J Bioecon (2008) 10:307–314 DOI 10.1007/s10818-008-9041-x BOOK REVIEW

Charles J. Lumsden and Edward O. Wilson, Genes, Mind, and Culture: 25th Anniversary Edition World Scientific, Hackensack, NJ, 2005 [1981], 428 pp. $83.00 Adrian Viliami Bell · Peter J. Richerson

Published online: 12 November 2008 © Springer Science+Business Media, LLC. 2008

Originally published in 1981, Genes, Mind, and Culture extended the sociobiological analysis to human behavior via a pioneering, ambitious and wide-ranging analysis of gene-culture evolution. Lumsden and Wilson constructed a theoretical framework for the role of genetics in human behavior and culture. In principle, culture could be unanchored by any evolved genetic mechanisms, or the evolving genome could code for behavior directly, dispensing with culture. More likely, they plausibly argue, both genes and culture are important in humans and the two interact by gene-culture coevolution. The book expands upon the agenda that Wilson had laid out in his 1975 treatise Sociobiology: The New Synthesis and in his Pulitzer Prize winning On Human Nature in 1978. Sociobiology, they say in the original introduction: …has not coped with learning and cognition or the consequences of the deep, thorough socialization that characterizes the behavioral development of human beings. The theory of gene-culture coevolution is designed to address these issues and permit the entry of evolutionary theory into the study of the mechanisms that produce the human mind and behavior. As we will show in the coming chapters, it derives patterns of cultural diversity from biological ground rules (p. 24). Lumsden and Wilson aimed to synthesize the social-scientific study of humans with the Darwinian theory of evolution by developing a comprehensive theory of how genes interact with cultural variation.

A. V. Bell Graduate Group in Ecology, University of California Davis, Davis, CA 95616, USA e-mail: [email protected] P. J. Richerson (B) Department of Environmental Science and Policy, University of California Davis, Davis, CA 95616, USA e-mail: [email protected]

123

308

A. V. Bell, P. J. Richerson

The primary concept they use to effect a connection between genes and culture is the epigenetic rule. Epigenetic rules are genetically encoded rules that are instantiated in the nervous system and which select among the cultural variants that one is exposed to. The influence of genes on culture is thus expressed through epigenetic rules or bias curves that bias individual preferences across a set of behaviors. Citing the then available evidence from developmental psychology, Lumsden and Wilson argue that childhood development is the stage at which epigenetic rules are elaborated. Even infants and toddlers respond differently to various sound and visual stimulants (hearing, color classification, taste, and smell). These innate rules affect, for example, the short-term evolution of languages, and consequently, long-term cultural pressure has shaped the genetic basis for the epigenetic rules that affect language. They used the metaphor of the Leash Principle; the epigenetic rules are the leash that genes create during development to make culture do their bidding. Genes, Mind and Culture provided a clear explanation for how genes might affect learning and development. As such, it established one of the two influential pictures of how culture operates as an evolutionary system. Lumsden and Wilson assume that cultural variants, what they call culturgens, are stored in long-term memory or are readily observable in the external world, and are recalled or taken up and used as the epigenetic rules dictate. This is clearly illustrated by the structure of their formal models in which the cultural variants that individuals use is little if at all dependent upon the culture used by their parents or other members of previous generations (see pp. 286–288 for the critical discussion of parameter choices). In the terms that Cavalli-Sforza and Feldman (1981) adopted to explain the acquisition of cultural variants, their models consider vertical and oblique transmission to be weak. Lumsden and Wilson propose that ‘culturgen usage rates’ do not depend very much upon the previous generation’s usage rates, but mainly upon the usage rates of peers and upon the epigenetic rules. They assume culturgens to be transmitted by socialization but this fact plays little role in the dynamics that they model which keeps track only of usage rates. Consider an example from language. We all have some slang words in our vocabulary and can choose to speak slang or use the more formal equivalent when we speak. If we all know the same slang and same formal words and if this had always been true, then the only thing that can be dynamic is our rate of usage of the two forms. Based on Lumsden and Wilson’s assumptions, we would have to believe that word choices of this type are little affected by individuals’ environments but are strongly determined by the genes they have inherited; some genotypes prefer slang and some prefer formal speech. These assumptions lead to what Lumsden and Wilson call the thousand-year rule: …during a period of this length substantial genetic evolution can occur in the epigenetic rules of cultural transmission, resulting in such effects as the genetic assimilation of culturgen preference and the assimilation of bias toward specific decision heuristics (p. 295). For the authors, the thousand-year rule is one of the key results of their modeling exercise. But, as they realize (p. 296), its applicability is limited to cultural traits that follow their assumptions, such as incest avoidance, personal bonding, recognition of kin, and territory defense. This is an odd list, since all of these traits occur in other

123

Book Review

309

species without any involvement of culture at all. In the thousand-year rule analysis, all that culture adds is some tendency to make mistakes that genes must adapt to by evolving stronger epigenetic rules to filter them out. In this model, and the related model of the ‘complete coevolutionary circuit’, the adaptive advantages of having culture are difficult to discern. Lumsden and Wilson certainly believe that culture has given humans a tremendous adaptive edge, but just what this advantage is does not emerge from their models (Maynard Smith and Warren 1982; Boyd and Richerson 1983). The second major school of cultural evolution, to which we belong, pictures cultures as sets of complex, highly variable traditions that are transmitted much like genes (Boyd and Richerson 1985; Laland and Brown 2002). Returning to the same language example, we can, of course, choose between a slang and formal word, but most of the words we know were acquired from other speakers of our language. Even if we know more than one language, we still have a quite limited choice of words to use compared to all the words that humans, use or have used in that situation. Slang is often limited to particular groups. Even though the colorful slang of Black Vernacular English leaks into standard American English at a steady rate, few speakers of a Euro-American dialect can pass as BVE speakers. A trained linguist can trace our dialect to a relatively small subset of speakers of our mother tongue. And so it is with social organization, technology, and many other forms of culture. Individuals do not choose to be kayak makers or outrigger canoe makers, members of patrilineal or matrilineal kin groups, or specialists in an advanced division-of-labor economy rather than subsistence farmers any more than they choose their mother tongue. It is mostly a matter of the culture they were brought up in. In our view of cultural evolution, the decisions we make are far from trivial, but they are made at the margin. Individuals acquire most of their culture from others and only tinker around the edges. Simon (1959) famously argued that humans are rather severely constrained decision-makers. In speech, most individuals use their natal dialect though everyone has modest unique features in their pattern of speech that linguists call our idiolect. In our view, decisions about which cultural variants to adopt are influenced not so much by strong epigenetic rules of narrow scope but by weaker, relatively general purpose decision rules that are often subject to environmentally contingent reinforcement. These decision rules might be characterized as “epigenetic rules lite”. As the environment changes, people invent new concepts and words to describe the new elements. If other speakers find a new verb such as “to google” useful, it passes into the language. Weak general purpose decision-making forces act a lot like natural selection. If, on average, people make decisions likely to enhance their fitness when they choose which cultural variants to adopt, this decision-making can act as a surrogate for natural selection. In Lumsden and Wilson’s picture, epigenetic rules control culturgen usage patterns by the brute force of individual decision-making. In our epigenetic rules lite version, the control of culturgen frequencies is shaped by the whole population of minds over successive generations, creating a relatively modest decision-making burden for individual minds. We think that natural selection would have favored a mechanism which relieved individuals of the cognitive heavy lifting that would be involved if we attempted to rationally choose which of the available cultural variants to adopt. Humans learn thousands of concepts,

123

310

A. V. Bell, P. J. Richerson

generalizations, and skills in the course of becoming competent practitioners of their given culture. To vet more than a small fraction of all the alternatives involved by the use of strong epigenetic rules implies an impossible cognitive burden. In the long run, a population using a rules-lite method of choosing between variants can evolve quite complex cultural adaptations. We call this ‘the kayak effect’. The skin boats of the Arctic peoples are far too complex for any single person to invent, though once invented, human imitative skills manage to pass the craft on. They are the cumulative product of many independent innovations and modifications occurring over time. And so it is with all the advanced products of culture. We acquire tens of thousands of words from others, yet only a handful of us will have invented one, at least one that entered into the speech of any significant number of people. In this picture, cultural adaptation is typically more rapid than organic adaptation and is often very effective, witness kayaks. It is culture, not genes, that follows a thousand-year rule. In a thousand years, humans can run up a novel cultural adaptation of some sophistication. Thus most of the adaptive slack will be taken up by culture, not genes, leading to what we might call the one hundred thousand-year rule for genes. In any arena where cultural adaptations are effective, cultural change will be rapid and selection on genes will be correspondingly weakened. People with a recognizably modern capacity for culture seem to have evolved in Africa around 70 thousand years ago, just before they spread out of Africa to colonize the rest of the Old World, following a history of accelerated modernization that dates back 250,000 years ago or more. Cultural evolution is on the order of 100 times faster than genetic evolution, and only rules-lite forces are required to make it so. In Chap. 7 The Biogeography of the Mind, Lumsden and Wilson do consider the costs and benefits of different degrees of flexibility of epigenetic rules and the problems of imagining how a relatively small number of genes can rigidly program a brain. These are the same considerations that suggested the epigenetic rules lite. PJR remembers reading Genes, Mind, and Culture closely as soon as it came out in 1981 to see if Boyd and he had gotten scooped. No, the same considerations had led Lumsden and Wilson to a different conclusion. One of the most important features of the rules-lite picture is that as culture becomes more complex and the decision rules applied to manage it weaker, cultural variation itself begins to respond to natural selection in addition to decision-making forces. Culture is a system for transmitting heritable variation, and selection can affect any sort of heritable variation, not just that carried by genes. Natural selection directly on culture can transform the coevolutionary process from one in which genes tightly control cultural evolution into one in which culture plays a co-ultimate role with genes. For example, groups made up predominantly of people with cultural traditions facilitating cooperation between individuals would have been more successful than groups made up of more competitive individualists. Thus, selection on the basis of cultural variation between groups would favor the kind of social behaviors that make humans suited to live in large, cooperative societies. Cultural variation turns out to be more susceptible to group selection than genes partly because cultural evolution is faster than genetic evolution and can lead quickly to big inter-group differences. Selection on genes alone would never have favored the human skill at cooperating in large groups of unrelated individuals like tribes and nations. If selection is an important force in cul-

123

Book Review

311

tural evolution, the ‘full coevolutionary circuit’ as described by Lumsden and Wilson is incomplete. Culture cannot be kept on a tight epigenetic leash by genes responding to evolving culture because selection on cultural variation creates a cultural environment that selects for genes adapted to that environment, forcing epigenetic rules to evolve to accord with selection on culture rather than culture to accord with selection on genes. For example, most if not all cultures have rules creating obligations to social units composed of distantly related and unrelated individuals (“tribes” for want of a better general term). Innate epigenetic rules arguably do make learning tribal identities and behaving according to the rules of the tribe easy, if not automatic. One explanation for such tribal social instincts is that group selection on cultural variation was the leading mechanism in a coevolutionary process in which culturally mandated social selection culled out genotypes that were too belligerent or clung too tightly to the rules of kin selection. Think of the leash metaphor again, but make the dog big and sometimes quite stubbornly in pursuit of its own goals. The anniversary edition of Genes, Mind, and Culture includes a new 35-page introduction in which Charles Lumsden reviews scientific progress in gene-culture evolution since 1981. He cites considerable accumulated evidence that he believes supports the strong epigenetic rules picture. He includes only a polite nod toward the rules lite alternative. He cites evidence from personality studies by behavior geneticists claiming a large role for genetic variation, the studies of modularist evolutionary psychology like Leda Cosmides and John Tooby, and some findings from gene sequence data involving recently evolved genes affecting language and other cognitive functions. But evidence for epigenetic rules is not necessarily evidence for strong rules. Much evidence can be brought to bear suggesting that human cultures include masses of variation not backed up by variation in strong epigenetic rules (Richerson and Boyd 2005). Language is a good example. Surely some innate structures exist that allow and motivate people to talk, or else chimpanzees would learn to talk too. We have some clues about what the genetic substratum of language is from studies of a variant of the FOXP2 gene. The common variant of the human gene evolved quite recently, and the bearers of a rare mutation have a speech handicap. On the other hand, humans have evolved thousands of highly diverse, mutually unintelligible languages. No one suggests that we have thousands of different innate rules underpinning this diversity. As the ready bilingualism of children shows, every person can learn whatever language(s) they are exposed to. The weak epigenetic rules, including relatively general rules not peculiar to language, that govern this evolution are tolerably well understood (Deutscher 2005). We think it unfortunate that Lumsden and Wilson’s models are not better appreciated among evolutionary psychologists. Tooby and Cosmides (1992) famously argued that the brain consists of hundreds or thousands of cognitive modules that work like strong epigenetic rules. At the same time, they imagine that cognitive modules are species typical mental adaptations and suppose that little genetic variation will exist for them. Lumsden and Wilson’s models show that strong modules, like strong epigenetic rules, ought to come under selection and be variable according to the thousand-year rule. In other words, a politically correct innatism of the sort pioneered by Noam Chomsky in which genes play a strong, direct role in cognition yet do not vary among individuals and among races is probably not possible. The fact that human populations

123

312

A. V. Bell, P. J. Richerson

are behaviorally diverse without having much genetic variation at all, and especially only modest amounts of genetic variation between different populations, suggests that cultural adaptations are largely shielding cognition from strong selection. Which of these two views of cultural evolution, the Lumsden and Wilson version or the rules-lite version, describes the bulk of cultural variation is in the process of being tested. Human behavior changed substantially as modern humans evolved from archaic ancestors between perhaps 250,000 years ago and 50,000 years ago. Human behavior has evolved from a world dominated by hunter-gatherers 10,000 years ago to one dominated by industrial economies today. The human genome project and its follow-on sequencing of the chimpanzee genome and the haplotypes of a growing number of people from around the world are rapidly giving us a picture of just what genes have changed in the course of human evolution. These genes contain internal evidence about when these changes have occurred. If culture is dependent upon strong selection on epigenetic rules along the lines of the Leash Principle and the ThousandYear Rule, we might expect to see appreciable change and differentiation in genes that plausibly underlie strong epigenetic rules as agriculture, urbanization, and industrialization have proceeded. On the other hand, if the 100,000-year rule for genes is more correct, we should see few if any new epigenetic rules in the last 10 millennia, but perhaps a considerable number before 50,000 years ago. On the other hand, agriculture changed diets drastically suggesting recent selection for metabolic adjustments. Population increases and the keeping of domestic animals exposed our agricultural ancestors to a suite of new virulent epidemic diseases, suggesting selection of genes involved with resistance to these diseases. Sabeti et al. (2006) recently reviewed the accumulated evidence for recent selection on nearly 100 genes and gene families. Unfortunately, a functional understanding of the roles of most of these genes is unknown or known only via the pathological effects of rare mutants. The adaptive role of the major alleles is still mostly unknown. Only seven of the genes found to have undergone selection recently seem to have anything to do with the brain at all. Two of them influence brain size (mutations in the genes cause microcephaly), one is related to language, one is related to cerebellar function, one affects a dopamine receptor implicated in personality variation, and one is a gene family related to pain responses. Most recently evolved genes are related to physiology and to infectious disease. For example, LAC is a regulatory gene whose function is well known. A high frequency derived allele in many dairying populations causes adults to secrete lactase as adults, almost certainly a recent response to the culturally evolved practice. Other very recently selected genes may be coevolved responses to agriculture. ADH is associated with alcohol metabolism. MC1R affects skin and hair pigmentation, and alleles common in extra-tropical population probably reflect the expansion of African populations to higher latitudes around 40,000 years ago. A total of five loci are involved with resistance to malaria including HBB whose function in malaria protection (and sickle cell disease) is well known functionally. Malaria is thought to have become a much worse disease in Africa and elsewhere after population density increases due to agriculture made it an epidemic disease. The genes related to personality and pain reception would seem to be the best candidates to be strong epigenetic rules. Brain size probably coevolved with culture and the language related gene (FOXP2) with symbolic expression, an important subset of culture. One of the

123

Book Review

313

brain related genes, ASPM, seems to have originated in the Near East about the time that agriculture became important. The FOXP2 gene evolved around 40,000 years ago roughly the same time modern humans first migrated out of Africa. Although, as Sabeti et al. note, our sample of recently selected genes is still quite crude, we are struck by the paucity of candidate strong epigenetic rules that have evolved in the last few thousand years and the rather larger number that are good to excellent candidates for coevolved responses to complex cultural adaptations like crop cultivation and dairying. The current plunge in fertility rates, now below replacement in many societies and dropping rapidly in most former population explosion countries, is an interesting current test case for the strong epigenetic rules picture. Modern culture is exerting very strong selection pressure on human genomes for epigenetic rules that would support high fertility. Modern healthy, wealthy women could achieve completed family sizes of eight or so children as against the less than two that prevail in most rich countries today. If any genetic variation for resistance to the demographic transition exists, it should be increasing a very rapid rate. So far as we are aware, demographers have not yet brought to light any evidence for the existence of such a gene. Interestingly, the Amish and Hutterite societies, among the very few in industrial societies that do maintain something like the fertility that natural selection would favor, appear to do so by cultural not genetic means. Edward Wilson, the ardent conservation biologist, would no doubt hate to see a high fertility epigenetic rule emerge as the best exemplar of the Thousand-Year Rule! In the face of the building avalanche of evidence from gene sequence and functional genetics studies, no sensible person will be dogmatic about what is a quantitative empirical question. Much less could any of us clearly foresee the outcome of the projects launched 25 years ago. The rules-lite picture seems to us to be necessary to account for cultural diversity in at least some domains like language. On the other hand, strong rules might conceivably dominate in some domains of culture, such as reproductive behavior, though obviously not strongly enough to maintain behavior which efficiently converts resources into offspring in modern societies. The tragedy of science, T.H. Huxley quipped, is that beautiful hypotheses are slain by ugly facts. No matter whose lovely models end up sinking on the new reefs of facts, we salute Lumsden and Wilson as important and tireless pioneers in the pursuit of a synthesis of biology and social science whose contribution is fittingly remembered with this 25th Anniversary Edition.

References Boyd, R., & Richerson, P. J. (1983). Why is culture adaptive? The Quarterly Review of Biology, 58, 209–214. Boyd, R., & Richerson, P. J. (1985). Culture and the evolutionary process. Chicago: University of Chicago Press. Cavali-Sforza, L. L., & Feldman, M. W. (1981). Cultural transmission and evolution: A quantitative approach, Princeton: Princeton University Press. Deutscher, G. (2005). The unfolding of language: An evolutionary tour of mankind’s greatest invention. New York: Henry Holt. Laland, K. N., & Brown, G. R. (2002). Sense and nonsense: Evolutionary perspectives on human behaviour. Oxford: Oxford University Press. Maynard Smith, J., & Warren, N. (1982). Models in cultural and genetic change. Evolution, 36, 620–627.

123

314

A. V. Bell, P. J. Richerson

Richerson, P. J., & Boyd, R. (2005). Not by genes alone: How cultural transformed human evolution. Chicago: University of Chicago Press. Sabeti, P. C., Schaffner, S. F., Fry, B., Lohmueller, J., Varilly, P., Shamovsky, O., Palma, A., Mikkelsen, T. S., Altshuler, D., & Lander, E. S. (2006). Positive natural selection in the human lineage. Science, 312, 1615–1620 + Supplementary Material online. Simon, H. A. (1959) Theories of decision-making in economics and behavioral science. American Economic Review, 49, 253–283. Tooby, J., & Cosmides, L. (1992). The psychological foundations of culture. In J. Barkow, L. Cosmides, & J. Tooby (Eds.), The adapted mind: Evolutionary psychology and the generation of culture. New York: Oxford University Press.

123