Way Out of Africa

Just as the “Cradle of Language” conference was celebrating our African origins last week, serious questions about the “Out of Africa” theory were beginning to take hold. Most important is a new study available from the Proceedings of the National Academy of Sciences (abstract here; full text here). It is too early to rewrite human history, but neither are matters as certain as they once seemed.

The “out of Africa” theory has made matters reasonably tidy for this blog. That theory holds that the human line of descent evolved in Africa and that over a million years ago an earlier species in the line, Homo erectus, spread over much of Africa and Asia. Then Homo sapiens emerged in Africa about 200,000 years ago and that species (us) spread over the whole world, replacing the archaic erectus and their descendants wherever they were found.

From the point of view of speech origins studies, this story keeps the plot line straight. It doesn’t matter what was going on with erectus outside of Africa because they were all replaced. The language faculty and speech communities that mattered were all taking shape in Africa.

Personally, I love a clean plot, so I am sorry to report that questions are emerging, actually never went away. The biggest difficulty with Out-of-Africa has always been that people are just too sexy. We are not that picky about our partners. Wherever people travel today, they soon start mixing with the locals, no matter how different they are in appearance, customs, and speech. Are we really supposed to believe that when sapiens met erectus they maintained a modest distance at all times? And that they did so for tens of thousands of years?

The illustration below contains a map from a presentation last April by Dan Dediu of the Language Evolution and Computation Unit at the University of Edinburgh. It shows the location of some million+ years ago fossil sites outside of Africa. The blue names mark very early sites, red names are for early sites, and green for the relatively more recent sites. Archaic peoples had spread very far, very early.

Of course there may have been biological reasons for the separation of the species. Matings between the two species may have been frequent but infertile, or the hybrids may have been unable to reproduce. These outcomes are common when species are crossbred for agricultural purposes. That kind of result would probably have helped keep interspecies unions to a minimum.

The popular press has certainly accepted the Out-of-Africa story but scholars have always had questions, Dan Dediu’s April presentation summarized the arguments. The chief source of skepticism has been the absence of decisive evidence pointing one way or the other. The facts are subject to interpretation. (Dediu’s paper is here and his slides are here).

A few days ago I had a brief post about a fossil that was said to be a cross between a Neanderthal and Homo sapiens, but that kind of noise on the screen is commonplace. More persuasive is a genetic study that suggests a common H. sapiens allele arose among one of those archaic lines over a million years ago and only entered our genome 37,000 years ago.

A team led by Bruce Lahn at the Howard Hughes Medical Institute has been investigating the gene microcephalin that regulates brain size. One allele is spread through 70% of the human population. Analysis indicates that the last common Homo sapiens ancestor for those with the gene lived only 37 thousand years ago, long after the worldwide distribution of Homo sapiens was well underway. Most archaic species had disappeared by then, although Neanderthals were still found in Ice Age Europe.

This emergence of the allele was also long after the human brain’s great size increase that occurred during the long years of Homo erectus. People with a non-functioning version of this gene have much smaller brains, three to four times smaller, although the brain’s architecture is unchanged. Plainly, the human brain did not suddenly grow three or four times in the past 40 thousand years, so it is obvious that the 30% of the population without the new allele do have one that provides similar enlargement services. The last common ancestor for the 30% with the different allele lived 900 thousand years ago, putting that ancestor deep into the erectus line.

Statistical analysis of the distribution of the two populations, with different functioning alleles, indicates the best explanation for the data is that the now commonplace allele first appeared 1.1 million years ago, on one of the archaic Homo lines that did not lead directly to H. sapiens. Over a million years later, sapiens and a carrier of the ancient gene met and successfully interbred. For unknown reasons, the gene was positively selected and has now spread throughout much of our species. The people least likely to carry the gene are sub-Saharan Africans, suggesting that the interbreeding occurred outside the African continent.

From this blog’s perspective, the story forces caution on seeking too neat a story. With so many uncertainties we cannot start insisting on some alternate theory, but we should refrain from building too much on the assumption that Out-of-Africa is fact.

Comments

[Dan Dediu]

Indirect support for these claims has also came from a recent study (Patterson, N., Richter, D. J., Gnerre, S., Lander, E. S. & Reich, D. (2006), Genetic evidence for complex speciation of humans and chimpanzees, Nature 441:1103-1108), where it is suggested that the separation between the chimp and humans lineages was far from a "clean" speciation event, involving gene exchanges a long time (millions of years) after their "separation".
Another interesting twist came very recently with a preliminary anouncement concerning the sequencing of Neandertal nuclear DNA (as opposed to just mtDNA) - see, for example the review in Pennisi, E. (2006), The dawn of stone age genomics, Science 314:1068-1071. One of the teams involved (the Max Planck Institute for Evolutionary Anthropology in Leipzig) seems to have found some hints of admixture (but everything is still preliminary). As S. Pääbo puts it: "Taken at face value, our data can be exaplained by gene flow from modern humans into the Neandertals" (cited in Pennisi, 2006:1070), suggesting, thus a directional geneflow. However, until more reliable data and details on the techniques and associated risks (contamination, decay, etc.) are available, one must refrain from speculating too much.