Replies

I expect it to be very like weather prediction: a nice chaotic dynamical system which gives accurate prediction only in the short-term, but also gives general predictions like punctuated equalibrium rather than gradualism, trends toward greater intelligence in certain niches, trends toward other characteristics in other niches. Small versions of the theory for confined ecological systems should be testable by experiment.

The point, however, is laws--laws of fitness which allow testable a priori predictions, not a posteriori explanations. Natural selection as actually used can be used to explain both mortality (of say humans) and immortality unless slain (of sponges), prolific reproduction (of cockroaches) and small numbers of offspring (of humans or certain predators), etc.

In a previous exchange on the subject, jennyp offered as a possible falsification of Darwinian theory an observation of an increasing incidence of a trait which in its homozygous form killed the organism before it could reproduce, but bestowed no survival advantage in its heterozygous form. I replied that it would be certain that the response to such an observation would be to postulate an as-yet-unknown and ill-understood survival advantage associated with the heterozygous form of the trait.

While Darwinism holds out in possible outline form a sketch of what a scientific theory of biological diversity might look like, in practice so long as the outline is not filled in, and natural selection continues to be used as a tautology, it isn't science. (I again remind you I am a Popperian.) Particular, falsifiable theories of natural selection in particular settings are good science. Darwinism as used polemically as an argument-from-no-design is not.