thinking makes it so

There is grandeur in this view of life…

The one about the English fox and the Australian rabbit

leave a comment »

Red fox

Red fox

Thank you Jerry Fodor and Massimo Piattelli-Palmarini (F&P-P) for keeping Australian rabbits safe from English foxes:

The number of rabbits in Australia is unaffected by the number of foxes in England. That’s because the predations of the one on the other are all merely counterfactual, and possible-but-not-actual events do not exert selection pressures.

[Follows Just-as story as third in a series on Jerry Fodor and Massimo Piattelli-Palmarini’s What Darwin Got Wrong, which began with Smear campaign.]

Of course counterfactual events cannot exert selection pressures. Of course English foxes don’t kill Australian rabbits – unless they emigrate.

It makes me wonder if F&P-P have based their critique of the theory of natural selection on an obstinate refusal to understand it.

We need to be careful about the role we want to allow counterfactuals to play.

If counterfactuals are disallowed in biology then they should be disallowed in other sciences. And in that case John Stuart Mill’s ‘method of difference’ would be illegitimate. Scientific theory – in most if not all sciences – is based on counterfactuals. We accept explanations of what is actually going on in the physical world based on a particular scientific theory, but that physical world is by definition not counterfactual. Or, as Wittgenstein put it: The world is everything that is the case. [Ludwig Wittgenstein, Tractatus Logico-Philosophicus, 1922]

F&P-P seem to be confusing a theory about what actually happens in the world with our methodological principles about what we are entitled to say (and/or entitled to say we know) about what actually happens in the world. For some reason they think they can impose stricter constraints on biological explanations than on other types of scientific explanations. But their justification for this is not obvious.

Metallic sodium in oil

Metallic sodium in oil

Consider the following statement in physical chemistry:

Sodium is a highly reactive metal.

If we restrict ourselves to what actually exists in ‘nature’ we might doubt this. Sodium is a component of salt, but we wouldn’t call salt particularly reactive. Why would we even call sodium a metal? Salt isn’t shiny and malleable for example.

Ah, but just because we cannot see sodium in its elemental state in ‘nature’ doesn’t mean we can’t manufacture pure sodium electrolytically, at high temperatures. Then we would see how reactive it is.

There was no doubt a time in the history of our planet when the electrolytic manufacture of sodium was practically impossible. So any statement about the reactivity of pure sodium would have been counterfactual.

Now take the pumping and thumping heart, discussed in Just-as story and also previously in Not thumping but pumping.

It is theoretically possible to manipulate the genome of (say) a mammal to create a phenotype possessing a heart which pumps but does not thump. Or if that kind of intervention by a designing mind offends F&P-P’s methodological sensitivities, then try the alternative thought experiment of waiting however many millions of years for a mammal to evolve with a heart which pumps but doesn’t thump. We could then (theoretically) see if the mammal survives as well as the mammal with the heart which thumps as well as pumps. In theory we could also wait for the mammal to evolve with a heart which thumps but doesn’t pump, and then assess its survival prospects too.

I cannot see any significant difference between this case and the case of sodium prior to its isolation as an element. Anyone who objects that the mammal with a heart which thumps but doesn’t pump will never evolve in nature would have to explain why they are so sure, if the distinction between what is selected for and what is a free rider is so incoherent.

Let’s dig a bit more to see why F&P-P might have thought there was a difference between the case of sodium and the case of the heart.

F&P-P characterise the theory of natural selection as saying that natural selection causes the existence of specific features in organisms. They say the theory claims that ‘selection for X’ (eg selection for a pumping heart) causes the existence of X (a pumping heart) in a particular organism. They reject this, because you can only distinguish X from Y and Z (where Y and Z are free-riding associates of X, like thumping) by introducing counterfactuals. ‘Selection for X’ therefore has a counterfactual component which cannot be analysed away. And something which is counterfactual cannot exert a selection pressure, and cannot cause anything.

But F&P-P seem to have put causation in the wrong place.

Downs Cell for producing sodium

Downs Cell for producing sodium

In the case of sodium the counterfactual event was the construction of technology to perform electrolytic separation. In the case of the heart the counterfactuals were the mammal with a heart which pumps without thumping and the mammal with a heart which thumps without pumping. The electrolytic separation technology was part of how we come to know that elemental sodium is reactive. The survival prospects of the mammal with a heart which pumps without thumping (versus the mammal with a heart which thumps without pumping and the mammal with a heart which both thumps and pumps) is how we come to know that it was the heart’s pumping, not its thumping, which was selected for.

Gould and Lewontin (see Just-as story) were right to advise caution when speculating on the adaptational origin of biological traits. But that doesn’t mean every adaptational explanation is unsound. It’s like extrapolating from advice about the design of controlled experiments to the conclusion that all experimental results are invalid.

What does it mean to say that natural selection causes the existence of specific biological traits? We need to be careful about where we locate the causation, and what we mean by ‘natural selection’.

Natural selection is not a thing, or even a set of forces. ‘Natural selection’ is a label for the differential survival of populations of living things, and the consequences of that differential survival. What actually causes the biological trait (eg the heart which pumps and thumps) is a complex process of catalysed embryogenesis starting with genetic material in the fertilised ovum. What caused the fertilised ovum to have that genetic material is that it was inherited from the ovum’s parents.

A necessary condition of that ovum’s existence is that its parents survived long enough to reproduce. If (for example) the ovum’s parents’ genetic material had been such as to produce a heart with a slightly defective pumping action, then in the struggle for survival those (potential) parents might have fallen by the wayside long before sexual maturity. But if, on the other hand, one of the parents’ genetic material had mutated slightly so as to cause the offspring to have a slightly more efficient heart, then in the struggle for survival the offspring would have been more likely to survive and reproduce, other things being equal. Natural selection is the generic name for that ‘survival filter’ – but what makes up the filter is the aggregate populations of competing individuals in the context of their natural environment. The filter does not itself cause the relatively more efficient pumping heart, but it filters out the less efficient ones.

Another reason for saying natural selection cannot be the cause of biological traits is the tautology at the heart of the theory. The individuals which survive and replicate are the individuals which survive and replicate. What causes some individuals to survive and replicate are relative advantages which matter in a competitive context. Some of these are obvious, some less obvious, and some are very obscure. But the concept of competitive advantage itself is completely transparent.

Black AND white rabbit

Black AND white rabbit

Take two rabbits. One is Australian, white and sterile and the other is English, black and fertile. The Australian rabbit has no offspring while the English rabbit has many. Is the difference due to their nationality, their colour or the fact that one is sterile and the other is fertile?

Change the example slightly. The Australian white rabbit is genetically incapable of having more than one offspring per litter. The English black rabbit is genetically capable of having litters of six or more. After six generations there are more English black rabbits than Australian white rabbits. Can we explain this or is it a complete mystery? Does anyone think that nationality or colour might be the explanation? Are we not allowed to say that ‘having litters of six or more’ has been selected for, and that Englishness and blackness are free-riders?

Change the example again. This time both rabbits are genetically capable of having normal sized litters. But the English black rabbit has a genetic disorder which reduces its digestion efficiency by 50% compared to the Australian white rabbit. After six generations there are more Australian white rabbits than English black rabbits. Are we not allowed to say that ‘normal digestion efficiency’ has been selected for and whiteness and Australian nationality are free-riders?

In its essentials the theory of natural selection is extremely simple. If F&P-P are right that the theory of natural selection is incoherent then its incoherence should be evident in extremely simple illustrations like these. But F&P-P have got it breathtakingly wrong.

© Chris Lawrence 2011.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: