What prevents evolution?

[arahant]

[ QUOTE ]
In order for A1 and A2 to be mutually infertile (the sense of "speciation" used here), either or both of them must become infertile with A.

[/ QUOTE ]
This is simply incorrect. Imagine a system in which an egg can combine with a sperm using either of two proteins, P1 and P2, which must match in the egg and sperm. Members of population A have both proteins. This is entirely reasonable, as such a system would confer redundancy to the process. Now population A1 loses the protein P1 to mutation, and population A2 loses the protein P2. Both A1 and A2 can reproduce with A, but not with each other.


[ QUOTE ]
If you say the propagation of such mutations within the isolated group will preserve its homogeneity, that's fine. However, if a mutation carries with it a tendency to reduce fertility without otherwise improving adaptation (and there's little reason to assume otherwise), merely duplicating the basis of that tendency in a potential reproductive partner does not necessarily correct the problem. In other words, two wrongs don't make a right.

[/ QUOTE ]

You are making the same error here. The mutations of interest here don't flat out 'reduce fertility'. They reduce fertility WITH RESPECT TO A CREATURE WITH ANOTHER GENOTYPE. As per the detailed mechanism in my earlier post, there will always be a selection for interfertility in a given population. It is the physical separation of A and A1 which allows the mechanisms of fertility to change WITH RESPECT TO EACH OTHER, but NOT within a population group.

Language makes an excellent analog for reproductive differentiation. Both English and German share a common linguistic root. No modern english or German speaker is capable of understanding this progenitor language, and neither can they understand each other. That doesn't mean I can't understand an english speaker. Reproduction is not a one way street...it is like communication. Both sexes need to use the same protocol.