OK, the title is somewhat of an exaggeration, but not much. Out of Africa, Not Once But Twice:
Modern humans are known to have left Africa in a wave of migration around 50,000 years ago, but another, smaller group -- possibly a different subspecies -- left the continent 50,000 years earlier, suggests a new study.
While all humans today are related to the second "out of Africa" group, it's likely that some populations native to Israel, Jordan, Lebanon, Syria, Australia, New Zealand and Indonesia retain genetic vestiges of the earlier migrants, according to the paper's author, Michael Schillaci.
The paper is forthcoming in The Journal of Human Evolution. The article weirdly keeps saying this was a genetic analysis, which I think will confuse people since it seems clear that the author was doing cladistics on the 28 fossil samples. I'll wait for what the bones people have to say when the paper comes out, but I have to admit a bit of fatigue about the whole issue about whether they would be a subspecies or not. I dread a lumper-splitter fight if more data comes to light of relict alleles or traits in modern populations from archaics. Why the fixation on vehicles?
I haven't read the paper yet, but based on the Discovery writeup this is a paper about biological distance. There are a number of techniques such as Euclidean distance and Mahalanobis D2 that are used.
I have to admit a bit of fatigue about the whole issue about whether they would be a subspecies or not
Especially, since, without a rigorous definition of what a "subspecies" is, it's not even a scientific debate. Cases like this make it pretty clear to me that the idea of a "subspecies" being an entity with some sort of basis in reality is kind of ludicrous, at least in some cases.
However...
There are a number of techniques such as Euclidean distance and Mahalanobis D2 that are used.
That's interesting. Have there been any papers proposing how distance techniques may be applied to taxonomy? (I've been thinking of writing such a paper with a different sort of metric in mind--assuming I can get it to yield semi-practical results.)
If you name it, it becomes real.
http://en.wikipedia.org/wiki/Borneo_Elephant
None that I am aware of. Usually such methods (in this case the authors used a technique developed by Relthford) look at gene flow, drift, selection, skeletal variability where population structure has an impact, etc. They can also be used to look at, as in this paper, evolutionary history, although this is one of the few papers I have seen that actually tries to do so.