The Na-Dene migrations just described were mostly terrestrial phenomena, undertaken by land-based hunters and gatherers. Perhaps some ancestral Na-Dene once lived along the coastlines of Beringia and western Alaska, but nowadays almost all the coastal populations of these regions speak Eskimo-Aleut languages (Figure 5.3) and the only truly coastal Na-Dene live further south. The northern Athapaskans are completely terrestrial, and we have no good evidence that they ever seriously colonized the ice-bound coasts of northern Alaska, northern Canada, or Greenland. There is a suggestion that Na-Dene languages spread with the Arctic Small Tool tradition which records the first human colonization of the Arctic coastline of North America (Dumond 2010), but the vast majority of North American linguists and archaeologists associate this tradition with ancestral Eskimo-Aleuts, or Paleoeskimos.
Robert McGhee (2005) provides an excellent description of what must have been one of the most challenging environments to be occupied by prehistoric humans. The Arctic Circle offers a good living for people who can hunt off treeless tundra and sea-ice, with very high biomasses of large mammal species such as seals, whales, walrus, polar bears, musk oxen and caribou (reindeer). It is a land of spring and summer plenty, and the key to survival is to withstand the winter, via the use of efficient technology for clothing, hunting, food storage and shelter. The first human migrations into these North American high latitudes occurred around 5000 years ago, during the period of maximum mid-Holocene warming.
The Arctic Small Tool tradition, sometimes referred to as the Paleoeskimo Tradition, also made its first appearance about 5000 years ago, initially as the Denbigh archaeological complex on Seward Peninsula in Alaska (Figure 5.3).14 Denbigh/Paleoeskimo descendants spread east along the northern coastline of Canada, to appear as the Pre-Dorset and Dorset cultures of Baffin Island, and the Saqqaq and Independence I cultures of Greenland, all evidently in place by 2500 bc (Ellis 2008). The Saqqaq people have recently hit headlines with the extraction of a 79% complete DNA profile from male hair preserved in the Greenland permafrost (Rasmussen et al. 2010). The owner, surprisingly, revealed little genetic similarity to existing Native American populations, including modern Eskimo-Aleuts. It matched most closely the modern Chukchi and Koryak populations of northeastern Siberia and the Uralic-speaking Nganasans of the Taymyr Peninsula far to the west. These are all indigenous Siberian hunters and gatherers and recent herders of reindeer.
The archaeological material culture and economy behind the Arctic Small Tool tradition and its many local descendants, from St Lawrence Island in the west to Greenland in the east, was particularly complex, befitting the colonization of a frozen landscape. It included microblades and small blade-like flakes, burins (engraving tools), ground stone adzes and knives, soapstone lamps, fiber-tempered and stamped pottery, portable skin tents weighted down by boulders, tailored skin clothing, bone and antler harpoons and arrowheads, skin-covered boats like ethnographic kayaks, and the spear hunting of fish, seals, caribou, and plentiful musk oxen.15 Dogs were also present.
Figure 5.3 The distributions of Paleoeskimo archaeological complexes (circa 2000 Bc), superimposed by the linguistic migrations of Eskimo-Aleut populations, especially the major Thule Inuit migration after ad 1200. Background map by Multimedia Services, ANU, with original data added by the author from maps in Fortescue 2013; Friesen 2013.
The pottery is interesting, given its manufacture in much of the world by sedentary groups, mainly farmers. It was being made by 5000 bc in the Lena Basin of Siberia, but did not spread into Alaska or northern coastal Canada until much later.16 In fact, the aforementioned items did not all appear together in the archaeological record, but most seem to have been present by the time of the Norton culture of Alaska, circa 500 bc.
One very interesting attempt to place the Arctic Small Tool tradition into a broader historical perspective is that by linguist Michael Fortescue (1998). He regards the major Uralic, Eskimo-Aleut and Yukaghir language families as being all related in a macrofamily termed Uralo-Siberian. I noted the existence of Nganasan (Uralic) and Saqqaq (Eskimo-Aleut) genetic links earlier. Whole-genome analysis of single nucleotide polymorphisms also relates modern Uralic and Eskimo-Aleut speakers quite closely, at least when compared to other Native Americans and Eurasians (Reich et al. 2012). The origin of the Uralo-Siberian macro-family is postulated to have been in southern Siberia, in the Lake Baikal region between the Yenisei and Lena rivers, prior to the extensive migrations through here of Turkic and Tungusic-speaking hunters and pas-toralists. Fortescue believes that ancestral Eskimo-Aleuts split from their Uralo-Siberian linguistic relatives at about 4000 bc and ultimately reached Beringia, and then Alaska, carrying the Arctic Small Tool tradition. As discussed earlier (page 99), the ancestral Uralic languages presumably started to move west at about the same time.
Uralo-Siberian is completely separate from Dene-Yeniseian, and the two macrofamilies show very little sign of ancient contact or borrowing. The unrelated Na-Dene languages were probably present in Alaska at the time of Uralo-Siberian expansion, as were similarly unrelated Chukotko-Kamchatkan languages in Beringia and Kamchatka. These distributions suggest again some kind of leap-frogging movement, as well as continuous territorial expansion, by the early Eskimo-Aleuts.
Other groups of unknown ethnolinguistic identity penetrated the northeastern coastline of Siberia at around the same time as the Paleoeskimos settled in the Arctic coastline of Canada, reaching Zhokov Island at about 76°N in one of the most northerly forays of hunter-gatherers in prehistory17 The Zhokov people constructed semisubterranean houses, used stone microblades, mammoth ivory picks and ground stone adzes, and hunted reindeer and polar bear. The excavator reports radiocarbon dates of 9000-6000 bc on wood and animal bone, but also notes that the island contains much driftwood preserved by permafrost. These dates can perhaps be questioned since radiocarbon analysis dates the death of the tree, not the date of human usage of the wood, so one wonders if this movement could actually have been contemporary with that of the Paleoeskimo settlers of Alaska, closer perhaps to 3000 bc.
The Thule Migration and the Inuit
The Eskimo-Aleut language family as a whole developed in coastal Alaska, and has a relatively short time depth of 5000 years, or less.18 From a purely linguistic perspective, one of its first recorded spreads (the Paleoeskimo languages are extinct, so we cannot know what languages they spoke) appears to have been that of ancestral Aleut speakers from the Alaska Peninsula into the Aleutian Islands.19 Elsewhere on the coastlines
Around Bering Strait, the ancestors of the living Eskimo languages diverged ultimately into the four present-day members of the Yupik subgroup, from which the Inuit dialect chain separated during the thirteenth century ad to migrate eastwards in a remarkably rapid 5000 km movement that reached as far as Greenland (Figure 5.3).20 The southern and western coastlines of this huge Arctic island were already settled at this time by Norse from western Europe, who arrived historically from Iceland in ad 985. This would have been one of the first meetings21 between Native Americans and Europeans, brought about by remarkable migratory achievements from both directions. It is perhaps worth remembering, however, that only the Native Americans survived in Greenland to the present, the Norse having withdrawn during the fifteenth century.
The Thule migration appears to have retraced some of the routes taken by Paleoeskimo migrants over 3000 years before. By ad 1000, technology had advanced to include an increasing use of polished slate tools, larger boats perhaps similar to the ethnographic umiak, dog-drawn sleds, more advanced forms of bone or ivory harpoon heads, and the use of iron tools. The iron was obtained either from meteoric sources in northwestern Greenland, or from the Norse hunters and traders who came to Disko Bay in western Greenland. McGhee (2005) notes that iron was also in use in coastal Alaska from about 2000 years ago, perhaps obtained from Chinese or Korean sources.
The entire migration from Alaska to Greenland perhaps took only a few decades, reflecting the vast distances between exploitable resources in the landscape. What was the stimulus? Archaeologists variously suggest that the movement was encouraged by a desire for Greenland iron, hunting of caribou and musk oxen, and by the mobility required to hunt Arctic bowhead whales. As Michael Fortescue (2013: 340) notes, the migration was rapid enough to be traceable through linguistic studies alone:
In fact, the Thule migrations from North Alaska as far as East Greenland about a thousand years ago represent a paradigm case for the rapid expansion of a language into virtually uninhabited regions, with ensuing gradation of innovations and losses away from its original homeland.
By the time of the Thule migration, few of the earlier Paleoeskimo populations, such as the genetically distinctive Saqqaq referred to earlier, appear to have survived in the eastern Arctic (hence Fortescue's reference to virtually uninhabited regions). One isolated non-Inuit group, the Sadlermiut, survived until 1902 in northern Hudson Bayt22 The Inuit presumably incorporated most preceding populations linguistically and genetically into their own communities. There are few traces of language mixing or convergence between Inuit and other unrelated languages, and language shift rather than intensive multilingualism was the main route towards assimilation (Fortescue 1998).
Why were the Inuit such successful long-distance migrants? By the thirteenth century ad, the 'Great Warming', described rather lucidly by Brian Fagan (2008), was releasing coastal resources again from the ice. One question that arises, therefore, is whether the two successive Eskimo-Aleut migrations about 4000 years apart - the Paleoeskimo and the Thule Inuit - were the results of the successive episodes of warm
Mid-Holocene and Medieval climate, allowing easier access to what would otherwise have been largely ice-bound coastal resources?23
I suspect they were, and return again to the suggestion by Krantz (1976), introduced in Chapter 4, that hunters and gatherers rarely migrated within landscapes that were already occupied by human populations of similar technological and demographic capacity They would have done so only into relative or absolute population voids. The migrations discussed in this chapter, especially the Thule and Apachean migrations, fit this perspective very well. But while there is truth in Krantz's observation, I would also point to the apparent ability of hunters and gatherers to respond to situations of technological innovation. Thule material culture was more complex than that of the Paleoeskimos. The skin boats, harpoons, polished slate knives and dog sleds that apparently distinguished the Thule Inuit all come to mind in this regard. They surely gave their owners some advantages, even though there might have been few surviving Paleoeskimo populations still around to experience the impact.