Saturday, January 23, 2010

The Predictive Power of Evolution

Some years ago, I recall part of the creationist argument against evolution as a valid scientific theory was:
It makes no testable predictions.
This was said, of course, to blunt similar criticism of creationism as a theory. Observations that don't match up with the creationist story are waived away with divine intervention. Evolutionary theory, in this view, is no better because terms like "survival of the fittest" are non-predictive (how do you tell is fittest? It survives!). Your theory, they say, is no better than ours.

Except of course that's nonsense and a recent Nova episode, What Darwin Never Knew, did a masterly job of proving this. I cite but one example.

When Darwin wrote On The Origin of Species the mechanism of inherited traits was not understood at all. While the basic concepts were there and some of the ways that dominant and recessive traits were passed on had been observed, molecular genetics was a century in the future.

We now know that cells contain DNA and that this controls how living things develop. We are now able to sequence DNA reasonably rapidly, and have sequenced the DNA of multiple species. And they're finding that many of the key changes in DNA between species are not in the protein-coding genes. There are other genes that are "switches", which turn these genes on and off. The "switches" are further controlled by "body-plan" genes, which control when and how long switches are thrown. And a minor set of changes to the "body-plan" genes can be responsible for big changes in characteristics.

This is entirely consistent with evolutionary theory. If a change to, say, arm length or brain size required the coordinated changes of hundreds of genes (or indeed, creating entirely new genetic information), then the probability of such a change occurring, being successful, and being transmitted to a new generation is vanishingly small. Yet now it can be seen that these changes can be caused by a very small set of changes, which is far more likely.

Did it have to be that way? In a creationist world, we could fully imagine that every species could have had unique genes, totally unlike any other. One would not expect that, say, the genes that control the size, shape, and placement of a fin in a fish would be the same ones that control the size, shape, and placement of a chimp's arm. And yet, that is what we find.

1 comments:

  1. The great thing about evolutionary theory is that it's -interesting-. Suppose we look at two species on different continents, and they look very similar.

    From an evolutionary viewpoint, we wonder if they're related, and if so how closely, and we study that. These days we would sequence the genome and figure it out from there, and then try to match up the findings with the geological history to see how descendants of the original single species ended up on the two continents.

    From a creationist viewpoint, I suppose we'd shrug and say, "Isn't it interesting that God liked that model enough to put similar copies on two continents." End of study.

    A science-killer, indeed.

    ReplyDelete