Not a lasting last for the Neandertals

The latest in a long line of "last known Neandertal" sites is now Gorham's Cave, Gibraltar. Of course, if this were actually a continuing string of "latest" sites, you would expect we would eventually either reach the present day, or some mathematical limit. There seems to be little danger of that happening for a while, though, since the previous "last known Neandertal" sites keep turning out to be older than their "first known" radiocarbon dates!

The current paper by Clive Finlayson and colleagues has a good short review of this issue:

The sequence of radiocarbon dates presented, including 14 dates at or statistically younger than 30 kyr bp, are the only currently reliable ones that establish the persistence of Neanderthals and associated Mousterian technology after 30 kyr bp. Earlier claims are now dismissed or are uncertain for a variety of reasons and in particular after the revision of dates on bone with the use of ultrafiltration treatment, a treatment only meaningful for dates on bone. Hyaena Den (UK) is now considered older than 30 kyr bp; the Vindija (Croatia) Neanderthals have been re-dated to between 32 and 33 kyr bp or older; Zafarraya (Spain) is now discarded for several reasons; the Mezmaiskaya, Russia, Neanderthal is now dated to at least 36 kyr bp. The single AMS date on Cervus bone for Caldeirão (Portugal) will require revision and is likely, given the result for Hyaena Den of similar age, to be older than 30 kyr bp. Finally, the single 14C date, from Patella shells, from Figueira Brava, Portugal, is not statistically younger than 30 kyr bp (Finlayson et al. 2006, references omitted).

So are the current radiocarbon dates for Gorham's Cave any better? Or, to put it another way, why exactly should we believe any new claims about recent dates, given the long list of dates that we are now supposed to forget about?

Now, I'm not an archaeologist, nor am I a geochronologist. So maybe I'm missing something. But look at this figure from the paper (Figure 1c):

Section from Gorham's Cave, showing points of radiocarbon sampling, Figure 1c from Finlayson et al. (2006).

Notice sampling points 16, 17, and 20. Those are the key samples for the paper's conclusion:

Thus, three samples (16, 17 and 20; Fig. 1) came from in situ Mousterian superimposed hearths. These three dates provide a stratigraphic sequence from 24,010 +- 320 to 30,560 +- 720 yr bp. Taken together, all the dates show that Neanderthals occupied the site until 28 kyr bp and possibly as recently as 24 kyr bp. The evidence in support of the 24 kyr bp date is more limited than for 28 kyr bp, which is taken as the latest well-supported occupation date (Finlayson et al. 2006).

OK, so we have three samples from the same place in the cave, over a short vertical distance, that appear to represent successive occupations over a few-thousand-year interval. The authors interpret conservatively that maybe the 24,000-year date is too young to be Neandertal -- although they don't describe just what makes the evidence "more limited," considering each date is supported by a single radiocarbon sample.

But look at sample number 11 in the figure. It appears to have been taken from directly above the putative hearths. So why does it have a date of 27,020 +/- 480 years?

I think we begin to detect why there is "more limited" evidence for the 24,000-year date. It is directly controverted by the sequence.

Moreover, we have to doubt the 26,000-year date, considering the evident contamination and/or turbation of the sample directly above it.

Am I saying the Neandertals weren't in this cave after 30,000 years ago? Well, if you look at the samples in the figure, and their locations, almost all the samples taken from the brown zone (layer IV) have dates between 28,000 and 32,000 years BP. But there are several with dates between 26,000 and 23,000 years, and these are mixed in amongst or below earlier dates in the 28,000-30,000 year BP range.

Please check out the dates yourself: Sample 23 (23,360 BP) is directly below sample 22 (29,720 BP). Sample 9 (26,070 BP) is directly adjacent to sample 10 (28,360 BP). Sample 15 (23,780 BP) appears to be stratigraphically below sample 14 (30,310 BP), although these are more spatially distant. Sample 28 (28,170 BP) is immediately below sample 27 (31,850 BP). Sample 29 (29,210 BP) is directly below sample 25 (31,780 BP). In all cases these discrepancies are outside the reported confidence limits.

There seems to be clear evidence of widespread movement of material or contamination in this sequence.

So, does the sheer weight of dates between 32,000 BP and 28,000 BP lead to the conclusion that the cave was occupied by Neandertals during that time range? Maybe so, but I think the paper raises a lot more questions than it answers. I have to think that we'll be hearing about how this date is equivocal or problematic, instead of it being the "latest Neandertal."


Brill D. 2006. Neanderthal's last stand. Nature News 13 Sept. 2006. DOI link

Finlayson C, and 25 others. 2006. Late survival of Neanderthals at the southernmost extreme of Europe. Nature, advanced online publication doi : 10.1038/nature05195

Pendergast DM. 2000. The problems raised by small charcoal samples for radiocarbon analysis. J Field Archaeol 27:237-239.