Back to the Grotte des Fees - Mellars et al. respond

[Daryl Habel]

In a continuation of the debate surrounding the question of Aurignacian/Chatelperronian interstratification at the Grotte des Fées, discussed earlier on the Palanth Forum HERE,  Paul Mellars, Brad Gravina, and Christopher Bronk Ramsey, who published their original paper in Nature in 2005, have responded to criticism published by Zilhão, d'Errico et al. in PNAS in 2006.

To be published in the 27 February 2007 issue of the Proceedings of the National Academy of Sciences, but published online 21 February from the PNAS Early Edition website:

Confirmation of Neanderthal/modern human interstratification at the Chatelperronian type-site

Paul Mellars , Brad Gravina , and Christopher Bronk Ramsey

www.pnas.org/cgi/doi/10.1073/pnas.0608053104

The nature of the replacement of Neanderthal by anatomically and behaviorally modern populations in Europe is currently a topic of lively debate in human evolution. In an earlier paper [Gravina B, Mellars P, Bronk Ramsey C (2005) Nature 483:51-56], we published a series of radiocarbon accelerator mass spectrometer measurements for the site of Châtelperron in central France, which had been claimed to show a clear "interstratification" of successive levels of Neanderthal and modern human occupation, on the basis of excavations carried out by Henri Delporte in the 1950s. This interpretation has recently been challenged by Zilhão and colleagues [Zilhão J, d'Errico F, Bordes J-G, Lenoble A, Texier J-P, Rigaud J-P (2006) Proc Natl Acad Sci USA 103:12643-12648], who suggest that the deposits excavated in the 1950s consisted largely, if not entirely, of the unstratified "backdirt" of the earlier, 19th century excavations on the site. We show here that the excavation backdirt interpretation for the Châtelperron stratigraphy can be refuted from many different aspects of the stratigraphic, radiocarbon, and archaeological evidence. We reassess the significance of this site for current models of the coexistence and interactions between Neanderthal and anatomically modern populations in western Europe.

Currently available from PNAS Early Edition HERE.   I have the pdf of the paper, and I've made a quick read through it.  But I'll have to review the whole debate before I can say much about whether anything is likely to be settled one way or another.

Dar

 

[lagarvelho]

Dar:

I, too, just downloaded the pdf of this paper a little while ago.  I'm going to have to read them all today and tomorrow.  I won't comment any further, other than to say that a response to Zilhao et al, by Mellars et al, could only be expected.  It probably won't end the debate, though.
Anne G

[Daryl Habel]

I’ve now read through the new paper by Mellars et al. (2007) and re-read the criticism of Zilhão et al. (2006), in conjunction with the original paper claiming Aurignacian occupation in the levels B4-B4a at the Grotte des Fées, interstratified between earlier Chatelperronian levels B5 and later Chatelperronian levels B1-B3a (Gravina et al. 2005).

I’m not totally convinced by either side of the debate.  Although both sides agree that level B4-B4a contains a total of 5 “diagnostic” Aurignacian artifacts, that same level also contains a total of 51 “diagnostic” Chatelperronian artifacts and 191 non-diagnostic artifacts.  In other words, of the total 237 artifacts in level B4-B4a documented by Delporte from his excavations at the Grotte des Fées in 1951-56 and 1962, only 2.1% are “diagnostic” Aurignacian.

In total, there are only 11 “diagnostic” Aurignacian artifacts in the entire Delporte inventory.  Two are from the disturbed surfacial level A, two more from levels B1-3, and one from the basal level B5 (plus one more tabulated in Table 4 of Zilhão et al. 2006 as from levels 4-5 , which, if from level 4-4a would raise the percentage of “diagnostic” Aurignacian artifacts from that level, to slightly below 3%).   

Mellars et al. (2007) do agree that the archaeological levels are palimpsests, but it is their contention that the AMS dating results presented in their original argument (Gravina et al. 2005) show the Chateperronian occupation in level B5 centered on ~42,000-43,000 cal. BP (calibrated to calendar years), while the later Chatelperronian levels are calibrated to calendar years as ~40,000-41,000 cal. BP.  This, they say, means the Aurignacian interstratification occurred between these two levels, and from inference about 41,000-42,000 cal. BP.

However, the two AMS dates from level B4 are 35,540 +/- 280 and 39,780 +/- 390 14C BP, and taken individually, the older date falls within the range of the dates for the earlier level B5 Chatelperronian occupation, while the younger date falls within the range of dates given for the overlying Chatelperronian occupation levels B1-B3. 

While the dating appears to be relatively consistent with the stratigraphic levels reported by Delporte, the entire sequence from level B1 to B5 only dates in calendar years to between about 43,000-40,000 cal. BP, about 3,000 years, during which time the great majority of artifacts from all levels are either Chatelperronian or undiagnostic. 

Much of the argument presented by Mellars et al. (2007) in this new paper is an attempt to refute the argument of  Zilhão et al. (2006) that “at least levels B1-B3, and in all likelihood B1-B5 sequence, represent reworked sediments and archaeological material, probably backdirt from the excavations conducted on the site in the nineteenth century” (Zilhão et al. 2006:3657).    In this argument, Mellars et al. strongly defend Delporte’s archaeological competence and interpretation of the Grotte des Fées stratigraphy.  They need to, because Delporte’s 1951-56 and 1962 excavations effectively removed all remaining Paleolithic strata. 

As I mentioned earlier, I'm not totally convinced by either side in this debate.  The argument made by Zilhão et al. that the site is disturbed seems reasonable when the stratigraphic distribution of the (very few) "diagnostic" Aurignacian artifacts is considered, but on the other hand, Mellars et al. strongly defend Delporte's interpretation, implying that Delporte would have the competence to recognize the exact spatial extent of the nineteenth century excavations and the difference between disturbed and in situ deposits.

All of which brings me back to a comment made by Chatelperronian in Reply #3 in the original discussion (with full references to Gravina et al. 2005 and Zilhão et al. 2006), of the Grotte des Fées interstratification debate CLICK HERE: "...but sometimes I think that both of them talk about sites and stratigraphies without an actual look at the places in question. .."  And although perhaps both sides have visited the Grotte des Fées, there isn't much there now to look at, and the debate on both sides is being conducted from documentation more than 40 years old. 

So, I expect this debate to continue.

Dar