I've recently lost confidence in the assumption that of course humans and Neanderthals mated (Leaving open only the question as to whether these pairings were productive or not.)--the idea being that humans will always mate with their neighbors when given enough time.

What about the Andaman islanders? There have been two isolated and distinct lineages present there for at least the last 20k years. On a tiny little island.

islands are strange, and subject to way reduced gene flow. this is why sardinia and iceland pop-up as "genetic outliers."

What two lineages is Kosmo talking about?

actually, i don't know.

I no doubt oversold my case a bit, as the timing is still in dispute and there does seem to be some mixture, but the Andamese islands appear to have been settled in two early and distinct waves, the remnants of which still strongly echo in the native populations today.

http://en.wikipedia.org/wiki/ And...Andaman_Islands

"male Onges and Jarawas almost exclusively belong to Haplotype D, which is also found in Tibet and Japan [Ainu], but is rare on the Indian mainland" and "Male Great Andamanese, unlike the Onge and the Jarawa, have a mixed presence of Y-chromosome halpgroups O, L, K and P, which places them between mainland Indian and Asian populations"

http://www.ncbi.nlm.nih.gov/pmc/...cles/PMC378623/