Co-Evolution is Real

A couple of years ago Derek Bickerton published his book, Adam's Tongue, on language origins. His main thrust was that humans had created their own niche and that many human traits were adaptations to an environment we ourselves had created. Since I had written that in a very early book of mine 30 years earlier, I was underwhelmed, but Bickerton has had an impact. These days I notice many glancing references to niche construction, as adaptation to one's own environmental changes is called. Britain's Royal Society, for example, has devoted a complete issue of one of its journals (Philosophical Transactions of the Royal Society B) to human niche construction. Of particular interest in that issue is a review paper by Herbert Gintis, which considers "Gene-culture coevolution and the nature of human sociality." Culture, Gintis writes, is a "special case of niche construction" [p. 878].

Niche construction is not particularly rare. Beavers build dams that change the environment and beavers themselves are adapted to living in the ponds their dams cause. Bees live in and are adapted to hives that they've built. Coral lives on reefs created by their ancestors. What's unusual about people is the immensity of the environment we have created and the fact that our special adaptations tend to be cultural rather than genetic.

Another difference between human and beaver niche construction is that besides adapting to our environmental changes, we adapt to cultural changes. For example, when one vowel sound changes we may have to alter other vowel sounds to maintain distinctions and ultimately the whole vowel system may switch as happened to English during the medieval Great Vowel Shift. Also, certain behaviors have cultural explanations. Some people wear crosses around their neck. The action is explained by another cultural reference, not by referring to a biological need. There can be very complex events to, such as the convergence of the Roman political structure with Christianity, leading to massive changes in both. Again, the changes are caused by culture.

This kind of culture is very different from ape social learning in which one troop makes tools one way and another troops makes them a different way. The explanations remain biological. They alter the niche a bit, if you want to understand the changes in ape culture, look to changes in the environment or in their genes.

Beavers, apes, and other animals can change their environment. Over generations their genes will adapt to those changes. That's the basis of niche construction. Gintis proposes something analogous in the co-evolution of genes and culture. Humans change their culture and over generations their genes adapt to those changes.

The surprising thing to me was that Gintis feels it necessary to insist that such a co-evolution is possible. It is such a standard part of language origin theory that I had forgotten the idea could be controversial. The objection is that genes and the environment are physical things, but culture is a bit amorphous. This kind of fundamentalist materialism is always a bit frustrating. Behavior is as real as a rock and should be able to have just as sharp an impact on natural selection, but my experience has been that every kind of fundamentalist—political, philosophical, religious, or constitutional—is hard to reason with.

Gintis maps out the gene-culture co-evolutionary process. He shows gene, cultural, and environmental units. Each of these can change. Genes and cultural units can change in response to each other's changes. Thus, they can co-evolve, i.e. adapt to each other. The point is that our languages, religions, literatures, laws, and kin systems are at least partially independent of both our genes and our environments. Gintis quotes an evolutionary psychologist, David Buss, who argues, "Culture is not an autonomous causal process in competition with biology for explanatory power." [quoted on p 880]. Gintis then rebuts this in an astonishing way. He has a simple mathematical argument. He reduces his map of the co-evolutionary process to three functions and derives an equation that combines all the functions. Gintis shows that if, as Boss says, the explanatory value of culture for biological changes is 0, it must also be 0 in explaining cultural change. But this conclusion is absurd ("obviously not appropriate," is Gintis's delicate usage) because obviously cultural changes in humans do have cultural explanations.

I was quite surprised to find a mathematical refutation of such a theoretical social argument and even more surprised that I was able to follow it. Thank heavens I once wrote a book about Einstein. The point is that co-evolution is possible and genes can adapt to cultural forces just as culture adapts to genetic conditions.

Gintis does not discuss this idea in terms of language origins, but it is important to any accounts of language evolution. Co-evolution means that there is no reason to assume that human biological evolution—i.e., the transition from Homo habilis to erectus on down to sapiens—reflects only human genetic changes. The transition from one Homo species to another could indeed reflect genetic adaptations to culture. Thus, we are not confined to scenarios in which species H supports one culture and then evolves for environmental and genetic reasons to species I, which is able to support a more complex culture. If you really embrace the co-evolutionary idea, the transition from species H to I can be explained perhaps in very large part in terms of cultural pressures.

This idea is not going to surprise anybody who regularly reads this blog, or anyone who focuses on language origins. Terrence Deacon published this argument in 1997. Derek Bickerton, as I said, has come to the same conclusion, via niche construction. Tecumseh Fitch's recent book on The Evolution of Language also embraces co-evolution of culture and genes. But there are many scholars who are interested in human origins while paying only cursory attention to language origins. These include evolutionary psychologists, anthropologists, archaeologists, and paleontologists, and Gitnis is writing for this group.

Gintis's argument also clarifies for me why there is a school that thinks language is recent, despite the extensive biological evidence of long term shaping of the body to support speech. If you believe that culture can have no impact on our genes, then all that physical evidence of biological adaptation to speech is for naught. The bodily changes must have some explanation other than cultural ones. We cannot expect philosophical fundamentalists to buy into Gintis's argument for co-evolution, even in its intimidating mathematical form. At best, it will be viewed as a paradox, like Zeno's proofs that change is impossible. The logic looks good, but something must be wrong.

Most of Gintis's paper is devoted to showing the reality of social behavior, what this blog consistently calls communal behavior, to distinguish it from ape social behavior. Once again the conclusions he presents are not going to be news to readers of this blog, except perhaps for the form of the argument, which relies heavily on game theory. It was interesting to me because I am familiar with game theory chiefly as a way of arguing against the reality of community. Gintis describes many experiments showing that people will consistently behave in ways that promote community and cooperation rather than maximize their own individual gain.

One important difference between Gintis as this blog is that he believes altruism emerges from culture, whereas this blog has reported strong evidence that altruism preceded human culture and made it possible. He is responding to people who argue that "other-regarding and moral behavior in humans … [are] the results of misdirected attempts at maximizing long-term self-interest," [883] and he does not address the arguments based on multilevel selection, the speech triangle, and the effects of cooperation on brain size.

What we do see in this paper is that there is more than one path leading to positions that this blog has come to by deeply considering questions about speech origins. As the machine showed this week on Jeopardy!, when many paths lead to the same answer, the odds of the answer being correct improve.

Comments

[Alex Moore]

As someone who doesn't read about things like gene-culture coevolution much, I really enjoyed the Gintis paper. In particular, I found the distinction between genetics for encoding things that are constant and stable in the environment, learning capacity for environmental conditions that vary rapidly, and culture for things in between.

As I thought about it, though, it did make me wonder about the nature of gene-culture coevolution. In particular, I understand that genes and culture are sort of feeding back onto each other. What I don't understand is how this works considering that biological evolution seems to be such a slow process in comparison to cultural evolution (perhaps my assumption is wrong?). Perhaps it's that biological selection is only affected by aspects of our culture (as well as the environment it creates) that are long lasting and stable like agriculture and literature? Or am I looking at this the wrong way?

At any rate, thanks for introducing me to this very interesting concept. I'll be sure to keep reading this blog!
-----------------------------------
BLOGGER: Culture does evolve more quickly than biology, but biology can move with some speed too. (See the spread of lactose tolerant genes within historic times.) An example of culture-gene coevolution related to language is the control we have over our tongue, making very precise sounds possible. No other primate has such a complex system. By the way, on this blog I put a lot of stress on group selection. Meaning that control of the tongue, giving a wider range of possible sounds (and therefore words) was selected because of the advantage it gave a group able to use a richer language, not an advantage to an individual whose ability to outpace other speakers in the community.