Genes, Mind, and Culture - The Coevolutionary Process: 25th Anniversary Edition Anniversary Edition

Others may have coined a more popular catchword; meme rather than culturgen; however, Wilson acknowledged that the concepts were the same.

Human prehistory is difficult to understand without understanding the coevolutionary prosesses. A very good quide to
think culture as whole.

Genes, Mind and Culture was one of the first books published by scientists on the topic of cultural evolution. It came out about five years after Richard Dawkins had published The Selfish Gene - in 1976.

The book is pretty dated, and most people buying it will probably be doing so in order to get a historical perspective on the topic.

The book has over 400 pages. These are pretty densely packed with mathematical models, which render much of the book pretty unreadable by most of its audience. The models are not well presented. In some cases they may be acting partly to create a veneer of respectability.

Lumsden and Wilson make the case for a scientific study of culture based on biology. The book introduced the concept of a "culturegen". They define this as follows:

"A culturegen is a relatively homogeneous set of artifacts, behaviours or mentifacts (mental constructs having little or no direct correspondence with reality) that either share without exception one or more attribute states seelcted for their functional importance or at least share a consistently recurrent range of such attribute states within a polythetic set."

From this you might begin to detect something of Lumsden and Wilson's style. While their own definition of a culturegen is awful, readers can understand its usage in the rest of this review by considering it to refer to memes or meme products - in the form of socially transmitted behaviours or artefacts.

The book then focused heavily on the epigenetic rules by which genes influenced what culturegens were adopted by their hosts. Like modern evolutionary psychologists, the authors were interested in the factors that make cultures similar everywhere. However, they did try to go beyond these commonalities and account for cultural differences. Their approach is based largely on population genetics.

One thing their book became known for is its "leash" metaphor. They write on page 13:

genetic natural selection operates in such a way as to keep culture on a leash.

In a subsequent section titled "Can culture have a life of its own?" they claimed that the establishment of deleterious culturgens in the population for extended periods of time could be demonstrated to be impossible - suggesting that epigenetic rules favouring the adoption of beneficial culturgens would be violated and that they would exert some kind of pressure which would alter the culturgens into a more favourable form. This idea was subsequently identified by some opponents as pinpointing where Lumsden and Wilson had gone wrong in their analysis. The authors also wrote: "Culture slows the rate of genetic evolution". We now know that this is not correct either.

In one of the best parts of the book, the authors offer an analogy for understanding cultural evolution based on island biogeography. In this analogy, islands represent human minds and archipelagos represent societies. Culturgens act like organisms colonising the islands. The reader is thus invited to transfer their knowledge of the dynamics of island biogeography into the cultural realm. This idea is an excellent one - although picturing brains as islands makes them seem rather passive and picturing ideas as colonising organisms makes them seem perhaps too active and agent-like.

Island biogeography might seem as though it is an esoteric subject - but many students of evolutionary biology pick up a smattering of knowledge about the topic as part of the process of learning about evolution. Islands represent natural evolutionary experiments, and so are of particular interest to evolutionary theorists. Darwin's famous visit to the Galápagos islands has also helped to put island biogeography in the limelight.

However, although they do have a chapter devoted to it, Lumsden and Wilson don't do very much with this (excellent) analogy. Had they based more of their work on it, their book might have been a lot better.

Part of the book's problem involves failure to apply the principles of reductionism. Lumsden and Wilson obsessively pursue the idea of a gene-cultural cycle - corresponding to a series of cycles of ontogenetic development followed by acculturation. Only by looking at this complete cycle can the whole process be understood, Lumsden and Wilson apparently believed. Most others divided cultural evolution from organic evolution and treated these as two separate but partly-interacting processes. Because Lumsden and Wilson don't divide the topic up in this way, they get rather bogged down with the enormity of trying to understand everything all at once. The result is that they make relativelty little progress in actually understanding how cultures evolve.

Another way of looking at the book is as an attempt to shore up sociobiology against its critics. Sociobiology seemingly tried to explain everytrhing in terms of genes. Of course, that approach doesn't work too well for culture, which is not inherited via DNA genes - and culture is an important determinant of behaviour. So: sociobiology needed fixing, by applying a patch to deal with culture. However, the authors attempted to apply standard sociobiology strategies to the topic - by tracing how everything was affected by genes. While this approach is not a totally unreasonable one, it seems like a rather biased research strategy. Instead of looking at culture and considering how best to explain it, the researchers used their existing sociobiology toolkit and attempted to apply it to culture. While it is perfectly possible to ask after the basis of social learning in DNA genes, it turns out that there's another highly-productive approach to studying how culture evolves = which involves considering culture as a partially independent instance of an evolutionary process, following the rules of universal Darwinism. Lumsden and Wilson totally missed this approach - apparently through their eagerness to apply their existing sociobiology toolkit and look at the genetic basis of cultural phenomena. Much the same approach was used by Cosmides and Tooby a decade later, with much the same messed-up result.

These problems put Lumsden and Wilson's work off the main path that lead to the modern understanding of how cultures undergo Darwinian evolution and coevolve with human genes. The models Lumsden and Wilson presented did not do much useful work. Most subsequent authors have not built significantly on their efforts - and those that did mostly went off the rails in a similar way.

This review will stop here. If you want a more in-depth review than this one, the late John Maynard-Smith wrote a good review of this book in 1982, which was republished in his book "Did Darwin Get It Right?" You can probably find his review online.

It is difficult to decide whether to praise this book for its (at the time) innovative and novel approach to gene-culture coevolution or criticize it for its endless slurry of ad hoc models with groundless or unspecified assumptions. One can do both. The idea they present, that there is a positive feedback mechanism between biological and cultural evolution, is by far the best working hypothesis for why human society "took off" after millenea of paleolithic stasis. The theory central to the book is that genetic constraints shape culture, which in turn becomes the cultural environment in which an individual's Darwinian fitness is determined, forming a positive feedback between cultural and evolutionary change. This posits a specific mechanism for the role of genetic change in cultural evolution, going far beyond the intellectually vacuous "resolution" of the nature-nurture debate by those who say "it's both." However, none of the models they present can be regarded as anything but mathematical playthings, in few cases are any of the variables or parameters quantities that can actually be measured or therefore tested. Often, it is not entirely clear what the dependent variables in the system correspond to in nature. Worse still, some of the models are completely ad hoc: first positing a dynamical behavior, then presenting an apparently arbitrary dynamical system which exhibits the property as "proof" of the theory. In essence, an uneven work, but one which I think will be at the foundation for further work in the area, at least as a basis for concepts and theory (provided the specifics are taken with a large grain of salt).