I’ve been reading Christopher Peacocke’s “The Realm of Reason”:http://www.amazon.com/exec/obidos/ASIN/0199270724/caoineorg-20?creative=125581&camp=2321&link_code=as1, and I was rather struck by one of the moves in it. Unless I’ve really badly misinterpreted what he says in Chapter 3, he thinks you can come to justifiably believe in, and perhaps even know the truth of, theories of natural selection by looking really hard at a kitchen table and reflecting on what you’re doing.

First a little background. Peacocke is a rationalist and a foundationalist. He thinks that whenever you have a justified belief, its justification can be grounded in a chain of rationally acceptable transitions which bottom out in either conscious states or a priori knowable truths. Moreover, the transitions have to be (a) justifiable a priori and (b) have their justified status explained in part in virtue of the contents of the states involved in the transition. There are many problems for this position, the salient one being how do we ever get to know that the external world is anything like how it looks, given such meagre stuff to work with.

Descartes famously appealed to God at just this point. We can tell a priori (via the ontological argument) that a benevolent God exists, and He would not let us be massively deceived. Hence we are a priori justified in going from “That looks flat” to “That is flat”.

In keeping with the spirit of the age, Peacocke replaces God with Darwin. We can justify the transition because we’re probably evolved by natural selection and that means (probably) that we were selected for having accurate representations at least of certain fundamental things. Now as it stands this all seems pretty reasonable, but remember Peacocke is a rationalist. The justification has to work a priori. The mere fact that we’re evolved won’t do it. This is where the tables come in.

The table I’m looking at now has straight edges and round corners. Or at least so it looks to me. That is, I represent straightness when I look at its edges, and roundness when I look at its corners. What explains how I could be the kind of creature that has the capacity for spatial representation? This is a surprising, as Peacocke says Complex, fact and it cries out for explanation. Some say God’s handiwork explains my representational capacities. Others say a mad scientist. But the simplest explanation, says Peacocke. is that I’m the product of a process of natural selection. Unlike the other explanations, this does not need to appeal to representational capacities to explain representation. So _a priori_ we can tell the best explanation of representational capacities is natural selection. And since accurate representations are selected for, we probably get basic things right most of the time.

This is pretty ingenious, but there are at least three things you could say against the argument.

First, it isn’t clear in just what sense natural selection is a particularly simple explanation of the existence of representation. The amount of complex interactions needed to generate a selective process is rather staggering compared to what you need to mirror nature.

Second, inferences to the best explanation should consider all possible explanations, not just the ones that are current in the philosophical literature. And I don’t see why natural selection does best among all possible explanations at explaining the existence of representations. For instance, the hypothesis of a young earth created by a miracle with creatures with representative capacity seems to avoid some of the messy details of evolutionary theory.

Third, no practicing biologist would seriously consider arguing for natural selection on the basis of careful reflections about tables. If they did, there would be a lot more to complaints that natural selection isn’t better supported than creationist fables and hence doesn’t deserve pride of place in schools. Maybe we shouldn’t be completely deferential to scientists, but this kind of consideration seems to have some force to me.

I have quite a bit of sympathy for Peacocke’s overall rationalist program, but this part of it (and it’s a big part) really needs repair and/or replacement.