Saturday, June 23, 2012

Pinker on Group Selection

Steven Pinker’s attack on group selection is getting some blog cred. It caught my attention mainly because one of his targets is the kind of group selection Jonathan Haidt makes a featured player in his Righteous Minds. I should note that Pinker is not just the aging pretty boy front man for evolutionary psychology, but an exceptionally lucid, penetrating, and witty writer on language and thought. His critique should be taken very seriously.

He has divided that critique into three parts, the first of which I have no argument with. The second, which he calls 2. Group selection as an explanation of the traits of individuals is more to the point that interests me. Here Pinker hangs his entire argument on one sentence:

It's only when humans display traits that are disadvantageous to themselves while benefiting their group that group selection might have something to add.

Well you can win any argument if you get to re-define enough of the terms, but I don’t think this is the crux of the group selection argument. The point, not mine, but Darwin’s, is that traits which would in isolation appear detrimental, may not be in the context of other traits. Once again, as Darwin noted, the key associated trait is ability to punish defectors, that is, those who fail to appropriately reciprocate altruistic acts. That is a very non-trivial ability, since it requires being able to remember who did what and also to understand something of the intentions of others.

Arguments about definitions are infamously fruitless, but let that not deter me. Pinker has picked a fight with Jonathan Haidt and E O Wilson, so it might make sense if he used the same definition that they (and Charles Darwin) used. My understanding is that that group selection exists if some or all of the selection takes place at the level of the group – that is, the members of the group survive or perish based on what group they are members of.

To continue, Pinker thinks the following is an argument for his cause:

In tribal warfare among non-state societies, men do not regularly take on high lethal risks for the good of the group. Their pitched battles are noisy spectacles with few casualties, while the real combat is done in sneaky raids and ambushes in which the attackers assume the minimum risks to themselves.[14] When attacks do involve lethal risks, men are apt to desert, stay in the rear, and find excuses to avoid fighting, unless they are mercilessly shamed or physically punished for such cowardice (see, for example, the recent meticulous study of Turkana warfare by Sarah Mathew and Robert Boyd).[15]

Actually, though, this precisely fits the Darwin and Haidt version of enforced group loyalty via punishment and shaming.

Pinker has a lot more to say on similar themes, but the core of his argument remains a simple redefinition of “group selection.” Such arguments are rarely profitable, but there remains one issue: does the notion of multi-level selection add to our understanding of what’s going on?

Pinker’s conclusion:

None of this prevents us from seeking to understand the evolution of social and moral intuitions, nor the dynamics of populations and networks which turn individual psychology into large-scale societal and historical phenomena. It's just that the notion of "group selection" is far more likely to confuse than to enlighten—especially as we try to understand the ideas and institutions that human cognition has devised to make up for the shortcomings of our evolved adaptations to group living.

I would accuse Pinker of being the chief agent of confusion here, by conflating a bunch of unrelated notions with the Darwin and Haidt notions of “group selection.” More on that later.