Across Atlantic Ice

by Dennis J. Stanford

  EXTRAORDINARY BIFACE CACHES

  Extra-large, extremely well made bifaces have been found buried in groups, sometimes with other kinds of artifacts and frequently with concentrations of red ocher. This caching of extraordinary artifacts has been found in only two more-than-13,000-year-old archaeological cultures: Solutrean (figure 7.8a) and Clovis (figure 7.8b). For both cultures, researchers have suggested that some of the artifacts were made for non-utilitarian purposes and that their very manufacture held special symbolic and ritual meaning.17

  In all of western Beringia there is only a single reported biface cache, so called even though the cluster of eight bifaces was found scattered around a small area in a final stage Dyuktai occupation level at the Tumular Site in the Aldan River Valley, Siberia.18 The bifaces in this cache are likely preforms for microblade cores and in any event not comparable to the caches of large bifaces found in Solutrean and Clovis contexts. They are not of special manufacture and probably represent provisioning behavior: leaving a collection of microblade precores for future retrieval. Two similar provisioning caches of precores have been reported from eastern Beringia, one at Onion Portage on the Kobuk River and other near Point Barrow.19 A radiocarbon date from the Onion Portage Cache suggests that both caches are likely post-Clovis in age.

  FIGURE 7.8.

  Oversize bifaces from caches: (a) Solutrean; (b) Clovis. Outlines show the extremely thin sections.

  SHELTERS

  Remnants of shelters or structures have been described in western Beringia, specifically at Ushki Level VI. These take the form of shallow circular depressions, often with a fire pit near the center. Some artifact clusters in eastern Beringian sites have been interpreted as indicating outlines of former shelters.

  Evidence of constructed shelters is absent from the Solutrean, but this is probably because most of the excavations have been in caves and rock shelters, where they were not necessary. Purposefully laid cobble floors found at early Magdalenian sites in southwestern France are, however, possible remnants of dwellings (figure 7.9a).20 These features consist of cobble pavements laid out in square or rectangular patterns measuring a few meters on each side. Some also exhibit small extensions in the center of one or more sides. Artifacts tend to be present in concentrations in and around the features. In addition, all but two are aligned with the cardinal directions.

  FIGURE 7.9.

  Comparison of structure pebble floors: (a) Lower Magdalenian, France; (b) Clovis/pre-Clovis, Gault Site, Texas.

  The rectangular pebble floor at the Gault Site (figure 7.9b) is highly reminiscent of these European features. It measures about 2 by 2 meters, is oriented in the cardinal directions, and has a small portico square centered at the north end. Use-wear studies of blade tools found around this floor indicate that reed matting was manufactured here, perhaps for flooring or external coverings.21 To our knowledge this type of floor does not appear in other Paleo-Indian campsites and may represent a long-lasting European Paleolithic construction technique.

  FIRE HEARTHS

  A distinguishing feature of Paleo-Indian sites is that their fire hearths were not stone lined. By contrast, many Solutrean fire hearths are lined with cobbles, although examples of unlined fire hearths were found in the lower Solutrean levels 4, 5, and 6 at La Riera Cave.22 It is probably not coincidental that these are the levels that produced all of the indented base projectile points found in situ, as well as exotic cherts, seal bones, and major accumulations of limpet shells and fish remains. This suggests that these levels represent temporary hunting camps used by people who were based on the coast and likely exploited the resources of the Atlantic Ocean. Their artifacts and camp debris are most comparable to the American pre-Clovis and Clovis camp situations. If Clovis descended from this branch of Solutrean peoples, perhaps unlined Paleo-Indian hearths were a continuation of the latter’s practice.

  MORTUARY PRACTICES

  While it is difficult to use lack of evidence as evidence of a relationship, we must point out that no Solutrean or Clovis burials have been found. Unlike with other Paleolithic European, Siberian, and later North American cultures, which all have evidence of mortuary customs, we have no idea how the earliest Paleo-Americans or Solutrean people buried their dead. Both groups apparently used a form of burial that did not preserve the human remains. Their customs may have been similar to the historic scaffold type of burial, wherein the body was left in an open environment to facilitate transformation to another dimension. Some investigators think the remains of two individuals recovered at Anzick Site in Montana are associated with the Clovis artifacts there.23 However, radiocarbon dating of the human bones indicates that they were placed near the cache 400 or more years later.

  The lack of human remains from both cultures renders it impossible to assess their paleo-genetic relationships.

  CONTRARY EVIDENCE IN ARTIFACT TYPES

  We would be remiss if we did not discuss typical Solutrean artifacts that are not present in the New World. Given the changing environmental conditions and adaptations to these shifts during the time span represented by the Solutrean–pre-Clovis–Clovis continuum, it is no surprise that some tool types would fall out of favor and new types be introduced. The resulting differences among the assemblages can be as interesting and explanatory as the similarities.

  PROJECTILE POINTS

  Some Solutrean projectile point styles are absent from pre-Clovis and Clovis assemblages, including several varieties of shouldered points primarily defined by their method of manufacture, willow leaf points, and stemmed and corner-notched points.

  In France, blade-based shouldered points occur in Solutrean deposits after 19,000 years ago. They exhibit unifacial pressure flaking, although many examples are partly flaked on both faces and some are completely bifacial. In northern Spain, Solutrean shouldered points were mainly made by limited, abrupt, probably anvil-supported percussion on blades. The conclusions of several studies, including replication and use experiments, are that they were used to tip projectiles, but whether for darts or arrows is still being debated.

  If pre-Clovis and Clovis tool forms descended from the Solutrean, why are shouldered points absent from the former? There are several possible answers. First, it is common for some artifact types to drop out of a cultural inventory, especially but not necessarily if functional requirements change. Second, not all Solutrean assemblages include shouldered points. If the particular Solutrean people who eventually found their way to North America came from a group or groups that didn’t use shouldered points—perhaps, based on radiocarbon dates from the Chesapeake Bay sites, because they left Europe before the advent of this technology—one wouldn’t expect these artifacts in North American assemblages.

  In some Solutrean sites several point types occur in the same occupation. Such a mix may have many explanations, particularly in an area geographically restricted for thousands of years like the Vasco-Cantabrian region. Different bands may have used socially identifiable, unique symbols such as point types and other markings, so middens with mixed types may be the result of such circumstances as cooperative hunting ventures or occupation by different contemporary groups operating in overlapping territories.24

  Some French Solutrean sites have yielded a form of artifact called a willow leaf point. These resemble the finer pressure-flaked Solutrean shouldered points, but they lack the shoulder. To our knowledge these points have not been subjected to use-wear studies, but along with many of the thin laurel leaf points, they may have been used as knives. Again, this form developed late in the French Solutrean and may not have been in the inventory of the people who adopted a maritime economy and eventually traveled to North America.

  Stemmed and corner-notched projectile points have been found in Mediterranean and Portuguese Solutrean sites. These weapon tips, thought by some researchers to be arrowheads, are not present in northern Spain or southwestern France, so there is no reason to expect Vasco-Cantabrian descendants in North America to have used them. C
onversely, indented base points do not appear south of the Cantabrian Range.

  In our hypothesis, technological innovations also occurred after people arrived in the New World. Even though many of the Spanish indented base points are basally thinned and a few might be classified as multiple fluted, we do not subscribe to the idea that there is an Old World antecedent of fluting technology. Fluting is not a common means of basally thinning bifaces in other times and places and is probably not a major technological advantage. It is unclear how and why it developed, and since it does occur in some pre-Clovis bifaces, we suspect that this particular technique was an American invention.

  BACKED BLADELETS

  Small blades and bladelets were part of most Upper Paleolithic assemblages in western Europe. Some consider the paucity of backed bladelets in Clovis assemblages significant, but like shouldered points they are not universal in the Solutrean either. For example, at La Riera there are no backed bladelets in nearly a third of the lower occupation levels; moreover, 81 percent of the artifacts at La Riera came from occupation levels that did not contain any stone projectile points, and roughly 75 percent of them came from the two upper occupation levels.25 These data suggest that by later Solutrean times slotted bone or antler projectile points were the primary weapon tip for hunting in this upland setting.

  Backed bladelets may have been part of a variety of items other than projectile points but were mainly used like razor blades and may have been inserted into slotted bone knives. Sagaies slotted for inset blades appear during the final stages of the Solutrean occupations of northern Spain, likely as a result of influence from other European Paleolithic technologies. Since slotted sagaies have not been found in any Clovis context, it appears that Solutrean people adopted this technology after the ancestors of pre-Clovis found North America, or in subgroups not directly related to the New World descendants. On the other hand, slotted sagaies armed with microblade insets became the dominant weapon technology in western Beringia long before any putative Clovis ancestors might have crossed the Bering Sea.

  BURINS AND ADZES

  A burin is a tool created by striking a flake or flakes from the edge of an artifact to produce a sharp 90 degree edge for working hard substances such as ivory, antler, and bone. The edge fragment, usually triangular or square in cross section, that was removed to make the burin is called a spall. Although a few burins have been found in Clovis artifact assemblages, they are rare. A burin and a spall were found at Miles Point, and a burinated blade was recovered from Jefferson Island. A burin-like tool was also recovered at the Coats-Hines mastodon butchering site in Tennessee. By contrast, at La Riera Cave burins make up about 10 percent of the stone tools recovered from the Solutrean levels. They are also extremely common throughout the Siberian and Beringian Paleolithic cultures. Apparently, for some reason burins lost their importance and were dropped from Clovis technology, regardless of whether their Paleolithic origins were in Iberia or Siberia. However, burin-like tools that may have functioned in the same way have been recovered. These were produced not by edge removal but rather by radial breakage of thin, mostly biface flakes.26

  Woodworking adzes occur in some Clovis and pre-Clovis sites but not in Solutrean or Beringian collections. We believe that this tool type has not been found in Solutrean assemblages because the sources of wood in the upland areas where Solutrean material culture has been defined were limited. Driftwood from North America that washed up along the coast would have been the only source of large timber for the Solutrean people. Consequently, evidence of adzes and other tools that may have been used to make boats or for other construction projects are now underwater.

  CLOSING THE CHRONOLOGICAL GAP

  The impression that the Solutrean and Clovis are separated by approximately 5,000 years is often cited as the fatal flaw of our hypothesis. Unless there exists a reasonable series of dated archaeological cultures with intermediate bridging technologies to fill the gap, like those before us who saw the similarities of Solutrean and Clovis technologies, we would have to consider the shared traits as independent inventions rather than historic relatives, whether related to Beringia or the Solutrean. The oldest of these bridging technologies would need to have artifacts that are nearly identical to the parent Solutrean technology, and the youngest should begin to closely match Clovis technology.

  FIGURE 7.10.

  Radiocarbon date ranges of the Solutrean, Mid-Atlantic Early Paleo-American, Southeast Early Paleo-American, and Clovis.

  This is where the pre-Clovis evidence from the eastern United States fits in (figure 7.10). The radiocarbon date of the Cinmar mammoth indicates that the Solutrean-style laurel leaf biface associated with it may date to at least 22,760 years old (see chapter 4). The Cinmar age is consistent with the date of greater than 21,000 radiocarbon years before the present for the occupation of Miles Point. The artifact assemblage from Miles Point includes biface projectile points, blades, scrapers, and burins that are technologically close to artifacts found in Solutrean levels 4, 5, and 6 at La Riera Cave that date to 20,970±620 RCYBP. The 16,940±50 RCYBP date for the pre-Clovis occupation at Cactus Hill also overlaps with the dates of later Solutrean occupations of the Vasco-Cantabrian area of Spain. However, the pre-Clovis artifact assemblage from Cactus Hill does not include the late Solutrean inset bladelet technology. At Page-Ladson, dated to circa 12,388 RCYBP, and related sites such as Jefferson Island, we see the continuation of blade and burin technology, but the projectile point technology now includes larger, unfluted “Clovis-like” projectile points. At the Johnson Site in Nashville, Tennessee, early attempts at systematic basal thinning are evident, but the Clovis technique of platform preparation for fluting has not been perfected, and there is a high rate of failure during final fluting attempts (figure 4.11). The Johnson Site bifaces are associated with Solutrean-style artifacts including a plane face point (figure 7.1m) and Solutrean scrapers (figure 7.1a–c) and blades. This evidence suggests that Johnson was occupied by proto-Clovis people, provides a date for the development of techniques to flute Clovis points, and serves as a transition between early Mid-Atlantic Paleo-American and Clovis technologies.

  The Cinmar date of 22,760 RCYBP pushes the presence of people in North America back to the time of the Solutrean culture. Clearly the chronological gap has been closed, and the high number of corresponding flaked stone and bone tool forms and technologies make a historical connection feasible.

  COMPARATIVE SUMMARY

  We have compared the chronology, tool types, and technology of the archaeological remains in late Pleistocene Beringia, continental North America, and southwestern Europe in an attempt to identify historical relationships. When we had large enough samples of artifacts, we compared tool assemblages and flaked stone technology and submitted them to cluster analysis. This allowed us to see what assemblages clustered, how they clustered, and what distinguishing attributes determined the clustering.

  Our cluster analysis distinguished between different types of sites in fluted point and eastern Beringian assemblages. However, it also indicated to us how inconsistently archaeologists have recovered, analyzed, and reported sites and assemblages over the past century. This is a major obstacle to investigating the origins of people in the New World, but our initial efforts show that some reasonable conclusions can be drawn.

  Comparisons of less quantifiable aspects of the archaeological record, including rarely recovered artifact types and various behaviors, have shown strong similarities between the fluted point and Solutrean traditions and a near lack of correspondence between fluted point and early Beringian traditions. The Nenana Complex shows the strongest similarities with Clovis, but Nenana is at best contemporaneous with Clovis and significantly younger than early Mid-Atlantic sites, raising the question of which may have been derived from which. Based on the available data, we would argue it is likely that certain aspects of Nenana and the Arctic Paleo-Indian assemblages were derived from people to the south, perhaps related to a northern br
anch of Solutrean descendants, rather than the other way around.27

  There are sufficient similarities between all of the circa 12,000-to-14,000-year-old archaeological cultures of eastern Beringia to suggest that they originated from the Upper Paleolithic microblade traditions of Eurasia. At the same time, in our opinion, the technological, behavioral, and dating evidence overwhelmingly support the theory that the fluted point traditions in North America derived from a regional and chronological variant of the Solutrean cultures of southwestern Europe. The only alternative, as opined by the archaeologist Gary Haynes, is to assume that the complex technologies of the fluted point traditions sprang out of an entirely alien Asian technology without transitional phases.28 As flintknappers and lithic technologists, we seriously doubt that such a scenario ever happened; complex technologies require developmental antecedents even if introduced by foreigners.29

  Although we have used general Solutrean artifacts, technologies, and behaviors in this analysis, we think it unlikely that all areas where Solutrean culture existed were directly involved in the colonization of eastern North America. It is more likely that only one relatively small area within the Solutrean sphere contributed to this process. The best fit we can find for the site of pre-Clovis ancestors, considering both the artifact assemblages and the known environment at the time of greatly lowered sea levels, is the area adjacent to the Bay of Biscay, where the Pyrenees meet the sea. Here the ancient beach line would have been clear of ice at times and would have extended almost straight along the Celtic continental shelf to several hundred kilometers west of glacier-covered Ireland. In other words, we hypothesize that the people who brought Solutrean traditions into the Americas were from the modern Aquitaine of France and Vasco-Cantabria of Spain—the Basque areas of southwestern Europe, and perhaps more specifically the adjacent submerged areas in the Bay of Biscay.