Saturday, 27 October 2012

Genetic similarity theory and ethnic nepotism

In the 1980s, genetic similarity theory was proposed:

We present genetic similarity theory (GST), which incorporates the kin selection theory of altruism under a more general principle. GST states that a gene ensures its own survival by acting so as to bring about the reproduction of any organism in which copies of itself are to be found. Rather than behaving altruistically only toward kin, organisms are able to detect other genetically similar organisms and to exhibit favoritism and protective behavior toward these "strangers," as well as toward their own relatives. In order to pursue this general strategy, an organism must, in effect, be able to detect copies of its genes in other organisms. We order several data sets with this theory including (a) kin recognition studies in animals raised apart, (b) assortative mating, (c) intrafamilial relations, (d) human friendship and altruism, and (e) ethnic nepotism.
This theory was frowned upon by most kin selection enthusiasts. As Richard Dawkins put it:
As to distancing myself from the National Front, that I did! The National Front was saying something like this, "kin selection provides the basis for favoring your own race as distinct from other races, as a kind of generalization of favoring your own close family as opposed to other individuals." Kin selection doesn't do that! Kin selection favors nepotism towards your own immediate close family. It does not favor a generalization of nepotism towards millions of other people who happen to be the same color as you.
Some of the points in genetic similarity theory are good. It is true that kin recognition is based on perceived relatedness - and perceived relatedness can be quite variable. It can extend beyond blood relatives. Also, one brother may appear to be more closely related to a father than another one - and this may well affect how they are treated.

However, probably the key point of controversy is that, in kin selection, relatedness drops off relatively quickly as the degree of relatedness increases - 1/2 for a brother, 1/8 for a cousin, etc - while in genetic similarity theory, genetic similarity is portrayed as dropping off more slowly - covering entire ethnic groups, for example.

This is partly an empirical question. Kin selection's metric wins quite convincingly empirically. Yes, all humans are genetically similar, but in Price's formulation, they can be expected to behave slightly spitefully towards half of the population. However it should be acknowledged that there are some effects - such as ethnic nepotism - which do seem to require an expanded version of relatedness to explain. So: what is actually going on?

Let's list some of the possible explanations:

  1. Kin recognition mechanisms are mis-firing - producing the effect;
  2. Humans have ingroup/outgroup detection, which is largely independent of kin recognition, and which causes this effect;
  3. High-level selection (e.g. group selection or species selection) is responsible;
The first theory probably has some truth to it. The fellow from down the street could easily be your brother, while the stranger with the different racial background probably is not. Indeed, they would probably represent an unrelatedness superstimulus. However, the theory doesn't really explain why kin recognition mechanisms would produce an incorrect slope on the niceness/relatedness curve. The idea doesn't seem very plausible as a complete explanation of the phenomenon.

The second theory seems to have a lot going for it. Humans live in groups - and, in ancestral times, group members had bonds involving kinship and reciprocation, while non-group members were likely to screw you over and/or bash your head in with a rock. Not just because of kinship issues, but because they would be unlikely to face sanctions from the rest of the group for bad behaviour against group members. Members of other groups tend to look and act differently - largely because they carry different cultural symbionts. Projected forwards into modern society this "groupishness" could easily resemble ethnic nepotism and xenophobia. A similar idea probably applies to other social animals.

The third theory is probably wrong. Group selection is essentially the same thing as kin selection - and so its effects follow the same "relatedness" curve. Species selection is probably a relatively weak force. High-level selection seems likely to be too weak to explain this effect.

Genetic similarity theory is still alive and kicking in some areas - so it is important to be clear about what is wrong with it.

Kin selection is given as the primary explanation for ethnic nepotism on the Wikipedia page on the topic. However, we should not just consider kin selection based on DNA genes. Cultural kin selection is an important influence on humans. Memes often exaggerate racial division cues for their own ends. There's more to the topic than DNA genes.

Update 2014-11-10: Similarity selection seems like essentially the same thing. These days, I think a sympathetic reading of genetic similarity theory is the more useful one.

Tuesday, 2 October 2012

Neither kin selection nor group selection are much use in explaining altruism

In biology, altruism is typically defined rather like this (from Trivers, 1971):

Altruistic behaviour can be defined as behaviour that benefits another organism not closely related while apparently detrimental to the organism performing the behaviour, benefit and detriment being defined in terms of contribution to inclusive fitness.

Note the last section - about inclusive fitness. In systems whose evolution has been influenced by group selection or kin selection agents act to increase their inclusive fitness. They are not behaving atruistically.

It is possible to define altruism without measuring costs and benefits in terms of contribution to inclusive fitness - but such definitions are not of very much use. Inclusive fitness is the currency of evolution - its natural yardstick.

Reciprocity isn't much use in explaining altruism either. Though many insist on calling it "reciprocal altruism" it isn't a form of altruism at all.