Wednesday, April 22, 2009

Coevolution

Can historians learn anything from biologists? Jared Diamond’s 1997 book Guns, Germs, and Steel sparked a flurry of interest among at least some historians, who published a special forum in the American Historical Review and held panels at the annual AHA meetings. A more common response to Diamond, however, if I go by numerous conversations with historians, has been disapproval tinged with an almost visceral rejection. I surmise that this disdain derives from the book’s underlying biological premise: as with other species, Diamond argues, humans’ fates have been determined by the environment and the availability of natural resources. It may also have to do with Diamond’s emphasis on the long term and his relative lack of interest in individuals and events. None of these things - the biological roots of human behavior, environmental determinism, the disregard for particularities – historians can abide. In fact, however, most historians have probably not taken a stance one way or the other in regard to Diamond’s book – or to the growing number of intellectual encroachments by natural scientists onto terrain usually reserved for historians. Whether due to parochialism or indifference, we historians remain, as Daniel Lord Smail has put it, in the “grip of sacred history.” We still conceive of history as starting with civilization and written records some 5,500 years ago in Sumer.
The relatively generous attention paid to Diamond in fact only confirms the extent of the problem: Guns, Germs, and Steel was a gripping, popular (but not unserious) read. If it hadn’t been a best-seller, it almost certainly would not have earned the AHA’s attention. Less visible, but in many cases even more important, works by natural scientists usually go unnoted by historians. Since the 1980s, for example, several schools of biologists and anthropologists have been developing ambitious theories of “coevolution.” These approaches treat human culture as an evolutionary system in its own right and investigate its properties and its interactions with its genetic counterpart. They thereby hope to develop comprehensive, indeed potentially revolutionary theories of human behavior, something one might think would be of interest to historians. Yet a jstor-search reveals that none of these books received even one review in a historical journal.
They deserve better. The following is a review of one of these projects, Robert Boyd and Peter Richerson’s 1985 book Culture and the Evolutionary Process. (The others are Luigi Cavalli-Sforza and Marcus Feldman’s 1981 Cultural Transmission and Evolution: A Quantitative Approach and William Durham’s 1991 Coevolution.)
Culture, as Boyd and Richerson (B & R) define it, includes all episodes of social learning of ideas and behaviors, whether by teaching or imitation. B& R provide considerable evidence that social learning of this type – and not individual learning or rationality, as neo-classical economics and rational choice approaches assume – plays a predominant role in human behavior. They point to the numerous cases of “cultural inertia,” in which people don’t respond to new circumstances, even for generations (see, for example, David Hackett Fischer’s excellent book, Albion’s Seed, on the persistence of different English folkways in America), and to psychologists’ plentiful evidence that people operate by various, often inaccurate, rules of thumb.
According to B & R, culture shares with genetic inheritance the three crucial ingredients necessary for an evolutionary system: variation, inheritance, and selection. I.e. people have different ideas and act in a variety of ways; they pass along these cultural traits to “cultural offspring,” who usually include their biological offspring, but can also include friends, students, etc.; finally, some cultural traits get passed on more often than others (for reasons to be discussed below). B & R therefore call their model a “dual inheritance” theory.
Crucially, the two inheritance systems, while similar, are not identical. Cultural evolution allows for “acquired variation,” i.e. it is Lamarckian. In genetic evolution the behavior of an individual has no effect on the genes he passes on. With culture, however, an individual can learn something on his own or otherwise pick and choose from his cultural heritage. What he passes on to his cultural offspring has been changed. Additionally, in cultural evolution there can be many “parents,” not just the two of biological reproduction.
B & R distinguish between several different “forces” giving cultural evolution its directions. The first two, which belong together, they call “guided variation” and “direct bias.” Guided variation involves the interaction of individual learning, or innovation, and cultural evolution by social learning. Despite having culturally inherited certain ideas or behaviors, individuals are also capable of assessing their surroundings and options and developing a new response, one they did not inherit. For example, a medieval farmer stumbles upon a different way to plow his fields. If he can ascertain that this is an improvement over traditional methods – something that may not be easy to evaluate - he is then likely pass on this new variant to his cultural offspring, in this case primarily his sons but perhaps also neighbors.. Direct bias, on the other hand, is less innovative: the person does not invent a new response, but adopts one of the various options she has inherited from various cultural parents. However, a certain predisposition may favor – directly bias – one kind of cultural alternative over the others. In the cases of both guided variation and direct bias, criteria are needed to make individual judgments. And these criteria, B & R argue, must come from our genes, i.e. from biological natural selection. For this reason, they refer to these two forces as socio-biological. That is, cultural inheritance will track and reinforce biological inheritance.
With the other forces – which, B & R argue, are likely to be more important than guided variation or direct bias - this is not necessarily the case. The socio-biological forces depend on individual learning or discrimination: even in the case of direct bias, the individual has to make judgments about the available options, which bias to apply, and how to do so. But gathering such information has costs, which opens the door to other, less costly “forces” affecting social learning. Two of these are “indirect bias” and “frequency bias.” With the first, one individual identifies another whom he deems successful – an older brother, a village headman, a movie star - and copies many behaviors from him. Overall, this process is less costly because the first individual is not trying to assess which behaviors of the cultural parent have caused the latter’s success; he simply copies many or all of them. However, in some instances, B & R argue, costly “runaway” processes can ensue: people go to great lengths to dress like rock stars they admire, efforts that could never be justified in terms of clothes’ evolutionary selective power. The process is akin to the evolution of the peacock’s tail, in which an arms race over sexual attraction may impair the creatures’ survival. Frequency bias means that people simply copy the most frequent cultural variant, which will often prove to be a simple, efficient strategy.
A final force is natural selection, not of genes, in this case, but of cultural variants. This arises because genetic and cultural evolution are asymmetrical. We inherit our genes from our mother and father and the same two individuals are often important for imbuing us with our ideas and behaviors. However, we often inherit cultural variants from many other sources as well (siblings, teachers, friends, religious leaders, public figures). These non-parental sources will become relatively more important as we age. If we only inherited culture from our parents, B & R argue, we might expect that those ideas and behaviors would track or conform to the biological impulses we inherited from them: for example, we would imbibe the idea that having large families is a good thing. However, the existence of asymmetrical strands of cultural inheritance means that ideas and values can spread that may run counter to our biological imperative (and hence to what our biological parents on their own would teach us). Thus, teachers and other professionals may spread the message that professional success - something they themselves have achieved, and which requires sacrifices of the time and energy necessary for physical reproduction – is of great value. A Darwinian competition would then ensue – between biological parents and teachers over whose ideas and values would spread faster. B& R make a convincing case that this kind of asymmetric inheritance and the resulting natural selection of cultural variants probably lie at the root of the current, extraordinary demographic revolution. People, especially in affluent countries, are having fewer and fewer babies. Biology and biological Darwinism would predict just the opposite: as resources increase – as they have for humans over the last century or more, especially in industrialized countries – birth rates should steadily increase. In these cases, B & R say, the cultural variant “enjoy your own life, be successful professionally, don’t acquire these noisy, troublesome little creatures” has undermined the biological imperative to reproduce as much as possible.
Because of these final three forces – indirect bias, frequency bias, and natural selection of culture – cultural evolution will often come into tension with the dictates of biological evolution. They help to explain the internal conflicts that individuals experience much the way Freud described the struggles between id and superego. They also distinguish B & R’s approach from a strictly socio-biological one and from William Durham’s 1991 Coevolution, which foresees greater – though still not complete - congruence between biology and culture.
Finally, B & R ask how cultural evolution itself could have arisen in the first place? This is especially acute given cultural evolution’s frequent (biologically) maladaptive consequences. The generic answer is that as long as culture is overall biologically adaptive, its benefits outweigh its considerable costs. More specifically, culture may be expected to arise under particular environmental circumstances. If the environment remains constant for long periods, the best strategy is to hard-wire behavior in genes. This eliminates the costs associated with learning, either of the individual or social kind. If, on the other hand, the environment changes significantly quite frequently, then not only is genetic hardwiring the wrong strategy. Social learning, with its inertia, is as well. Under these circumstances, individual learning is the best option. Social learning – which allows for limited individual learning and variation – is best when the environment remains fairly constant but changes to some degree. In later work, B & R suggest that this was precisely the environment during the ice ages starting 2.5 million years ago and lasting until 12,000 years ago.
Culture and the Evolutionary Process is a challenging work. B& R rely frequently on mathematical models, which will not always be easy to follow unless one already has considerable facility with such methods. However, the authors always take the trouble to walk the reader through the main steps and, most important of all, the conclusions of the models. They also offer tangible examples from history and other social sciences to illustrate their points. The book should be required reading for anybody interested in “big” or “deep” history. Even at smaller time scales, the book offers a very stimulating framework for analysis, especially for thinking about broad patterns of social and cultural development. So, can historians learn anything from biologists? Yes – if they are willing.