A Tale of Two Migrations: Reconciling Recent Biological and Archaeological Evidence for the Pleistocene Peopling of the Americas
Bonnie L. Pitblado 2011 Journal of Archaeological Research 19:327–375
This gem of an article, currently available as a free download from the JAR website, avoids being dogmatic while reviewing the past decade of scientific research.
What does the recent scientific writing reveal? “that the peopling of the New World occurred in two pulses, both originating in southern Siberia. The first proceeded along the Pacific Rim and coast of Alaska via watercraft 16,000–15,000 years ago, and the second 1,000 years later on land by way of Beringia and the ice-free corridor.” (p. 329) The groups in the first pulse used boats, were focused on coastal and other water-based resources, and eventually made it down to the Monte Verde site in Chile. The members of the second pulse, as they were moving down an interior ice-free corridor, were obviously not coastal-oriented, and it is the latter group that developed the distinctive Clovis technology.
Pitblado begins with a light rebuttal of the Clovis First/Pre-Clovis manufactured dichotomy, saying “in fact, the number of archaeologists who claim that there is a ‘‘traditional’’ peopling model far exceeds the number of archaeologists who have ever advocated the model in print.” (p. 329) Instead, she (following Yesner et al. 2004 and Turner 2002) recommends “a peopling hypothesis that is maximally holistic, parsimonious, and offers the greatest concordance of disparate lines of evidence. We should endeavor to reconcile as many well-established data points as we can, being explicit about what we consider solid evidence and where we perceive gaps that could guide future research or that might even be deal-breakers for our own scenarios” (p. 330), adding later that “this might add a dose of humility to the enterprise that could help us move beyond the gratuitous nastiness that surfaces increasingly often in peopling-related manuscripts. Adovasio and Page (2003), Fiedel (2002), and Grayson and Meltzer (2003) are just a few of the otherwise outstanding practitioners whose occasionally toxic rhetoric undermines their scientific positions and repels some readers.” (p. 354) Nice. And she’s not afraid to name names, though several more could probably be added to her list.
It’s worth quoting at length what Pitblado considers unequivocal evidence:
“(1) A number of sites in North and South America have been convincingly shown to predate Clovis, but none of the most convincing or widely accepted more than slightly predates c. 15 k cal. B.P. We can debate which sites make the list, but even the most skeptical among us must recognize that such a list exists. (2) Some sites unequivocally shown to date to Clovis time do not share the Clovis predilection for mammoth hunting (e.g., marine-oriented Quebrada Jaguay, Peru). (3) Genetically, Native Americans today express a limited number of mtDNA and NRY haplogroups, but most are widespread in the hemisphere. They share that range of haplogroups with precious few Old World populations. First Nations people also share a few nuclear DNA markers with populations in western Beringia and nowhere else. (4) The oldest New World skeletal remains show a generalized morphology akin to that of Australians and Africans but different from Native Americans and East Asians. (5) The Pacific coast of Beringia and the Northwest was ice free-by 16,000-15,000 years ago. A corridor between the Laurentide and Cordilleran Ice Sheets deglaciated later, c. 14,000-13,500 years ago.” (p. 330)
My first impression of this list is positive, although re: #2, I suspect a lot of Clovis sites don’t actually show much evidence for mammoth hunting, either. Those Old World groups in #3 are all from Siberia (a.k.a. North and/or East Asia), with different genetic researchers sometimes choosing different regions within this broad area. Those groups of people who did actually cross over into the New World were likely descended from people who had been living in western Beringia (a.k.a. eastern Siberia) for at least a few thousand years (see page 334).
Because of the nature of genetic analysis, the timing of the movement to the New World is more uncertain. Significantly, “most researchers are deriving significantly later dates for colonization of the New World than the 20,000-30,000 or more years ago often asserted in early studies” (p. 334), with most centering around 18-15k cal. B.P. Pitblado does not discuss whether these genetic estimates can actually determine when populations crossed over to the New World, as opposed to when populations only became genetically isolated from other Old World ones (e.g., while still remaining in western Beringia).
Regarding the number of discrete “founding populations”, most genetic researchers pick one or two, and “Few, if any, geneticists argue that the peopling of the Americas occurred over an extended period of time” (p. 335), not surprising given that genetic analyses are predicated on identifying divergences. Therefore, it’s rather notable that “some proponents of the single-founder model (e.g., Goebel et al. 2008; Hey 2005) have noted that if the same source population generated a migratory pulse more than once, this could have yielded descendant genetic signatures and distributions that are essentially identical to one another, particularly given the gene flow that likely took place among such populations before and after the Americas were settled.” (p. 336) To her credit, Pitblado “conclude[s] that based on genetic data alone, the number of founding populations is not resolvable, or in any event has not yet been resolved.” (p. 336)
Moving onto the osteological evidence from the New World, Pitblado does not say precisely how many Late Pleistocene or very early Holocene skeletons have been found (the number is very small). Those that have been found have some morphological characteristics that differ from modern Native Americans and are similar to contemporary or older skeletons from Asia and Europe, suggesting, not surprisingly, that populations in both the Old and New Worlds continued to evolve after the first groups of people came to the New World.
Finally, we get to the archaeological data. There is currently a gap in western Beringia between about 32 kya and 14 kya in which no well-dated sites are known. The earliest sites in eastern Beringia (a.k.a. Alaska) date to at least 14.2kya, and similar-aged sites are also known in western Beringia.
There are still two possible routes into the New World: coastal and inland (She handily dismisses the Atlantic migration proposed by Bradley and Stanford in two sentences [p. 355]). The coastal path “was open and populated by plants and animals by 16 k-15 k cal. B.P.” while “The interior ice-free corridor… did not open and attain ecological viability such that people could have traversed it until 14,000–13,500 cal. B.P.” (p. 346; see article for references)
Sites mentioned along the coastal route include Monte Verde (actually 58 km inland) in Chile, Arlington Springs, on one of the Channel Islands off of California, which has human remains dated to 13,100-13,000 cal B.P., and Quebrada Jaguay, among others, in Peru. The latter two sites are roughly contemporary with Clovis. There is also Paisley 5 Mile Point Caves in Oregon, where human coprolites dated 14,270-14,000 cal. B.P. have been found (p. 349). Meanwhile, “in some areas Clovis people whose ancestors entered the New World via the IFC [ice-free corridor] were, in fact, first” (p. 351) In other words, “By 13,100 cal. B.P. two migratory waves of hunter-gatherers with fundamentally different lifeways are visible in the archaeological record at the continental scale.” (p. 352) Pitblado notes that this does not mean these two waves necessarily differed genetically.
Pitblado acknowledges sites from the eastern and midwestern United States raise problems for her narrative. Topper (South Carolina) and Cactus Hill (Virginia) “both yielded absolutely dated Clovis components, and both have components underlying Clovis dated by radiocarbon or OSL to c. 20,000-16,000 years ago” (p. 355); Meadowcroft Rock Shelter (Pennsylvania), of course (somehow I missed that “A minimum date on carbonized basketry places humans at Meadowcroft by 18,500 cal. B.P.” [p. 355]); Page-Ladson in Florida; Schaefer and Hebior (Wisconsin), both mammoth sites dating to 14,800-14,200 cal. B.P.; and Lovewell (Kansas) and La Sena (Nebraska), also mammoth sites that date to between 22 and 19k cal. B.P. Of the four mammoth sites, it sounds like only Schaefer has any possible human made stone tools, and Lovewell and La Sena have no human artifacts or human-made marks on the mammoth bones (the researcher argues that spirally fractured mammoth bones with evidence of dynamic loading indicate human activity). Pitblado adds that “Goebel et al. (2008, p. 1500) summarized the challenges: insufficient reporting for Page-Ladson, potential natural origin of alleged stone tools in early Topper levels, charcoal translocation at Cactus Hill, contamination at Meadowcroft, and the possibility that natural forces caused the purported human alteration of mammoth bone at Hebior, Schaefer, La Sena, and Lovewell.” (p. 356) If all these sites are truly archaeological and accurately dated, they imply migration from Beringia through the ice-free corridor sometime prior to 32 kya.”
So, as always, more research needs to be done, but Bonnie Pitblado shows that our knowledge of the first peopling of the Americas has truly increased over the past ten years.