Kin selection and ethnic group selection

Sometimes I interrupt the normal day-by-day progression of Logarithmic History to cover my own work. Here I introduce a just-published paper, “Kin selection and ethnic group selection.” It’s about what, if anything, ethnicity has in common with kinship – evolutionarily speaking that is, on the assumption that human psychology has been shaped by natural selection. The paper doesn’t have anything to do with galaxy formation or nucleosynthesis, recent topics on the blog, but it would have been a good fit on August 5 last year, when I wrote about cultural group selection, population genetics, and prehistory, or December 15, when I wrote about nationalism in Europe at the end of the Cold War.

The paper itself is behind a paywall, but here’s a link to an earlier uncorrected, unpublished draft.

As a starting point, take the concept of ethnic nepotism. If you look up the term on the web, one thing you’ll find is an array of sources arguing that ethnicity is kinship on a large scale, and that the theory of kin selection, developed in evolutionary biology to explain altruism, cooperation, and conflict in families, is also a key to understanding such things at the level of ethnic groups. In the paper, I cite academic publications that take this position, including some from my late colleague at the University of Utah, Henry Harpending. And here is a non-academic link.

But you’ll also find people arguing the opposite, that ethnicity can’t be equated with kinship, at least as far as the theory of kin selection is concerned. Again I cite academic publications in the paper, and here, here, and here are some non-academic links.

The nay-sayers win the first round of the argument. I cover this in the first part of the paper. The theory of kin selection is concerned with r, the coefficient of relatedness, the expected number of genes that one organism shares with another as a result of common descent. Natural selection favors altruism between family members in proportion to their r’s, as a gene’s way of making more genes. So we’re told by William Hamilton, the biologist who figured this out. As it turns out, we can calculate r values not just for families, but for large groups – nations, continent-scale races. Does this mean we can plug these r’s into the standard formula and predict altruism between ethnic group members accordingly? No, because we’re now violating something called the weak selection assumption (see the paper for details). A physics analogy: at Earth’s surface, a falling object accelerates at a constant 9.8 meters per second per second. So we’re told by Galileo. This works for heavy objects over short distances. But we run into problems if we try to apply this law to lighter objects and longer distances without allowing for air resistance. Assuming weak selection in the theory of kin selection is like assuming no air resistance in physics, a simplifying assumption that can get us in trouble.

Eppur … even if ethnicity can’t simply be equated with kinship, it’s still theoretically possible to rescue the idea of ethnic nepotism, with the help of two further principles.

Socially enforced altruism. Suppose you decide, on your own, to help somebody at some cost to yourself. (If we’re thinking about evolution, we’ll want to count benefits and costs as fitness increments and decrements.) This is an instance of individual altruism. Discussions of kin selection commonly begin and often end here. But now imagine that you are part of a group that decides collectively to help another group. You and your fellow villagers, say, vote to tax yourselves to help a neighboring village recover from a flood; you don’t expect them to pay you back. This is socially enforced altruism. It’s not altruism at the individual level – you pay the tax to avoid a penalty – but it’s altruism at the village level – y’all could have kept the money for yourselves. In an earlier paper, I analyzed a variant on this, a reputation-based system where you help the needy not so much out of pure kindness, but to get the benefits that go with having a good reputation. I showed how the social enforcement of charity via reputation can amplify altruism toward distant kin. (Here’s the article, and a blog post about it, Beating Hamilton’s Rule, and an earlier article, Group nepotism and human kinship, and another post on the Brothers Karamazov Game, a simple three-person version of group nepotism.)

Ethnic group relatedness. The earlier paper was concerned with socially enforced altruism at the scale of local kin groups. Socially enforced altruism might also work at the level of ethnic groups. In this case, however, genetic similarity among segments of an ethnic group may reflect something other than just shared descent. In this case, two segments of an ethnic group may be genetically similar because they have shared a common culture for some time, resulting in similar selection pressures on genes contributing to the maintenance of that cultural regime. The basic principle behind kin selection can still operate here – you (or y’all; see above) help others because they share your genes, even if they can’t pay you back. But the expected number of shared genes – the ethnic coefficient of relatedness – no longer tracks the standard r’s based on genealogy or genetic similarity over the genome as a whole.

So ethnic group nepotism resulting from ethnic group selection* is a theoretical possibility, and I lay out the theory in the middle part of the paper. Whether it actually occurs I consider in the last part of the paper, which reviews some population genetics and political psychology.**

 

* Depending on how we define our terms, selection for socially enforced altruism may or may not count as group selection, but either way the usual objections to group selection for pure altruism don’t hold here.

** The social science literature on ethnicity and nationalism, including Conor, Gat, and Horowitz, is a topic for another day.