[whatshisface]

Large interbreeding populations are "locked in" and recessive traits are rare. Small isolated populations can make recessive traits permanent. This is well-known to everbody who has stopped to consider what conditions produce major genetic changes in human morphology (inbreeding). Large gene pools protect against morphological changes in humans and animals alike.

Species are sent back and forth between those two extremes by changes in the environment brought on by our planet's constantly changing climate, topography and ecosystems. Sometimes, when the isolation ends, the combination of made-permanent recessive traits turns out to be harmful and the isolated population is replaced by the larger one (if the larger one survived), and other times it is helpful and they spread out into the larger environment. It isn't a contradiction any more than first and second gear on your bike are - you can't have them both at once, but they are switched between by systems external to the bike itself (the rider).

The idea is that every species is descendant from a highly inbred population of another species. Curiously, this happens in the Bible (Noah's ark).

The molecular basis is the same as that of Darwinian evolution (transcription errors plus whole-organism natural selection), combined with the usual laws of population genetics.

[seandhuine]

Right at the end of all that, there's an IOU for a molecular explanation of puntuated equilibrium. How is the IOU cashed out? Oh, "transcription errors" plus some non-molecular magic. Transcription errors...think about that. Gradual errors (which are nearly always either neutral or deleterious -- in the real world, mind) that suddenly operate over a very short period to create a new species. Let me attempt to paraphrase the motivation for PE: "We can't see it happening because it's gradual, but it needs a catastrophe oh there it just happened!" Or it must have, because reasons. Errors, or something. Fairy dust.