Tuesday, 14 August 2012

A critique of Richard Dawkins’ The Greatest Show on Earth: The evidence for evolution

5. Smoke and Mirrors - trying to obscure the challenge of embryology

As I said previously, it's evident that Dawkins realises the complexity of embryological development presents an insuperable challenge to an evolutionary origin; but, being committed to the evolutionary cause, he tries various diversionary tactics to dumb down that complexity.


First he goes to considerable lengths to argue that the popular portrayal of DNA being a blueprint is wrong. For example he says that, although a house can be built from a blueprint and a blueprint drawn from the house, it's not possible to deduce the DNA sequence from the form of the host body. Of course no-one is suggesting that DNA is a graphical representation, even in coded form, of the developed body; and this is so obvious that one might wonder why Dawkins takes the trouble to make the point. His reason is that, because it's obvious a blueprint must have a designer, he seems to think that by showing that DNA is not a blueprint he is showing that there is no need for its having a designer. As if the only way the input of a designer might be inferred is if they leave some sort of blueprint - which is obviously not the case.

DNA may not be comparable with a blueprint, but certainly it contains the information for developing an organism; so, quite apart from how that information is encoded, the important question is, How did that information originate?


He then turns to starlings to try to support his premiss that biology does not have a designer. Starling flocks may appear to act as a whole - perhaps resembling a troupe of ballet dancers; but, he emphasises, in the case of starlings there is no choreographer - each bird is merely acting individually (not independently, as it responds to those nearby). And so he extrapolates from this that there is no director behind the embryological development of organisms - he argues that each cell is merely following local rules.

To try to reinforce his argument, he describes how the behaviour of birds in a flock can be modelled mathematically, just by building in local rules for each bird to follow. But actually this serves to reinforce the point that, even with something like a flock of birds (whose behaviour is so much simpler than that of cells in a developing body) at bottom it rests on rules. Rules which, even in the case of his simple model, must but be conceived and formulated, coded into a computer programme, and then fed into a complex machine that can apply the rules for all the individuals - each stage requiring intelligence.

I shall explain later why his flock analogy falls woefully short of embryological development anyway.


This is taken even further with his other ‘analogy for development’ - origami. Dawkins accepts that, even though not a blueprint, DNA does contain the instructions for growing an organism. And, because embryological development includes (among many other things) the folding of tissues, he compares this with origami - using the well-known construction of a junk as an example.

Even at a very superficial level, this analogy falls well short of illustrating his evolutionary premiss, because he would have to assume that an origami junk had arisen simply through random foldings of randomly shaped pieces of paper - which of course is totally wrong. On the contrary, an artisan will first have conceived of the end product and then devised a way of constructing it - with the desired object in mind as he does so.

But the really amazing thing about embryological development is the way in which all of the development takes place internally - there is no external agent doing the folding etc. Dawkins acknowledges the importance of self-assembly, but yet again thinks that by modelling the process with a computer he is explaining it away.

He wrongly says that the scientists in question have ‘deciphered’ the embryological process whereby tissues can fold (p229). They have not done this at all; all they have done is mimicked one aspect - and that in only two dimensions rather than three. But the key point, yet again, is that, even with this considerably simplified system, it requires formulation and programming of the rules, and then using a computer to implement them.

It’s all very well for Dawkins to argue that development proceeds through the implementation of local rules (and does not require a grand plan, but I will contest that next); but even the implantation of local rules requires considerable input of information. And none of his examples even attempts to explain whence or how that information is derived.

No comments: