The Life History of Culture Learning in a Face-to-face Society

This paper identifies the temporal sequence during which important cultural beliefs (food taboos) are transmitted to individuals in an oral society living in the Democratic Republic of Congo. Results are based on patterns of correlation in scaled measures of cultural similarity between same-household and same-clan members. In the first phase of the cultural life history (0-10 years-of-age), individuals are innocent of food taboos. In Phase Two (11-20 years-of-age), when most taboos are acquired, parents dominate the enculturation of the young. Phase Three (20+ years-of-age) consists of a less intense period of changing beliefs transmitted from individuals outside the family. This adult learning is usually ignored in socialization studies, but has a significant impact on cultural population dynamics. Since this sequence of early family-based and later more broadly-based cultural influences reflects the temporal expansion of the social universe of individuals everywhere, this life history is probably representative of how cultural traits are generally acquired. Further, the timing of this ontogenetic process mirrors changes in the influence of family environment on cognitive ability in behavioral genetic studies. This parallel suggests that both cultural and psychological traits evolve similarly in response to age-related changes in the social environment. Finally, comparing studies based on statistical models of cultural similarity with those based on reported transmission pathways suggests that the latter, which indicate overwhelming parental influence, probably reflect normative response biases. The Life History of Culture Learning 3 Conventional ideas about socialization into cultural life are quite rudimentary. While it is recognized that individuals are not born with knowledge of how to behave like a member of their cultural group and therefore must acquire this competence, the means by which they do so typically continues to be conceptualized as a disembodied and abstract process which magically endows succeeding generations with the group’s cultural heritage.1 Strauss (1992:9) calls this the "FAX model" of enculturation. This is obviously unsatisfactory as a theory of how cultural knowledge persists over time. In particular, there is a conspicuous lack of attention -in standard social theory generally -to information as a distributed resource (Thompson 1995). Recognition of this simple fact has many ramifications. I will concentrate here on three. First, it urges us to identify the agents behind cultural transmission. Second, it implies not everyone has equal access to cultural knowledge, which in turn suggests that intra-cultural variation may be significant. Third, by emphasizing the need for beliefs and values to spread, it forces attention on the psychology of information acquisition, since only internalized knowledge is likely to be further transmitted. In the following, I first discuss each of these implications in turn. I then detail a study designed to address many of the issues arising from the recognition that knowledge must diffuse through cultural groups. Whom should we expect to ensure that the accumulated wisdom of previous generations is perpetuated? The presumption in developmental psychology is that parents are all-important early in life, but that peers become increasingly significant role-models as children go to school, and particularly during adolescence (e.g., Schaeffer 1996). However, this is based primarily on the general observation that the amount of time spent with parents versus age-peers decreases with age, and secondarily from self-reports about the relative importance of these social relationships. A small number of empirical studies have suggested that, at least in easier-to-study traditional populations, socialization takes place primarily within families, largely through the agency of parents (e.g., Hewlett and Cavalli-Sforza 1986; Ohmagari and Berkes 1997). The general conclusion that parents should dominate in the teaching of culture, especially early in life, is consistent with an independent viewpoint: evolutionary theory. Parents and offspring share genetic interests, so there should be cooperative signalling between them. Since parents have similar microenvironments to offspring, but greater experience, information should be passed from them to the younger generation (Cronk 1991). More particularly, in species with low certainty of paternity mothers should show greater concern with the well-being of offspring than fathers. Consequently, in most animals, it is mothers that dominate education of the young (Caro and Hauser 1992). Thus, cultural transmission should mimic genetic transmission, with information passing "vertically" from parents to offspring. Further, if there is any critical period for learning culture, and this period occurs relatively early in life, then it is reasonable to suppose that much of this learning comes from those who dominate the social environment of individuals at such ages: parents and other close relatives. The cultural alternatives then presented to learners would not be random but artfully put together by these experienced kin, who would profit from adaptively constraining their children’s choices. Perusse et al. (1994) call this hypothetical process "teaching bias." However, Harris (1999, 1995) has recently mounted a significant attack on the received wisdom that parents largely bear the burden of indoctrinating the young, arguing that the literature on parental influence has been oversold. She contends the evidence is highly conflated with genetic factors and excludes the possibility of reverse socialization of parents by children. Further, most such studies only examine behavior within the home, where parental constraints are most effective. Indeed, her most important argument is that all socialization is context-dependent: when at home, children behave according to one set of 1 For example, a recent text (Schaffer 1996:247) suggests that “socialization refers to the transmission from one generation to another of a society’s standards of beliefs and behavior.” The Life History of Culture Learning 4 rules; when at school, they follow another, because the rewards and punishments differ in these varying environments. To her mind, the reason behavioral genetic studies show such a trivial effect of common family environment is that much of social modeling looks to individuals outside the household, from the peer group. There are evolutionary reasons to expect such "horizontal" or intra-generational transmission as well. Peers are highly similar in their social positions, so if there is rapid change in the social or ecological environment, they are the individuals most likely to have tracked such changes, and so should provide the best information about current conditions (Boyd and Richerson 1985). There seem to be substantial areas in which the expectation of non-parental transmission is fulfilled. For example, parents generally seem unwilling to educate their offspring regarding sexual matters -as exemplified in one subsistence-level society, where men almost always report learning about sexually-transmitted disease from peer group members, particularly older male sibs or cousins (Bailey and Aunger 1995:210). In more complex societies, professionals (e.g., school-teachers) have taken over the role of enculturation to a greater or lesser degree. It therefore seems that no general proposition concerning socialization can be applied to all kinds of knowledge in all kinds of societies; in fact, responsibility for indoctrinating the young may be distributed throughout the local community. Nevertheless, there is considerable interest in determining the roles various sources of cultural information play in society. This is because the distribution of knowledge depends on both access to, and variation in abilities to interpret, alternative cultural choices. Cultural dynamics in turn depend on the distribution of knowledge, at least in structured populations (Lowen and Dunbar 1997). We therefore need to address the second implication of knowledge as a distributed resource: intra-cultural variation. Such variability has been significant wherever it has been studied (Pelto and Pelto 1975). How are we to explain it? Harris’ model of socialization by peer groups would suggest a high degree of cultural uniformity, at least within age groups, and so is not likely to characterize all cultural learning. On the other hand, if it is allowed that occasional mistakes creep into the cultural transmission process, any such differences would then persist within the family-based lineages produced by parent-offspring learning. Over time, the accumulation of such mistakes could lead to significant intra-cultural differences. It therefore seems reasonable, from this basis alone, to argue that most societies rely to some degree on the parental transmission of cultural beliefs. But this conclusion still does not indicate what role individuals might play in their own education. Education involves not just teachers, but also pupils. Exposure to novel ideas is not sufficient for their adoption. Schwartz (1981), for example, recognized that individuals are active participants in the transmission of cultural knowledge, and incomplete as human beings until they acquire this inheritance. However, the "acquisition of culture" is not perfect, so the idioverse of each individual can be somewhat idiosyncratic, distilled from their unique set of cultural experiences. This is because cultural knowledge is not just transmitted information but the internalized derivatives of others’ social inputs. This internalization depends on the entire personalities of each individual: cognitive, evaluative and affective. Through this process, some cultural information acquires emotional and directive force, and thus determines an individual’s behavior. Other anthropological research has shown, however, that the nature of culture acquisition is also determined by the social context in which transmission occurs. In particular, the social practice school (e.g., Lave 1988, Suchman 1987) emphasize that internalization is itself incomplete; much knowledge is implicit, and can only be acquired through practice. So personality alone is not an accurate predictor of how cultural acquisition "takes," as Schwartz and others have argued. Competence is slowly acquired, and not just in episodes where information is transferred from the expert to the novice. For information to become embodied knowledge, the individual must engage in the everyday use of that new knowledge, so that feedback from experience can produce understanding. Thus, over time, socialization (or FAX theory) has given way to an emphasis on the active filtering of cultural inputs (internalization), which in turn has been replaced by activityin-context as the dominant paradigm within which the reproduction of social systems is understood. The picture has become progressively more complex as new types of The Life History of Culture Learning 5 considerations have been added. The unit of analysis has advanced from the abstract group, to the passive individual (the recipient of culturally transmitted information), to the actively appraising individual (internalization theory), to the socially situated individual, to a cluster of behaving individuals (novices, experts and their tools) within a field of practice. The notion of culture itself has followed these changes in perspective -going from being a bucket poured into empty mental reservoirs, to the product of an active engagement between individual minds and a circulating complex of knowledge. Individuals are seen as gaining access to this knowledge within a specific social context and incorporating it in their own inimitable fashion. Existing empirical studies of socialization or enculturation are not well-suited to addressing the issues raised by these theoretical advances. Several problems can be identified. First, few take a life-span perspective; in particular, cultural learning among adults is almost universally ignored. This is because socialization has traditionally been presumed to end at adolescence. However, significant changes in social roles and self-perceptions continue into adulthood, as individuals enter new social arenas (e.g., work), and become spouses and parents (Durkin 1995:629). Levinson’s (1978, 1986) influential account of stages in adulthood (including the "mid-life crisis") points out that within each decade of an individual’s life reassessments of social place and a revision of the life narrative can take place, depending on differences between expectations and actual achievements (e.g., "I never wrote the Great American Novel"). As a result, it is difficult to determine the relative significance of the various possible pathways of cultural transmission during the life course, since the sources of learning later in life may be different from those characteristic of the early years. Second, socialization studies almost exclusively rely on either association patterns or reported transmission events. However, the former does not measure transmission -because attention to, and absorption of, information can take place during a nonrandom portion of the time spent in the presence of others -while the latter may not reflect true patterns of cultural learning if there are biases in the recall or social acknowledgement of transmission episodes. Third, interest has concentrated primarily on the development of personality and social role-playing abilities (e.g., emotional control, gender roles, and prosocial behaviors), or the competences necessary for everyday life (e.g., parenting and childcare), rather than the transmission of cultural beliefs and values per se. Such proficiencies tend to involve physical or social skills rather than simply semantic knowledge. As a result, such traits can exhibit significant genetic rather than cultural inheritance. In fact, behavioral genetic models tend to find low vertical cultural transmission within families of traits such as social attitudes (e.g., Martin et al. 1986), emphasizing instead positive assortative mating between parents. It is only when cultural beliefs themselves (such as religious affiliation) are investigated that parental cultural transmission appears significant (e.g., Eaves et al. 1990). Thus, the few behavioral genetic studies of cultural traits are in accord with the scant anthropological evidence. Secondarily, if skills are acquired through individual practice, questioning an informant about the transmission of such skills from others becomes confusing. My objective in this paper is to present a case study of cultural transmission which alleviates these problems. The study concerns an oral society of horticulturalists and foragers living in the Ituri Forest of the Democratic Republic of Congo (DRC, formerly Zaire). First, by focusing on a set of beliefs themselves (avoidances concerning the consumption of particular foods), any confound with behavioral practices -and hence genetic influence -is avoided. Second, transmission is inferred using statistical methods based on partitioning variation in the beliefs themselves rather than simply on calculating proportions of informant recalls concerning where they learned something. This avoids possible biases in the reporting of transmission pathways (e.g., in favor of normative authority figures). Third, a nearly complete sample of the local population is used, including individuals from a wide range of ages. I cannot assess the role of practice in the acquisition of this domain of cultural knowledge since the relevant social activities were not rigorously observed. Nevertheless, this approach permits an estimation of the overall significance of different social relationships in the transmission of information, and particularly the importance of learning in later life. In fact, results from the pattern of correlations between members of households and within villages in the study population suggest the degree of non-parental transmission is The Life History of Culture Learning 6 insignificant in this belief system, at least during the early years of life when most food avoidances are acquired. Thus, it is true that parents are important figures in the maintenance of these cultural traditions. This may be particularly the case for aspects of culture which are closely tied to personal identity, such as food avoidances. Some avoidances are also linked to a norm that such beliefs should be acquired specifically from parents. However, even here, it is possible to see a discrepancy between norms and practice: especially as individuals age and come under the influence of people outside their close family, they continue to learn about their culture, obliterating to some degree the traces of knowledge acquired earlier from parents. In effect, socialization is not a simple process and continues throughout life, with only the early, normative part involving high parental input. I further argue that other empirical studies reporting overwhelming parental influence probably reflect normative biases in reported transmission. Finally, the fact that individuals turn to age-peers for guidance just when they are supposed to play the role of teacher to their own children has important implications for cultural change. It suggests that children may not learn what their parents learned when they were young (from grandparents), but instead what the parent more recently heard from someone outside the family. The population dynamics of this latter possibility are quite different: rather than resulting in independent family-based lineages, cultural relationships become blurred, with exchanges of knowledge producing complex networks of cultural affiliation. This is the situation found here in the analysis of cultural correlations between members of the same village. Further, the occurrence of some intra-generational ("horizontal") transmission between episodes of inter-generational transmission introduces the possibility of non-adaptive learning, as documented elsewhere for the Ituri food avoidance system (Aunger 1994a). Thus, while non-vertical transmission may not be the means by which a high proportion of all beliefs learned, it nevertheless has a strong influence on the distribution of beliefs in the population. The significance of non-vertical transmission is thus intimately tied to its timing in the life history of cultural learning. I therefore conclude that the distribution of cultural information depends not only on the relative strengths of the various links along which information is exchanged in a social network, but also on network dynamics -the temporal relationship between episodes of cultural learning (or relearning) and of transmission by individuals in that network. I conclude by discussing the social and cultural context within which transmission takes place in this society.