Jos Gibbons, the important thing is not whether a gene is a "property". The point is that the gene makes its "group", the genome, more successful.

Of course genomes "do not exist in allelomorphic variations on a locus". But a genome can certainly become increasingly common in successive generations because it outperforms a rival. That's pretty much what happens when organisms that reproduce by cell division compete with each other. Sure, they change a bit over time. But as long as they are sufficiently stable I don't see why you couldn't calculate genome frequencies of, for example, the bacterial strains in your gut flora. Sexual reproduction makes genomes unstable, so that decreases the importance of that level in the organisms that use it.

Yes, it is insufficient for the occurrence of group selection that entities co-operate. And that's why I have discussed necessary conditions: 1) An advantage of diversity and 2) stability of the group. Regardless of whether the group is a group of genes or a group of individuals, the principle is the same. In practice, human groups are of course much less stable than cell lines, but I'm after whats theoretically possible here.

So for that reason, it doesn't matter if many civilizations have risen and fallen. If a religion or culture lasted long enough, and kept sufficiently stable over time, group selection would be possible. I repeat, possible. That's all I'm saying. If they stop being stable, group selection would cease to function and it would be competion between genes again. Cross-cultural marriage here works on the group/gene pool level a bit like sexual reproduction on the genome level.