A publication of the Archaeological Institute of America
The debate about Monte Verde raises several important questions. First, what does it say about how archaeologists prove, or disprove, particular claims for sites? Second, where does it leave us in terms of when and how the Americas were colonized?
THE BURDEN OF PROOF
The ongoing debate about Monte Verde raises questions about how sites are interpreted. After all, it would seem to be a simple thing to show that a site was or was not occupied before the ~11,500 B.P. Clovis barrier. Two sorts of trials have been used: sets of criteria that must be met and examination of the site by outside experts.
From the time of the Folsom discoveries of fluted points with the bones of extinct bison (Bison antiquus) in 1926 and 1927 to today, various criteria have been proposed to demonstrate first the great antiquity of people in the New World and later to demonstrate a pre-Clovis occupation. C. Vance Haynes (1969: 714) summarized the criteria as follows: "For establishing man's presence, the minimum requirements met for the Folsom site still apply for future excavations. The primary requirement is a human skeleton, or an assemblage of artifacts that are clearly the work of man. Next, this evidence must lie in situ within undisturbed geological deposits in order to clearly demonstrate the primary association of artifacts with stratigraphy. Lastly, the minimum age of the site must be demonstrated by primary association with fossils of known age or with material suitable for reliable isotopic age dating." E. James Dixon (1999: 48), based on Haynes and on Stanford (1983: 65), suggests the following questions must be answered:
1. Are the artifacts clearly the product of human manufacture?
2. Is the recovered material within clear stratigraphic context?
3. Are there reliable, concordant, and stratigraphically consistent radiocarbon dates from the deposit?
4. Are paleoenvironmental studies consistent with ages assigned to the site?
5. Are there human remains that are reliably dated older than 11,500 B.P.?
If there are no human skeletal remains or indirect evidence of humans such as preserved footprints, the first four questions must all be answered positively if a site is to be considered pre-11,500 B.P.
David Meltzer of Southern Methodist University has observed (1993: 62-63) that these criteria are not too different from those used in Africa "to identify the earliest (and often the most ambiguous) traces of human behavior found with pre-modern humans--traces which seem uncommonly uncontroversial. Yet...the criteria have proven extraordinarily difficult to meet. So difficult, in fact, that proponents cry foul. The criteria, they charge, are excessively rigid, and were they generally applied to sites of all ages we would be left wondering whether, for example, anyone ever occupied Eastern North America after 11,500 B.P. These are serious charges: Are the criteria so perversely narrow that they blind archaeologists from seeing legitimate evidence? A closer look is in order."
Meltzer then notes some of the difficulties facing those trying to demonstrate a valid pre-Clovis occupation according to such criteria: many artifacts would have been made of perishable materials and did not survive, less modified tools (like a river cobble used to smash open bones for marrow) might be difficult to recognize as tools or to distinguish from naturally made counterparts (like a cobble flaked while tumbling in a stream). Context, he says, can make a difference: a river cobble by a hearth at a campsite not near stream might be more convincing as an artifact than one found in a riverbed.
At the same time, Meltzer(1993: 55-56) admits that in the late 1800s and the first decades of the twentieth century, "Skepticism about a genuine Pleistocene human presence was not arbitrary, but forged in the face of repeated cases which failed to withstand scrutiny--cases often based, as proponents themselves admitted, upon unverified finds made under appalling field conditions by untrained farmers or collectors." I think the same standards of proof, or higher since we have better excavation and analytical techniques available to us, must be applied today as those 100 years ago. Is it difficult to meet them? Yes. Should it be? Yes. Lowering the bar will not get us closer to understanding the peopling of the Americas.
Over the past 150 years the examination of archaeological sites by outside experts has been seen as an important tool in establishing their validity (Meltzer 1993: 44-48). Site visits played an important part in the first acceptance of humanity's great antiquity when British geologist Sir Charles Lyell traveled to Abbéville, France, in 1859 to examine firsthand the findings of Jacques Boucher de Perthes who for decades had collected stone tools and fossils from sites in the Somme River Valley. In the Americas, site visits were critical in the debunking of the supposed Palaeolithic tools found beginning early in the 1870s by physician Charles Conrad Abbott near Trenton New Jersey.
Meltzer asks (1993: 55-56), "Could it possibly be that easy nowadays? Is the price of resolving the ongoing human antiquity dispute in America no more than the cost of the plane tickets to fly half a dozen skeptics to a corner of South America or the outskirts of Pittsburgh for an on-site evaluation?" His own experience should convince him, and others, that the answer is no. Meltzer was part of the panel that validated Monte Verde in 1997 and was one of three Paleoindian specialists who conducted a similar visit to Pedra Furada, a Brazilian another pre-Clovis hopeful, in 1993. (The others were James Adovasio of Mercyhurst College, excavator of Meadowcroft Rockshelter, and Tom Dillehay of the University of Kentucky, excavator of Monte Verde). Meltzer et al. (1994) critiqued the evidence from Pedra Furada in the journal Antiquity, concluding that "Obviously, we are sceptical of the claims for a Pleistocene human presence at Pedra Furada, and in our view the concerns raised here must be resolved before this potentially important site is accepted (at least by us)." Given the authors' stature in American Paleoindian scholarship, this conclusion was damning and Pedra Furada has been virtually written off. The French excavators of the site countered in a subsequent issue of Antiquity, claiming that the critique was based on "partial data and false information." Their response, not in my mind convincing, was pointless anyway because most had already accepted Meltzer, Adovasio, and Dillehay's verdict. Today, with Meadowcroft not free of all doubt as pre-Clovis and Monte Verde under fire, one might excuse the excavators of Pedra Furada if they looked askance at all of this.
What can a site visit accomplish? Today much of the archaeologist's work is done not in the field but in the laboratory, and not in one place by a single scholar but in many places by dozens of specialists. One can look at the site during excavation to check the quality of the fieldwork and to examine firsthand the stratigraphy, but beyond that the critical body of evidence has to be the final site publication in which the excavation is set forth in sufficient detail so that anybody can review the data and interpretation of it. In that respect one can note that Tom Dillehay has done just that for Monte Verde, made a review possible through his publication. And on his part, Stuart Fiedel has undertaken just such a review.
BEYOND MONTE VERDE
Even if doubts are cast on Monte Verde as being a bona fide pre-Clovis site, there are other contenders out there, notably Cactus Hill in Virginia and, possibly, the Topper Site in South Carolina. Less well known, but perhaps stronger candidates are the Schaefer and Hebior sites in southeastern Wisconsin. So even if Monte Verde goes from certain pre-Clovis status to question mark, the Clovis barrier itself may still be broken and a return to the Clovis-first model of New World colonization may not be possible.
Cactus Hill. The following description of Cactus Hill is based E. James Dixon (1999: pp. 73-75). Four distinct localities (designated A, B, C, and D) have been identified at this site on the Nottoway River in southeastern Virginia. In area B, artifacts have been recovered beneath a Clovis level. The small assemblage includes triangular projectile points, waste flakes, retouched flakes, retouched blade-like flakes, and cores of local quartzite (McAvoy and McAvoy 1997). Radiocarbon dates from hearth-like features in this level are 16,670 ± 730 B.P. and 15,070 ± 70 B.P., while the Clovis layer above it was dated to 10,920 ± 250 B.P.
Topper Site. Excavations have revealed a deep stratum with apparently pre-Clovis artifacts at the Topper site on the Savannah River near Allendale, South Carolina. Albert Goodyear, of the University of South Carolina's Institute for Archaeology and Anthropology, was surveying chert sources in 1981, when a local man (named Topper) led him to this site, which is on a hillside near the Savannah River (500 feet from main branch). Testing in 1984 revealed side-notched points, dated elsewhere to 10,000 (radiocarbon) years before present, were found at 70 to 80 cm and fluted blanks (Clovis preforms) were found at 80 to 100 cm. Later excavations never went beyond the one meter mark. At the time, no site had been accepted as older than Clovis (10,800 to 11,200 radiocarbon years), and there was therefore no reason to expect deeper culture-bearing deposits existed.
Inspired by potential pre-Clovis sites like Monte Verde, Chile, and Cactus Hill, Virginia, Goodyear decided to dig deeper in 1998. After some 40 cm of essentially barren deposits, the excavators began finding small flakes and microtools. Goodyear recalls that he "kind of went into shock. I had no idea we'd find artifacts." This year's excavations have confirmed that discovery.
The lower level, now exposed over a total of 28 square meters, has yielded some 1,000 waste flakes and 15 microtools (mostly microblades). The excavators also found a pile of 20 chert pebbles plus four small quartz pebbles, possible hammerstones. The same yellow chert was used in the upper and lower levels, but apparently in the upper levels the people had access to large pieces of chert extracted from the hillside and cobbles of it from the riverbed, while in the deeper level only small pebbles of it were used. Because artifacts of the types in the upper level are not found in the lower level and vice versa, Goodyear does not believe the flakes and tools were pushed into the lower level by tree roots or burrowing animals.
For now, dating of the artifacts depends on the stratigraphy and comparison with other sites. There is little organic material preserved in the sandy matrix making radiocarbon-dating difficult. Samples for carbon-dating and OSL (optically stimulated luminescence) dating are now being analyzed. Goodyear thinks the site was used for the exploitation of chert pebbles sometime between 12,000 and 20,000 years ago. No evidence of bifaces and unifaces typical of later Clovis have been found in the lower level, and Goodyear looks to Siberian microblade industries for parallels. Artifacts from possible pre-Clovis sites, including Topper, Cactus Hill, and Meadowcroft (in Pennsylvania), will be shown to Asian scholars at the Smithsonian this August. Plans call for excavations at Topper, which stopped at the end of May, to resume next spring.
Schaefer and Hebior. The final reports on these are not yet out; the following description of them is from E. James Dixon (1999: pp. 66-67):
Excavations began in 1992 at the Schaefer site, where there was a pile of disarticulated bones of woolly mammoth (Mammuthus primigenius). The bones exhibited cut marks, fractures, and striations suggesting they had been butchered and piled at the edge of an ancient lake that gradually dried and became a bog. The bones were subsequently covered with 130-140 cm (51-55 inches) of peat that accumulated over the site (Overstreet et al. 1993).
Subsequent excavation at the Schaefer site in 1993 uncovered chipped-stone implements directly associated with the mammoth bones. A radiocarbon date on bone collagen run on mammoth bone produced a 14C date of 10,960 ± 100 B.P. (Beta 62822), suggesting that the Shaefer site might be a kill site associated with the Chesrow complex. The Chesrow complex is an early archeological complex identified at several sites in southeastern Wisconsin. It is characterized by a variety of lanceolate projectile points characteristically manufactured from local cherts and is contemporaneous with the Clovis complex on the Great Plains (Overstreet 1993; Overstreet et al. 1995).
However, two radiocarbon dates on wood deposited at the site shortly after the mammoth bones dated to 12,220 ± 80 B.P. (Beta 62823) and 12,480 ± 130 B.P. (Beta 62824). This suggests the first bone collagen date was too recent. A highly purified sample of the mammoth bone collagen was subsequently dated to 12,310 ± 60 B.P. (CAMS-30171), and is consistent with the earlier dates run on wood.
The Hebior site is located near Kenosha, Wisconsin. Approximately 25 m2 (270 ft2) of the site was excavated in 1994. The Hebior mammoth was an adult male, and like the Schaefer mammoth, the bones exhibit signs of butchering. Several stone artifacts including flakes, a chopper, and two flaked-stone bifaces were found among the bones (Hall 1995). Two AMS C14 determinations run on purified samples of mammoth bone collagen from the site produced dates of 12,480 ± 60 B.P. (CAMS-28303) and 12,520 ± 50 B.P. (CAMS-24943). These results led Overstreet and Stafford (1997) to conclude that both the Schaefer and Hebior sites resulted from human hunting or scavenging of mammoths in the southwestern Lake Michigan basin by at least 12,500 B.P.
Both the Schaefer and Hebior sites occur in low energy depositional environments, and are located in primary depositional contexts. In these types of depositional environments, there is little reason to attribute either the bone modifications or the lithic artifacts to noncultural processes. Both sites have been professionally excavated, recorded, and reliably dated by the 14C method, Based on the available evidence, both the Schaefer and Hebior sites appear to provide evidence indicating human interaction with mammoths prior to 11,500 B.P.
Review of E. James Dixon's Bones, Boats, and Bison