Across Atlantic Ice

by Dennis J. Stanford

  BERINGIA AND THE LAND BRIDGE

  Beringia is the name given by scientists to a massive extent of land that surrounded the Bering Strait, the narrow gap of ocean that separates the northeastern tip of Siberian Russia from the western point of Alaska.1 During the last ice age, sea levels were much lower than today, exposing large sections of ocean floor. As they emerged, the world’s continental shelves created new coastal landforms, including the Bering land bridge, which linked Asia and North America. The surrounding region, Beringia, extended east as far as the edge of the continental glacier near the present-day Mackenzie River and south to the northern terminus of the Cordilleran ice sheet in the mountains of southern Alaska (figure 3.1). Westward it stretched across Siberia to the headwaters of the Lena River and then angled southeast into the Russian Far East. Beringia is nearly twice as wide to the west of the Bering Strait as it is to the east. People traveling by land from Asia to North America would have had to cross this vast, treeless Arctic plain, which was icy during the long, dark winter and wet and buggy during the short summer months.

  FIGURE 3.1.

  Map of Beringia showing sites mentioned in the text.

  Quite a few reasons and quite a few assumptions make the land bridge crossing the dominant theory of the peopling of the New World. Some are based on sound science, others on intuition. For example, it seems reasonable to think that because modern Native Americans look Asian and possess Asian genetic traits, their ancestors must have all come from Asia. But this assumes that all ancient Americans looked like modern Indians and that the earliest people who migrated from northeast Asia must have looked like those who reside there today. Since virtually no late-glacial-age human remains have been found in either North America or northeastern Asia, it is difficult to test this assumption.

  Another supporting idea is based on early research indicating that near the end of the last ice age, the two major North American continental glaciers had melted far enough apart to create a passable ice-free corridor from Alaska to the plains of southern Alberta.2This date for an opening would have allowed enough time for the establishment of an environment that supported plants and animals as well as the Clovis ancestors working their way southward to reach the Lower Forty-Eight. Recent paleoecological and geological research has provided persuasive evidence, however, that the biogeographic corridor was not established until sometime between circa 12,000 and 14,000 years ago.3 In our opinion this date for the opening would have been much too late to allow for the pan-continental distribution of Clovis technology by 13,000 years ago. Nevertheless, the impact of the original concept lingers, as in this recent statement by the Arctic archaeologist F. H. West: “Had there been no corridor southward or had that corridor lain to the east of the continental ice sheet, it seems certain that the discovery of America would have had to await the coming of navigators from over the seas.”4

  This statement by West includes another assumption, namely, that early hunters could have crossed the Bering Strait only by dry land, because they had no watercraft. Thus they would have been unable to cross even a small stretch of water—or, in West’s view, to walk across a frozen expanse of sea. This, of course, has not been the case in more recent times. Many Holocene groups have crossed the Bering Strait, and Inuit people have gone back and forth frequently in historic times, either in skin boats or over the ice by foot or with dog teams.

  Also implicit in the land bridge concept is the conviction that the High Arctic landscape would have been habitable and amenable to foot travel during the Last Glacial Maximum (LGM). Reconstructions by vertebrate paleontologists portray much of the area as being a shrub tundra environment with a diverse community of large game animals.5 Some palynologists, by contrast, read the plant evidence as indicating a bleak, polar desert.6 Other researchers envision much of the region as cold, moderately dry to moist environments with heaths, dry meadows, and shrub tundra interspersed with marshes and small ponds.7 No matter what the climate, the low-lying topography of the land bridge would have been poorly drained in summer and covered with meandering rivers, lakes, ponds, and marshes, making it one of the least desirable land surfaces known for foot travel. Regardless of its specific physiography, most researchers concede that the land bridge was probably harsh and difficult to traverse.

  Even today northeastern Siberia is the coldest place on earth, with prolonged and severe winters with temperatures as low as -94°F, exceeding the average low temperature of Antarctica, and short summers with average July temperatures of 59°F.8 In the Late Pleistocene the climate was even more extreme. And along with adverse paleoecological conditions, there were many large rivers, rugged mountains, local glaciers, and numerous other obstacles to foot travel. We can only imagine what challenges this area must have presented during the Late Pleistocene.

  Newspapers, books, magazines, and museum exhibits have repeatedly illustrated ice age hunters as primitive people who dressed in rudimentary animal skins and eked out their survival with spears and clubs, creating the comic strip conception of the caveman and the die-hard notion that cavemen coexisted with dinosaurs. But even in today’s climate, people who dressed and lived like that would not live long. Therefore, we suggest that ice age people had cold-weather technologies at least as sophisticated as those required for survival during the warmer conditions of the Holocene, a geological epoch that began around 12,000 years ago. At the very least, tailored, waterproof clothing was a necessity, along with substantial all-weather shelters. With the skills to create these items, they no doubt also had the ability to construct sewn-skin boats. One has only to ask whether the peoples who live in Arctic conditions today could survive as they do without adequate clothing, shelter, or boats to cross and travel on rivers and lakes. Lacking good archaeological evidence to the contrary, we assume that this knowledge was integral to successful Arctic colonization.

  Another relevant aspect of Beringia is that because much of it is above the Arctic Circle, at 66 degrees north, cultures there must adapt to months of darkness alternating with months of continuous daylight. Why would anybody move into such a region? To answer this question, we need to examine the late Pleistocene archaeological record on the Siberian and Alaskan sides of Beringia. Fortunately, several relevant works are available. One of the most complete, American Beginnings, pulls together much of the information, and we take advantage of this source in this chapter’s review, unless otherwise indicated.9

  Virtually everybody who has attempted to summarize the archaeological data from Beringia has employed many qualifiers, because the research is characterized by different theoretical approaches and analytical methods.10 This is partly because of the length of time over which the work has been accomplished but also because of major differences between Soviet and American archaeological approaches to survey coverage and intensity of excavation, as well as what questions are asked and what analytic criteria employed.

  Another difficulty is that we are looking at a part of the world that has never been densely occupied. Most archaeological sites were temporary and probably used by small groups of people. As elsewhere, durable material culture represents only a small fraction of what people used, and little of it was left behind at any one place or time. This means that most ancient sites have small artifact inventories, and these represent few of the activities undertaken by the inhabitants. Furthermore, with the exception of rare sites where artifacts and other cultural remains are preserved by permafrost, the Arctic environment is as destructive today as it was during the ice age. Freezing and thawing, solifluction (the movement of saturated sediments on top of frozen deposits), ice wedges, and other frozen-ground phenomena, as well as tectonic uplift, isostatic subsidence and rebound (movements of deposits related to glacial compression and release), erosion, rising and falling sea levels, and shallow sedimentary deposition, are all destructive forces prevalent in the Arctic. All of this results in either archaeological sites with little to no stratigraphy or soil levels that tend to be disturbed and
mixed. Thus, most Arctic sites are scatters of artifacts from multiple time periods found mixed on the surface or in shallowly buried contexts. Inaccessibility, short field seasons, and relatively high expense have contributed to the difficulty of doing fieldwork in the north. With all of these handicaps, it is fortunate that we can say anything about the prehistory of Beringia.

  We commonly hear that Beringia is so vast and inaccessible that our current knowledge is based on a small fraction of what could be discovered. This is obvious, but not unusual in archaeology. The American southwest has great preservation and a hospitable climate, and its sites are readily accessible compared to those of the Arctic. Yet we have estimated that even in southwestern Colorado, where fieldwork has been conducted for more than one hundred years, less than 1 percent of the available record is well enough known to contribute to modern research interests. Even so, archaeologists must not refrain from interpreting the available record if we expect to make any scientific headway.

  What do we know about the late Pleistocene and earliest Holocene archaeology of Beringia? Although it is a vast geographic area, it has been significantly explored and archaeologically surveyed. Investigations tend to focus on the more accessible areas, but if they are convenient today, they were probably used in the past. This is especially true of prime spots on rivers and in river valleys, where fordings, portages, and other transportations were easier, then as now. In eastern Siberia, surveys and less formal explorations have recorded archaeological sites along most of the major rivers and their tributaries. This is also true of much of Alaska and the ice-free part of the Yukon. Of course, a lot of the area has not been examined intensively, and unknown cultural traditions may be waiting to be discovered. New discoveries are being made: for example, the Yana Rhinoceros Horn Site of far western Beringia, some 2,200 kilometers from the Bering Sea coast.11 Dating to 30,000 years ago, this site is the first evidence of pre-LGM habitation in the High Arctic. This is an important discovery, but does it relate directly to the New World colonization issue? At present the answer is unknown, but based on the flaked stone technology found at Yana, a relationship to Clovis seems highly unlikely.

  The archaeological data from Beringia can be compiled in several ways. One is to group sites according to artifact assemblages, a common approach since many sites lack radiocarbon dates. Those with similar tools and technologies are considered related historically and presumed to be of a similar age. Their chronological placement is established by comparisons with similar archaeological sites that have been dated. Considering the geographic distribution of sites related technologically creates additional clusters, which archaeologists describe as “cultures” based on various criteria such as tool forms, settlement patterns, and technology. These are not the same as what we might consider cultures today, and there is no assurance that prehistoric people would have recognized our archaeological cultures. For example, the Clovis culture has meaning in terms of its archaeological characteristics, but this doesn’t necessarily mean that Clovis people would all have spoken the same language or believed the same things. Cultures are just a convenient way for archaeologists to communicate. Once established, they can be plotted in time and space and their distributions used to reconstruct the culture history of the region.

  WESTERN BERINGIA

  With the exception of Yana, the earliest artifact assemblages in Beringia belong to a generalized Asian microblade tradition. (Tradition in this case refers to a shared technology but not necessarily other traits that would qualify it as an archaeological culture.) These early assemblages are classified as the Early Dyuktai Culture. Five sites with Early Dyuktai material have been dated, and quite a few other sites have been classified as Early Dyuktai by the composition of their assemblages. With the exception of Ust Mil 2, where some anomalous older dates were recorded, all of these sites fall between 18,000 and 20,000 years BP (before the present), within the LGM, and all are in far western Beringia. Straight-line overland distances from them to the Bering Strait range from 2,200 to 3,200 kilometers and average more than 2,750 kilometers.

  Artifact assemblages vary across these sites, but all are dominated by microblade technology, with wedge-shaped cores made on flakes (figure 3.2a) or bifacial precores (figure 3.2b). The accompanying tools are retouched microblades and blade insets, while the next most common tools are burins, made from both flakes and blades (figure 3.2c). Less common are bifacial knives (figure 3.2d), drills/borers, pebble tools, and side scrapers. Bone and ivory artifacts are rare. They include an ivory needle without an eye, encountered in the 18,000-year-old Verkhne-Troitskaya Site. A possible antler hammer from Ezhantsy is around 17,000 years old, and a bone awl from Ikhine 2 may be as much as 30,000 years old but is more likely around 20,000 years old.

  Late Dyuktai assemblages, dating between 10,000 and 13,000 years BP, differ from Early Dyuktai assemblages primarily in an increase in bifacial knife and point production. Microblade technology still dominates, and there is an increase in the presence of microblade insets (figure Intro.3.a–b). Burins are present (figure 3.2f–i). Bifacial manufacture of microblade cores and precores becomes more common (figure 3.2 j–m), and bifacial techniques are used to create finished tools (figure 3.2e). Percussion bifaces are reduced proportionally (see chapter 1 for a description of this method), with rare cases of thinning. Bifacial points were initially percussion shaped and frequently pressure finished, producing thickened cross sections. Fine pressure retouch is rare, and lower-edge grinding of projectile points is unreported.12 The only bone artifacts yet found in a Late Dyuktai collection are a possible projectile point flaked from mammoth ivory and an antler hammer from Dyuktai Cave.

  Four Late Dyuktai sites have been dated in western Beringia, but like Early Dyuktai sites they are far from the Bering Strait—2,100 to 3,400 kilometers away, with an average of 2,825 kilometers. Several Late Dyuktai–like sites described by the archaeologist Anatoli Derev’anko are very close to the Bering Strait, but they might just as easily be related to the Denali tradition of eastern Beringia (described below).13 In spite of the intervening Bering Strait, they are geographically closer to Denali sites then they are to any of the dated Late Dyuktai sites.

  Several sites in western Beringia do not fit in the generalized Dyuktai tradition. Either they lack microblade technology or it is only a minor component of their flaked stone repertoire. Most of these sites are near Vladivostok, in the part of the Russian Far East known as Primorye, more than 4,000 kilometers by land from the Bering Strait. Radiocarbon assays of several of these sites indicate that they were occupied between 9,000 and 12,000 years ago, and little evidence has been found that their inhabitants spread north and east before 9,000 years ago, if at all. Their technology was based on large blade production (figure 3.3g), although flake tools are also present in significant proportions. Retouched tools, most of which were made on blades, make up only small percentages of the assemblages, dominated by end scrapers (figure 3.3c–d) and burins (figure 3.3e–f). Rare tools include perforators, possible unifacial points, and an occasional proportional biface. A few microblade cores (figure 3.3a–b) and some microblade core debitage are present in these assemblages, but it is not clear if the Primorye flintknappers used the Dyuktai technique to produce pressure microblades.

  FIGURE 3.2.

  Early and Late Dyuktai artifacts. Early Dyuktai artifacts from Dyuktai Cave: (a) wedge-shaped core with microblade; (b) obverse, cross section, and reverse sides of precore; (c) obverse and reverse sides of burin, with arrows showing direction of spall removal; (d) biface knife. Late Dyuktai artifacts from Level V, Leten Novyy Site: (e) biface; (f–i) burins; (j–k) microblade precores; (l) microblade core. Late Dyuktai artifacts from Kurung Site Layer V1: (m) microblade core.

  FIGURE 3.3.

  Primorye artifacts: (a–b) microblade cores; (c–d) end scrapers with cross sections; (e–f) burins, with arrows showing direction of spall removal; (g) conical blade core.

  Two sites stand out as having n
on-Dyuktai components. Ushki, on the Kamchatka Peninsula, has a typical Late Dyuktai component in Level VI (figure 3.4a–i) underlain by Level VII, which contains a unique flaked stone assemblage (figure 3.4m–t). It is characterized by small, thick, bifacially flaked stemmed projectile points, which were finished by pressure flaking (figure 3.4m–p), and also contains burins (figure 3.4s), foliate bifaces (figure 3.4a–c and r), and unifacial knives (figure 3.4t). N. N. Dikov, the original investigator, reported that microblade technology was absent, but the presence of percussion bladelets suggests to the Siberian archaeology experts Z. A. Abramova and Roger Powers that a microblade core technology was indeed used at the site (figure 3.4q).14 This assemblage is like no other described in the region, but a few stemmed points similar to those from this Ushki collection have been found in small numbers at undated surface localities (figure 3.4j–l). Level VII was originally dated to around 14,000 years BP, but a recent reanalysis clearly demonstrates that it is only 13,000 years old.15

  The other non-Dyuktai site, located above the Arctic Circle and known as Berelekh after the river nearby, has a bone assemblage that dates to between 12,500 and 13,000 years BP. It is best known as a paleontological locality that produced large quantities of mammoth and other Late Pleistocene animal remains. Most of the cultural artifacts were collected during large-scale bone-digging operations, and few were found in good context. It is probable that more than one archaeological component is represented, given the heterogeneity of the flaked stone collection. Though bifaces dominate, a single microblade core was recovered from the mining spoil. The bifaces include well-made fragments that were probably used as knives. There are also some small-point fragments pressure-flaked from thin flakes that closely resemble the Chindadn-style points found in central Alaska (see below). The recovery methods—mainly high-pressure water blasting—and probably mixed nature of this site make it difficult to place chronologically. The Russian archaeologists Y. A. Mochanov and S. A. Fedoseeva conclude that it was probably a very late Dyuktai occupation.16