Modern humans older than thought?

I can understand that National Geographic wants to promote news from researchers who take National Geographic money. It’s only natural, and as a publicity organ for paleoanthropology, they haven’t done so badly in the past.

But this story titled, “Humans 80,000 Years Older Than Previously Thought?” is just silly. To be clear about our knowledge of the ESA/MSA transition in Africa, here’s a paragraph from a well-cited review article by Sally McBrearty and Alison Brooks (2000):

Late Acheulian and early MSA dates cluster between 200 ka and 300 ka, and the Acheulian seems to have disappeared in most of Africa by about 200 ka. The Acheulian at Isimila, Tanzania has been dated by U-series to ca. 260 ka (Howell et al., 1972). Acheulian occurrences in the Kapthurin Formation, Kenya, underlie volcanics dated by K/Ar to ca. 240 ka (Leakey et al., 1969; Tallon, 1976, 1978; McBrearty et al., 1996), but now estimated by 40Ar/39Ar at ca. 280 ka (Deino & McBrearty, under review). The Acheulian in the western desert of Egypt is also thought to end at either about 200 ka (McHugh et al., 1988) or perhaps as much as 350 ka (Wendorf & Schild, 1992). A U-series date of ca. 174 ka on a late Acheulian occurrence at Rooidam, South Africa (Szabo & Butzer, 1979) seems somewhat anomalous in this context (McBrearty and Brooks 2000:488).

Here’s a passage from the introduction of paper described in the National Geographic News article, by Leah Morgan and Paul Renne:

Most well-dated and documented MSA sites postdate 130 ka. However, a few sites hint at a much earlier origin of the MSA. These include Malewa Gorge, Kenya, dated to before ca. 240 ka by K-Ar (Evernden and Curtis, 1965); Twin Rivers, Zambia, dated to before 265 ka by U-series on speleothems (Barham and Smart, 1996); and Cartwright's site at the Kinangop Plateau, Kenya, also dated by K-Ar to before 439 ka (Evernden and Curtis, 1965). Although the old age from Cartwright's site is intriguing, its accuracy is questionable, as stratigraphic relationships at this locality are not clear (Evernden and Curtis, 1965) and the K-Ar method is highly susceptible to contamination by older grains. The oldest MSA archaeology whose age is well documented occurs in the Kapthurin Formation in Kenya, where single crystal 40Ar/39Ar dating indicates that it is older than 284 24 ka (Deino and McBrearty, 2002). (Uncertainties are given at the 2? level here and throughout, with the possible exception of K-Ar ages published by Laury and Albritton [1975] and Wendorf et al. [1994], for which confidence levels were not reported.) In contrast, hominids at Herto in Ethiopia utilized what appears to be a technology transitional between the Acheulean and MSA as recently as 160 ka (Clark et al., 2003). The appearance of MSA-typical technology may indicate a shift in tool-making abilities, whereas the persistence of Acheulean technology might be expected to be spatially heterogeneous, depending on such variables as available source materials and cultural circumstances.

That’s a perfectly reasonable introduction to the issue, albeit short, and basically the same as McBrearty and Brooks’ earlier review, except for the addition of more recent finds. In other words, they start from the same place.

Here’s part of the conclusion of the new paper:

The Middle Stone Age has previously been shown to extend back to before 284 ka in the Kapthurin Formation. Although only limited data were published for the tuff dated to 284 24 ka in the Kapthurin Formation (Deino and McBrearty, 2002), there is a suggestion of evidence for excess 40Ar in the 40Ar/39Ar data in the form of a correlation between 40Ar/36Ar and apparent age for individual analyses. An isochron fit to all the data for this tuff yields an age of 279 40 ka, with initial 40Ar/36Ar = 304 38 and MSWD = 1 (A. Deino, 2008, personal commun.). Our significantly more precise new ages for Gademotta and Kulkuletti render the oldest MSA at these sites indistinguishable in age from that found in the Kapthurin Formation. Furthermore, MSA artifacts from the upper part of unit 9, which underlies unit 10 and is thus older than 276 4 ka, display tool variability and complexity that rival assemblages from many much younger Upper Paleolithic sites (Schild and Wendorf, 2005). As seen in Figure 3, comparison of typologically equivalent tools (blades and points) shows that Gademotta unit 9 artifacts more closely resemble the much younger (80100 ka) MSA from the site of Aduma in the Middle Awash of Ethiopia (Yellen et al., 2005) than the contemporaneous Kapthurin ones. It should be noted, however, that the vast majority of artifacts from Gademotta and Kulkuletti, and many at Aduma, were made from obsidian, while Kapthurin has relatively few tools made from this material. Since obsidian is among the raw materials most conducive to Levallois technology, the advanced appearance of the very early MSA at Gademotta and Kulkuletti may be more related to the simple circumstance of proximally available raw materials than to the developmental stage of the toolmakers.

They emphasize that the date is consistent between these Ethiopian sites and the early MSA from Kenya, and that these sites have some artifacts that would be consistent with later MSA sites moreso than the earlier Kapthurin assemblages. Pretty simple.

These results are pretty cool, and we need many more like them. It would be great to have a better chronology of the end of the Acheulean and the early MSA, and in particular a better understanding of these issues of regional heterogeneity and progression over time within different African regions. But given the small number of early MSA sites, I don’t really think we know whether the heterogeneity is real. More informative would be the date of final occurrence of Acheulean/ESA in these regions. But really, since the two overlap in many elements, the first/final question tends to be answered in terms of type fossils like handaxes and blades. And these are heterogeneous themselves.

Don't believe the press

I really don’t understand why the news report goes so far overboard in describing the paper as “pushing humans back”. The record simply does not show that. There are no fossils here, and no reason to believe that the initial MSA was produced by anything other than the kind of people of which we have fossils in the Middle Pleistocene of Africa. There aren’t very many of them, and the good crania tend to be a bit earlier (e.g., Kabwe, Ndutu, Salé) and variable, so we can’t be sure what we’ll find next week.

But why would we expect that the transition to MSA would have a major anatomical effect, when the adoption of Middle Paleolithic/Mousterian industries in Eurasia did not? Why would we assume that behavioral and cultural traits have to be linked to anatomical features that we can recognize in fossils?

Now, you may remember that I had a big problem with another National Geographic News release last spring, which also concerned the appearance of modern humans. I wouldn’t ordinarily be so concerned about a given press release, which are always misleading in some ways because of their brevity, except that these keep coming from one funding source. And it’s clear that a large fraction of professionals are taking this information from the press, without even reading the secondary literature.

So I’m doing my best to point out these misleading aspects. National Geographic needs to get a better editor!

You really know there’s nothing to a story when the final quote goes like this:

If anything, the story has now become more complex, added Laura Basell, an archaeologist at the University of Oxford in the U.K.
"The new date for Gademotta changes how we think about human evolution, because it shows how much more complicated the situation is than we previously thought," Basell said.
"It is not possible to simply associate specific species with particular technologies and plot them in a line from archaic to modern."

Well, that’s entirely correct. The story is possibly complex, and if it’s complex, then it follows that every new piece of information will make it look more complicated. So no knock on Basell – her answer is fine.

But it really helps to show the lack of new information here. We might just as easily say this:

You know, the story is really quite simple. Around 300,000 years ago, hominids in several regions started making tools using a slightly more complicated reduction sequence. These were all "pre-modern" humans, if you like, since they all occur at least 100,000 years before the first appearance of "modern" traits in their region. Their behavior evolved.