John Hoffecker replies to John HawksThe archaeological site of Kostenki, Russia, was reported in Science magazine on January 12, 2007, and a brief article on it appeared here. Kostenki contains evidence for the arrival of Early Modern Humans in eastern Europe at the same time or slightly before EMH appeared first in the rest of Europe. Popular science blogger John Hawks responded to the Science article, expressing doubt about the dating of the occupation layers. In the following contribution, one of the co-authors of the original Kostenki paper, John F. Hoffecker of the University of Colorado at Boulder, responds to Hawks' comments.
The Middle/Upper Paleolithic Transition in Europe at KostenkiJohn Hawks of the University of Wisconsin has posted some comments on a paper published in Science on January 12, 2007 on which I was a co-author. These comments warrant responses, and I am grateful to Kris Hirst for providing me with an opportunity to post them on her website.
The paper was entitled "Early Upper Paleolithic in Eastern Europe and Implications for the Dispersal of Modern Humans" and it was authored by M. V. Anikovich, A. A. Sinitsyn, and others. It reports on artifacts assigned to early Upper Paleolithic industries that are buried below a volcanic tephra horizon at several open-air sites at Kostenki on the Don River in Russia.
Chronological QuestionsHawks begins with a discussion of the dating of the archaeological layers at Kostenki, noting that many radiocarbon dates have been obtained on various layers at these sites. He observes that calibration of even the oldest radiocarbon dates from Kostenki does not suggest an age for the earliest layers as great as that proposed in the paper (45,000-42,000 years ago). This is correct, and it is noted in the text of the paper that calibration of dates obtained on wood charcoal (dates on bone were excluded) yield age estimates that are roughly 1,000-2,000 years younger than estimates based on other methods. Two recently published calibration curves were used, including CalPal (Cologne Radiocarbon Calibration Programme).
The chronology for the early Upper Paleolithic layers at Kostenki is based primarily on other methods, which indicate that the layers buried below the volcanic tephra horizon antedate 40,000 calendar years. The anchor for the Kostenki chronology is the tephra itself, which is now identified as the Campanian Ignimbrite (CI) Y5, derived from southern Italy. Although Hawks states that this is not new, the statement is misleading. The tephra at Kostenki was tied to a source in Italy (Campi Flegrei) in the 1980s, but not to a specific eruption--only a range of possible eruptions prior to 30,000 years BP. It was identified as the CI Y5 tephra by David Pyle (Cambridge University) on the basis of samples collected at Kostenki 14 and other localities in August 2002. Pyle's results have only just been published in Quaternary Science Reviews. In the fall of 2005, Biagio Giaccio (CNR, Rome) independently confirmed this identification using a sample that I collected in 2004 from Borshchevo 5 (one of a related group of sites located several km southeast of Kostenki). The CI Y5 tephra exhibits a highly characteristic geochemical signature, and is readily distinguished from other ashes.
The Middle/Upper Paleolithic Transition in EuropeThe CI Y5 represents a massive eruption that is now dated by Ar/Ar in the eastern Mediterranean to ~40,000-39,000 calendar years ago. Its stratigraphic and chronologic position is well established in Italy, where it is correlated with the Greenland ice core record, paleomagnetic stratigraphy, and other sequences for this time period (uncorrected radiocarbon dates in these sites are similar to the dates from Kostenki). In my view, the chronology put together by our Italian colleagues is the most reliable and informative in Europe for the Middle/Upper Paleolithic transition. The Italian sequence indicates that "proto-Aurignacian" assemblages (Paul Mellars has suggested that we should call these "Fumanian" because they differ from the classic Aurigancian in many respects) immediately underlie the CI Y5 tephra and date to about 42,000-40,000 years. Their appearance is correlated with warm phases GIS 10 & 9 in the Greenland ice core record (GISP2). Below these lie transitional assemblages assigned to the regional Uluzzian industry and--like the Chatelperronian and Szeletian--probably were produced by local Neandertals. Typical or classic Aurignacian assemblages overlie the tephra.
Because the CI Y5 tephra was deposited across a large portion of south-central and eastern Europe, it offers a useful chrono-stratigraphic marker for the Middle/Upper Paleolithic transition in places outside Italy as well. The CI Y5 ash is identified at Temnata Cave in Bulgaria, where it overlies early Aurignacian assemblages that some believe reveal a sequence of development from an older local industry.
At Kostenki, the CI Y5 tephra overlies several early Upper Paleolithic layers that contain a variety of assemblages. The tephrochronology is supported by paleomagnetism and a series of OSL dates. In fact, even the calibrated radiocarbon dates suggest that the layers below the tephra are more than 40,000 years, but the OSL dates suggest that the lowest occupation levels at Kostenki 12 and 14 are at least a few thousand years older. On the basis of recent palynological analyses at Kostenki 12, G. M. Levkovskaya (Institute of the History of Material Culture, RAS) suggests that the lowest levels may be correlated with the GIS 12 warm interval in the Greenland ice core record, which is dated at ~45,000 years.
Transitional vs Non-transitionalAlthough the various assemblages recovered from the layers underlying the tephra at Kostenki cannot be readily fit into Upper Paleolithic cultural classifications devised for western and central Europe, they do fall broadly into the categories of "non-transitional" and "transitional," which parallels a pattern seen throughout the continent. In his comments, Hawks states that our discussion of the two assemblage types is contradictory, but I think this is due largely to an unfortunate choice of verb ("represent" rather than "contain") in the opening sentence of the discussion. What is significant about the assemblages in the lowest levels at Kostenki is the presence of an industry that exhibits a sharp break with the Middle Paleolithic. The contents of the assemblage in Layer IVb at Kostenki 14--worked bone and antler tools, perforated shells imported from distances of at least 500 km, a carved piece of mammoth ivory (whether it represents a piece of art or technology)--these are items that are completely lacking in the Middle Paleolithic of eastern Europe. The stone artifacts, which include small prismatic blade cores, bladelets, and burins, also are not typical of Middle Paleolithic or transitional assemblages.
Although associated human skeletal remains are limited to a couple of isolated teeth, we assume that modern humans made these artifacts, and this leads to the key conclusion of the paper--that modern humans appeared on the central East European Plain at least as early as they appeared in places like Bulgaria and Italy.
The presence of transitional assemblages in the form of those assigned to Strelets does not contradict or negate this conclusion. The Strelets assemblages represent a classic transitional industry that is dominated by Middle Paleolithic technology and tool types and devoid of elements exclusively associated with industries produced by modern humans (e.g., figurative art). But the makers of these assemblages, which are found both below and above the tephra at Kostenki, is not clear. As noted in the Science paper, they may represent an activity variant (e.g., kill-butchery tool kit and associated debris) and a prime example is found in Layer III of Kostenki 12, where the Strelets tools and few waste flakes are found with a mass of horse and reindeer bones. Recently, a similar assemblage was recovered in an early Upper Paleolithic context at Mamontovaya Kurya on the Arctic Circle in northern Russia. I doubt that Neandertals made the tools at Mamontovaya Kurya. Nevertheless, the problem of the Strelets assemblages remains unresolved; they may or may not represent a true analog of the Chatelperronian or Uluzzian in eastern Europe.
Chronological FrameworksIn his final comment, Hawks notes that the limitations of radiocarbon dating the Middle/Upper Paleolithic transition revealed at Kostenki have implications for other sites and regions. On this point, I fully concur. Sites and assemblages dating to the period of the transition cannot be dated by radiocarbon alone, because both transitional and non-transitional industries are present in Europe prior to 40,000 calendar years ago. The chronological framework must be based primarily on other methods, such as tephrochronology, paleomagnetic stratigraphy, and OSL-preferably as many as possible at any one location.
John F. Hoffecker
Institute of Arctic and Alpine Research
University of Colorado at Boulder
16 January 2007