Yesterday's post has me all worked up about evolutionary just-so stories, so I thought I would give a quick example today. My target here is not extrapolative evolutionary psychology as professional evolutionary psychologists practice it, but rather commonsense evolutionary psychology. I've met some very intelligent people who find commonsense armchair evolutionary psychology absolutely irresistible. I'm not sure if it's the snug fit between prejudice and scientific authority, or the ease with which such theorizing produces elegant solutions to complex problems, or what, but there is something about armchair evolutionary psychology that turns it into a mental tick for some people. Once you get going with this stuff, you can get to feeling that you never have to stop. But you do. You do have to stop. Now.
I know the temptation myself, to be honest. Raised as an atheist, I grew up pondering how to fit my own observations and intuitions about people into a Darwinian framework. Now, I'm still an atheist, and I'm still very much a fan of Darwin, but I've come to see that armchair evolutionary psychology is cheap trick. And because it's so widespread, and for some people so irresistible, I think it is vitally important that people come to see that they're not being clever, or scientific, or especially rational when they indulge in armchair evolutionary psychology. They're just being fucking wankers.
This is supposed to be a short post, because I have work to do, so I confess that if you're that kind of wanker, I won't be able to save your soul today. But I can get us started by talking about a just-so story, and explaining where it goes off the rails. It's important to see that I will not be disputing the theory of human nature that the just-so story purports to explain or confirm. My point is that the evolutionary story adds nothing in the way of justification to the theory.
Let's take an old favourite, xenophobia. The story goes like this: In the environment of evolutionary adaptation (the EEA), behaviour favouring kin over non-kin would have conferred some selective advantage. So considerations of optimality suggest hostile behaviour towards non-kin in the EEA. Now human behaviour(of this sort) is the product of human desires, beliefs, drives, etc. So we need to posit an underlying psychological mechanism to explain hostile behaviour to non-kin in the EEA. And the most obvious underlying psychological mechanism to produce hostile behaviour to non-kin in the EEA is . . . hostility to non-kin, or xenophobia. So evolutionary pressures would have favoured early humans whose psychological makeup was partly xenophobic. And that would explain a lot about contemporary humans too, wouldn't it? After all, modern humans are essentially gussied up early humans, and we might wonder if we carry to this day the same xenophobic psychology that carried us through thick and thin in an earlier age. And - lo and behold - if we turn on the television they're killing each other, yet again, in Yugoslavia, or Rwanda, or Iraq, or down the block, or wherever it happens to be. What better explanation for the xenophobia we do see in the world than that contemporary humans are acting on an innate drive? Sure, we can work against the drive, but you'll be working against primordial passions which are built deeply into the grain of human nature. Evolutionary theory suggests that this is part of the human condition.
Well perhaps. But if it's so, it can't be claimed on the basis of a just-so story like the one I very briefly sketched above. To see why, let's break it down into steps. Here's are the basic moves driving the story above:
1. Construct an optimality model predicting a certain type of behaviour in the EEA.
2. Posit an underlying psychological mechanism which would reliably produce the behaviour in the EEA.
3. Posit that the underlying psychological mechanism is a part of human nature, i.e. that it is innate and non-malleable.
4. By 3, extrapolate the lesson to contemporary humans.
There is a vigorous debate in the literature about optimality modeling, which I'm not going to pretend to be familiar with. The upshot seems to be that it's methodologically very tricky, trickier than you might think if you're an armchair theorist. I'm sure that many just-so stories never make it past #1, but let's let that point go for now. I'll just say that it's at this point that armchair theorists start speculating about sex without remembering that in the EEA women would be either pregnant or lactating for most of their lives (which really throws off your calculations, if you're not paying attention), etc. etc. etc. If you're going to do this, get help or you'll end up looking like an ass. It's harder than it looks.
Step #3 also gets into terrible trouble over unsophisticated confusions about innateness and malleability. But let's let those worries go too. Just let me point out that even if we know that some tendency or capacity is innate, we still know virtually nothing about how malleable it is. To borrow an example from a guy whose name escapes me at the moment, it's awfully natural for tigers to kill dogs. But if you raise a tiger in a zoo, so that it suckles at the teats of its dog surrogate-mommy from infancy, and it is socialized as a dog, then the last thing it's going to do is kill that bitch (or perhaps any dog - I'm not sure). And despite the unnaturalness of the arrangement, working against nature in this case doesn't produce any psychological strain on the tiger. So even if we knew that in the environment of the EEA, humans were naturally and innately xenophobic, this would tell us nothing about how malleable we ought to expect the tendency in contemporary humans to be. But, as I said, set aside this point too. I want to focus on the apparently obvious and natural move from #1 to #2.
The basic problem with the move from #1 to #2 is that it illicitly reads the evolutionary function of the behaviour described in #1 into the content of the beliefs and desires that we posit as the psychological mechanism underlying the behaviour - illicitly because there's absolutely no reason to do that.
What the hell do I mean? Here it is more slowly: We agree for the sake of argument that the evolutionary function of the behaviour is to provide a selective advantage by being part of a complex of behaviours that favour kin over non-kin. It's therefore vital that any satisfactory description of the behaviour in an evolutionary context is going to involve some statement about non-kin. But the beliefs and desires that produce that behaviour don't need to have anything to do with non-kin in order to reliably produce that behaviour. They might. But they certainly need not. To see this, consider the following alternatives:
a) A hatred of non-kin because they are non-kin.
b) A hatred of non-kin because they are unfamiliar.
c) A hatred of non-kin because they are potentially dangerous.
d) A hatred of non-kin because speak a different language.
e) A hatred of non-kin because they have different initiation rites.
f) A hatred of non-kin because they are outsiders.
g) A hatred of non-kin because . . .
. . . And we can do this all day. The move from #1 to #2 above assumes that a) provides the only possible description of the content of the beliefs and desires which we need to posit in order to explain xenophobic behaviour to non-kin. But b) through z) will do just as well in the EEA, because in the EEA non-kin will be unfamiliar, potentially dangerous, speaking a different language, outsiders with different initiation rites, and so on and so forth.
Now there are two very important points here. The first is that because of this we simply can't move from #1 to #2. So many different possible underlying psychological mechanisms might underwrite any given behaviour that we can't infer safely from the fact that our optimality model predicts a given behaviour to any particular psychological mechanism that would have to underwrite it. The inference only seems like a no-brainer because you're not thinking hard enough.
The second point is that although all these underlying psychological mechanisms would do an equally good job reliably producing xenophobic behaviour in the EEA, they would not do an equally good job producing xenophobic behaviour outside of the EEA. That is what makes the extrapolation to contemporary humans so dubious. Suppose that the underlying psychological mechanism is actually c). That is, in the EEA, strangers were an unknown quantity, and since it's much better to be safe than sorry, hatred towards potentially dangerous people is quite enough to produce xenophobic behaviour towards non-kin in the EEA. But once we leave the EEA, then it becomes clear why we're typically so comfortable with non-kin, and in a way that we couldn't possibly have been in the EEA: they're probably not dangerous. (And when they are, we get ready to fight back or run away.)
This is all very brief and schematic. I've got to leave it at that, though, or my mother will begin to worry about the progress of my dissertation (on a topic wholly unrelated to any of this). Just a few closing points: First, we could run through the same difficulties with some other old favourites, including evolutionary stories about sexual behaviour. I just don't have time. Second, it's been several years since I read much of the literature in this field, so I'm a bit rusty. If you go this far, you might really enjoy the name of the paper of Richard Boyd's (the Richard Boyd in the Cornell philosophy department) I recommended the other day, without being able to remember the title or citation. I'll post that when I get a chance.
Finally, I'll just admit that I find these discussion a bit upsetting. As a good friend of mine remarked the other day, people who indulge in this crap just don't seem to see how intellectually irresponsible it is. They don't seem to understand that, especially when it touches on gender issues, unwarranted and lazy speculation about evolutionary theory can affect people's lives, and shape their self-understanding in ways that are really unhealthy. This kind of lazy speculation is part of a culture that remains deeply sexist and cynical about human beings, even if some of the cruder elements have been swept away (or undercover) in the last generation or two. If the theories promoted in this way turn out to be true, so be it. I'll post them on the damn blog if they're true. But until you know them to be true, on the basis of something a bit more solid than armchair theorizing, then don't be such a fucking wanker, ok?
Posted by Chris at January 20, 2005 10:09 AMBravo. That's one of the best post I've read on here. It deserves a Cy Young, or a Koufax, or whatever award the blogosphere is yapping about these days..
Posted by: Paul at January 20, 2005 12:05 PMThe idea, as you say, quickly goes off the rails when kinship is identified with the cultural unit of the tribe, or, even worse, the state. From what we know about tribes, they often have exogamous marriage rituals -- and they often extend kinship in ways that depart from the rules of Hamiltonian rules of kinship selection that apply to ants. For instance, slaves can become part of the tribe. Now, you could come up with a just so story about this -- for instance, there might be an evolutionary advantage to increasing the chance of variation (in fact, this is what capitalism is all about) -- but you'd then face a conflict with the just so story about xenophobia.
However, instead of a totally determining just so story that crudely connected xenophobia to kin selection, one could devise a modification that would say: the unity of the group would be expressed most efficiently through some production of kin "signs." Hence, the idea of being a 'brother' in a tribe, or a 'sister'. And the political use of the trope of the family to preserve or enhance group feelings.
Even here, though, one has doubts. One of the things that kin selection suggested was that ants would develop some markers (an odor, for instance) that would express the chromosonal order of kin -- markers that would then be subject to selection pressure. I believe these markers have been discovered among some species of ants. But given that human beings are not haploid, the expression of such signs is unlikely, and the traces of kin selection among humans should be rather faint, if present at all.
Posted by: roger at January 20, 2005 12:19 PM