Maybe Chinese fire drills explain the date discrepancies

3 minute read

I noticed that the cover of the most recent New Scientist is a story about modern human origins by science writer Dan Jones. It's headlined, "Going global: how humans conquered the world."

I think Jones has done a nice piece of work here -- at 2700 words, the story is easy to read, and it illuminates a certain kind of current consensus. It touches on everything from the Herto hominids, to the Blombos The underlying theme is the idea of a "coastal route" dispersal of modern humans from Africa, coupled with some detail about Paul Mellars' Afro-Indian connection, Spencer Wells' Y-chromosome story, the early Herto and Omo Kibish remains, the relevance of Oase and Tianyuan to early dispersal scenarios, and the "megadroughts" of the African Late Pleistocene.

I'll tell you one thing: The piece succeeds at making me feel like a member of the Neandertal Underground, standing on the side of the road as the march of the "Human Revolution" goes by.

The thing is that none of these separate elements fit together. It's not hard to figure that tracing the Y chromosome genealogy of Eurasia to a divergence in the Middle East 40,000 years ago doesn't match up very well with the idea of an "early coastal route dispersal" 60,000 years ago, or an initial colonization of Australia 50,000 years ago. Placing "modern human anatomy" earlier and earlier in time -- back to 200,000 years ago -- isn't exactly helping to explain the behavioral record in the last 70,000 years. And the archaeology that places "modern human behavior" increasingly into the Middle Stone Age doesn't explain why the same behaviors should be found in Neandertals.

Sometimes the contradictions are so glaring that Jones almost can't help but juxtapose them:

"The similarities between Africa and India are not coincidental, and fit in beautifully with the DNA evidence," says Paul Mellars, an archaeologist at the University of Oxford. Although none of these artefacts is more than 35,000 years old, that may simply reflect the fact that sea levels are about 100 metres higher today than they were 50,000 years ago. Any artefacts or bones left by the first coastal migrants are now buried beneath the sea.

I never credit someone with quotes taken from a news article -- every nuance of the evidence is simply not that important to the casual reader. But it's sort of obvious that some of the DNA evidence poses a problem here. And the dates are entirely discordant.

Mellars has emphasized in print (e.g., 2006) the material similarities between early Upper Paleolithic assemblages of India and the Howieson's Poort industry of Africa. The similarities are there, but the dates are quite different. "Lower sea levels" is only arm-waving: Sure, the lack of earlier evidence of similar industries is a problem, but a much bigger problem is explaining the 40,000-year persistence of these "similar" industries in the constant adjacent presence of other patterns of material culture.

The obvious alternative is that the similarities are coincidental -- or at least don't reflect a lineal cultural relationship between 70,000-year-old Africans and 30,000-year-old Indians. That doesn't argue against dispersal: after all, the abilities represented by the material remains may have dispersed, early or late, even if the tools themselves didn't.

But we should also consider the similarities with the cultural remains of late Neandertals and even earlier peoples of Europe, including the pigment use, engraved lines, pendant drilling and blade manufacture.

What we have here is a clown car of a hypothesis: everything thrown in but the bearded lady. No hypothesis is ever tested: Consistency rules. This is no discredit on Jones at all, who clearly does the best job possible of fitting together all these recent papers. The problem is that when you see them all next to each other, you can't help but see that these 115,000-year-old Eritrean shellfish, 40,000-year-old Y chromosome divergences, 65,000-year-old mitochondrial haplogroups, 30,000-year-old Indian blades, 35,000-year-old Romanian skeletons, 70,000-year-old ochre engravings, and 190,000-year-old African skulls really can't fit together to tell a story of a single human dispersal at a single time.

Either the hypothesis is wrong, or some of the data are. Or both.


Jones D. 2007. Going global: how humans spread across the world. New Scientist, Oct. 27, 36-40.

Mellars P. 2006. Going east: new genetic and archaeological perspectives on the modern human colonization of Eurasia. Science 313:796-800. doi:10.1126/science.1128402