Monday, April 22, 2013

Evidence Of 22,000 Year Old Human Habitation In Brazil Is Weak

In the extraordinary claims require extraordinary evidence department, I am deeply skeptical of the claim that archaeologists have found human made stone tools in a Brazilian rock-shelter that date to more than 22,000 years ago. (The linked story's source is: C. Lahaye et al., Human occupation in South America by 20,000 BC: The Toca da Tira Peia site, Piaui, Brazil, Journal of Archaeological Science. (March 4, 2013)).

While there very likely were pre-Clovis modern humans in the New World, the evidence that there were humans in Brazil nine thousand years pre-Clovis is not strong.  The new evidence at the Toca da Tira Peia rock-shelter is in the same Brazilian national park as the site of a previous claim at Pedra Furada alleged to be 50,000 years old by their investigators.

Skeptics have argued that the "unearthed burned wood and sharp-edged stones" dated to such ancient time periods, "could have resulted from natural fires and rock slides." 
[The new] site’s location at the base of a steep cliff raises the possibility that crude, sharp-edged stones resulted from falling rocks, not human handiwork, says archaeologist Gary Haynes of the University of Nevada, Reno. Another possibility is that capuchins or other monkeys produced the tools, says archaeologist Stuart Fiedel of Louis Berger Group, an environmental consulting firm in Richmond, Va. 
Other researchers think that the discoveries are human-made implements similar to those found in Chile and Peru at the Monte Verde site 14,000 years ago and at another possibly as old as 33,000 years ago (the dating method for the older dates is likewise widely questioned).

The dating methods are also suspect. 
Dating the artifacts hinges on calculations of how long ago objects were buried by soil. Various environmental conditions, including fluctuations in soil moisture, could have distorted these age estimates, Haynes says. . . . An absence of burned wood or other finds suitable for radiocarbon dating at Toca da Tira Peia is a problem, because that’s the standard method for estimating the age of sites up to around 40,000 years ago, Dillehay says. But if people reached South America by 20,000 years ago, “this is the type of archaeological record we might expect: ephemeral and lightly scattered material in local shelters.” 

The dates of 113 putative human artifacts were made with a "technique that measures natural radiation damage in excavated quartz grains, the scientists estimated that the last exposure of soil to sunlight ranged from about 4,000 years ago in the top layer to 22,000 years ago in the third layer. Lahaye says that 15 human-altered stones from the bottom two soil layers must be older than 22,000 years."

Fundamentally, the dates are questionable because:

* There is no historical precedent for modern humans moving into virgin territory on a sustained basis for thousands of year without expanding exponentially and leaving an obvious sign of their presence.  If the site showed the signs of a marginal community that lasted a few hundred years and collapsed that might be imaginable.  But, this site purports to show continuous habitation for eighteen thousand years or more.

* There is an absence of intermediate sites between South America and any possible source of humans in the appropriate time frame.  (There is one alleged older site in the American Southeast with similar dating and hominin identification issues).

* There is no skeletal evidence matching the remains definitively to modern humans prior to 14,000 years ago or so.  Even if the dates were undisputably that old and made by hominins, in that time frame, a small band of archaic hominins with less of a capacity to dominate their surroundings might be more plausible.

* No radiocarbon dating has been possible and it is not well established that the dating method used is really accurate to the necessary degree of precision.

8 comments:

Maju said...

"There is no historical precedent for modern humans moving into virgin territory on a sustained basis for thousands of year without expanding exponentially and leaving an obvious sign of their presence".

Nearly all Asia... In East Asia particularly the archaeological record is very irregular and confuse.

"There is an absence of intermediate sites between South America and any possible source of humans in the appropriate time frame. (There is one alleged older site in the American Southeast with similar dating and hominin identification issues)".

You essentially contradict yourself in this paragraph. Whatever the case, lack of evidence is not evidence of lack.

"There is no skeletal evidence matching the remains definitively to modern humans prior to 14,000 years ago or so".

Pointless. In Europe there is no such skeletal evidence prior to Gravettian but still nearly everyone agrees that the H. sapiens occupation is some 20 thousand years older. Even worse is the case of South Asia with nearly certain AMH occupation since c. 80 Ka BP but no human bones till c. 35 Ka BP.

"No radiocarbon dating has been possible and it is not well established that the dating method used is really accurate to the necessary degree of precision".

That's the key issue in fact. The authors used a method of luminiscence dating which may or not be accurate. I'd like to know why you think this method is uncertain.

Maju said...

Actually for what I can find OSL is very widely used and acknowledged: Mungo Man, Arabian OoA archaeology, pottery, etc.

Which is the older site in SE North America you mention?

DDeden said...

I agree with the cautious view on the evidence. Capuchins use oyster shells to pry open mangrove oysters, but I haven't seen evidence that they use sharp stones.

Nets, thread (spindles) have been found older than this date, so possibly early sailing vessels crossed the Atlantic to Brazil, but skepticism is required.

Maju: "Even worse is the case of South Asia with nearly certain AMH occupation since c. 80 Ka BP but no human bones till c. 35 Ka BP."

False, 2 pygmy fossils found in Narmada River valley, one ~80ka, other ~200-300ka. Maju's comment reflects both shortsightedness and short-term memory loss, since he himself wrote comments on it.

andrew said...

Topper, South Carolina is the only other Eastern American outlier site, and there is one outlier site on the Pacific coast. The two sites in the Brazilian park and the Topper, South Carolina site are alleged to be much, much older than anything else in the Americas. As the Wikipedia entry on the Topper site explains:

"In 2004, Albert Goodyear of the University of South Carolina announced that carbonized plant remains, found as a dark stain in the light soil at the lowest excavated level at the Topper Site, had been radiocarbon dated to approximately 50,000 years ago, or approximately 37,000 years before the Clovis people. Goodyear, who began excavating the Topper site in the 1980s, believes that lithic artifacts at that level are rudimentary stone tools. Other archaeologists dispute this conclusion, suggesting that the artifacts are natural and not human-made. Other archaeologists also have challenged the radiocarbon dating of the carbonized remains at Topper, arguing that 1) the stain represented the result of a natural fire, and 2) 50,000 years is the theoretical upper limit of effective radiocarbon dating, meaning that the stratum is radiocarbon dead, rather than dating to that time period. Goodyear discovered the artifacts by digging 4 meters deeper than the Clovis artifacts readily found at the site. Before discovering the oldest lithics, he had discovered other artifacts which he claimed were tools dating around 16,000 years old, or about 3,000 years before Clovis."

The Pedra Furada sites are discussed at the link. There is no doubt that humans were there a long time ago (perhaps 11,000 years ago or more), the questions are whether the oldest dated items are artifacts or mere "geofacts" (i.e. natural occuring items that mimic human sourced ones) and whether the dating was done correctly.

Note also that I don't have a problem with pre-Clovis instances per se which are pretty well established.

The older Monte Verde date of 33,000 years ago which is not widely accepted is the only other besides Topper and this Brazilian National Park area.

The absence of evidence absolutely can be evidence of absence so long as you define well the absence of evidence proves.

Maju said...

Before writing my own blog entry, I made some searches and found this article that claims Topper's lowest layers being c. 23 Ka (C-14). I also found some references to sites in Alberta of similar or older dates.

The Monte Verde date may well be caused by a natural fire.

terryt said...

"I agree with the cautious view on the evidence".

So do I.

"Nearly all Asia... In East Asia particularly the archaeological record is very irregular and confuse".

Not really true. We have pretty good dates for the expansion of 'modern humans' through much of the world. The reason for the patchy evidence in East Asia is most likely a product of a small, scattered population rather than indicating 'absence'. At his bog Maju wrote:

"The example of New Zealand is very different because the evidence is much more recent, i.e. better preserved, easily dated with radiocarbon and other methods and shallower in the geology (easier to locate)".

Therefore more to be taken seriously.

"Also Neolithic peoples are always more numerous, settling the countries in a much denser way".

I have tried to point out to you before that the pre-European Maori can hardly be considered 'Neolithic' in any meaningful sense of the word. Sure, in the north they did grow a few edibles but their main sustenance was as hunter-gatherers.

terryt said...

This is not relevant to the post here but you wrote at Maju's site:

"It is also striking how thin the archaelogical record is in Iran (presumbably due to lack of research and not lack of evidence out there to be discovered), given its key central location ... Another possibility, given the small number of finds marked in Central Asia where I know there has been a lot of archaeology done, is that there is a wealth of information, including Iranian data (a Russian ally for much of the post-1979 period), that is in Russian (or Soviet) scholarly journals that are not widely available in the West".

Some of those Russian papers are available if you search, but I too was thinking about Iran the other day. I realised that in fact if you're going to claim that basal haplogroup diversity is evidence of origin surely you have to be inevitably drawn towards Anatolia and the Iranian Plateau to as far north as Altai. Although most modern haplogroups almost certainly did not originate in the region everything has passed through it at some stage. But many basal mt-DNA haplogroups could easily have originated there: such as N2, N3, N1'5, X and some R haplogroups (R1, R0, R2 and U), and possibly even M and N themselves. As well as several basal Y-DNAs such as G, F3,IJ, T and possibly C5 as well as some R haplogroups (Q, R1a, R1b). It seems to be fashionable to ignore Iran's role in eatly human expansion. For political reasons?

andrew said...

Political reasons are quite plausible. On the Iranian side, the theocracy may not favor allowing people to conduct sciences that seek to directly contradict the regime's religious doctrines related to human and angel and jinn origins. What waste money trying to prove wrong what God has already told you?

On the Russian side, in the face of a collapsing post-Cold War economy, disseminating Soviet era archaeological work in Iran and Central Asia (much of which may have disputed ownership in connection with the break up of the Soviet Republics or may be owned by governments that have no reason to care about work done outside their jurisdiction) may not be a priority.