I was reviewing recently Stuart Kauffman’s critique of the Darwinian selection mechanism and thought I would share the upshot of it here, especially in light of the recent discussion at UD concerning Haldane’s Dilemma:
If selection could, in principle, accomplish “anything,†then all the order in organisms might reflect selection alone. But, in fact, there are limits to selection. Such limits begin to demand a shift in our thinking in the biological sciences and beyond. We have already encountered a first powerful limitation on selection. Darwin’s view of the gradual accumulations of useful variations, we saw, required gradualism. Mutations must cause slight alterations in phenotypes, But we have now seen two alternative model “worlds†in which such gradualism fails. The first concerns maximally compressed programs. Because these are random, almost certainly any change randomizes the performance of the program. Finding one of the few useful minimal programs requires searching the entire space ÂÂrequiring unthinkably long times compared with the history of the universe even for modestly large programs … But the matter is even worse on such random landscapes. If an adapting population evolves by mutation and selection alone, it will remain frozen in an infinitesimal region of the total space, trapped forever in whatever region it started in. It will be unable to search long distances across space for higher peaks. Yet if the population dares try recombination, it will be harmed on average, not helped. There is a second limitation on selection. It is not only on random landscapes that evolution fails. Even on smooth landscapes, in the heartland of gradualism, just where Darwin’s assumptions hold, selection can again fail and fail utterly. Selection runs headlong into an “error catastrophe†where all accumulated useful traits melt away…. Thus there appears to be a limit on the complexity of a genome that can be assembled by mutation and selection!
Stuart Kaffman, At Home in the Universe: The Search for Laws of Self-Organization and Complexity (New York: Oxford University Press, 1995), 183-184.