Introduction

This paper considers how a large part of northern Canada was interpreted as having been a part of the Shield Archaic culture for most of prehistory before archaeological work was ever undertaken there. The genesis of this paper was a review of the archaeological literature on the James Bay region of northern Quebec.

The archaeology of eastern James Bay is usually considered within the broader context of the Canadian Shield, a physiographic landform that covers over half of Canada (Fig. 1). Geologically, it consists of a base of largely metamorphic rock, covered by an extremely thin layer of soil, the result of the soil having been scraped off by glaciers in the last glaciations, as well as low soil accumulation ever since.

Figure 1 

Physiographic regions of Canada. Contains information licensed under the Open Government Licence – Canada.

In 1968 the prevailing paradigm of the time for this region was summarized in James V. Wright’s chapter on “The Boreal Forest” in the volume Science, History and Hudson Bay:

Within the Archaic Stage, the Middle Woodland period and the Late Woodland period, there exists a degree of spatial homogeneity unknown farther south except at a very early time level. This condition appears to prevail throughout the Boreal Forest of western Quebec, northern Ontario, northern Manitoba and, in part, adjacent Saskatchewan. A number of factors are probably involved in this unusual cultural continuity over large areas but, in my opinion, the dominant factor is environment. Referred to by a colleague as “the infertile crescent” the region under consideration placed stringent demands upon its occupants and the small nomadic bands ranging over large tracts of land in order to survive are a necessary prerequisite for the apparent cultural similarities seen over extended areas… (59).

Wright was by no means alone in his assumption of homogeneity concerning the Algonkian speakers of the Canadian Shield (Brochu 1970: 29; Cooper 1946: 280; Martijn & Rogers 1969; Spaulding 1946: 146–147; Willey 1966: 448–449). He saw most of the Shield’s prehistory as belonging to the Shield Archaic tradition, a term he coined himself but which was later widely used (Wright 1972: 1–2).

Although the Shield Archaic tradition defined by Wright included the James Bay part of northern Quebec, it was in fact designated as an area included in the Shield Archaic a year before any archaeological excavation began there in 1973. The entire Shield is a difficult environment for archaeological work. Before salvage work began as a result of hydroelectric development, only a handful of excavations had been undertaken in northern Quebec.

Despite the paucity of data, however, over the first half of the twentieth century there developed an astonishingly clear conception of the prehistory of the Canadian Shield, especially the period known as the “Shield Archaic”. The Boreal Zone in northern Quebec was included in these interpretations, and it was using this framework that archaeology in eastern James Bay was developed. As a result, the history of archaeology in James Bay and northern Quebec begins long before any actual excavations were done.

The construction of prehistory in northern Quebec began with the earliest contact of Europeans with Native Canadians and developed from religious explanations to Classical Evolutionary ones to Culture-Historical ones to Neoevolutionary scientific ones. However, although the interpretations were different, the content of the conclusions hardly changed at all.

Early Contact and Christian Perspectives in James Bay

Some of the earliest accounts of the Cree in northern Quebec were written by missionaries and fur traders. The characterization of the Cree by the HBC traders has been dealt with extensively by Toby Morantz (1983), and so will not be discussed in detail here, other than to say that a careful reading of these documents leads her to conclude that anthropologists have been incorrect in their interpretations of Cree social organization as being extremely simple at the time of contact (155). It seems clear from this that anthropologists were much more influenced by the admittedly more colourful descriptions furnished by early missionaries.

This early literature characterized the Cree population of James Bay as dispersed and simple. What is interesting is the emphasis on mobility as the main factor mandating a ‘savage’ or ‘barbarian’ culture, and limiting the ability to properly embrace Christianity.

The Jesuits were not a large presence in the effort to convert the Cree of James Bay, however their influence on both the philosophy and practice of missions in New France was extensive. They had almost exclusive control of missionary activities until 1657, and clearly set the patterns of evangelism for future groups (Ronda 1972: 386). In addition, their manifold reports on their work in New France were replicated, and influenced the assumptions and expectations of the next generation of Europeans in North America (Healy 1958: 144; Dorsey 1998: 405). From the beginning they were concerned with the influence of secular practices and behaviours, particularly mobility, on the convertibility of their target populations to Christianity. Early Relations express the Jesuits’ inherent assumption that a mobile, or as they saw it, wandering, population is both socially and morally inferior to sedentary agricultural society:

They are, I say, savage, haunting the woods, ignorant, lawless and rude: they are wanderers, with nothing to attach them to a place …(Jesuit Relations I, 173, cited in Healy 1958: 149).

Although they live in the woods, they are nonetheless Men (Jesuit Relations XXIX, 281, cited in Healy 1958: 150).

These observations were accompanied by a good deal of frustration, and concerted efforts to mould various indigenous groups into a European-style agricultural population:

Father Le Jeune wrote that missionaries might “work a great deal and advance very little, if we do not make these Barbarians stationary (Jesuit relations XI, 149, cited in Ronda 1972: 390).

The Jesuits were, nonetheless, optimistic about the potential of native peoples to become Christian, and therefore sedentary. They saw the native population of New France as crude, but still potential Christians (Ronda 1978; Dorsey 1998).

In the chapter “Christians and Cree” of Home is the Hunter: The James Bay Cree and their Land, Hans M. Carlson (2008) presents an insightful history of the struggles of missionaries in James Bay. Although Christianity was adopted fairly widely early on, both Jesuit and Anglican missionaries were unable to control or direct the manifestations of faith. He gives the following example:

[The Anglican missionary] Horden… spoke to a man who claimed he had visited heaven, and who planned to follow the same path again. His vision did not become a movement, though he urged others to follow the path he had found. This speaks to the Cree’s continued experimentation with Christian ideas and how they adapted them to a context in which communication beyond the material world was a necessary part of life. The fact that these experiences were spoken of as a “path” is also provocative (110).

Many of the missionaries’ problems centered on mobility. As the Cree adopted syllabic writing very early on, missionaries could not control what doctrine the Cree chose to interpret from the Bible. This took the control of faith out of the hands of the clergy, who were at a double disadvantage: they could not follow their congregation into the bush, where they spent most of their time, and the structure the church was at odds with both mobility and independent thought. The entire authority of the clergy depended on their privileged position as interpreters of God’s word. And that depended on sedentism, as the first bishop of Rupert’s Land, Reverend Lofthouse, was keenly aware. Carlson quotes him as stating:

I am convinced that the work will be more permanent, and the gospel take deeper root, when the people are settled with something to bind and connect them to the soil. (Carlson 2008: 103)

Lofthouse refers to the people’s connection to the ‘soil’, not the land. Lofthouse came from a tradition of Christian doctrine that, like the structure of the church it supported, was based in sedentism and agriculture. The interpretations of the Catholic and Anglican Churches were not only steeped in over a millennium of mostly sedentary existence, they also associated mobility with punishment: Cain’s punishment for killing Able is that the ground will not yield food for him, and he will be forced to wander (Genesis 4:12, New English Bible). The Israelites wandered in the wilderness in punishment for their disbelief for forty years (Numbers 14: 26–38, New English Bible). This doctrine had also become intertwined with anti-Semitism.

The idea that the Jewish people were dispersed and homeless as a result of God’s displeasure with them seems to have originated with St. Augustine in the 5th century (Augustine 2004: 828). Anti-Semitism continued to influence both theology and public policy throughout the 18th century. This point was embodied in the folkloric figure of the Wandering Jew, who was condemned to roam the earth until the second coming for having insulted Jesus on his way to being crucified (Brichetto 2006). Nomadism was associated with moral degeneration, exile from civilization, and food privation.

This association was later made explicit by several authors in claiming that the Native Americans were one of the lost tribes of Israel. In his 1816 text, Star in the West Or a Humble Attempt to Discover the Long Lost Ten Tribes of Israel Preparatory to Their Return to Their Beloved City, Jerusalem, American statesman Elias Boudinot drew a number of parallels as proof of this assertion, including relating the truculence of Native Americans to become civilized to the ungrateful behaviour towards God shown by the chosen people in the wilderness after their escape from Egypt (2003). The implicit message was that they shared in the punishment of the Jewish people.

A nomadic way of life was condemned by the early missionaries in James Bay as being both impractical and morally degenerate. These are the contact narratives on which early anthropologists based their conceptualization of prehistory in northern Quebec.

Classical Evolutionary Assumptions

There do not appear to be any studies written specifically about northern Quebec from the Classical Evolutionary perspective, however information derived from contact narratives was used as analogies for the reconstruction of European prehistory. The following quote, from a volume called Fossil Men and their Modern Representatives by J.W. Dawson (1888), demonstrates this use:

By thus sitting at the feet of the red man, we may chance to discover some truths which the learned archaeologists of the old world have not yet attained and in any case may hope to present some interesting and instructive pictures of primitive man in the old world and the new. (17).

Early theories of Classical Evolution were radical explanations that human society and history were not predetermined, but the result of more universal rules governing the interaction of humans and their environment, similar to the rules identified in the burgeoning fields of, geology and biology (Carneiro 20031-4, 14). These rules showed that human societies developed from simple to complex in a similar fashion, and therefore that all living cultures were located somewhere along this line of development.

A broad racial homogeneity was attributed to the native populations of Canada. Using craniology, Daniel Wilson (1882), a Canadian anthropologist, broadly grouped the Indians of Canada into a ‘dolichocephalic’ type, which contrasted with the ‘brachycephalic,’ type of southern North America. Wilson was in fact quite critical of his intellectual forbearers and peers for ignoring diversity in North American craniology by grouping the entire continent into a single race (1882: 10–11), and he notes the political influences on interpretations of earlier periods:

The literature of the fifteenth and sixteenth centuries abounds with evidence that it was much easier to persuade the men of that age that Calibans and monstrous Anthropophogi peopled the strange regions beyond the Atlantic, than that these were inhabited by human beings like themselves (14).

While rejecting some of the more extreme interpretations of the early contact narratives, Wilson (1865: 19–20) nonetheless concluded that the nomadic groups of the Canadian subarctic represented the most primitive stage of society.

Although the concept of social reversion existed (Carneiro 1973: 80), there was no evidence to suggest that these groups had ever been more ‘advanced’. In fact, there was little evidence at all. Within the classical evolutionist paradigm developed by E.B. Tylor (1871), Herbert Spencer (1881), Dawson and others, this meant that, given the very low level of development observed, there was little possibility of the prehistoric Cree having changed much at all. This was compounded by political motivations. Trigger has argued that,

…the spread of Enlightenment rationalism among more educated Euro-Americans was creating the need for nonreligious explanations of Indian inferiority. As a consequence, racial myths eclipsed religious ones as a justification for seizing Indian lands and violating Indian treaty rights. It was widely maintained that the Indians were brutal and warlike by nature and biologically incapable of significant cultural development (2006: 159).

The land of the Cree was not yet at risk, but the interpretation of their prehistory would be affected by the wider political history of the incipient social sciences in North America. Alexis de Tocqueville’s chapter on “The Present State and the Probable Future of the Indian Tribes Inhabiting the Territory of the Union” in Democracy in America (1840) is a good example. De Tocqueville, writing in the mid-nineteenth century, describes in detail the deplorable ways in which settlers trick the Indians out of their lands, describes Europeans as tyrannical, and considers the official government policies oppressive; yet he sees very little hope:

It is easy to foresee that the Indians will never want to become civilized, or when they do so, it will be too late. Civilization is the result of prolonged social endeavour taking place on the same spot, an endeavour which each generation bequeaths to the next. It is harder for civilization to establish its sway over a hunting people than over any other. Pastoral tribes move from place to place, but there is always a regular system in their migrations and they continually retrace their steps; the dwelling place of the hunter changes like that of the animals he hunts (327. Emphasis added).

The assumptions of Classical Evolution were challenged in the early twentieth century, and yet we will see that they continue to influence archaeological theory. In the archaeology of the Shield Archaic, classical evolutionism is reflected in a characterization of prehistory as mobile, simple and almost unchanging. In its rejection of Divine explanations for social change, the theory of social evolutionism posited itself as objective. Later scientists were influenced by the picture of prehistory constructed under its assumptions.

The prehistory of northern Quebec that was developed in the early twentieth century can be understood in light of the factors discussed above. Homogeneity of culture across large areas was assumed. The primary cultural data was not archaeological or even anthropological, but the data came instead from contact narratives generated by highly biased sources which saw nomadism as immoral. The lens through which this data was interpreted was Classical Evolution, a teleological approach which saw Western Europe as the epitome of social evolution and to which had been added a distinctive rectilinear twist that did not allow for Native Americans to have ever been ‘more developed’ or even much different than they were at contact.

Early Culture-Historical Studies in Northern Quebec and the Canadian Shield

The influence of Classical Evolution can be seen in the works of early scholars of prehistory focusing both on northern Quebec and on the Canadian Shield. Although a new paradigm, Culture-Historical archaeology, explicitly rejected many aspects of Classical Evolution, this had little effect on interpretations of the prehistory of northern Quebec. It remained mobile, unsophisticated, and unchanging.

Academic work on prehistory in northern Quebec before 1948 involved virtually no archaeology (Martijn & Rogers 1969: 313). Among other reasons, it was considered unnecessary. Frank G. Speck, an ethnologist trained by Franz Boas, explained why in 1926:

We can hardly think of these natives of the northeast otherwise than as populations who have followed the same mode of existence for a very lengthy period, tribes which, to do this, cling to the habitat of the caribou and moose, the beaver, the seal, and the bear, the canoe-birch and the long semiarctic winter… The area had perhaps achieved its optimum of population; the balance may have been reached, beyond which the natives did not increase against the rigors of an exacting and infertile environment. Yet another reason for the stability of population seems to stand forth, in that early native ingenuity was not sufficient to overcome the difficulties and unresponsiveness of barren nature by inventions which would have enabled the people to exploit to a fuller degree the resources of the [Quebec-­Labrador] peninsula. Then, above all in importance, is the further consideration that the region was out of the path of culture diffusion. And so it remained untouched by the advancing frontiers of agriculture, the agricultural arts, and those of native political and social change which moved from certain centers in the southeastern, central and southwestern portions of the continent transforming un-ruled, unorganized hunting units into settled and socialized farming groups as we see them among the central and eastern central Algonkians (272–274).

Here, Speck listed and combined most of the reasons that have ever been put forward to explain and promote the uniformity of Canadian Shield populations through space and time: cultural tradition, environmental restriction, lack of intelligence and lack of diffusion from other areas. Moreover, these characteristics, which Speck attributed to the contemporary population, not only explained a fossilized culture, but compelled it.

Speck was trained by Boas (Hallowell 1951), and the result of Boasian relativism in the hands of archaeologists was clearly a focus on identifying ‘cultures’ in archaeology through material remains, specifically artefact typologies. This rendered most of northern Quebec singularly uninteresting, because of the extreme paucity of data to interpret. Speck himself believed that archaeologists were unable to say anything about non-material behaviour in human prehistory (Trigger 2006: 365). Prehistorians fell back on the Direct-Historical approach, trying to project prehistory backwards from the known evidence of the historical period. However, since almost no material evidence of prehistory had been discovered in this area, it had the effect of projecting early historical descriptions and ethnography back onto nothing. This created an essentially unchanging prehistory, maintaining many of the assumptions of Classical Evolutionism about the prehistory of the Canadian Shield, as is clear from Speck’s description.

In 1946 the Peabody Foundation published a volume entitled Man in Northeastern North America, edited by Frederick Johnson. In his preface, Johnson discusses the disagreements and contradictions to be found in the volume, saying: ‘It is clear that these present discrepancies are of extreme value; they are some of the problems to be faced in the future’ (vii). From a history of archaeology point of view, this volume demonstrates a lack of consensus within and between disciplines when dealing with indigenous peoples and indigenous archaeology. Looking at the archaeology and anthropology papers in particular, the volume shows an increasing dissatisfaction with the Culture-Historical approach, but little rebellion against the underlying evolutionism. Disagreement is evident on several issues within the volume, including the cultural homogeneity of the area (McKern 1946; Spaulding 1946), the role of diffusion in its prehistory (Cooper 1946; Spaulding 1946: 147–148), the relative weight of individual versus group decision-making (Fisher 1946: 223–234; Hallowell 1946) and the level of cultural sophistication (Cooper 1946; Flannery 1946).

There is little distinction by any authors between present, contact, and prehistoric data. There does not seem to be any coherent pattern between theory and conclusions. For example, Spaulding (1946: 146) says that the lack of natural barriers and similarity of environment in the boreal zone creates a uniformity in culture. McKern (1946: 36), on the other hand, sees the lack of natural boundaries as allowing for the broad diffusion of cultural traits from all directions, creating a ‘kaleidoscope of complexity’.1 This was the state of knowledge when the first real archaeological survey of the Boreal Forest area of Quebec by Rogers and Rogers took place in 1948.

Prehistory in Northern Quebec: Excavation and Early Syntheses

The earliest professional field archaeology work in the Boreal zone of northern Quebec was done in 1947 and 1948 by Edward Rogers and Murray Rogers in Mistisini-Albanel and along the Rupert River. In their reports, they generally discuss sites located inland.

The excavation of two quarry sites, and the discovery of artefacts beneath a substantial leeched layer, led them to conclude that there was a stone industry of ‘respectable antiquity’ in the area (Rogers & Rogers 1950: 336). They also concluded that this same industry continued to the present almost unchanged (336). They thought that the tools were similar to lower and middle Paleolithic tools found in Siberia and Alaska, and that all other tool technologies in North America were ‘less primitive’ (Rogers & Rogers 1950: 336). They did not offer explanation for this geographical discontinuity, but were explicit in stating that they did not ascribe it to degeneration theory. In fact, their theoretical explanations were cautious and minimal:

Until our knowledge of the ecology of the region is more complete we can only assume that the taiga invaded the tundra which covered the region at the close of the Pleistocene. The people who moved into the region are assumed to have been taiga hunters, the theoretical and as yet unproved origin of which has usually been located to the west. As yet no clear typological similarities among various groups of cultural materials in the north can be seen (337).

Perhaps the most surprising aspect of their analysis is the comparison with old world artefacts and typologies. Not only is there a disjunction in time and space, there is also the consideration of available material. Whereas typologies from Europe are usually based on flint or chert tools, the majority of material available locally in the Canadian Shield is quartz or quartzite, which is nowhere near as high quality a material. The more ‘primitive’ characteristics of these tools is just as likely to be derived from the material as from the skill of the producers. Rogers and Rogers were operating under the assumption that all technologies were derived from the Old World, and therefore that the explanations for archaeological materials is to be found in a theory of the peopling of North America.

Frederick Johnson, of the Peabody Foundation, was asked to review the lithic material brought back by Rogers and Rogers (Johnson 1948). Although he acknowledged the comparability of the material to that of the Middle Paleolithic, he questioned the usefulness of applying this typological system in North America:

It has been the custom to convert a tenuous inference into a dogma which states that American stone tools are derived from the Neolithic. Time and energy have been expended in the attempt to correlate types of American tools with types found in Siberia and even Europe. These efforts have led to arguments, based on typology, which tend to question or even deny stratigraphic and geographic factors, to say nothing of ignoring the more intangible factors which are responsible for the development of human culture (92).

He continued, pointing out that this material is quite different even from that found in the adjacent Tadoussac region (95–96), and concluded that, ‘The place the Rogers’ collection occupies in the culture history of the region is as yet quite unknown’ (95).

Taken together, these articles demonstrate the beginnings of what would become important characteristics of field archaeology in Northern Quebec. These include intensive descriptive detail, conclusions only being applied to specific geographic areas, and conservatism in interpretation. This conservatism was not so much in adherence to established theory, which Johnson challenged, as it was an unwillingness to generalize beyond the data or even speculate much about its significance. In short, no theory was yet developed to challenge the one articulated by Speck.

In 1955 Gordon Lowther published a summary of archaeology in Quebec in a volume entitled Survey of the Aboriginal Populations of Quebec and Labrador. The volume and this chapter were intended to summarize the state of knowledge at that point, as opposed to presenting anything new. Lowther’s contribution was a strange piece. He was insightful and clear about how archaeology could be practiced better in Quebec, advocating interdisciplinary training and approaches (69–73). His theory of prehistory, however, was a throwback to Speck. He highlighted the fact that most of Quebec archaeology has focused on the very far North or the Iroquoian societies in the south, but spent most of the article interpreting the lack of information in between (67). He argued:

It is reasonable to assume that the Indians who inhabited the interior of Quebec before the sixteenth century and back as far as perhaps the first century A.D. lived in similar social conditions to those obtaining amongst the Indians of that area in the historic period. This assumption is based on the evidence of the accounts written by the earliest European explorers, on our knowledge of the rate of change amongst present day Indian societies—and on negative evidence (67).

The ecological environment of Quebec has altered but little since the last positive phase of the Quaternary ice-age, and for the last eight thousand years the physical environment must have been almost as it is today… This environment is one of a peculiar harshness…

The climate and topography over most of this territory exclude abundant agricultural activity, and the Indian population is compelled in consequence to rely on hunting and collecting for subsistence.

Societies dependent upon a hunting economy rarely achieve an advanced social organization, and mobility is a prime factor in their economy (68).

In Lowther we find the same associations of mobility, primitive organization, and fossilized culture that are found in the contact period literature and earlier academic writings, but explicitly combined with ecological determinism. This is where the biases of the contact narratives and the stages of social organization defined by Classical Evolution became part of the scientific narrative that eventually produced the concept of the Shield Archaic.

Lowther was taking a new, scientific approach to an academically well-established theory: that mobile people were rarely socially advanced. The emerging dominance of ecological and environmental explanations was part of a wider debate in archaeology (Trigger 2006: 386–390). The difference between Johnson and Lowther is an example of the shift from Culture-Historical to Neoevolutionary archaeology and New Archaeology. Cultures were being interpreted as functionally integrated systems. As Trigger (2006) argued, ‘Changes in all aspects of cultural systems were therefore interpreted as adaptive responses to alterations in the natural environment, changes in population pressure, and competition with adjacent cultural systems’ (394). There was now a scientific explanation to link ‘primitive’ subsistence strategies to ‘primitive’ social complexity, ‘primitive’ religion, indeed, to ‘primitive’ everything.

The ‘Infertile Crescent’ and the Shield Archaic

In 1972 Wright suggested that the ‘broad spatial homogeneity and temporal continuity’ of the Canadian Shield ‘permit far more extreme extrapolations than are possible in areas to the south which have more complex developments’ (1–2). This is obviously a circular argument.

The concept of the Shield Archaic, as it was defined by Wright, is more the result of the history of thought described above than of systematic archaeological investigation throughout the region. Wright himself observes, when considering excavated material, that archaeological sites in the Shield show temporal and spatial variation. A broad and technologically mixed buffer zone between the Shield and Laurentian cultures is suggested to explain the differences in material culture of the southern Shield during the Archaic period, while ‘unexplained’ connections with the Keewatin culture of the Northwest Territories deal with the same problem in the north (1968: 57).

The Shield Archaic dominated the study of prehistory in the boreal zone of northern Quebec for thirty years. It is only within the last fifteen years that use of the term has tapered off, in recognition of both the increasingly evident variability of the archaeological record and theoretical critiques (Clermont 1998; McCaffrey 2006). Excellent work has been done on the influence of biased contact narratives in academic writings about First Nations in North America generally (Trigger 1980; 1982) and in the subarctic in particular (Holly Jr. 2002). Donald H. Holly Jr. (2002) summarizes the problem succinctly:

So long as our categories of study and interpretation remain spatially broad and historically shallow, subarctic prehistory is doomed to a timelessness of our own design (17).

One of the biggest problems with the Shield Archaic was that it was applied to such a broad region. This made it very difficult for any single archaeologist or any single site to challenge the overarching paradigm, although attempts were made within Quebec (Martijn & Rogers 1969: 329; Séguin 1985: 93; Ethnoscop 1988: 15).

Conclusion

The study of prehistory in northern Quebec went through religious, classical evolutionary, culture-historical, and Neoevolutionary scientific explanations with hardly an alteration in their conclusions. Figure 2 is a summary of how many sources, interpretations and motives all led to the same conclusions.

Figure 2 

Summary of conclusions.

The seeds of alternative explanations are present, however: in the variability through contact identified by Wright; the temporal and regional sub-divisions suggested by the later Quebec archaeologists; and even in Lowther’s ecological explanation. More recent understandings of the prehistory of northern Quebec, and of the Canadian Shield as a whole, are shifting away from homogenizing explanations (Clermont 1998: 56; Denton 1998; Dumais & Poirier 1998; McCaffrey 2006:162). Yet a new paradigm has yet to emerge. Rather than creating another situated narrative about the prehistory of northern Quebec as part of the Canadian Shield, old and new data need to be understood in light of the broader theoretical discourse on hunter-gatherers, a discourse which acknowledges the variability of hunter-gatherer adaptations (Ames 2004). A new paradigm for the prehistory of northern Quebec must be part of a larger theoretical project that seeks to explain the processes by hunter-gatherer societies are shaped, rather than classifying them in restrictive categories.

In his book The Invention of Primitive Society, Adam Kuper (1988) tracks the persistence of the idea of ‘simple’ and ‘primitive’ society through the history of anthropology as a whole. He describes how anthropological arguments about the existence and nature of primitive societies are torn apart and inverted by every generation of scholars while the central assumptions remain intact. He argues that primitive societies were in fact the original object of social anthropology, the reason the tools of anthropology had been developed, and ultimately the background against which any other societies were compared (240–241). This last point is the most important in the case of earlier academic characterizations of the prehistory of the Cree in northern Canada. What was needed was a baseline for North America, against which anthropologists could compare what seemed to be the much more interesting cultures south of the Shield. By the time anthropologists had become interested in the prehistory of the Canadian Shield in itself, a theory was already firmly established not only despite a lack of data, but probably because of it. It took thirty years of data accumulation to overthrow a paradigm based on almost no evidence.