for some of the activities most important to adults: religion, trading expeditions, ritual exchanges, and some aspects of kinship relations.On the other hand, Raum (1940: 145–146, 243–255) describes in detail the imitative social role playing of Chaga children from Mt. Kilimanjaro in Tanzania, which appears to cover the whole range of adult activities of which children are aware. In cultures as diverse as those in New Guinea, India, and France, children learn adult behavior by observing and imitating it (Whiting 1941:44–47;Wolfenstein 1955:113–114; Minturn and Hitchcock 1966:113, 124, 128).

Imitative play is a common way of practicing adult activities (e.g., Fortes 1970 [1938]:58–74;Hogbin 1970 [1946]:136–140;Read 1959:82–85; Nydegger and Nydegger 1966:142–145;Leis 1972:53–54).In Okinawa, children learn by observing adult activities and initiating participation in progressively more complex tasks; even a four-year old may become skillful with a sickle without his parents ever instructing him in any way (Maretzki and Maretzki 1966:144–148, 152–157). As Fortes describes the Tallensi, play

has a noteworthy role in their social development. In his play the child rehearses his interests, skills, and obligations, and makes experiments in social living without having to pay the penalty for mistakes. Hence there is always a phase of play in the evolution of any schema preceding its full emergence into practical life. Play, therefore, is often mimetic in content, and expresses the child’s identifications. But the Tale child’s play mimesis is never simply mechanical reproduction; it is always imaginative construction based on the themes of adult life and of the life of slightly older children. He or she adopts natural objects and other materials, often with great ingenuity, which never occur in the adult activities copied, and rearranges adult functions to fit the specifically logical and affective configurations of play.(Fortes 1970 [1938]:58–59)

Western ethnographers expect to find adults telling children about the culture: teaching how to do things, explaining the reasons for things, instructing the child about basic precepts. They are surprised to find that very little of this takes place; children learn most of their cultures on their own initiative, without pedagogy (see Atran & Sperber, 1991). Nor do children in most societies commonly ask for explanations; Western ethnographers often note the absence of “why” questions (e.g., Mead 1975:126). Almost every ethnography of children or socialization comments on the paucity of instruction and, conversely, on the fact that children take the primary initiative and responsibility for working out for themselves how to participate in their culture.For example, in Okinawa: 

There are no complex systems of training in skills. Adults rely heavily on observation and imitation on the part of children; they seldom “teach” them to do things systematically. Parents were surprised and amused when question such as “How do you teach children to transplant rice, harvest rice, or otherwise help in the fields?” were put to them. “We don’t teach them; why they just learn by themselves,” was the usual answer. 

Children learn by observing and experimenting. Whatever adults are doing, children are present to watch their activities and overhear their conversations.(Maretzki and Maretzki 1966:144;see Williams 1970:168 for a similar observation.)

In Oaxaca, Mexico, Romney and Romney (1966) describe similar patterns of socialization of children aged five and under: “Most of their simple tasks are more in the nature of imitative behavior of the older siblings and cousins. Helping by young children often takes the form of apparently spontaneous help within being asked or with any kind of formal or overt instruction” (Romney and Romney 1966:114). 

Mc Phee (1955) says of Balinese boys age 6 to 11: “Their early life is based upon imitation of their elders; their play is partly reproduction in miniature of various adult activities, carried out with great regard for detail” (p. 74). Balinese children are avid patrons of the performing arts and, like children in most traditional societies, they have access to the settings of most adult activities. Balinese children learn ritual and dance by observation and by having adults move them like puppets: “Verbal directions are meager; children learn from the feel of other people’s bodies and from watching, although this watching itself has a kinesthetic quality” (Mead 1955:43). They make simple puppets and masks to wear as they mimic adult performances.Mc Phee (1955:76–77) describes how one group of neighborhood boys, listening all their lives to gamelan (a type of orchestra formed around percussion instruments), learned to play on their own. Making their own two-person barong tiger costume, this group put on some remarkably polished performances. Then when Mc Phee procured a full set of instruments for the boys and hired a teacher, he was surprised to observe that from the very beginning the instructor, Nengah, simply modeled the parts the children were to learn:

The teacher here does not seem to teach, certainly not from our standpoint. He is merely the transmitter; he simply makes audible the musical idea to be passed on. The rest is up to the pupils.... No allowance was made here for youth; it never occurred to Nengah to use any method other than that which he uses when teaching an adult group. He explains nothing, since for him there is nothing to explain. If there are mistakes, he corrects them, and his patience is great.(Mc Phee 1955:89)

Occasionally adults may intentionally model correct behavior. Schieffelin (1991) describes daughters learning from their mothers, who never provide verbal instruction but occasionally do explicitly model tasks by segmenting their performances for their daughters to imitate. But intentional demonstration with pedagogical intent seems infrequent.Even in societies where such modeling occurs, it is infrequent and unimportant for cultural competence in most domains. As Bloch (1994:278) notes, citing some additional early sources, 

“In nonindustrialized societies most of what takes people’s time and energy—including such practices as how to wash both the body and clothes, how to cook, how to cultivate, etc.—are learned very gradually through imitation and tentative participation.... Knowledge transmission tends to occur in the context of everyday activities through observation and “hands-on” practice,There is a minimum of direct, verbal instruction.” 

Bloch argues further that much of cultural knowledge is not formulated in sentential or other linguistic form.Hence, he reminds us, it can only be learned by participant observation, in which the fieldworker’s learning is measured by her capacity to function in the community, especially in social relations. 

In many, perhaps most societies, people regard young children—especially under age 6 or 7—as pretty much incapable of knowing, understanding, or having common sense; as unable to exercise moral control over their own behavior; and as incapable of taking responsibility (e.g., Fortes 1970 [1938]:24–25;Read 1959:88;Maretzki and Maretzki 1966:114–115, 120;Nydegger and Nydegger 1966:146;Romney and Romney 1966:118–119;Lucy and Gaskins 1997).Consequently, adults and older children make little effort to train young children or explain anything to them. Rogoff, Newcombe, Fox, and Ellis (1980) found that in many cultures children aged five to seven are assigned and perform family chores such as tending animals or younger children, but they do these tasks with supervision. Their participation in the culture is guided by older mentors. However, by age eight to ten children can assume independent responsibility for many such tasks.

Learning theory, along with the importance of commendation and recognition of children’s accomplishments in the United States, might lead us to assume that, even without explicit instruction per se, parents and other caretakers are training children by rewarding through praise. But this assumption seems to be misguided, again mistakenly assuming that adults are directing the process of cultural transmission. Most observers of socialization in traditional societies have reported that correct performance of expected skills is almost never praised or rewarded.[3] Robert LeVine’s (1989) characterization of socialization matches my own observations among the Moose of Burkina Faso:

As Gay and Cole (1967) describe childhood among the Kpelle of Liberia, and as I observed it among the Gusii of Kenya and other peoples, children grow up without experiencing praise from their parents or others for behaving in a socially approved way or for learning a desirable skill. In contrast with the familiar [white middle class] American sequences of a child’s performing well, calling the performance to adult attention, and being praised by the adult, the African child learns through another sequences: observe the approved task (as performed by an older sibling), imitate it spontaneously, and receive corrective feedback only for inadequate performance. There is no expectation of recognition for good performance in learning or carrying out a task, yet tasksare learned and performed with skill. (LeVine 1989, p. 63;see also LeVine et al 1994:216)

This resembles this my own experience learning to cultivate millet fields along side Moose farmers, but it is important to add that “corrective feedback” typically involves nothing more than modeling the correct behavior, without explaining what is deficient in the performance of the child or novice (cf. Maretzki and Maretzki 1966:144–145). Indeed, Moose often laughed at my incompetence, saying, “You don’t know how to do that!” and simply took the hoe away from me. Similarly, Moose often prevent children from continuing with a task that they are performing incorrectly, without demonstrating how to do it right (see also Minturn and Hitchcock 1966:153 on similar practices among Rajputs in India).When a child does anything wrong, Moose mothers tell the child to stop or threaten punishment, just as Gusii mothers do (LeVine and LeVine 1977:148). They do not explain, discuss, or attempt to persuade by reason. In fact, in virtually all of the cultures in which child rearing has been described, commands and negative feedback supplement imitation. Parents and other adults and older children send children to fetch things or tell them to perform tasks, with no instruction, and then tell them if they are doing something wrong. For example, Moose and Gusii children learn to do everyday tasks by participating in adult activities, principally by being ordered to fetch and carry, by observing and by asking to be allowed to help (LeVine and LeVine 1977:163-165). When children fail to perform adequately, adults say “no,” tease, ridicule, punish, or threaten—sometimes with bogey men or supernatural punishments (see, e.g., Spencer 1899:80–81; Leighton and Kluckhohn 1947:51–52). Shaming is common, and in many traditional societies it is sufficient to indicate that the untoward behavior is “not what we ­­­­­­­­­­­­­______ do,” or else observe that it is the way some outgroup behaves.LeVine et al (1994) found that imperatives comprised over half of all caretakers’ utterances to children from 3 to 27 months old among the Gusii of Kenya; beginning at nine months, negative utterances were more common that positive ones. 

In short, even when they are reacting to a child’s failure, adults or older siblings virtually never explain the exact nature of the deficiency in a child’s performance, much less the reasons for the way it should be done. Caretakers’ awareness of a child’s incompetence does not lead to instruction. Children generally have to deduce for themselves precisely what they have done wrong if they are unable to imitate successfully.

As Mead (1975:120–124), Fortes (1970 [1938]:40 ff., 58–59) and Raum (1940: 255–259) point out, imitation is not simple replication. Children’s mimetic actions are never mere copies of what they have observed (or heard about). Imitation is mediated by implicit motoric representations (models, syntaxes) that are selective and creative. Using these representations of objects, activities, relationships and roles, children’s imitation is generatively constructive.But we actually know very little about how people imitate, and we should not take the process for granted; it is a very subtle, skilled capacity. Even the simplest mimicry of a molecular action is a complex perceptual-motor task. Moreover, most of what we call imitation involves generative productivity entailing subtle complementarities among actors and between actors and objects. It is no trivial or mechanically obvious task to observe another’s actions, discover the patterns or syntax, develop the practical competence that underlies them, and then generatively reproduce not the mechanical actions themselves, but meaningful patterns that correspond to them in a meaningful way.This needs to be carefully studied. 

Furthermore, imitation arises from, is embedded in, and constitutively creates social relationships. Imitation often is focused on high status persons, but it is also a means of borrowing across cultures (Taussig 1993).Imitation of adults and older children seems to result from identification, and perhaps it cements that identification as well (see Fortes 1970:56–57). Mead (1975:135–150, 154–157) observed that Manus children identified with and adopted the social personalities of their natal or adoptive parents; and of course they develop gender identities. Most ethnographies of childhood clearly show the desire of children to emulate and then take on adult activities and roles. Consequently, imitation gradually develops into real assistance with sharing of tasks and responsibilities. Children generally want to help; they want to participate; they want to do what their older kin and neighbors do (see, e.g., Fortes 1970:37–39).Gradually they move from mimetic play to peripheral apprentice-like participation to full task performance and responsibility (Lave and Wenger 1991;Rogoff, Mistry, Göncü, and Mosier 1993;Rogoff, Baker-Sennett, Lacasa, and Goldsmith 1995).

Imitation is not the only means available to children for learning their cultures. Complementing and supplementing imitation, there are numerous other media in which children acquire the capacity to construct their cultures. Language always has some role, and in many cultures proverbs, folk tales, myths, and gossip are instruments of socialization, especially with regard to morality. At initiation or marriage, adults in some cultures may admonish or give general instructions. Of course, explanations and abstract analyses may be available in some schools. However, in the African primary and secondary schools that I have observed in Malawi, Zaïre, and Burkina Faso, as well as those that I have read about elsewhere in the third world, a common method of instruction is to repeat the teacher’s words in unison or copy the lesson off the blackboard, word for word (see, e. g., Nash 1970:307). In the Koranic schools I have observed in Burkina Faso, boys learn the Koran by rote, in Arabic, sometimes memorizing major segments of the text without any exegesis or discussion of its meaning—and, it appears, often without much understanding of the Arabic language. So while certain kinds of schooling may entail a dramatic shift from imitation toward explicit conceptual transmission of declarative knowledge and certain formal skills, the shift may be limited within schools, and may not transform the mimetic transmission of more fundamental cultural practices outside of school. 

Linguistic competence itself is not a result of simple imitation alone. However, modern techniques of language instruction involve rote learning of phrases in context, without translation and consequently without the learner initially being able to conceptualize the meaning. Adults virtually never define words for children, and small children often mimetically reproduce sentences without much understanding of them. Such practices show that complex utterances can be mimetically acquired with little or no conceptual articulation. 

The Concept of Culture and the Objective in Ethnography

If I could tell you what it meant I wouldn’t have had to dance it.(Isadora Duncan, quoted in lectures by Gregory Bateson, cited by Levy 1996) 

In the normal business of life it is useless and even mischievous for the individual to carry the conscious analysis of his cultural patterns around with him. That should be left to the student whose business it is to understand these patterns.(Sapir 1949/1927:558)

If people learn their cultures in large part by observation, imitation, and incremental participation, ethnographers should do likewise. In these terms, then, “participant observation” should aim at the learning of practices in the same manner that members of the culture acquire them. The goal is for the fieldworker to operate in the same medium as informants, reproducing the manner in which they normally learn, remember, reproduce, retransmit, contest, and transform the relevant practices. That is, the best ethnography aims at acquiring practical competence the same ways informants do.This goal may be difficult—sometimes impossible—to attain, but it should be the standard for judging fieldwork and ethnographic data. 

In fact, however, anthropologists often rely on interviewing and related verbal methods; psychologists use rating scales and sociologists use questionnaires. Social scientists privilege articulate verbal concepts and propositions in part because every scholar has been instructed in classrooms and studied in libraries for 20 years through the medium of conceptual language and then writes, teaches, and conducts conferences primarily in a verbal medium. Because social science is conducted in this medium, it seems natural to use verbal methods in research. Furthermore, informants’ linguistic behavior is easy to record, translate, digest, and convey in articles, books, and lectures. But the relative ease of recording and communicating the concepts and propositions of linguistic discourse should not lead us to suppose that the motives and causes of human action correspond to what any informant says, or could possibly report. Verbal methods are not a valid substitute for participant observation because we cannot expect informants to explain or describe the many aspects of their cultures that no one has ever explained or described to them.If we require informants/respondents/subjects to tell us about their cultures, we get responses that are valid representations of only a small portion of the culture. With respect to the nonverbal domains and aspects of the culture, informants’ responses must inevitably contain a great deal of confabulation generated on the spot for the investigator. Asking people to communicate information or attitudes in a mode other than the mode in which they themselves acquired, think about, and communicate them produces invalid or distorted responses. People cannot produce an accurate, valid, meaningful verbal account of how they dance, conduct a ritual, or make attributions about others’ motives—because they do not learn, remember, transform, or reproduce these skills in a verbal medium. Naive attempts to transform the knowledge/competence of one cognitive-semiotic system into another produces a misrepresentation that is liable to be very incomplete, distorted, or simply false. For the performer, most motoric and social skills are inexplicable. To understand the human mind, action, and social processes, researchers have to access the appropriate channels, in the medium in which the relevant competence normally operates. 

If we do participant observation in order to learn from informants in the manner that our informants learned, then we face the difficult problem of recording, analyzing, and conveying what we have learned implicitly. As researchers, we have to begin by observing and attempting to imitate the practices of informants, and then meticulously transforming our practical knowledge into abstract conceptual understanding. Learning cultures implicitly by observation and participation does not mean that we have to represent what we learn in a corresponding medium, in mime, experiential narrative, or film.If is perfectly appropriate for researchers to analyze and express their non-conceptual practical competencies in abstract language (as I am attempting to do here).But the goal then is not “translation” (which only applies across linguistic texts) but transformation of procedural knowledge into conceptual frameworks that can be articulated formally and analyzed abstractly. This extremely difficult task is properly ours, as researchers; we can ask our informants for some help and corroboration of our interpretations, but it is a mistake to expect them to simply verbalize non-verbal competence.

Of course, the necessity for participant observation does not in any way imply that adequate fieldwork can be done without a command of the local language, or without ever asking informants for accounts of their behavior. Observing Moose rituals I typically found that the rituals were segmented, separated by non-ritual interludes. Each segment of ritual often had a name, and people would often say things to each other (or to me) such as, “Hey, let’s get going—it’s time to feed the dead man.”There were always verbal labels for categories of participants and ritual objects: “Come on, all you sisters’ sons; come drink your funeral beer!” Occasionally they would say with opprobrium, “That’s wrong—you messed it up!”Or they would say, “The Nakomse offer up sheep; we offer up goats.”This kind of verbal labeling, critique, and commentary on practices are part of the normal process of organizing activities, and provide the ethnographer with an invaluable orientation to salient entities and issues that are otherwise inarticulate. 

Language is a crucial channel for acquiring culture. But language is not the only channel; in many domains it is not the primary channel; it is never a sufficient channel; and linguistic information is not equivalent to or interchangeable with information conveyed through other channels. 

In principle, there is no theoretical reason why we cannot eventually devise more or less adequate non-naturalistic methods for investigating many specific aspects of implicit cultural knowledge. But the criterion for the validity of such methods must be that they corroborate and match (or illuminate) what is learned by imitative participant observation or other methods based on learning cultures as informants naturally acquire them. Of course, ethnographic fieldwork should not rely on participant observation alone, unsupported by other means of observing and collecting data. Additional methods should be used to supplement and provide convergent evidence to compare with the results from participant observation. While in certain cases it may be pragmatically expedient to use alternative methods for certain limited purposes as makeshift substitutes for participant observation, long-term participant observation is the key criterion against which other methods should be evaluated.

“Culture” can be defined as whatever people acquire, do, use, produce, feel or think by virtue of participation in a group or network of communicative social interactions, and as the means for participation in such a group or network.Thus culture is comprised of those processes which are simultaneously the presuppositions for social interaction, the mechanisms people use to interact, and the social consequences of these communicative relationships. This concept of culture implies that participant observation is the fundamental method for studying culture: If we want to understand that which is the prerequisite, means, and consequence of social interaction, then we ourselves should engage in the relevant social interactions. People learn their cultures by participating in the social relationships that the culture makes necessary or possible. At the same time, people reproduce, transform, and invent their cultures though these culturally-mediated communicative interactions. Hence the most direct, natural, valid means for learning a culture is for the researcher to participate as fully as possible in the widest possible range of culturally constituted (and cultural constituting) relationships. 

If ethnographers learn a culture in approximately the same manner as informants they may be able to acquire approximately the same implicit representations of that culture. Indeed, if they become able to participate fully in the culture, they have roughly the competence that characterizes informants. Participants in a culture must acquire mediating devices (implicit models, scripts, syntaxes, or whatever) that enable them to construct, interpret, coordinate, evaluate, contest, and sanction meaningful, affective, motivated interactions. That capacity to participate fully is the proper social test of the objectivity of knowledge about a culture. If implicit knowledge permits fully meaningful coordination of social interaction, appropriately motivated and evaluated, it is objective.Thus the true test of objectivity is the complementarity or ‘fit’ of the ethnographer’s actions, affects, motives, and ideas vis-à-vis diverse informants’ actions, affects, motives, and ideas.This participatory competence must be assessed with respect to a wide range of domains and aspects of the culture, and the adequacy of the participation must be assessed with respect to many very different social relational criteria.But the implication is that, like children, immigrants, or spouses marrying into a community, ethnographers can acquire a great deal of objective cultural competence, in some cases more or less approximating that of the native. Furthermore, different ethnographers can acquire comparable cultural competence if they learn a culture in the same manner as informants, though the same channels. This competence will inevitably be limited by the constraints of participation that begins in adulthood, when many of the evolved mechanisms for acquiring a culture may no longer be fully operable. Furthermore, adults trying to function in a second culture have the difficult task of unlearning or ignoring proclivities acquired in their first culture. (Because of such factors, most adult language learners cannot duplicate the linguistic competence of people who were immersed in a language before age eleven; Johnson & Newport, 1989;Mayberry, 1993;Newport, 1990, 1991.)But even the deficiencies in the ethnographers’ social relationships can be used with great effect to understand just what native proficiency is. 

Multiple Forms of Competence: Evidence from Memory Systems

The effects of past events on current experience and performance can be expressed not only via explicit remembering, but also by subtle changes in our ability to identify, act on, and make judgments about words, objects, and other stimuli—changes that are frequently independent of the ability to engage in conscious recollection of a prior experience (Schacter 1995:821). 

To this point we have only distinguished between explicit, conceptual, verbal knowledge and whatever is implicit, motoric, or practical. But this is a crude dichotomy. Are there only two kinds of cultural competence, the one transmitted via conceptual language, the other by observation, imitation, and participation in bodily practices? Can we go beyond this dichotomy between language/abstract/explicit and body/experience/implicit?[4]What kinds of cultural knowledge, skills, and generative potentials are there? How do people transmit, apprehend, and transform these kinds of culturally patterned competence? 

Chomsky (1980, 1988), Fodor (1983), and others have framed a general case for the specificity of distinct modes of perception, learning, knowledge, and competence.Chomsky’s modularity argument focuses on the specificity of the prior constraints and structured potentials necessary for learning in different domains.Fodor’s modularity argument emphasizes the relative independence of different perceptual systems, each with their corresponding, distinct forms of mental representation; modules are informationally encapsulated, meaning that information in one is not fully or reliably accessible to others.Sperber (1994) shows how the modularity of thought underlies the epidemiology of cultural representations (see also the many other relevant chapters in Hirschfeld and Gelman 1994).Keil (1981, 1990), Gelman (1990), and Brown (1990) point out that learning requires domain-specific constraints that provide an initial set of structured assumptions and hypotheses and that focus attention on relevant features. For example, people seem to have distinct ways of learning about and understanding animate beings, material artifacts, numbers, and music. Some evidence suggests that these respective kinds of knowledge may be based on anatomically distinct brain structures (Caramazza, Hillis, Leek, & Miozzo 1994).

Research under the rubrics of learning and memory represents some of the most sophisticated, empirically grounded analysis of the diverse and distinct ways in which experience affects behavior, capacities, and knowledge. While little of this research has focused on the everyday acquisition of culture, it is nonetheless extremely informative regarding the media in which people learn and the modalities of knowing. A wide range of research with humans and other animals has invalidated earlier theories of a unitary mechanism for learning: there are many distinct processes of acquiring or developing competence. As Gallistel (1995) puts it, there is no more reason to expect any organism to have one generalized learning mechanism than there is to expect an organism to have a single general-purpose sense organ.It is now clear that humans and other organisms have many specialized adaptations for solving evolutionarily important problems (Barkow, Tooby, and Cosmides 1992;Tooby & Cosmides 1992;Gallistel 1990, 1995). These diverse capacities are content-dependent and involve mechanisms that are quite specific to their proper domains. For example, a person may have great difficulty adding and subtracting negative numbers, yet find it easy and obvious to keep track of a formally homologous social domain such as taking turns driving in a car pool or figuring out who owes dinner invitations to whom.

Researchers studying human memory systems have made considerable progress in characterizing and differentiating among several distinct types of memory. One kind of evidence for the distinctness of these memory systems comes from studies showing that species differ in the kinds of memory they have. Other studies demonstrate that patients with lesions that cause loss of one memory system may retain others intact; for example, certain patients with damage to the temporopolar cortex may be completely unable to recall any events they have experienced in their own life, while demonstrating normal semantic knowledge of the meaning of ideas and things (Markowitsch 1995). Other evidence comes from imaging studies showing differentially localized brain activity related to tasks requiring different kinds of retrieval. Some of the most important evidence comes from studies of retrieval and encoding: research showing that measures of memory in one system are uncorrelated with measures of memory in other systems, and research demonstrating that different types of memory are acquired in different ways (Schacter 1995). These kinds of experimental and clinical research on human memory systems have extensively corroborated the theory that conscious explicit recall (declarative knowledge) is quite distinct from learning that takes place without awareness of the events, process, or result of the learning—but that nonetheless affects responses in many ways (Schacter 1995).

Recent research has shown that in fact people have at least five or six distinct systems for encoding, storing, and retrieving information (Schacter and Tulving 1994; Umilta and Moscovitch 1994). Tulving (1985) contrasts episodic, semantic, and procedural memory, comparing this scheme with many other related taxonomies of memory systems. Procedural memory is knowledge of how to do something, based on actual practice. Semantic memory is knowledge of facts or meaning (e.g., the fact that my great-grandmother is dead), while episodic memory is memory of events that the person has experienced (I remember my mother talking about her grandmother). Episodic memory is what is usually accessed by verbal methods that require descriptions of events and autobiographical reports, and Tulving states that episodic memory is the basis of self-awareness, personal identity, and the sense of a personal history. Tulving argues that episodic memory depends on semantic memory which in turn depends on procedural memory; hence semantic knowledge implies procedural and episodic knowledge implies both semantic and procedural. Conversely, procedural knowledge can function independently of the other two, and semantic memory can be present without episodic memory. Similarly, Connerton (1989:22–23), a social historian, distinguishes among personal memory of one’s life history, cognitive memory of the meanings of cultural entities, and the habit-like capacity reproduce a certain performance.In accord with Tulving, Connerton observes that the second two types of memory often exist without personal memory of the events or experiences in which the learning occurred. 

In contrast to semantic and episodic memory, procedural memory is acquired and expressed primarily though overt action; hence we might better call it procedural competence. Procedural memory involves a number of different mechanisms, each capable of operating more or less independently. Tulving (1995) distinguishes among four subsystems of procedural knowledge: simple associative learning, simple conditioning, motor skills, and certain cognitive skills. Semantic memory includes two subtypes: spatial and relational subsystems. 

Alongside procedural, semantic, and episodic memory, Tulving (1995) later added two other systems. The fourth basic type is the perceptual representation evident in sensory priming that entails the capacity to quickly recognize familiar entities based on prior exposure to their perceptual forms (auditory, visual, tactile, etc.). Other cognitive research has revealed that the human capacities to recognize faces, objects, and words are dissociable: patients may lose one capacity without impairment of the others; see Young 1988. Subsequently Tulving has added to his taxonomy a widely recognized contrast among these long-term systems and the kind of short-term memory used in executing a task or solving a problem (Baddeley 1995). In Tulving’s 1995 taxonomy, primary (working or short-term) memory includes visual and auditory subsystems. Amnesiac patients with no long term memory retain short-term working memory, and the distinction is clear-cut in other animals as well (Squire and Knowlton 1995). Thus Tulving’s major systems are episodic, semantic, procedural, perceptual priming, and working memory, each with their various subsystems. Tulving (1985, 1991, 1995) describes the evidence that memory in these distinct systems is serially encoded, stored in parallel, and retrieved independently from each system.

[Figure about here.]

Squire and Knowlton (1995) make very similar distinctions, dividing declarative (explicit) memory into memory for facts and memory for events. They divide nondeclarative (implicit) memory into skills and habits, priming, nonassociative learning, and simple classical conditioning.(See Figure; they further subdivide classical conditioning into emotional responses and skeletal musculature responses.).Each system can be more or less localized in specified brain systems. Markowitsch (1995) discusses the evidence that episodic and semantic memory structures are located primarily in the cerebral cortex, while priming and procedural memory are located primarily in the telencephalic nuclei (various subsystems linked with distinct regions of the cerebellum).

This research has direct and important implications regarding how people acquire culture and the methodologies we should use to should study cultural knowledge and capacities. The distinction between semantic knowledge and episodic memory tells us that people may be fully able to report personal life events they have experienced without having a corresponding conceptual representation of the meaning of these experiences in the abstract.Conversely, people may be able to convey ideas without being able to describe the circumstances in which or the point in time when they learned or formulated the ideas.These two kinds of explicitly articulated knowledge are in no way equivalent or interchangeable. Furthermore, each of them may operate independently of cognitive skills such as the ability to make complex social inferences and attributions. A person may have elaborately articulated ideas about how a social system operates, without having the (Implicit/procedural) cognitive skills to function effectively in it.Or—more likely—the reverse.Additionally, a person may remember many important biographical events and observed interactions without having acquired the social skills or made the social inferences that might seem relevant to this collection of experiences. And it is entirely possible that any person’s autobiographic memory, their ideas and their cognitive skills may be mutually inconsistent in various respects. 

Likewise, it is obvious that a person may be able to make an elaborate work of art or conduct a complex ritual without having a semantic representation of its meaning in conceptual terms—or may have ideas about meaning without the procedural capacity to construct the relevant entities. Another distinction, the contrast between associative learning and perceptual priming, indicates that people may be skilled at recognizing some entities without knowing when or where they are likely to occur. On the other hand, people may have been conditioned so that they are anxious when they encounter some entity that has previously been followed by frightening or painful experiences. Yet they may be unable to use this knowledge in a rational or calculative way to solve the problems they face, unable to recall the aversive experiences, and unable to articulate the (semantic) meaning of the entity.The autonomy of working memory indicates that people may be able to represent aspects of a situation in order to deal with it when it occurs, yet have no enduring representation of it after the fact. Furthermore, their long-term semantic knowledge regarding aspects of an experience may not correspond to the way that they represented the immediate experience when dealing with it at the time.In addition, processes of non-associative learning such as habituation to constantly repeated stimuli mean that people can take for granted and ignore their most regular, invariant experiences: this background of perceptually taken-for-granted and hence ignored features may not be directly represented in episodic, semantic, procedural, or working memory. 

The dissociability among these distinct ways of learning/knowing means that the fieldworker who learns only in one mode is failing to discover what participants in the culture learn in each of the other modalities. Cultural competence is compartmentalized, such that knowledge or capacities are often limited to one modality. What one system does is not fully or directly available to other systems, and different aspects of any experience affect different systems. People perceive, encode, retrieve, reformulate, and enact different kinds ofpractices in different modalities. The processes of storage and retrieval differ among these systems, so that what goes into them and what can come out depends on the specific system. What people are capable of acquiring, transforming, and retransmitting in one modality they may be quite unable to acquire, transform, or retransmit in another modality. Furthermore, capacities and proclivities in one modality may not be consistent or harmonious with those in other modalities. Diverse kinds of knowledge may be logically contradictory, even if they are pragmatically compatible:dissociable ways of knowing permit paradoxical knowledge. However, we may presume that the capacity to participate fully in culturally organized social relations typically depends on the complementarity and sometimes coordination of distinct modalities. 

We have to adapt our methods to this diversity among the modes in which people acquire and construct the various aspects of their cultures. This means familiarizing ourselves with people’s perceptual worlds, experiencing events and constructing conceptual meanings, listening and learning to converse, developing inarticulate cognitive and motor skills, learning associations and conditioned emotional and behavioral responses, while becoming inured to familiar sensations until we take them for granted. We should learn other people’s cultures in each of the ways that they themselves learn them—which are the ways we learn our own.


Baddeley, Alan

1995Working Memory. In Michael S. Gazzaniga (Ed.), The Cognitive Neurosciences (pp. 755–764). Cambridge, MA: MIT Press.

Baldwin, James Mark
1895Mental Development in the Child and the Race; Methods and Processes.New York and London: Macmillan.
1897Social and Ethical Interpretations in Mental Development: A Study in Social Psychology. New York and London: Macmillan.
1911The Individual and Society; or Psychology and Sociology.Boston: Gorham Press. 

Bandura, Albert, and Richard H. Walters
1963.Social Learning and Personality Development.New York: Holt, Reinhart and Winston. 

Bargh, John A.
1990Auto-Motives: Preconscious Determinants of Social Interaction. In E. Tory Higgins, and Richard M. Sorrentino (Eds.)Handbook of Motivation and Cognition: Foundations of Social Behavior, Vol. 2. (pp. 93-130). New York: Guilford Press.

Bargh, John A.
1992 The Ecology of Automaticity: Toward Establishing the Conditions Needed to Produce Automatic Processing Effects.American-Journal-of-Psychology (Special Issue: Views and Varieties of Automaticity) 105(2): 181-199
1994The Four Horsemen of Automaticity: Awareness, Intention, Efficiency, and Control in social Cognition. In Robert S. Wyer Jr. and Thomas K. Srull (Eds.) Handbook of Social Cognition, Vol. 1: Basic Processes(2nd ed.) (pp. 1-40). Hillsdale, NJ:Lawrence Erlbaum Associates.
1997Automaticity. Chapter in press in Daniel Gilbert, Susan Fiske and Gardner Lindzey (Eds.)Handbook of Social Psychology, 4th Ed.New York: McGraw Hill.

Barkow, Jerome H., Leda Cosmides, and John Tooby
1992The Adapted Mind: Evolutionary Psychology and the Generation of Culture. New York: Oxford University Press. 

Bloch, Maurice
1994Language, Anthropology, and Cognitive Science. In Robert Borofsky (Ed.) Assessing Cultural Anthropology (pp. 276–283). New York: McGraw Hill. 

Borofsky, Robert
1993On the Knowledge and Knowing of Cultural Activities. In Robert Borofsky (Ed.) Assessing Cultural Anthropology (pp. 331–348). New York: McGraw Hill. 

Bourdieu, Pierre
1990/1980The Logic of Practice.Stanford, CA: Stanford University Press. 

Brown, A. L.
1990Domain-specific principles affect learning and transfer in children.Cognitive Science, 14, 107-133.

Caramazza, Alfonso, Argye Hillis, Elwyn C. Leek, and Michele Miozzo
1994The organization of lexical knowledge in the brain:Evidence from category- and modality-specific deficits.In Mapping the Mind: Domain Specificity in Cognition and Culture.Lawrence Hirschfeld and Susan Gelman eds. Pp. 68-85. New York:Cambridge University Press.

Chomsky, Noam
1980Rules and representations.New York: Columbia University Press.
1988Language and Problems of Knowledge.Cambridge: MIT Press .

Connerton, Paul
1989How Societies Remember.Cambridge: Cambridge University Press. 

Csordas, Thomas J. (Ed.)
1994Embodiment and Experience: The Existential Ground of Culture and Self. Cambridge and New York: Cambridge University Press.

Desjarlis, Robert R.
1992Body and Emotion: The Aesthetics of Illness and Healing in the Nepal Himalayas.Philadelphia: University of Pennsylvania Press. 

Devisch, Rene
1993Weaving the Threads of Life: The Khita Gyn-Eco-Logical Healing Cult among the Yaka.Chicago: University of Chicago Press.

Donald, Merlin
1991Origins of the Modern Mind: Three Stages in the Evolution of Culture and Cognition.Cambridge: Harvard University Press.
1993Précis of Origins of the Modern Mind: Three Stages in the Evolution of Culture and Cognition (with commentaries and reply).Behavioral and Brain Sciences 16:737–791.
1995The Neurobiology of Human Consciousness: An Evolutionary Approach.Neuropsychologia 33:1087–1102. 

Eckerman, Carol O., J. L. Whately and L. J. McGhee
1979Approaching and Contacting the Object Another Manipulates:A Social Skill of the 1-Year-Old.Developmental Psychology 15(6): 585-593.

Eckerman, Carol O., Claudia C. Davis and Sharon M. Didow
1989Toddlers' Emerging Ways of Achieving Social Coordinations with a Peer. Child Development 60(2):440-453.

Eckerman, Carol O. and Mark R. Stein
1990How Imitation Begets Imitation and Toddlers' Generation of Games.Developmental Psychology 26(3):370-378.

Eckerman, Carol O. and Sharon M. Didow
1996Nonverbal Imitation and Toddlers' Mastery of Verbal Means of Achieving Coordinated Action. Developmental Psychology 32:141-152.

Elias, Norbert
1993The Civilizing Process. Translated by Edmund Jephcott. Oxford: Blackwell. 

Ericsson, Karl Anders, and Herbert Alexander Simon
1993Protocol Analysis: Verbal Reports as Data (rev. ed.). Cambridge, MA: MIT Press.

Feld, Steven
1982Sound and Sentiment: Birds, weeping, Poetics, and Song in Kaluli Expression.Philadelphia: University of Pennsylvania Press. 

Fiske, Alan Page
1991Structures of Social Life: The Four Elementary Forms of Human Relations.New York: Free Press (Macmillan).

Fiske, Alan Page, and Nick Haslam
1996Social Cognition Is Thinking About Relationships. Current Directions in Psychological Science: 5: 13­1–148. 

Fiske, Donald W.
1978Strategies for Personality Research.San Francisco: Jossey-Bass.
1980When Are Verbal Reports Veridical?In Richard A. Shweder (Ed.) Fallible Judgment in Behavioral Research (pp. 59–66).New Directions for Methodology of Social and Behavioral Science, 4.San Francisco: Josey Bass.
1981Editor’s Notes.In D. W. Fiske (Ed.) Problems with Language Imprecision (pp. 1–2). San Francisco: Jossey-Bass.
1986Specificity of Method and Knowledge in Social Science. In Donald W. Fiske and Richard A. Shweder (Eds.) Metatheory in Social Science: Pluralisms and Subjectivities. (pp. 61–82).Chicago: University of Chicago Press. 

Fiske, Susan T.
1995Words! Words! Words! Confronting the Problem in Observations.In Patrick E. Shrout and Susan T. Fiske (Eds.) Personality Research, Methods and Theory: Festschrift for Donald Fiske (pp. 221–240). Hillsdale, NJ: Lawrence Erlbaum Associates. 

Fodor, Jerry
1983Modularity of mind.Cambridge, MA: MIT Press.

Fortes, Meyer
1938Social and Psychological Aspects of Education in Taleland.London: Published by the Oxford University Press for the International Institute of African Languages and Cultures. Reprinted in John Middleton (Ed.) From Child to Adult: Studies in the Anthropology of Education (pp. 14–74). Garden City, NY: Natural History Press for the American Museum of Natural History.

Freud, Anna
1973The Ego and the Mechanisms of Defense. Rev. ed.New York:International Universities Press.

Gallistel, Charles R.
1990The Organization of Learning. Cambridge: MIT Press.
1995The Replacement of General-Purpose Theories with Adaptive Specializations. In Michael S. Gazzaniga (Ed.), The Cognitive Neurosciences (pp. 1255–1267). Cambridge, MA: MIT Press.

Gay, J., and Cole, M.
1967The New Mathematics in an Old Culture.New York: Holt, Rinehart and Winston. 

Gelman, Rochel
1990First principles organize attention to and learning about relevant data:Number and the animate-inanimate distinction as examples. Cognitive Science, 14, 123-124.

Heyes, Celia M., and Bennett G. Galef, Jr.
1996Social Learning in Animals: The Roots of Culture.San Diego, CA: Academic Press. 

Highwater, Jamake
1985Dance: Rituals of Experience.New and revised edition.New York: Alfred van der Marck. 

Hirschfeld, Lawrence, and Susan Gelman eds.
1994Mapping the Mind:Domain Specificity in Cognition and Culture. New York, NY:Cambridge University Press.

Hogbin, H. Ian
1970 (1946)A New Guinea Childhood: From Weaning till the Eight Year. In John Middleton (Ed.) From Child to Adult: Studies in the Anthropology of Education (pp. 134–162). Garden City, NY: Natural History Press for the American Museum of Natural History.

Jackson, Michael
1995.At Home in the World.Durham, NC: Duke University Press.
1996Introduction: Phenomenology, Radical Empiricism, and Anthropological Critique.In Michael Jackson (Ed.) Things as They Are; New Directions in Phenomenological Anthropology (pp. 1–50).Bloomington: Indiana University Press. 

James, E. O.
1917Primitive Ritual and Belief: An Anthropological Essay.London: Methuen.

Johnson, Allen
1997.Repression.In press, Ethos. 

Johnson, Jacqueline S., and Elissa Newport
1989Critical Period Effects in Second Language Learning:The Influence of Maturational State on the Acquisition of English as a Second Language.Cognitive Psychology, 21, 60-99.

Kardiner, Abram
1939The Individual and His Society: The Psychodynamics of Primitive Social Organization. (With Ralph Linton .)New York: Columbia University Press. 

Keil, Frank
1981Constraints on knowledge and cognitive development. Psychological Review, 88(3), 197-227.
1990Constraints on constraints:Surveying the epigenetic landscape. Cognitive Science, 14, 135-168.

Kluckhohn, Clyde
1943Covert Culture and Administrative Problems.American Anthropologist 43:413–419.Reprinted in Herbert Applebaum (Ed.), Perspectives in Cultural Anthropology (pp. 173–180).Albany: State University of New York Press (1987). 

Kurtz, Stanley N.
1992All the Mothers Are One: Hindu India and the Cultural Reshaping of Psychoanalysis.New York: Columbia University Press. 

Lave, Jean, and Wenger, Etienne
1991Situated Learning: Legitimate Peripheral Participation.Cambridge: Cambridge University Press. 

Leighton, Dorothea, and Clyde Kluckhohn
1947Children of the People: The Navaho Individual and His Development.Cambridge: Harvard University Press. 

Leis, Philip E.
1972Enculturation and Socialization in an Ijaw Village.Case Studies in Education and Culture. New York: Holt, Rinehart and Winston. 

LeVine, Robert A.
1989Cultural Environments in Child Development.In William Damon (Ed.), Child Development Today and Tomorrow. San Francisco and London: Josey Bass. 

LeVine, Robert A. and Barbara B. LeVine
1977Nyansongo: A Gusii Community in Kenya.Six Cultures Series. Huntington, NY: Robert E. Krieger. 

LeVine, Robert A., Sarah LeVine, P. Herbert Aldermen, T. Berry Brazleton, Suzanne Dixon, Amy Richman, and Constance H. Keefer
1994Child Care and Culture: Lessons from Africa. Cambridge and New York: Cambridge University Press. 

Lévi-Strauss, Claude
1953Social Structure.In A. L. Kroeber (Ed.), Anthropology Today: An Encyclopedic Inventory (pp. 524–553).Chicago: University of Chicago Press. 

Levy, Robert I.
1996Bateson Reflected.Paper presented at the American Anthropological Association Meetings, San Francisco, 1996. 

Lucy, John, and Suzanne Gaskins
1997Grammatical Categories and the Development of Classification Preferences: A Comparative Approach.Forthcoming in S. Levinson & Melissa Bowerman (Eds.), Language Acquisition and Conceptual Development. Cambridge & New York: Cambridge University Press.

Malinowski, Bronislaw
1961/1922Argonauts of the Western Pacific: An Account of Native Enterprise and Adventure in the Archipelagoes of Melanesian New Guinea.New York: Dutton.
1954/1926Myth in Primitive Psychology.In Bronislaw Malinowski,Magic, Science and Religion and Other Essays (pp. 93–148).Garden City, NY: Doubleday.
1927Sex and Repression in Savage Society.Chicago: University of Chicago Press. 

Marett, R. R.
1929/1909The Threshold of Religion.London: Methuen and Company. 

Maretzki, Thomas, and Hatsumi Maretzki
1966Taira: An Okinawan Village.Six Cultures Series. New York: John Wiley and Sons. 

Markowitsch, Hans J.
1995Anatomical Basis of Memory Disorders.In Michael S. Gazzaniga (Ed.), The Cognitive Neurosciences (pp. 765–780). Cambridge, MA: MIT Press.

Mauss, Marcel
1973/1936.Les Techniques du Corps. In Marcel Mauss, Sociologie et Anthropologie (pp. 363–386).Paris: Presses Universitaires de France.

Mayberry, Rachel I.
1993First-language Acquisition After Childhood Differs from Second-Language Acquisition:The case of American Sign Language.Journal of Speech and Hearing Research, 36, 1258-1270.

McDougal, William
1908An Introduction to Social Psychology.London: Methuen. 

Mc Phee, Colin
1955.Children and Music in Bali. In Margaret Mead and Martha Wolfenstein (Eds.) Childhood in Contemporary Cultures (pp. 70–98).Chicago: University of Chicago Press.

Mead, Margaret
1975/1930Growing Up in New Guinea: A Comparative Study of Primitive Education.New York: William Morrow.
1955Children and Ritual in Bali. In Margaret Mead and Martha Wolfenstein (Eds.) Childhood in Contemporary Cultures (pp. 40–51).Chicago: University of Chicago Press. 

Merleau-Ponty, Maurice
1962/1945Phenomenology of Perception.Charles Smith, trans.new York: Humanities Press. 

Miller, Neal E., and John Dollard
1941Social Learning and Imitation.New Haven: Yale University Press for the Institute of Human Relations. 

Miller, Peggy J., and Hoogstra, Lisa
1992Language as tool in the socialization and apprehension of cultural meanings. In Schwartz, T., White, G. M. and Lutz, C. A. (Eds.) (1992).New directions in psychological anthropology. Cambridge and New York: Cambridge University Press (pp. 83–101). 

Minturn, Leigh, and John T. Hitchcock
1966The Râ­­jpûts of Khalapur, India.Six Cultures Series. New York: John Wiley and Sons.

Nagel, Katherine, Raquel S. Olguin, and Michael Tomasello
1993Processes of Social Learning in the Tool Use of Chimpanzees (Pan troglodytes) and Human Children (Homo sapiens).Journal of Comparative Psychology 107:174–186. 

Nash, Manning
1970Education in a New Nation: The Village School in Upper Burma.In John Middleton (Ed.) From Child to Adult: Studies in the Anthropology of Education (pp. 301–313). Garden City, NY: Natural History Press for the American Museum of Natural History. 

Newport, Elissa L.
1990Maturational Constraints on Language Learning.Cognitive Science 14, 11-28.
1991Contrasting Concepts of the Critical Period for Language.In Susan Carey and Rochel Gelman (Eds.), The Epigenesis of Mind: Essays on Biology and Cognition.Hillsdale, NJ:Erlbaum.

Nicholaisen, Ida
1988Concepts and Learning among the Punan Bah of Sarawak.In Gustav Jahoda and I. M. Lewis (Eds.) Acquiring Culture: Cross Cultural Studies in Child Development (pp. 193–222). 

Nisbett, R. E., and Wilson, T. D.
1977Telling more than we can know:Verbal reports on mental processes.Psychological Review 84: 231-259.

Nydegger, William F., and Corrine Nydegger
1966.Tarong: An Ilcos Barrio in the Philippines. Six Cultures Series. New York: John Wiley and Sons.

Obeyesekere, Gananath
1990.The Work of Culture.Chicago and London: University of Chicago Press. 

Ochs, Eleanor
1988 Culture and language development: Language acquisition and language socialization in a Samoan village. Cambridge and New York: Cambridge University Press

Ochs, Eleanor, and Bambi Schieffelin
1984Language Acquisition and Socialization: Three Developmental Stories and their Implications. In Shweder, R. A., and LeVine, R. A. (Eds.) (1984). Culture theory: Essays on mind, self, and emotion. Cambridge and New York: Cambridge University Press (pp. 276–322).

Parsons, Anne
1969Is the Oedipus Complex Universal?The Jones-Malinowski Debate Revisited. Belief, Magic, and Anomie: Essays in Psychological Anthropology.New York: Free Press.London: Collier?Macmillan.

Rabain, Jacqueline
1979L’Enfant du Lineage: Du Sevrage à la Class d’Âge chez les Wolof du Sénegal.Paris: Payot.

Raum, Otto Friedrich
1940Chaga Childhood : A Description of Indigenous Education in an East African Tribe. London: Published by the Oxford University Press for the International Institute of African Languages and Cultures.

Read, Margaret
1960Children of their Fathers: Growing Up among the Ngoni of Nyasaland.New Haven: Yale University Press. 

Rogoff, Barbara, Jacqueline Baker-Sennett, Pilar Lacasa, and Denise Goldsmith
1995Development Through Participation in Sociocultural Activity.In Jacqueline J. Goodenow, Peggy Miller, and Frank Kessel (Eds.), Cultural Practices as Contexts for Development (pp. 45-66). San Francisco: Josey Bass. 

Rogoff, Barbara, Jayanthi Mistry, Artin Göncü, and Christine Mosier
1993Guided Participation in Cultural Activity by Toddlers and Caregivers.Monographs of the Society for Research in Child Development, Serial No. 236, Vol. 58, No. 8. 

Róheim, Géza
1943The Origin and Function of Culture.New York: Nervous and Mental Disease Monographs. 

Romney, Kimball, and Romaine Romney
1966The Mixtecans of Juxtlahuaca, Mexico. Six Cultures Series. New York: John Wiley and Sons.

Rowe, Paul M.
1997The Science of Self Report.Observer 10(1):35 ff. Washington, DC: American Psychological Society. 

Sapir, Edward
1949/1927The Unconscious Patterning of Behavior in Society.In Selected Writings of Edward Sapir in Language, Culture, and Personality (pp. 544–559).David G. Mandelbaum, Ed.Berkeley: University of California Press. 

Schacter, Daniel L.
1995Implicit Memory:A New Frontier for Cognitive Neuroscience.In Michael S. Gazzaniga, (Ed.),The Cognitive Neurosciences (pp. 815–824).Cambridge,MA: MIT Press.

Schacter, Daniel L. and Endel Tulving
1994Memory Systems 1994.Cambridge: MIT Press.

Schieffelin, BambiB.
1990The Give and Take of Everyday life: Language Socialization of Kaluli Children. Cambridge and New York: Cambridge University Press.

Schutz, Alfred
1977/1951Making Music Together; A Study in Social Relationship.Social Research 18:76–97.Reprinted in Janet L. Dolgin, David S. Kemnitzer, and David M. Schneider (Eds.) Symbolic Anthropology; A Reader in the Study of Symbols and Meanings. New York: Columbia University Press. 

Shweder, Richard A., Lene Arnett Jensen, and William M. Goldstein
1995Who Sleeps By Whom Revisited: A Method for Extracting the Moral Goods Implicit in Practice. In Jacqueline J. Goodenow, Peggy Miller, and Frank Kessel (Eds.), Cultural Practices as Contexts for Development (pp. 21–40). San Francisco: Josey Bass.

Smith, Eliot R.
1989Procedural Efficiency and On-Line Social Judgments.In John N. Bassili (Ed.) On-Line Cognition in Person Perception (pp. 19-37). Hillsdale, NJ: Lawrence Erlbaum Associates.
1994Procedural Knowledge and Processing Strategies in Social Cognition. In Robert S. Wyer Jr. and Thomas K. Srull (Eds.)Handbook of Social Cognition, Vol. 1: Basic Processes, 2nd ed. (pp. 99-151). Hillsdale, NJ:Lawrence Erlbaum Associates.
1997 Mental Representation and Memory. Chapter in press in Daniel Gilbert, Susan Fiske and Gardner Lindzey (Eds.)Handbook of Social Psychology, 4th Ed.New York: McGraw Hill.

Smith, Eliot R. and Nyla R. Branscombe
1987Procedurally Mediated Social Inferences: The Case of Category Accessibility Effects. Journal of Experimental Social Psychology 23(5): 361 382. 

Spencer, F. C.
1899The Education of the Pueblo Child: A Study in Arrested Development.Columbia University Contributions to Philosophy, Psychology and Education, Vol. 7, No. 1. New York: Macmillan. 

Sperber, Dan
1994The modularity of thought and the epidemiology of representations.In Mapping the Mind:Domain Specificity in Cognition and Culture.Lawrence Hirschfeld and Susan Gelman eds. Pp. 39-68. New York: Cambridge University Press.

Spielman, Lisa A., Felicia Pratto, and John A. Bargh
1988 Automatic Affect: Are One's Moods, Attitudes, Evaluations, and Emotions Out of Control? American-Behavioral-Scientist (Special Issue: Communication and Affect) 31(3): 296-311.

Spiro, Melford E.
1965Religious Systems as Culturally Constituted Defense Mechanisms.In Melford E. Spiro, Context and Meaning in Cultural Anthropology (pp. 100–103).New York: Free Press.
1982Oedipus in the Trobriands.Chicago and London: University of Chicago Press. 

Squire, Larry R., and Barbara J. Knowlton
1995Memory, Hippocampus, and Brain Systems. In Michael S. Gazzaniga (Ed.), The Cognitive Neurosciences (pp. 825–838). Cambridge, MA: MIT Press.

Stocking, George W., Jr.
1983.The Ethnographer’s Magic: Fieldwork in British Anthropology from Tylor to Malinowski. In George W. Stocking (Ed.) Observers Observed: Essays on Ethnographic Fieldwork (pp. 70–120).Madison: University of Wisconsin Press. 

Stoller, Paul.
1989The Taste of Ethnographic Things: The Senses in Anthropology.Philadelphia: University of Pennsylvania Press. 

Tarde, Gabriel
1900/1890Les Lois de L’Imitation; Étude Sociologique. Third Edition, revised and augmented. Paris: Félix Alcan. 

Taussig, Michael T.
1993Mimesis and Alterity: A Particular History of the Senses.New York: Routledge.

Tomasello, Michael, Ann C. Kruger, and Hilary H. Ratner
1993Cultural Learning (with commentaries and reply).Behavioral and Brain Sciences 16:495–552. 

Tooby, John and Leda Cosmides
1992The Psychological Foundations of Culture.In The Adapted Mind:Evolutionary Psychology and the Generation of Culture.Jerome Barkow, Leda Cosmides, and John Tooby eds. Pp. 19-137.New York:Oxford University Press.

Tulving, Endel
1985How Many Memory Systems Are There?American Psychologist, 40 (4): 385-398.
1995Organization of Memory: Quo Vadis? In Michael S. Gazzaniga (Ed.), The Cognitive Neurosciences (pp. 839-847). Cambridge, MA: MIT Press.

Turner, Victor
1967a/1959Muchona the Hornet, Interpreter of Religion.In Victor Turner, The Forest of Symbols; Aspects of Ndembu Ritual (pp. 131–150). Ithaca: Cornell University Press.
1967b/1960Ritual Symbolism, Morality, and Social Structure among the Ndembu. In Victor Turner, The Forest of Symbols; Aspects of Ndembu Ritual (pp. 48–58). Ithaca: Cornell University Press.

Uleman, James S. and Bargh, John A. (Eds.)
1989Unintended Thought. New York: Guilford Press. 

Umilta, Carlo, and Morris Moscovitch (Eds.)
1994Attention and Performance 15:Conscious and Nonconscious Information Processing.Cambridge, MA: MIT Press.

van Beek, Walter E. A.
1991Dogon Restudied; with commentaries.Current Anthropology 32:139–167. 

Young, Andrew W.
1988Functional Organization of Visual Recognition. In L. Weiskrantz (Ed.) Thought Without Language. A Fyssen Foundation Symposium (pp. 78–107). Oxford: Clarendon.

Watson-Gegeo, Karen, and Gegeo, D.
1977From Verbal Play to Talk Story: The Role of Routines in Speech Events Among Hawaiian Children.In S. Ervrin-Tripp and C. Mitchell-Kernan, (Eds.) Child Discourse (pp. 67–90). New York: Academic Press. 

Whiting, John W. M.
1941Becoming a Kwoma: Teaching and Learning in a New Guinea Tribe.New Haven: Yale University Press for the Institute of Human Relations.

Whorf, Benjamin Lee
1956a/1937.Grammatical Categories.In Language, Thought, and Reality: Selected Writings of Benjamin Lee Whorf (pp. 87–101).Cambridge: MIT Press.
1956b/1940Science and Linguistics. In Language, Thought, and Reality: Selected Writings of Benjamin Lee Whorf (pp. 207–219).Cambridge: MIT Press.

Williams, Thomas Rhys
1970The Structure of Socialization Process in Papago Indian Society. In John Middleton (Ed.) From Child to Adult: Studies in the Anthropology of Education (pp. 163–172). Garden City, NY: Natural History Press for the American Museum of Natural History.

Wolfenstein, Martha
1955French Parents Take Their Children to the Park.In Margaret Mead and Martha Wolfenstein (Eds.) Childhood in Contemporary Cultures (pp. 99–144).Chicago: University of Chicago Press. 

Zemplini-Rabain, Jacqueline
1973Food and the Strategy Involved in Learning Fraternal Exchange among Wolof Children.In Pierre Alexandre (Ed.), French Perspectives in African Studies (pp. 221—233).London: Oxford University Press.

ALAN PAGE FISKE is Associate Professor in the Department of Anthropology at the University of California, Los Angeles, CA90095-1553 (afiske@ucla.edu);when he wrote the first drafts of this he was Research Associate in the Departments of Anthropology and Psychology at Bryn Mawr College. 

[1] I am very grateful to NIMH, who supported the workshop that stimulated me to write this paper, and whose grant, 5 R01 MH43857-07 made it possible for me to write it. Special thanks to Edison Trickett for organizing an exciting workshop, and Mary Ellen Oliveri for supporting it. For comments and suggestions on earlier drafts, I wish to thank Michael Agar, Steve Ferzacca, Siri Fiske, Susan Fiske-Emory, Byron Good, Nick Haslam, Michael Jackson, Allen Johnson, Robert LeVine, Spero Manson, Clark McCauley, Richard Shweder, George Stocking. Paul Stoller, Michael Tomasello, and Harriet Whithead.
[2] Tarde acknowledged the importance of invention as the source of acts subsequently imitated, and also recognized resistance or counter-imitation. But he regarded mutual imitation as the mechanism of creativity, resistance, and all other aspects of sociality.
[3] Socialization in Wogeo, on the north coast of New Guinea, seems to be an exception to these generalizations; Hogbin (1970 [1946]) describes a great deal of explanation and praise.
[4] We do not get much help here from most contemporary anthropology. The recent move toward “embodied” epistemology lacks a very sophisticated theory of the forms of somatosensory, kinesthetic, or mimetic representation; it often conflates, for example, mind and language. And it tends to contrast mind to body as if minds did not operate in brains and as if bodies acted and communicated with each other without perceptual or mental mediation. Some contemporary anthropological writing conflates personal, subjective, experience, sensory, body, tacit, implicit—indeed everything that informants or ethnographers cannot readily formulate in referential propositional language.