Across Atlantic Ice

by Dennis J. Stanford

  At present no undisputed remains of Clovis or Solutrean people have been found to provide genetic data for comparisons; thus, grand schemes that link the earliest Americans to either Asia or Europe via DNA evidence are at best only speculation. A mitochondrial genetic marker known as haplogroup mtDNA X has an intriguing distribution: evidence of this haplogroup is found in early pre-Columbian remains in eastern North America, but it does not occur in Asia in prehistoric times or east of the Altai Mountains in modern times. When fully understood, the X haplogroup may provide evidence of early pre-Columbian origins in Europe. The preconceived notion of geneticists and archaeologists that the Americas must have been settled by Asians has produced interesting models to get the Paleolithic X marker out of Asia without leaving a record in the modern eastern Asian population.2

  The distribution of the X haplogroup suggests that it is very old. Within X there are two major branches, X1 and X2.3 X1 is restricted to North and East Africa and the Near East. X2 is found in North Africa, Europe, the Near East, and central Asia, and there is a trace of it in some North American populations. Carriers of X2 are thought to have expanded into Eurasia through the Near East at or near the end of the LGM and therefore would not have been present to add to the North American gene pool before the end of the LGM.

  Native American X2 sequences have a further distinctive branch, known as X2a, which is not found outside the New World. Hence, it may represent the very early split of a population that had a limited distribution in western Europe and did not spread eastward. Traces of X2 are found in both living and prehistoric Basque populations, which are thought to be the descendants of Paleolithic Europeans.4 X2 also occurs among people of the Orkney Islands, an area that was part of the unglaciated North Sea shelf during the LGM. The discovery of LGM sites on the eastern shore of North America and the prehistoric occurrence of the X2 haplogroup in eastern North America suggest that scholars should consider a model of Paleolithic people carrying the X2 haplogroup into western Europe and eventually into the Americas during the LGM. It is interesting that the occurrence of mtDNA X among modern Native Americans has a distribution pattern similar to the occurrence of late Pleistocene sites, with the earliest occurrence and the highest incidence of both in eastern North America and progressively lower ages and incidences toward the west (figure 11.1). Is this only a coincidence?

  To sum it all up, we contend that the archaeological evidence that Clovis predecessors were immigrants from southwestern Europe during the LGM is stronger and more compelling than the evidence that their ancestors were from an Asian microblade tradition that came out of northeastern Asia at the end of the LGM. This is based on dating, comparisons of archaeological assemblages, and environmental reconstructions. Sometime during the last glacial epoch, a biface-thinning, blade-making culture figured out how to exploit the marine environment that was present along the ancient strands and ice edges off southwestern Europe. This may have been a small segment of what is known to archaeologists as the Solutrean. But not all Solutrean people or periods would have been represented in this founding group. We see them as deriving from the middle Solutreans (possibly between 19,000 and 23,000 years ago), who probably inhabited the area along the north shore of the Iberian Peninsula and the western shore of the continental shelf that stretches from southwest France northward along the British Isles, most of which is now under the ocean. These people adjusted their technologies and lifeways to exploit a marine environment, and through generations of exploration some found their way across the North Atlantic to make landfall in northeastern North America. We do not envision an immediate rush of immigrants to colonize the newfound lands. Why would people who had adapted to an adequate marine environment suddenly abandon it for the unknown promise of yet another impoverished terrestrial landscape? As long as the polar environment persisted, we see people focused on the ice edges, tethered at each end by land. Solutreans expanded their territory to include the ice edge, eventually incorporating the eastern shore of North America; they didn’t migrate as a group to the New World.

  FIGURE 11.1.

  World map showing haplogroup X mtDNA distributions. (Data from Izagrirre and de la Rua 1999; Brown et al. 1998; Reidla et al. 2003; Schurr 2000; Schurr et al. 1998; Smith et al. 1999; Stone and Stoneking 1998.)

  The durable tool kit of the founding Solutrean population in North America would have included both biface and large blade tools and a range of bone, antler, and ivory tools. The perishable material culture probably would have amazed us: it would have included cold weather and water-resistant tailored clothing, substantial portable dwellings, and boats.

  As people settled into the New World, we see explorations up rivers and changes in how they made their living as the Pleistocene wound down. Perhaps for a long time, some people maintained a sea margin lifestyle, moving back and forth between the continents. Their base sites would have been mostly along the now submerged coastlines. Climatic changes at the end of the ice age caused the glaciers to melt at a rapid rate. Meltwater from coastal and marine glaciers drained directly into the oceans, while meltwater from terrestrial glaciers collected into massive lakes behind retreating glaciers. Around 14,500 years ago, shortly before the advent of Clovis, receding glaciers opened up lowland channels, such as the Saint Lawrence waterway that drained the lakes and released massive amounts of water into the seas, causing sea levels to rise nearly 20 meters in 300 years. The rapid sea level rise no doubt reduced the productivity of coastal resources and pushed Paleolithic populations inland. The traditional ice edge exploitation of the North Atlantic became less and less tenable or desirable. The Magdalenian replaced the Solutrean in southwestern Europe as the populace in the New World was cut off from its ancestral homeland. While this may have happened relatively quickly in geological terms, the effects on individuals would have been less obvious. As the North Atlantic became more difficult to exploit and traditional areas were inundated, Solutrean peoples may have gravitated toward the western edge of their territory, leaving sparse representation on the European side, which eventually absorbed them into its shifting cultural scene.

  The archaeological evidence suggests to us that there were at least two technological adaptations to North American climes. One was the progenitor of Clovis: the people who used it continued to focus on littoral resources and worked their way down the East Coast, eventually spreading into Florida and possibly even the West Indies. Whatever the impetus, Clovis then developed in the southeast and colonized the region by traveling up the rivers and around the western edge of the Gulf of Mexico.

  The second adaptation stuck to the edge of the retreating glaciers. It may have followed the glacial front to the west, producing the Meadowcroft and Chesrow evidence and eventually developing into the Goshen Complex of the northern High Plains, the Plainview Complex of the southern High Plains, and perhaps the Nenana of Alaska. When Clovis people explored the west, these people were already there: Goshen in the northern plains and descendants of Denali and other Asian peoples in the Great Basin and the northwest. There are many scenarios of how they might have interacted as friend or foe. The key element of the later colonization of the New World is that people continued to arrive from Asia, but the European route was cut off. There would be no more genetic input from across the eastern ocean for at least ten millennia.

  The ideas in this book are not intended to be an explanation as much as a set of testable theories to help guide future research. If we have learned anything during this study, it is that thinking we know much about the archaeological cultures we so blithely speak about and use as the bases for our interpretations is the utmost arrogance. We are convinced that researchers continually underestimate the abilities, the vision, and the intelligence of our ancient cousins, whether in the east or the west.

  So what’s next? We don’t see a major change in the way archaeologists explore for new evidence of ancient cultures. The main modifications we contend have to be made are to open ourselves to new possibiliti
es and to develop underwater exploration of the continental shelves. Let us allow the data, in concert with theoretical development, to guide our research. There are many new lines of research developing right now, some of which will undoubtedly result in dead ends. Yet others, even some now considered to be completely crazy, may be just the approaches that will result in the next renaissance in the search for the predecessors of the earliest Americans.

  ACKNOWLEDGMENTS

  During the 1990s, new archaeological discoveries and changing ideas about ice age environments caused us to become skeptical of the traditional yet unsubstantiated hypothesis about New World origins. In light of new data, we thought it was important to consider evidence for alternative hypotheses. In particular, we wanted to evaluate the significance of the degree of technological similarities between Solutrean and Clovis artifacts to see if there was enough evidence to justify pursuing a hypothesis of a historic relationship between these two archaeological cultures. Our aim was only to write a short paper and stir the pot. However, we are especially indebted to Doris Kretschmer, a former acquisitions editor at University of California Press, who heard a talk by Dennis about the Solutrean-Clovis connection at the California Academy of Sciences in 1999 and, along with William Woodcock, a former acquisitions editor for Academic Press, convinced us to take on the issue fully. This book owes its existence to their encouragement.

  Funding for such a cutting-edge project was not available through normal academic sources, hence we are extremely grateful to Randy Best, Katherine Dimancesu, Jeff Meyers, and Tom Lennon and Charles Wheeler at Western Cultural Resource Management, who helped provide the necessary funds to conduct our research. We were blessed to have an excellent photographic record of our research by the late Chip Clark, the National Museum of Natural History’s exhibits photographer. Marcia Bakry, the illustrator for the Department of Anthropology at the National Museum of Natural History, produced all of the detailed artifact illustrations and maps. Ted Timreck, of Timreck Production, New York, produced video documentation of the research. Maggie Dittemore, librarian in the Department of Anthropology at the National Museum of Natural History, was most helpful in locating hard-to-find manuscripts and publications. We are deeply appreciative of the advice and comments of our colleagues Betty Meggers, Anta Montet White, Steven Beckerman, and Carolyn Rose, who read and critiqued various drafts of the manuscript. William Woodcock, John Thomas, and the anonymous reviewers selected by University of California Press also made valuable contributions.

  Special appreciation is due to the incomparable Juliana Froggatt for her ability to bring metamorphic change to our manuscript so that it is now easily readable. We are grateful to Blake Edgar, the acquisitions editor who took on our book after Doris left UC Press, and our sainted project editor Dore Brown for their help and patience bringing the project to fruition. Editorial coordinator Lynn Meinhardt and acquisitions coordinator Hannah Love deserve our deep gratitude, along with Alex Dahne, who will see our collective works gain maximum visibility.

  We received valuable assistance from many scholars who helped us at the National Archaeological Museum in Madrid, the Prehistoric and Archaeological Museum of Cantabria in Santander, the Museum of Altamira, and the Archaeological Museum of Asturias in Oviedo in Spain; the National Archaeological Museum of Lisbon in Portugal; and the National Museum of Prehistory in Les Eyzies and the Museum of Prehistory in Solutré in France. We are deeply grateful to the scholars and staff at these museums for sharing their Solutrean collections and knowledge; without them we could not have conducted our research. We are likewise indebted to Darrin Lowery, who generously allowed us access to his collections from the Eastern Shore of the Chesapeake Bay and provided some maps and photographs.

  Bruce wants to acknowledge the many people who had an early, profound impact on his interest in knapping and technology. François Bordes served as his mentor and friend and introduced him to Solutrean and Old World technologies. Don Crabtree, J. B. Sollberger, Bruce Huckell, Jacques Tixier, Evgenia Giria, Thierry Aubry, Bertrand Walter, and Jacques Pelegrin have greatly influenced his knapping and have always encouraged him to stretch his thinking. His early endeavors at applying knapping knowledge to archaeological questions were supported and promoted by Marie Wormington, Art Jelinek, George Frison, Dennis Stanford, C. Garth Sampson, Charles McBurney, Mike Collins, Hizri Amirkhanov, Sergey Lev, and countless others. All of these colleagues have his profound gratitude. Whatever they think of this work, their help, support, and guidance have influenced Bruce to continue his explorations into the application of technology to archaeological questions.

  Dennis offers his loving thanks and appreciation to his wife, Pegi (Margaret) Jodry, who supported not only Dennis but the fifteen years of research as well. She also provided excellent ideas about the direction of the research and subject content and helped assemble the database used in the analysis. Moreover, she is an articulate proponent of alternative hypotheses. Thank you, Pegi.

  Finally, we are extremely grateful to all of those people who have listened patiently to our theories and evidence for the ice age Europe origins of some people in the New World. Supporters have encouraged our investigations, and critics have ensured that we extended our research and presented solid evidence. We are greatly appreciative of the Smithsonian Institution and the University of Exeter, which have employed us throughout this long and interesting intellectual journey.

  APPENDIX

  CLUSTER ANALYSIS

  We did two cluster analyses, the first to see how the method would work with the kind of data we had, and the second to test our ideas about possible historical connections. Cluster analysis takes numerical data in the form of space-delineated columns and rows and runs varied comparisons on them. The output is known as a dendrogram. We particularly like this graphic because it is easy to visualize the relationships between categories, and the categories are not influenced by the order in which the program receives the data: the program, not us, identifies the clusters. It is of course gratifying when groups of items that we think are similar end up in the same cluster and equally welcome when those we see as different are in different clusters. But it is also possible that assemblages may cluster in unexpected ways. Such cases challenge us to look at the data with fresh eyes to search for explanations and help us to see around some of our biases.

  U.S. SITE COMPARISON

  Our first application of cluster analysis was to generate a table of U.S. fluted point assemblages by tool type proportions. The expectation was that if all of the sites were historically related and had the same function, they would cluster tightly together. In most cases, we used the published categories and counts, but on occasion it was possible to do some minor reclassification based on illustrations. Our greatest challenge was to identify artifact types shared by most of the site reports available to us, since different researchers had used different typologies. For example, in some reports bifaces are identified as preforms, in others as knives, and in still others simply as bifaces. The obvious thing for us to do was to always choose the simplest type category used in all of the analyses. Unfortunately, this ended up degrading the possible comparisons to the point that they were almost useless. We ended up running our cluster analysis on tool type percentages rather than raw counts. This equalized the assemblages so that clustering would not result simply because of large and small numbers of artifacts. There are many ways to manipulate data within cluster analysis. We tested several different combinations and found that the general order and clustering did not change. In this and the following analyses we used squared Euclidean distance for our proximity indicator and mean proximity as our clustering method.

  TABLE A.1 Tool Type Proportions of U.S. Fluted Point Assemblages

  Table A.1 shows our tool type categories from selected fluted point assemblages along with percentages by site. Although there are reports for other sites, they either lack totals for each tool category or contradict one another. We preferred to use origin
al site reports but on occasion deferred to other accounts because they had more comparable categories or gave actual frequencies.

  The results of clustering fluted point sites are indicated by levels, which indicate the degree of correspondence. At level 1 all sites cluster. At level 2 there are two major clusters. At level 3 a cluster of four sites breaks out of the main cluster, and at level 4 four distinct clusters are apparent. It is not surprising that in most cases the clusters seem to distinguish site functions and conform well to the interpretations of their investigators. The four functions are habitation (six sites), camps (sixteen), kill site/camps (twelve), and caches (four).

  There are a couple of exceptions explained by special circumstances. The Adams Site, a stone-processing locality with bifaces mostly fragmented during manufacture, clusters with caches simply because of its high proportion of bifaces. But although caches have high proportions of bifaces, most of them are complete, not fragments. Accordingly, a further site function discrimination might be possible if we could distinguish between complete and broken bifaces. Unfortunately, many site reports do not give such information. Another exception is the Drake Cache, which clusters with kill sites even though it is clearly a cache. This is explained by the fact that the cache contained primarily projectile points, as did most of the kill sites.

  At the sixth level of clustering, campsites separate into two groups, generally those in the east and those in the west. Interestingly, two clusters distinguish between northeast fluted and Clovis/Folsom sites. Equally intriguing is a cluster composed of Clovis and Folsom. Clustering has to be taken all the way down to the eighth level before kills are distinguished from kill site/camps.