Signs of human band and caribou herd affiliation have been accumulating since humans and caribou first entered the Barrenlands. Piled slab inukshuit made by Thelon and Dubawnt River hunters mark the route of the herd returning to the forest. That hunters followed it is seen in distinctive Eyeberry Lake quartzite from the heart of the range being carried several hundred kilometres south to Lake Athabasca. Archaeological levels separated by wind-blown sand contain caribou hunting and processing tools that became instrumental in testing the idea of herd following. Archaeological sites were preserved because sand from glacial lakes blown in from adjacent dunefields covered the tools of each culture. This separation between levels permitted the defining and radiocarbon dating of tools used by subsequent hunting bands, with those at the bottom or 2 m level being 8000 years old.
The affiliation of human bands with specific herds means a separation from other bands and herds, a segregation reflected in different but contemporaneous tool styles in the four herd ranges (Gordon 1975). This herd separation even exists in the rutting area used by both Beverly and Kaminuriak caribou, where 94% of all ear-tagged caribou return to their herd after mingling (Parker 1972). It follows, therefore, that as hunting camps of past bands are confined to the modern caribou range, band-herd affiliation would consequently have existed throughout prehistory, resulting in herd following (Gordon 1975).
The concept of herd following has been argued for many years. 19th century diffusionists and 20th century environmentalists used it to explain supposed artifact similarity over wide areas of Ice-Age Europe. It is seldom referred to in archaeological contexts as alternate game was often available. And when used it is confounded. Burch's (1972) Caribou Inuit study misconstrued herd following as herd accompaniment, which is clearly absurd given the sustained pace of migration. His evidence in that study was obscured by effects from past tribal animosity. Historic Caribou Inuit were prevented from herd following by their traditional Chipewyan enemies who were confined to the winter range by the fur trade (Gordon 1990:399-400). When they eventually starved on the tundra, the Inuit were relocated to coastal villages. Ironically, Burch (1976, quoted in Smith 1978) mentions Chipewyan hunters following the herd onto the tundra to the calving ground, which suggests they used the whole range. Had Burch known at that time that the Chipewyan were free to follow the herd for two thousand years without impediment from Inuit, he would have found herd following realistic. Unfortunately, he did not know of the many Chipewyan and earlier tundra camps that demonstrate the full seasonal cycle.
Given the time depth represented in this study, did herds always migrate along the same corridors? Several biological papers suggest variability in herd movement from year to year, based on aerial survey. I have little doubt that modern seasonal movements vary due to overhunting, forest fires and human impingement on caribou range, but calving ground locations have remained stable for centuries, based on aerial survey and the archaeological record, which shows that Beverly artifacts do not extend beyond them. Surface and stratified sites on the modern corridor contain tools and bone from all phases which prove long term use.
Herd following is rejected by some historians embellishing a feast and famine concept by focussing their research on rare fur trade accounts of chaotic winter subherd movement near forts. By exterminating local subherds to provide meat to traders, meat-hunters destroyed local migration, leading to the assumption that the main migration vanished and that herds are unpredictable. Away from forts, the major migration continued, with meat-sharing between bands established in central crossing camps surrounded by small camps. Ironically, predictability is even rebuffed by some Chipewyan villagers who avoid these crossings because they require boats, spears and long waits. Instead, they fly to open areas peripheral to the treeline and rely on distant shooting.
That cached meat could permit sedentery bands is unlikely because rock caches are absent on the tundra and tree caches are rare in the forest. I believe old and sick people and toddlers too heavy to carry and incapable of long treks summered in the forest, subsisting on berries, fish and forest game, but their number and energy needs are tiny compared to those of herd followers. In late spring, the Chipewyan followed the herd to the calving ground but did not accompany it; in autumn, both returned to the forest. Herd following was aided by hoof marks in mud and polished bedrock and verbal accounts of past movement. Strangest of all, the earliest most-cited Chipewyan observer, Samuel Hearne, documented herd following two centuries ago: "As their whole aim is to procure a comfortable subsistence, they take the most prudent means to accomplish it; and by always following the lead of the deer, are seldom exposed to the griping hand of famine" (Hearne 1958:83).
Since 1970, I have recorded a thousand sites with about 10,000 culturally assigned artifacts, most of which are on the tundra. The largest sites are just north of treeline and well beyond the winter range. They are also the most transiently occupied because the herd crosses them in 2-3 days. Their size is due to the intense activity occurring during their short occupation. If the band did not follow the herd, how do we explain these sites? That these transient camps were used over the centuries is apparent in stratified bone and tool levels to depths exceeding 2 m and ages to 8,000 years. Along the 600 km migration route, evidence of hundreds of ancient camps from all time periods where other game or caching were absent, prompted Burch (1991:439-444) to retract his denunciation of herd following.
Rather than canoe the rivers, the Chipewyan and their predecessors followed the herd overland through the north and central drainages in the summer and over frozen rivers and lakes of the south drainage in winter and spring. On May 20th, 1771, Hearne noted that the Chipewyan carried one-man canoes hundreds of km in order to intercept the herd at a crossing (Hearne 1958:62). Mobility had its costs, however, and invalids and the aged paid the dearest cost under severe conditions. The Chipewyan were also forced to contend with seasonal inconsistencies of abundant meat and fat in summer, and almost none in spring. To endure the trials of life on the move, possessions had to be light, as tents and poles, bedding, vessels, dried meat and baby carriers were carried by hand or sled. But herd following had its rewards: semi-permanent access to a travelling and seemingly endless meat supply, more access to seasonal and geographical resources such as berries, migratory birds and fish; soapstone, chert and quartzite for tool-making; birchbark for baskets, and spruce root for nets and line.
The affiliation of a band with a specific herd was also responsible for the development of distinctive local dialects and a kinship that was unique to the range (Gordon 1975). An example of these distinctions is evidenced by the Kaminuriak band which is comprised of local bands of Hatchet and Duck Lake and Barren Lands. Each band is sororal, where a woman may marry her sister's widower, leviradical, where a man may marry his brother's widow, and patrilateral, where women are exchanged through cross-cousin marriage. But the Caribou-Eater in the adjacent Beverly range are neither sororal nor leviradical, with rare patrilocality. Rather, they are bilaterally-related, with bride going to the groom's band (Smith 1978). Such distinctions support the concept of restricted past contact, a constraint also appearing in tool styles and trade goods which are suitable for testing past herd following.
It appears that human mobility is highest in summer, lowest in winter and sporadic at treeline as caribou enter the forest in August and return to tundra for October rut. Bands did not remain at treeline to conduct both tundra and forest forays because there is no evidence of long occupation. Instead, camps at small stream crossings were used sporadically to hunt subherds dispersed widely east and west. Spring forays north onto tundra had to align with the main migration path, while fall forays deep in the forest involved meandering animals. It was far better to hunt at the productive camps north of treeline at north-south water-crossings where the Thelon bends east towards the calving ground at Beverly Lake.
from Gordon 1996. People of Sunlight; People of Starlight: Barrenland Archaeology in the Northwest Territories of Canada. Archaeological Survey of Canada, Mercury Series Paper 154. Canadian Museum of Civilization, pp.11-15.