15,000? 22,000? 32,000?

As I was saying about the possible pre-Clovis finds at the Friedkin Site, the biggest problem with all of the possible pre-Clovis sites is that once you accept any of them it is hard to avoid accepting a lot of them, and together they make no sense. At the Lovewell Site in Kansas, paleontologists excavated a bunch of mammoth bones that date to 22,000 years ago, and they are now claiming that the bones could only have been broken by people smashing them with rocks. Beats me, I mean, I don't know anything about elephant bones. But if there were humans in Kansas 22,000 years ago, how did they get there, and where did they go? Even if you add together all of the remotely plausible pre-Clovis sites you get a very, very thin record. The 9,000 years between Lovewell and the population explosion of Clovis times is a long time; by 9,000 years after Clovis, North America was so densely populated that every decent camping spot is covered by thousands of artifacts. And, remember, the genetic evidence points to a single small source population that left Asia around 15,000 years ago.

If you want the first Americans to have arrived 22,000 years ago, you have to throw out the genetic evidence or postulate that a few hundred Clovis migrants somehow completely exterminated all those who were here before them. And then you have to assume that there are thousands of pre-Clovis sites all over the Americas that have been misidentified or not yet found. This is possible, especially if the hypothetical pre-Clovis people did not use any "diagnostic" artifacts that identify their collections. After all, most Clovis sites can only be identified as such by the presence of Clovis spear points. Still, it starts to be a stretch.

Actually there is another reason to doubt the Lovewell finds, which is that no stone tools of any sort were found around them. Stone Age people were very messy, and as they butchered large animals they tended to break or wear out a lot of stone tools, which they then resharpened on the spot. So at known mammoth and buffalo kill sites from Clovis and later times there are always at least a few stone artifacts. But Michael Waters is a fan, on the theory, I suppose, that if his site is pre-Clovis there must be lots of other pre-Clovis sites, so the more the merrier.

So I doubt the Lovewell claims will be much talked about a decade from now. I mention them because they illustrate the point I was trying to make about pre-Clovis archaeology. It is possible to reconcile finds from a thousand years or so before Clovis with a coherent story of human expansion and adaptation, and with the genetic evidence. But possible pre-Clovis sites seem to be randomly distributed in time, going back 30,000 years and more, and I simply cannot understand by what model humans could have been in North America for that long.