Watching the wrong race: a response to Paul Nesselroade

In his article 'Betting on all the horses', Paul Nesselroade (2003) suggests that Darwinian explanations are sometimes weak because Darwinists interpret whatever evidence they come across as support for their theory.

This article is a response to Nesselroade. I will argue that his argument fails because he misunderstands what Darwinism is about.

The thrust of Nesselroade’s argument shifts somewhat throughout his article. At times, he seems content to point out the psychological fact that people tend to interpret any evidence they come across as supportive of their existing beliefs, using the evidence rather “as drunks use lampposts – more for support than illumination”. Now, if Nesselroade intended his article simply as an epistle against sloppy thinking, I would have no complaint. But at other times, he makes specific claims about Darwinian theory – arguing not just that the proponents of this theory sometimes succumb to sloppy thinking, but rather that the theory itself is flawed. At any given point in his article, it can be hard to work out whether Nesselroade is making the weaker claim, that Darwinists are as guilty of sloppy thinking as the rest of us, or the stronger one, that Darwinian theory itself is weak because many of its claims are unfalsifiable – i.e. formulated to be compatible with any conceivable evidence.

What is clear, however, is that Nesselroade means to make the stronger claim at least sometimes, as when he writes:

I will concentrate on this stronger claim.

Nesselroade gives two examples to support his argument. The first refers to “biological order”:

And the second refers to the observed variation in phenotypic features:

If his diagnosis were correct, Nesselroade would be drawing our attention to a severe problem for Darwinism. His basic principle, that any theory which bets on all the horses isn’t worth bothering with, is sound.

But, thankfully for Darwinists, Nesselroade misunderstands what Darwinian theory actually says. The theory doesn’t make any direct predictions about phenotypes; it simply talks about the processes that produce phenotypes. It’s the interaction between the theory and historical variables that gives rise to the phenotypic variation we see.

In other words, like most scientific theories, Darwinism is an if-then theory: if a certain set of selection pressures happen to apply in a given historical context for a given amount of time, then a certain phenotype will result. When it comes to explaining how a particular species has turned out, Darwinism only tells half the story. The other half is the particular set of contingencies that the species has encountered over its long evolutionary history.

So it isn’t a problem that Darwinism is compatible with both exquisite design and kludgy design, or with both bright plumage and subtle camouflage. The theory is neutral between these phenotypic outcomes. It simply says that either outcome is possible, and the one which happens to obtain will depend on the contingent evolutionary history of the organism in question. Given a range of organisms with a range of evolutionary histories, we’d expect a range of different phenotypes: some exquisite, some kludgy; some vivid, some camouflaged.

So is Darwinism is compatible with every logically possible phenotype? Nesselroade suggests that it is when he writes, “Is there any doubt, then, why you never hear a Darwinian say, ‘Now if I saw that, I’d know it was designed’?” (and here he’s using ‘designed’ to denote non-Darwinian origins).

But this is just plain wrong. Darwinism predicts that a given phenotype will only arise if its evolutionary history includes the appropriate selection pressures, so any phenotype which goes against this prediction would be straightforward evidence against Darwinism. Such counter-evidence could take two forms:

1. Unexpected features: an observed phenotypic feature leads us to postulate a selection pressure that it turns out never actually existed.

2. Unaddressed problems: a known selection pressure leads us to postulate a feature that doesn’t actually exist.

In other words, if a combination of Darwinian theory and the specific historical variables of an organism’s evolutionary history were to predict an outcome other than the phenotype which we actually observe, then we know that Darwinian theory is wrong (or, of course, our understanding of the evolutionary history is wrong, or both).

Of course, much of biology isn’t yet rigorously mathematical, so biologists can’t make precise quantitative predictions about outcomes, but they can make distinguishable qualitative predictions and test them. The fact that we haven’t yet observed any outcomes incompatible with Darwinism is good confirmation for the theory.

What science in general, and Darwinism in particular, try to do is come up with generalisations about the causal interaction of variables. So scientific theories are necessarily if-then in structure: given specific values of conditions x and y, we can predict specific outcome z; given ranges of values of conditions x and y, we can predict the range of possible outcomes z. When variables vary, as they’re wont to do, so do outcomes. A theory is falsified not simply by pointing to a particular outcome, but by showing how a particular outcome does not follow from the combination of theory and variables.

Nesselroade misses the point when he claims that Darwinism bets on all the horses. He’s been watching the wrong race.

 

 

Portfolio | Academic | Misc | Web design