ELSEVIER Culture and the Evolution of Social Learning Mark V. Flinn Department of Anthropology, University of Missouri Applications of modern evolutionary theory to human culture have generated several different theoretical approaches that challenge traditional anthropological perspectives. “Cultural selection” and “mind parasite” theories model culture as an independent evo- lutionary system because transmission of cultural traits via social learning is distinct from transmission of genes vla DNA replication. “Dual-inheritance” and “co-evolution” theories model culture as an intermediary evolutionary process that involves informa- tion from two inheritance systems: genetics and social learning. “Evolutionary psychol- ogy” theories emphasize that the evolutionary history of natural selection on mental pro- cesses links culture and biological adaptation; hence, cultural information is viewed as part of the organic phenotype and not an independent evolutionary system. Cross-cul- tural universals and scenarios of the “environment of evolutionary adaptedness” are used to identify characteristics of the “evolved mind” (human nature). “Behavioral ecol- ogy” theories examine relations between behavior and environmental context. Behav- ioral/cultural variations are viewed as products of flexible decision-making processes (evolved mind) that may respond adaptively to micro-environmental differences. It is difficult to devise empirical tests that distinguish among these theories, because they share many basic premises and make similar predictions about human behavior. In- deed, some of the apparent differences may be more semantic than substantive. Social learning is the key process underlying these evolutionary paradigms. Here I argue that human learning mechanisms are products of natural selection, and hence process information in ways that reflect evolutionary design. I review common objec- tions to this hypothesis, including (1) learning processes are uncoupled from genetics and biological adaptation; (2) culture (or its effects) is partly extrasomutic; (3) culture, by most definitions, involves mental phenomena, including conscious thought; (4) culture involves the use ofarbitrary symbols to communicate information; (5) culture appears to have emergent properties at the group level, such as shared values and beliefs resulting in political and religious institutions; (6) culture involves historical processes; and (7) com- plex culture is uniquely human-we need an explanation for why the human species alone evolved such extensive social learning aptitudes. I suggest that these controversies can be resolved, although empirical tests are difficult. Received October 16, 1996; revised October 24, 1996; accepted October 25, 1996. Address reprint requests and correspondence to: Mark V. Flinn, Department of Anthropology, University of Missouri, Columbia, MO 65211. Evolution and Human Behavior 18: 23-67 (1997) 0 Elsevier Science Inc., 1997 1090-5138/97/$17.00 655 Avenue of the Americas, New York, NY 10010 PII S10902-5138(96)00046-5 24 M.V. Fhnn I posit that social competition was a primary selective pressure on human mental abilities (Alexander 1989) and that this favored domain-general, constructivist learning capabilities (e.g., Quartz and Sejnowski, in press) that can manage context-dependent analysis and integrate information from domain-specific mechanisms (Hirschfeld and Gelman 1994; MacDonald 1991; Sperber 19%). Humans are unique in the extraordi- nary levels of novelty that are generated by the processing of socially learned informa- tion. Human culture is cumulative; human cognition produces new ideas built upon the old. To a degree that far surpasses that of any other species, human mental processes must contend with a constantly changing information environment of their own cre- ation. Cultural information may be especially dynamic because it is a fundamental as- pect of human social competition. Apparently arbitrary changes in cultural traits, such as clothing styles, music, art, food, dialects, and so forth, may reflect information arms races among social competitors. The remarkable developmental plasticity and cross- domain integration of some cognitive mechanisms may be products of selection for spe- cial sensitivity to variable social context. Human “culture” is not just a pool or source of information; it is an arena and theater of social manipulation and competition. Studies of human behavior-including language, kinship, mating relationships, subsistence, economics, and politics-generally are consistent with an evolutionary basis for social learning, but often they fail to add specific new knowledge about the mecha- nisms. Analyses of cognitive aptitudes underlying language, kinship, and so forth, often are inconclusive because cultural information (and consequent behavior) involves com- plex interaction among history, environmental variation, ontogenetic pathways of men- tal processes, and specific context. I suggest that empirical tests of evolutionary culture theory must build upon identification of apparent universals and examine individual variability by incorporating developmental psychology, environmental conditions, and social and historical context. This synthesis would benefit from enhanced cooperation between cognitive psychologists and cultural anthropologists. 0 Elsevier Science Inc., 1997 KEY WORDS: Culture; Evolution;Learning; Psychologicalmechanisms; Social competition. ecent modifications of evolutionary theory have kindled renewed inter- est in relations between culture’ and biology. New concepts of learning as an adaptation (J.L. Gould 1986; Johnston 1982) have inspired new mod- els of culture (e.g., Alexander 1979a, 1990a; Boyd and Richerson 1985; LumsdenR and Wilson 1981; Sperber 1996). Although there is general agreement that evolutionary theory is useful for understanding human behavior and culture, impor- tant issues remain unresolved, and directions for empirical research are enigmatic. Some models, beginning with a population genetics paradigm, posit that cul- ture is a distinct “co-evolutionary” system because transmission of cultural traits via social learning is separate from transmission of genes via DNA replication (e.g., Cavalli-Sforza and Feldman 1981; Dawkins 1982; Richerson and Boyd 1978; Durham 1991; Rindos 1986b). Culture and genetics are viewed as different, albeit interacting, types of inheritance processes. Grganically evolved (i.e., genetically controlled) “learning biases” may influence culture content, but the property of transmissibility of socially learned information is considered sufficient to distin- ‘By “culture” I refer to common (non-exclusive) usage of the term in social science. I do not find it important here to distinguish among the various definitions. I use “information” as a more specific term. Culture and the Evolution of Social Learning 25 guish it from other environmental effects on the phenotype. This approach repre- sents cultural evolution as analogous to organic evolution. Other models, originating from a Tinbergian ethological paradigm, emphasize that evolved mental or psychological processes constrain culture. Human psychol- ogy (including learning aptitudes) and, hence, behavior are posited to have been shaped by natural selection (e.g., Alexander 1979a; Barkow 1989b; Irons 1979a; Lumsden and Wilson 1981; Cosmides and Tooby 1995). Sources of controversy in- clude the genetic basis, degree of specificity, ontogeny, and plasticity (particularly in regard to modem social conditions) of psychological processes (e.g., Daly and Wilson 1995; Symons 1989). The objective of this paper is to examine a key issue underlying this new body of culture theory: the evolutionary basis for social learning. Here I attempt to inte- grate models of culture with knowledge of learning processes from behavioral ecol- ogy and psychology. Empirical studies of culture from evolutionary perspectives are briefly reviewed, and unresolved theoretical problems discussed. I argue that human learning mechanisms are products of natural selection, and hence process informa- tion in ways that reflect evolutionary design. I posit that social competition is a pri- mary selective pressure on human mental abilities (Alexander 1989; Humphrey 1984) that has favored some domain-general, constructivist learning capabilities (e.g., Quartz and Sejnowski, in press) that can manage context-dependent analyses and integrate domain-specific mechanisms (Hirschfeld 1994; MacDonald 199 1; Sperber 1996). This argument remains speculative because it lacks information about (1) the developmental ontogeny and epigenetics of cognitive processes, (2) the influ- ence of environmental conditions and historical context on information choice, and (3) the dynamics of the use and manipulation of information in social competition. COMMON GROUND Anthropology has identified several important relations between biology and cul- ture. First, human “aptitudes” for culture are products of natural selection, evi- denced by (1) increases in cranial capacity from australopithecines to modem humans (Holloway 1975; Spuhler 1959; Wolpoff 1995), and (2) cognitive differ- ences between humans and other primates (Byrne 1994; Cheney and Seyfarth 1990; Gibson and Ingold 1993; Griffen 1984; Holloway and de la Coste-Lereymondie 1982; Quiatt and Reynolds 1993; Parker and Gibson 1979, 1990). Second, “biologi- cal needs” are facilitated by culture, such as acquiring sufficient food, protection from predators and the elements, reproduction,