A research proposal

Douglas T. Robbins
Department of Anthropology
McGill University
1989

This proposal was prepared as part of my PhD studies at McGill

Near the close of the fifteenth century, Spanish, Basque and Portuguese fishermen exploiting the wealth of newly-discovered fishing grounds near the Island of Newfoundland became the first Europeans -- since the time of the Norse voyages -- to encounter the native inhabitants of what is now Atlantic Canada. Throughout the sixteenth and seventeenth centuries encounters occurred between Europeans and Beothuk Indians of Newfoundland, yet until the latter years of the eighteenth century writings pertaining to the Beothuks consisted of little more than fleeting accounts. Beginning in the 1760s and continuing into the early nineteenth century concern for the Beothuks increased, and as a result their mention in historical texts became more numerous. This was, however, but the last chapter in a poorly-understood native history, as the Beothuks were extinct, or extirpated from the Island, by the 1820's.

Since that time, the extermination of the Beothuks has been popularly attributed to the actions of aggressive English settler-fishermen who perpetrated an unofficial war upon the natives, murdering, pillaging and committing unspeakable atrocities without remorse; in fact, with pleasure (eg., Horwood 1959). This view has been repeated most recently (to my knowledge) by J.G. Bradley in his introduction to the 1974 edition of J.P. Howley's The Beothucks or Red Indians: The Aboriginal Inhabitants of Newfoundland, which constitutes the most exhaustive collection of historical references to the Beothuks. Recently, historians have offered alternative explanations of the Beothuks' demise. Francoy Raynauld (1984) and Ralph T. Pastore (1987) have argued against the significance of open conflict between Beothuks and whites as a factor in the extinction of the Beothuk population. The latter has cited the absence of European fur-traders, missionaries and Indian agents in Newfoundland as being in responsible for the course of events. In the absence of a fur-trade Beothuks acquired Europeans goods through pilferage from seasonally-abandoned fishing premises which, while satisfying short-term needs, locked them into an antagonistic relationship with whites that proved disasterous in the long term (Pastore 1987).

The end result of white-Beothuk relations was unique in northeastern native history. The setting for these events -- an island -- was also, to a degree, unique. To what extent was the extinction of the Beothuks related to their island existence? Was island-dwelling, for the Beothuks, a factor which figured prominently in the course of historical events? The little that is recorded regarding Beothuk culture indicates that they differed in several ways from natives of the neighbouring mainland, with regards to such features as language, the extensive use of red ochre, and their style of canoe. These traits -- especially language -- would seem to suggest long term isolation from neighbours. Turning attention to other insular populations, however, it may be noted that the physical confines of an island seldom present such insurmountable barriers as to effect total isolation.

This research proposes to examine firstly the relationship between the Beothuks of insular Newfoundland and natives of the neighbouring mainland, and secondly to explore possible effects of that relationship upon the course of historical events. The first question to be addressed concerns the connection between insularity and isolation: to what extent were Newfoundland Beothuks politically, economically and socially divorced from their mainland contemporaries? This question will probably never be answered in toto, due to the dearth of historical documentation and the absence of ethnographic accounts. A partial answer may be obtained through archaeology, however, by a comparative analysis of Indian material culture in Newfoundland and Quebec-Labrador. Artifact assemblages from several prehistoric and protohistoric Beothuk sites on the Island will be examined, to gain a knowledge of Beothuk toolkits and the methods and materials used in manufacturing them, and then compared and contrasted with assemblages from the mainland region. This should provide a measure of the degree to which Newfoundland Beothuk culture stood independent of contemporaneous Indian occupations of the neighbouring mainland.

In addition to this archaeological analysis, research of northeastern ethnography, prehistory and ethnohistory will be pursued, with the objective of exploring the nature of aboriginal interrelationships in this area, and the role they played in the European-white relations during the historical period. This will provide a background from which analogues may be drawn regarding possible relationships between Newfoundland and Labrador, and their significance in Beothuk and Beothuk-white history.

i. Recent Indian Archaeology in Newfoundland and Labrador

The archaeological remnants of Beothuk culture escaped the attention of early Newfoundland researchers (W.J. Wintemberg, Diamond Jenness, Elmer Harp Jr.) partially due to a geographical chance. These archaeologists concentrated their efforts (in Newfoundland) on the Great Northern Peninsula -- mostly along the Peninsula's western shore. To date, not a single site of Beothuk habitation has been discovered in this region of Newfoundland, although there are many sites of more ancient Maritime Archaic and Palaeo-Eskimo occupation. Whether this absence of Beothuk sites is real -- and significant, as will be discussed below -- or merely reflects insufficient exploration remains an open question.

Beothuk research thus fell to a later generation of archaeologists. Helen Devereaux (1970) and Raymond LeBlanc (1973) excavated and reported upon historic Beothuk sites along the Exploits River. Also along the Exploits, amateur archaeologist Don Locke collected at numerous Beothuk sites (Locke 1972), and in recent years has "reconstructed" a Beothuk village. During the 1970's Paul Carignan performed a number of excavations at sites in Bonavista Bay (Carignan 1975b, 1977). These excavations proved important as they were the first to explore prehistoric Beothuk culture: previous works had dealt with Beothuk sites of historic age. Carignan's fieldwork led him to suggest that Beothuks and Dorset Palaeo-Eskimos co-existed in Newfoundland during the first millennium AD, and that Beothuks derived from the Maritime Archaic tradition (1977:ii). This latter idea has not been refuted or supported in its entirety by subsequent research. It is now suspected that sites excavated by Carignan possessed mixed components, which led him to group together artifacts of different ages, and in some instances, derive inappropriate radiocarbon dates.

Work by Gerald Penney in southern regions of Newfoundland during the early 1980's (Penney n.d.a, n.d.b, 1980, 1981) consti­tuted an advancement in Beothuk prehistory, as it opened a prev­iously unrecognized chapter of Indian occupation on the Island. Whereas Carignan dealt with, in Bonavista Bay, a series of multiple component sites -- multiple in terms of repeated Indian and Palaeo-Eskimo occupation -- these south coast sites presented small but homogeneous assemblages dating to the early centuries of the first millennium AD. Penney named these assemblages the "Little Passage" complex, after the channel known as Little Passage between Bay d'Espoir and Hermitage Bay (Penney n.d.a).

The second major contribution to Beothuk prehistory during this decade stemmed from the work of Ralph T. Pastore at the Boyd's Cove site in Notre Dame Bay (Pastore 1983, 1984, 1985, 1986). From Beothuk house pits at Boyd's Cove, Pastore recovered spear- and arrowpoints and other items of native-worked European iron in association with stone implements nearly identical to those of the Little Passage complex. Hence, it was possible to link the historic Beothuk population -- as known from Boyd's Cove and sites along the Exploits River -- with the bearers of the Little Passage complex, and extend Beothuk prehistory back to about AD 1100.

With this new knowledge of Little Passage and historic Beothuk toolkits, it became possible to re-examine collections made by Carignan in Bonavista Bay. This made obvious the fact that some tools discovered by Carignan conform to neither Little Passage nor Beothuk types, but instead represent another period of Indian occupation. Labelled "The Beaches" complex, for The Beaches sites in Bonavista Bay, these tools display some resemblances to Little Passage ones, but also possess characteristics which, together with a few radiocarbon dates, indicate an older age. Most evidence of The Beaches complex comes from eastern and northeastern Newfound­land, where it is believed to date ca. AD 900-1100.

The third and final archaeological entity which composes what is known as the "Recent Indian" period of insular Newfoundland is the "Cow Head" complex. First identified at the Cow Head site in northwestern Newfoundland, Cow Head complex differs from the pre­viously-described complexes in possessing none of the notched projectile points characteristic of The Beaches and Little Passage. Instead, it is represented by a variety of large bifaces and a few stemmed projectiles, unlike other Recent Indian implements, but reminescent of Maritime Archaic artifacts. Radiocarbon dating at the Cow Head site puts this complex between 50 BC and AD 950. The relationship between Cow Head and The Beaches complexes remains unclear; it is convenient to suggest an evolutionary one, yet this remains undemonstrated. As well, the origin of the Cow Head complex is unknown. Tool forms suggest Maritime Archaic affiliations, yet there is an hiatus of more than 1,000 years between that tradition and the Cow Head complex, judging by the existing C14 chronology.

The Labrador coast presents a somewhat different picture of Indian occupation. Whereas insular Newfoundland appears to have been unoccupied following the demise of the Maritime Archaic trad­ition 3,200 years ago until the appearance of Recent Indians about 2000 BP, a population of "Intermediate Indians", likely descended from the Maritime Archaic, persisted along the central and southern Labrador coast during this time. Thus, a continuum of occupation, if not necessarily of population, occurred along the Labrador coast from about 7,000 years ago until recent times.

Surveys and excavations by researchers along the central Labrador coast have led to the identification of a "Point Revenge" complex, a Recent Indian occupation dated AD 1000-1650, and possibly as early as AD 700 (Fitzhugh 1978:146). Both the origin and even­tual fate of the Point Revenge complex remain unclear. Typological similarities with the preceding Intermediate Indians are few and do not indicate clearly an in situ development, although this possibility cannot be entirely ruled out. It may be argued that the frequent occurrence of distinctive Ramah chert in Point Revenge and Intermediate Indian assemblages suggests continuity, yet this is certainly not conclusive evidence. Fitzhugh (ibid. p. 172) has suggested that Point Revenge origins may be found in Middle Woodland cultures of the lower Gulf of St Lawrence, the southern Maritimes and northern New England, although it remains unknown whether migration or diffusion was the vehicle of this expansion. Recent Indian sites in Labrador dating to the first millennium AD are rare, and those which are known (one north of Hopedale and another at Pinware in southern Labrador) have produced few diagnostic artifacts.

The Point Revenge occupation in Labrador is contemporaneous with The Beaches and Little Passage complexes on the Island of Newfoundland. Typological similarities between artifact assemblages exist, including toolkit inventories and the forms of individual specimens. The degree of similarity is not known, however, nor is it known whether these resemblances reflect only a general Middle Woodland ancestry, "functional" convergence of tool types, or real interactions between Indian populations of Labrador and insular Newfoundland. It is these questions, among others, which must be addressed in order to gain an understanding of Recent Indian prehis­tory in this far northeastern region of North America.

ii. Beothuks as Island Dwellers

An island setting may provide unique circumstances for occupation, both for animals and humans. Biologists have had, perhaps, a more successful history in quantifying the effects of island geography (e.g., MacArthur and Wilson 1967). As oceans present barriers to expansions, islands are often "depauperate" with respect to the variety of mammalian wildlife; the number of species is less than continental regions, in proportion to island size. Newfoundland illustrates well this phenomenon. There are only 14 species of native land mammals, ten of which are distinct subspecies occurring only on the Island (another, Canis lupus beothucus, is now, like its namesake, extinct) (Mednis 1981).

In some ways it is tempting to draw an analogy between animal and human island occupation, and view human groups as being also limited and confined by physical barriers of insularity. A brief survey of the literature, however, reveals few instances of cultural isolation effected by island geography; the most striking case is perhaps that of the aborigines of Tasmania. These people were unable, or unwilling, to make the lengthy open-sea crossing to the Australian mainland. Their traditional boats, constructed of rolls of bark, retained buoyancy only for limited periods and were entirely unsuited for such a voyage (Jones 1976:239-48). Geography and technology in this case resulted in millennia of total isolation for the Tasmanian natives.

Such instances, however, appear more the exception than the rule. The case of the Polynesians might be viewed as a marked contrast to the Tasmanians. These people undertook voyages of considerable distances, and were able to make accurate landfalls on islands invisible over the horizon. This voyaging was undoubtedly aided by the well-designed, sea-going crafts the Polynesians possessed (Finney 1979).

It might be concluded that, while human populations were affected by maritime geography, the effects were different than upon animals species, as mankind's ability to travel was not necessarily confined by water barriers. A maritime environment may present efficient routes for cultural relations. The particular pattern (of relations) which may arise will be dependent, in part, upon local geography and environment, and the water-transport technology of the inhabitants.

Travel distance and transport technology are therefore two determining factors in the occurrence of relationships between island and mainland populations, or between two or more island groups. These factors interrelate, as relatively rudimentary vessels would permit water crossings of short distance (the Tasmanians exploited near-shore islets off the Tasmanian coast), while functionally more efficient vessels would be required for longer crossings. This presents an interesting chicken-and-egg conundrum with regard to maritime travel technology and cultural interrelations in prehistory: did such technology (developed, perhaps, for immediate resource-exploitation purposes) provide a means of increasing outside contacts, or did a desire for such contacts provide an incentive for developing more efficient boats?

In the case of Newfoundland, geographical considerations do not appear to impose insurmountable barriers to intercourse with the mainland. The Strait of Belle Isle, which divides Newfoundland from the continent, is, at its narrowest, 20 km wide. On a cloud-and fog-free day (admittedly rare) the opposite shore is clearly visible. From approximately late January to late April each year dense pack ice fills the Strait (Farmer 1981:69-74) making it possible, under ideal conditions, to cross without the aid of boats.

In this respect it is interesting to note that during the summer of 1988 it was discovered that a small population of coyotes had established itself on the Island of Newfoundland. Local wildlife biologists believe that the coyotes, mammals not indigenous to Newfoundland, crossed to the Island on the previous spring's pack ice.â During other times of year a crossing by boat would not be unduly impeded by maritime conditions. A general clockwise ocean current flows around the Island of Newfoundland and eastward through the Strait of Belle Isle, where it meets the southward-flowing Labrador Current (ibid. pp. 60-8). Prevailing winds are westerly or northwesterly throughout much of the year (Banfield 1981:103-9). Thus, voyagers crossing the Strait by boat would tend to be swept east by the current, and possibly the wind, although a northerly wind would partially counteract this effect. For travellers from Labrador, this drift could easily be allowed for by initiating the crossing at a longitude west of the intended landing point in Newfoundland.

A well-timed attempt to cross the Strait (avoiding, for instance, days of gale-force winds and high seas) would likely meet with success, even without the benefit of elaborate vessels. Beothuk Indians in Newfoundland during the historic period used a type of birch bark canoe similar in some ways to Algonkian ones, but possessed of a unique shape. If the account of one historical observer is to be trusted, the Beothuks travelled in these vessels to the Funk Islands, some 64 km off the northeast coast of Newfoundland:


It is not known when the birch bark canoe appeared in Beothuk prehistory. The fact that Newfoundland was occupied at all (during the Recent Indian period) suggests that the initial immigrants possessed boats, unless, of course they crossed to the Island on the pack ice. Marshall (1985), in an analysis of data pertaining to Beothuk canoes, argued that there were in fact two functionally different canoe forms. One type, which Marshall believed the more ancient, possessed attributes making it suited for river and lake travel. The second type had a strong rocker (a rising bow and stern), and appeared specialized for ocean travel (ibid. pp.127-9).

Another thread of evidence supports the contention that prehistoric Beothuks were employing boats for coastal voyaging. A few artifacts have been discovered on the island of St. Pierre, off the south coast of Newfoundland, which show marked resemblances to Newfoundland Recent Indian ones in terms of lithic material and form. If these artifacts are indeed evidence of links between Indian occupants of St. Pierre and Newfoundland, then their makers must have possessed some form of watercraft in order to cross this expanse of ocean, which usually remains ice-free year-round.

In sum, it seems that neither local geography and environment nor the available transport technology would negate culture contacts across the Strait of Belle Isle, between Newfoundland and Quebec-Labrador.

A consideration of voyaging like this invites comparison with other groups of the Atlantic region. In a recent paper, Charles Martijn (1989) has discussed Micmac voyaging among islands of the southern Gulf of St Lawrence, namely Cape Breton, the Magdelens and southern Newfoundland. Archaeological evidence from the Magdalen Islands (McCaffrey 1986) indicates that voyages of approximately 100 km occurred during prehistory. Martijn notes that in 1597 (Quinn 1979, vol. IV:69-70) there were 300 Amerindians (presumably Micmac) on the Magdalens. He suggests that this Micmac presence may indicate that the Magdalens "...constituted a seasonal extension of the geographic space normally exploited by the Eastern Micmac." (Martijn 1989). This extension may have occurred during times when aquatic resources in Cape Breton fluctuated and were unreliable. On the Magdalens, Micmacs were able to subsist upon harbour and grey seals, walrus, and a variety of fish and fowl resources.

Might an analogy be drawn between Micmacs voyages to the Magdalens and Beothuks journeys to the Quebec-Labrador coast during the late prehistoric period? Extensive coastal voyaging was likely not beyond the capabilities of the Newfoundland Beothuks. However, extended travel along the Great Northern Peninsula of Newfoundland, through a coastal region apparent not regularly occupied, would presumably have required fairly strong motivations. A question is, therefore, did such motivations exist?

As was the case with Cape Breton Micmac, Beothuks along Newfoundland's northeast coast may have experienced fluctuations in the availability of resources, which might have forced them to expand their territory. Harp seals, for example, may have become inaccessible in the short-term due to local weather and ice conditions. Yet it seems that in this instance Beothuk may have turned their attention towards southern Newfoundland, rather than Labrador. The former area offered herds of sedentary seals (harbour and grey), together with other resources, that would not be influenced by the adverse conditions on the northeast coast. It should be emphasized that I am referring to the late prehistoric period, prior to European settlement of this region of the Island.

i. Hypotheses Regarding Beothuks in Prehistory

The analysis will proceed with two general and alternative hypotheses in mind:

1. No significant differences exist between the (archaeological remains of) late prehistoric Indian occupations of insular Newfoundland, the southern Quebec-Labrador coast, and the central Labrador coast. If this is so, then one or both of the following conclusions may be warranted: (a) communication and exchange occurred throughout the region. This would appear to imply that extensive voyaging occurred, so that unoccupied zones (e.g., the Great Northern Peninsula) could be easily bypassed; (b) differences which may have existed (cultural, social, political, etc.) are not reflected in the archaeological record.
   
   
2. Demonstrable patterning exists, which relates to the network of communication and exchange that existed during the late prehistoric period. One of at least two patterns may be observable:
   
   
   1. a clear dichotomy between the Island of Newfoundland and the Quebec-Labrador mainland. This would appear to support the conclusion that, throughout much of the late prehistoric period, communication and exchange between these regions was minimal.
      
      
   2. a configuration whereby Island assemblages are seen to compare favourable with ones from southern Labrador and the Quebec Lower North Shore, and contrast with ones from central Labrador (to the north) and the Middle North Shore (to the west). The would appear to indicate interaction between Newfoundland occupants and peoples of the nearest, adjacent mainland region. The diversity of assemblages increases with distance to the north and west, which would appear to imply a concomitant decrease in communication and exchange.

ii. Beothuks in Newfoundland History

These hypotheses concerning Beothuk relations with mainland populations gain new significance with respect to the historical period, and consequently it is desirable to examine briefly the historical setting of Beothuk-white contacts on the Island.

The potential of the Newfoundland codfishery was recognized by the early 1500s, and throughout the 16th century fish stocks were being exploited by vessels travelling to Newfoundland each year from western European nations. During the seventeenth century, this migratory fishery faced increased competition from colonial endeavours such as those initiated by John Guy and George Calvert (later Lord Baltimore), and an unofficial settler population. Dried fish, produced by the settler population, competed with the wet salt fish brought to Europe by vessels engaged in the migratory fishery. The growing settler population and the market demand for dried fish resulted in increased coastal land use by whites, for the construction not only of dwellings, but also of stages, wharves and drying flakes for the prosecution of the in-shore fishery and the preparation of dried fish. The white presence in Newfoundland became more widespread, elaborate and permanent.

The fishery remained, however, a seasonal occupation in most of Newfoundland. During the dormant season, from the beginning of winter until spring, settlers turned to other activities to augment subsistence, creating the pattern of seasonal transhumance discussed by Smith (1987a, b) and referred to as "winterhousing". Smith has traced this phenomenon through historical references back to the 1730's. It may well have existed, perhaps in a less distinct form, prior to that time. During this "off-season" settlers in several regions of the Island moved from their coastal settlements to interior parts of the Island to exploit food resources, to gain access to a plentiful supply of wood for burning and construction, and to trap fur-bearing animals. The last of these provided another means for settlers to participate in a market economy: many became part-time furriers.

By the early 1700's regular contacts were likely occurring between Beothuks and white settlers. The Beothuks' seasonal subsistence cycle brought them to the coast each spring and summer, to hunt seals and birds, and take salmon during the spawning runs in rivers. Many traditional Beothuk coastal camps were situated close to settler communities; in some instances they coincided precisely. White furring and winterhousing opened up many interior regions of Island to settlers, bringing them again into contact with Beothuks, or Beothuk camp locations.

Despite this, white settlers made no attempts at creating an amicable relationship. The primary incentive for white-Indian alliances which existed in other northeastern regions -- fur trading -- was absent, as settlers were exploiting directly the fur resources of Newfoundland in a manner which complemented their fishing activities. In Newfoundland the most important economic endeavour was the codfishery, the potential of which far eclipsed that of fur-trapping. The codfishery operated in an economic, political and technological milieu completely foreign to traditional Beothuk culture.

The white settler likely saw no reason for cultivating relations with native Beothuks. Indeed, in some regards they may have been thrown into direct competition, for coastal land and interior fur trapping areas. Historical accounts (e.g., Prowse 1895; Howley 1915) contain numerous records of antagonistic encounters between Beothuks and whites. From these encounters has grown the most widely known, and popularly accepted, explanation of the Beothuks' extinction: that they were slaughtered by settler-fishermen outraged by acts of Beothuk plundering.

For over a decade historians (Upton 1977, Raynauld 1984, Pastore 1987, Marshall 1988) have offered other explanations for the extinction of the Beothuks, which have gained favour over the myth of wholesale slaughter of the Beothuks by white settlers. Both Francoy Raynauld (1984) and Ralph T. Pastore (1987) have argued against the significance of open conflict between Beothuks and whites as a factor in the extinction of the Beothuk population. Raynauld (1984:58) has viewed accounts of violence against the Beothuks in light of the changing political and economic environment of late 18th century Newfoundland. He has argued that an exaggeration of the violence perpetrated by settlers against the Beothuks served the purposes of the seasonal English fishermen in opposition to the growing settler population.

Ralph T. Pastore believes it is significant that a fur trade did not develop between Beothuks and whites in Newfoundland. In the absence of a fur trade, Beothuks obtained desirable European goods by theft from seasonally abandoned fishing premises. The long term effects of becoming "locked into a pattern of acquisition of goods through theft rather than trade" may have contributed significantly to the Beothuks extinction, as there were "no significant Europeans to whom a continued Beothuk existence was important" (Pastore 1987:59-60).

Another perspective on the question of the Beothuks' extinction might be gained by viewing it in light of relations -- or possible relations -- between Beothuks and mainland natives. What role might the presence, or absence, of an ongoing communication between Indian groups of Newfoundland and Quebec-Labrador have played in the course of historical events?

Before addressing that question, it may be pertinent to con­sider briefly some implications of European-Indian trade. Perhaps the earliest instances of such trade in the northeast were motivated, on the Indians' behalf, by a curiosity of and desire for the new and superior materials the Europeans possessed. As the opportunity for trade increased, however, different and more compelling motivations appeared. Kenneth Morrison has stated that, "The Algonkian who gathered annually at Tadoussac were the first American Indians to experience the long-term cultural impact of the fur trade.... The coastal peoples, especially Montagnais and Micmac, enjoyed advantages in the developing trade. Precontact trading patterns intensified as the eastern Algonkian traded metal for furs with more isolated tribes. (1984:21). (Emphasis added.)

Inter-tribal trade of European goods was a means of enhancing prestige and gaining wealth for tribes who played a middle-man role between European traders and more distant tribes. Also, a military alliance with whites presented new and powerful means of pursuing warfare with traditional enemies. In these respects, the nature of prehistoric relationships between Beothuks of Newfoundland and mainland Indians may have figured prominantly in the course of historic events on the Island.

If there existed a prehistoric alliance and trade between Montagnais and Beothuks, then Montagnais might have pursued trade with the Beothuks during the historic period, exchanging European goods for furs, as they did with tribes to the north. It seems reasonable that Beothuks would have readily entered into such an exchange, as it provided a means of obtaining desired European goods and at the same time avoiding whites. There is one recorded incident, in 1718, when Montagnais came to Newfoundland explicitly to trade with Beothuks, yet this expedition was motivated by a French trader at Bradore Bay rather than by the Montagnais (and the Montagnais were unable to find any Beothuks) (Pastore 1987:58). It seems that throughout much of the 18th century Montagnais came to Newfoundland to hunt furs themselves rather than enter into a "middle-man" position involving Beothuks, avoiding what might have been a profitable economic role.

If, on the other hand, a traditional enmity existed between Beothuks and Montagnais then this might have provided an incentive for Beothuks to seek trade with Europeans in Newfoundland, to obtain more efficient weapons for use against their enemies. Indeed, if such a state of affairs existed, the Beothuks might have actively sought an alliance with whites, especially English, in Newfoundland. Yet it is clear that the Beothuks did neither of these.

The prehistoric state of relations between the Island and the Quebec-Labrador mainland may have discouraged any intensification of Beothuk-Montagnais trade during the historic period. If little trade occurred prehistorically, and no active hostilities were present, then it can be argued that some of the primary motivations for pursuing and intensifying relations during the historic period were absent. As well, these same factors may have influenced the Beothuks' refusal to trade and interact with whites (after the initial period). Pastore (1987:48) has noted that the Beothuks did not need trade to acquire European goods. However, the Beothuks were acquiring these goods primarily for expedient use, e.g., for manufacturing iron counterparts of traditional hunting weapons, and not for barter to other tribes as a means of enhancing prestige or gaining wealth. Pilfering from white premises satisfied these immediate needs.

It may be argued from this standpoint, therefore, that if there was only sporadic interaction between the Island of Newfoundland and the Quebec-Labrador mainland during the late prehistoric period (prior to A.D. 1500), then the extinction of historic Newfoundland Beothuks was an understandable outcome. Lacking well-established trade partners and allies, and having no motivation for forming an alliance with whites on the Island, extinction was almost inevitable for Beothuks in the face of an expanding and unsympathetic white population. The only possibility for survival for Beothuks lay in a concerted and outspoken effort on behalf of whites to establish amicable and beneficial relations. What was done in this regard was too little, too late.

The analysis currently in progress should assist in clarifying questions of prehistoric relationships between the Island of Newfoundland and the Quebec-Labrador mainland during the Recent Indian period. I believe that the nature and status of these relations will prove illuminating with respect to the course of Beothuk history, subsequent to the arrival of Europeans in Newfoundland.

i. Archaeological Sites and Assemblages

Any regional examination of Recent Indian cultures in Newfoundland and Labrador is impeded by several characteristics of the sites and assemblages themselves. Recent Indian peoples occupied many coastal locations that Palaeo-Eskimo, and especially Dorset, populations had previously settled. Thus, many prehistoric habitation sites exhibit dual, and in some instances multiple, components. In many instances poor stratigraphic buildup caused by the very slow rate of soil formation has resulted in a mixture of culturally and temporally distinct artifactual remains. In some sites located on sandy ground, deflation has led to similar confusion. With respect to formal, diagnostic tools this does not cause undue problems as it is possible to distinguish readily most Recent Indian artifacts from Palaeo-Eskimo ones on the bases of form, manufacture technique and raw material. A degree of uncertainty persists, however, with a few classes of tools, most notably the end scraper. It is exceedingly difficult to distinguish between Recent Indian and Palaeo-Eskimo end scrapers in a mixed collection, because, it seems, of a convergence of forms. With regard to non-formal tools and manufacturing debitage the problem is intensified. In most mixed collections these artifacts can be assigned a cultural affiliation only on the basis of raw material. As little is known regarding the distribution of different raw materials, or their sources, this method remains highly suspect and amounts to little more than guesswork. A large amount of potential data that could be derived from non-formal tools and flaking debitage is lost (but, in the future, possibly may be regained) in these mixed assemblage sites.

One way in which this problem can be circumvented is by concentrating efforts on assemblages from single component sites, or sites where components are horizontally or stratigraphically separated. This limits the data base, but seems to be advisable, rather than risk deriving erroneous conclusions from contaminated collections. Formal and diagnostic Recent Indian tools from mixed sites might, as a later step, be appended to the analysis.

Sites which present "clean" Recent Indian assemblages are Winter Cove 4 (central Labrador coast), Saddle Island West (Strait of Belle Isle), Port au Port (western Newfoundland), Boyd's Cove (northeastern Newfoundland), and Frenchmen's Island (southeastern Newfoundland).

Winter Cove 4

The Winter Cove 4 site is the largest and best researched site of the Point Revenge complex on the central Labrador coast. It is situated near the entrance to Groswater Bay, on the bay's north shore. The site was first tested in 1969, and excavations were later performed by members of a Smithsonian Institution crew under the direction of Gilles Samson. Winter Cove 4 is unique among Recent Indian sites of Labrador and Newfoundland in possessing distinct structural remains, in the form of a four by eight metre tent ring composed of large rocks and inclosing a central, double-sided slab hearth (Fitzhugh 1978:154). Circular cobblestone hearths were discovered outside the tent ring structure, and contained fire-cracked rocks, charcoal, calcined bone, and stone tools and flakes. The lithic collection from the site consists of a number of formal tool types, including notched projectile points, small bifaces, flake scrapers/knives, and end scrapers, together with retouched flakes and flaking debitage. Charcoal from two of these features have given radiocarbon dates of 465 +/- 45 BP (AD 1485) and 435 +/- 45 BP (AD 1515) (ibid. p. 154).

Saddle Island West

This site is located on Saddle Island, in the mouth of Red Bay harbour on the Labrador coast of the Strait of Belle Isle. Extensive archaeological research as part of the Basque Project -- which also includes underwater excavations and documentary research -- has been ongoing each year since 1977. The search for and excavation of tryworks, cooperages and living sites pertaining to 16th century Basque whaling has brought to light several site of aboriginal occupation on Saddle Island and in its immediate environs. These include sites of Groswater and Dorset Palaeo-Eskimo, Labrador Eskimo, and Recent Indian habitation.A site of the last-named population was discovered and partially excavated during 1986 and 1987 by a Memorial University crew under the direction of the author. It consists of a series of small hearths composed of a few fire-cracked rocks, and scattered charcoal and calcined bone deposits. Associated with these hearths were projectile points, small bifaces, scrapers and knives, together with manufacturing debitage. Raw materials include Ramah chert, the material from which almost all Point Revenge artifacts in central Labrador are made, and a variety of coloured cherts. Non-lithic native artifacts discovered at Saddle Island West include a small fragment of an incised bone object -- possibly a pendant -- and several sherds of two pottery vessels. Fragments of Basque roofing tiles and iron nails incorporated in the hearth deposits may indicate that Recent Indians occupied this site during or shortly after the period of Basque activity on Saddle Island.

It seems most likely that Saddle Island West was occupied by Recent Indians during either the early historic period of the 16th century -- at a time when European goods were peripheral to their technology -- or during the late prehistoric period. Occupation may span the historic and prehistoric periods. The presence of Ramah chert suggests contacts with Point Revenge peoples to the north. Other cherts may indicate contact with groups to the south, in Newfoundland, and to the west, along the Quebec lower north shore.

Port au Port

The Port au Port site is located immediately north of the narrow isthmus which joins the Port au Port Peninsula with mainland Newfoundland. It was first explored by Paul Carignan (1975a). More recently work has been done by David N. Simpson, a graduate student at Memorial University of Newfoundland. Surveys and excavations in the region of the Port au Port formed the basis of Simpson's Master of Arts thesis (Simpson 1986).

The site at Port au Port was occupied by Palaeo-Eskimo people prior to Recent Indians. Despite this dual occupation, however, little or no mixing of assemblages occurs: they are, instead, horizontally separate. The collection of formal tools includes projectile points, small triangular bifaces, larger ovoid bifaces, scrapers, and linear flakes. These, together with flaking debitage, were associated with two concentrations of charcoal, bonemash and fire-cracked rocks. One of these charcoal deposits has been radiocarbon dated at 790 +/- 90 BP (ca. AD 1160) (Simpson 1984:129).

Two outcrops of chert were discovered in the immediate vicinity of the Port au Port site. This chert is, it seems, the material used by Recent Indians of the region. Simpson performed a comparative analysis of the chert from these outcrops and that from the site, the results of which appear in his thesis. It is possible that these chert sources supplied material for Recent Indians in western and southern Newfoundland.

Boyd's Cove

Excavations were carried out at Boyd's Cove each year from 1982 to 1986 by Ralph T. Pastore of Memorial University (Pastore 1983, 1984, 1985, 1986). The primary focus of Pastore's work at Boyd's Cove has been upon historic Beothuk, as illustrated by the numerous house pits and the material culture contained within them. During the course of excavations at Boyd's Cove, however, much prehistoric material comparable to the Little Passage complex has been recovered. The largest concentration of prehistoric material comes from an area outside the house pit features.The artifact inventory from this area closely parallels that from other Little Passage assemblages. Artifacts and flaking debris were associated with scatterings of fire-cracked rocks, suggesting hearths, yet little charcoal was recovered. No absolute date is available; it predates the historic occupation at Boyd's Cove, which Pastore suggests may have begun around AD 1670, and therefore likely falls within the period AD 1100-1600.

Frenchmen's Island

Excavations at Frenchmen's Island were performed during 1980 and 1981 under the direction of Clifford Evans of Memorial University. Frenchmen's Island is one of two Recent Indian sites located in Bull Arm, a small inlet in Trinity Bay. The other site, Stock Cove, was excavated by the author in 1981 (Robbins 1982, 1985, n.d.). Both sites have produced evidence of Dorset occupation in addition to Recent Indian components. The Stock Cove site is considerably larger than Frenchmen's Island, and proportionally richer in cultural remains. Serious problems arise from mixing of Recent Indian and Dorset deposits, however. Stratum 1, from which Recent Indian artifacts were collected, also contains Dorset and European material, making it very difficult to distinguish with certainty between non-diagnostic Dorset and Recent Indian artifacts, and flaking debitage, from this level. Hence the Stock Cove collection is not suited to the analysis proposed here.

While some mixing also occurs at Frenchmen's Island, Stratum 2 of the "eastern area" and Stratum 1 of the "western area" of the site (Evans 1982:213-14) appear to present "clean" Recent Indian components, not mixed with Dorset material. These assemblages include projectile points, small triangular bifaces, scrapers, linear flakes and flaking debitage. Distinctions with regard to raw material type and projectile point forms may be observed between these two assemblages, with the eastern area occupation possibly predating that of the western area. Both assemblages can be attributed typologically to the Little Passage complex.

Evans reports two radiocarbon dates from the site, of 1870 +/- 180 BP and 1320 +/- 100 BP. Both are far out of line with all other dates pertaining to the Little Passage complex in Newfoundland, and probably pertain to the Dorset Eskimo presence at Frenchmen's Island.The eastern and western area Recent Indian assemblages are believed to date between AD 1100 and 1600.

ii. Analysis of Assemblages

a. Objectives

The fundamental objective of this analysis is to determine the degree of similarity/difference between archaelogical assemblages from insular Newfoundland and Labrador, and to recognize patterns (if patterns exist) defined by these similarities and differences. "Pattern" in this context is synonomous with Albert Spaulding's concept of "structure" (1982:1-2). This in itself raises a point of some debate, regarding what it is that is being examined (or discovered). Are differences, if differences exist, representative of (consciously established) ethnic identities, or did they arise unconsciously, independent of perceived group identification? Alternatively, if no significant differences can be seen to exist, does this speak for the absence of ethnic divisions in the region?

For purposes here, I do not associated results of assemblage analyses with ethnicity, and consequently respond to all the preceding queries with a provisional "no". The purpose of this analysis is not to identify ethnicity. The purpose is simply to identify the pattern or patterns of similarity/difference among artifact assemblages dispersed throughout a fairly broad geographical area.

Briefly, at least three outcomes are possible: 1). Throughout the region, from central Labrador to southern Newfoundland, assemblages are similar, with differences being random, and largely attributable to individual craftsmanship. 2). Throughout the region, any one assemblage resembles most closely those assemblages geographically closest to it. Differences are patterned, and increase with distance. 3). Newfoundland and Labrador assemblages form two or more distinct clusters, which contrast.

b. Typology

Methodological Procedure:

Each of the assemblages included in the set to be examined is numerically small, presenting difficulties regarding the statistical determination of "types", and the usefulness such types might have for purposes of cross-comparisons. To take a hypothetical (but not unrealistic) example: one assemblage presents a collection of 24 notched projectile points, among which three "good" types are recognized, which include 18 of the 24 projectile points. The remaining six are deemed "transitional" or "deviant". In a second assemblage (of comparable size), two "good" types are recognized, with leftover transitional and deviant specimens. Neither of these two types replicates exactly one of the three types from the first assemblage. There are now, then, five "good" types, none of which is meaningful for cross-comparisons, as types are not directly comparable. When all assemblages have been examined in such a manner the number of "types" will have grown proportionally, as will the number of un-typed specimens ("deviant" and transitional" ones). As the "types" produced in this manner are not necessarily suited to cross-comparisons (comparisons between assemblages), little or no progress will be made towards the analytical goal -- the observation of similarity/difference and the determination of relatedness throughout a geographic area. As well, an increasing amount of potential data is lost in the discarded "transitional" and "deviant" categories: it is important to note that a specimen is considered "deviant" or "transitional" only in relation to other specimens in its assemblage.

As a means of circumventing these problems an alternative methodology is proposed, which is summarized here in two stages.

Stage 1:

Rather than begin with individual assemblages, all assemblages are lumped and analysed as if they constituted one. The resulting "types" (of projectiles points, in this example) will likely be numerically fewer (a desirable but not necessary outcome) and statistically better established. Also, the number of "deviant" or "transitional" specimens will be fewer, as each projectile point is considered in relation to all specimens in all assemblages. Thus the data potential is maximized. These "types" will be directly applicable to cross-comparisons ("Stage 2" of the analysis) as, logically, all or most projectile points in any particular assemblage will have been "typed", and each type will exist, or not exist, in any particular assemblage.

Stage 2:

What follows from this is a consideration of each assemblage in relation to the "baseline" typology established in the foregoing analysis. If, for example, 13 types (A to M) are established overall, Assemblage 1 might illustrate A,B, and C; Assemblage 2 B,C, and D; Assemblage 3 J,K and L; etc. In this manner each assemblage is described in relation to the overall typology. As well, any one assemblage may be compared directly to each/every other assemblage.

This procedure is very amenable to statiscal determination of "distance" between assemblages. Very simply put (as I have not explored sufficiently this avenue yet) the number of types held in common will produce a measure of similarity/difference. If Assemblages 1 and 2 hold two types in common (B and C), then it can be argued that there exists a relationship between them which is not shared with Assemblage 3 (which presents 3 alternate types, J,K and L, none of which is present in Assemblages 1 and 2).

Lithic materials. A first stage in the analysis will involve an examination of lithic materials. In this respect it is desirable to be able to distinguish materials native to the Island of Newfoundland from those of the Quebec-Labrador coast. The presence of Island cherts in mainland assemblages, and vice versa, would in itself speak for contacts across the Strait of Belle Isle. A variety of cherts occur in Newfoundland Recent Indian assemblages; however, few types have been traced to source locations. Definite sources for some cherts have been identified at Port au Port (Simpson 1986) and Cow Head (Tuck) on the Island's west coast. These and other Ordovician cherts native to Newfoundland may be identified by the presence of radiolari microfossils, easily visible with a hand lens. These microfossils are not present in cherts native to the Quebec-Labrador, east of Quebec City; they do occur on the Gaspe Peninsula (Robert Stevens, Department of Geology, MUN, pers. comm.). Therefore, the presence of these microfossils may serve to distinguish Island cherts from mainland ones in archaeological assemblages. Cherts from each of the sites mentioned above will be examined for radiolari.

Other cherts types may be readily identified by visual inspection, especially Ramah and, perhaps somewhat less so, Cow Head chert. The only known source locale for Ramah chert is in Ramah Bay, north of Saglek Bay on the Labrador coast. It is by far the dominant raw material used by Recent Indian peoples in central and northern Labrador, and occurs frequently at the Saddle Island West Recent Indian site in southern Labrador. Its precise frequency in Newfoundland assemblages is not yet known; preliminary examinations suggest it is exceedingly rare. Other lithic materials, e.g., Mugford chert, have been located and identified along the Labrador coast (Gramly 1978), and counterparts tentatively identified in the Saddle Island West assemblage (Gramly pers. comm., August, 1988). Examples of these materials are available for comparative purposes (ibid.).

In sum, an examination of lithic materials can provide useful information regarding interrelationships among Recent Indians of the area. Radiolari present a quick and easy method of distinguishing Island cherts from those of the mainland. This is of considerable importance as the sources of many cherts found in Recent Indian assemblages are not known. A few cherts have known source locations, both on the Island and the Labrador mainland, allowing a more precise examination of distribution patterns.

It is suggested that the lithic material of each assemblage be typed in a manner which utilizes existing knowledge of cherts and source locales. This will firstly involved the separating out of known cherts (e.g., Ramah, Cow Head). Following this, cherts with radiolari may be separated from those without. Each of these latter groups may then be typed by criteria such as colour and texture, which should provide a sufficiently fine typology for purposes of inter-assemblage comparisons.

Attributes of formal artifacts. A provisional list of attributes pertaining to projectile points includes nominal, ordinal and interval scale attributes. The list is presented in Table 1.

A brief dicussion of these attributes is provided here for purposes of clarity. Integrity refers to completeness or wholeness of the specimen, with possible values being "complete", "tip", "midsection", or "base". Integrity in itself will not assist in the establishment of patterns in the context discussed above. Rather, this attribute is included in light of studies (Keeley 1982) which have examined the effects of disposal of hafted tools on interassemblage variability.

Table 1: tentative list of projectile point attributes to be examined
(Scale is provided in parenthesis)

• Integrity (Nominal)
• Material (Nominal)
• Source (Nominal)
• Radiolari (Nominal)
• Notch form (Ordinal)
• Stem form (Ordinal)
• Edge A shape (Ordinal)
• Edge B shape (Ordinal)
• Blade length (Interval)
• Point-to-tang length (Interval)
• Tang width (Interval)
• Maximum stem width (Interval)
• Minimum stem width (Interval)
• Maximum thickness (Interval)

Notes:

• Material refers to lithic raw material, and source the lithic source. As discussed above, specific sources remain generally unknown in this study area.
• Radiolari is recorded as present / absent.
• Notch form describes the shape of each notch, as "narrow", "broad", and "deep" and "shallow" (with conceivably a third, intermediate value in each of these paired values; also, it may be advisable to split this attribute into two: "notch breadth" and "notch depth" for ease of manipulation).
• Stem form is recorded as "straight", "expanding", or "contracting".
• Edge A, B shape refers to the lateral blade edges, and is recorded as "straight", "convex" or "concave".

Metric attributes are selected to enable the quantification of projectile point elements which may vary through space (and time). Figure 1 presents in stylized form the precise locations where the values for these attributes are to be obtained. The following example demonstrates the manner in which the metrical analysis of attributes may procede.

Previous researchers have recognized that projectile point elements such as shoulder form vary through time, and perhaps also through space. However, confusion exists with respect to methods of recording and comparing an element such as shoulder form. The metrical attributes suggested above will provide a means of quantifying this observation, which previously was only recorded qualitatively.

A simple index may be created:

Shoulder Index = point-to-tang length + stem length / maximum length

Points with "straight" shoulders produce an S.I. equal to 1; for points with shoulders which "droop", S.I. > 1; points with "rising" shoulders have a S.I. < 1. Basic statistical procedures can then be employed in a search for order in the data base (the collection of projectile points) with respect to shoulder form. This analysis can then be placed in the context of the inter-assemblage comparison.

c. Implementation

As noted in a previous section, excavations at the archaeological sites under consideration have been performed in previous years, by a number of researchers. No further fieldwork for this Ph.D. research is planned. Data for the study will be obtained from these existing ar­chaeological collections. This will entail an analysis (and in some instances a re-analysis) of each collection. As more than a decade elapsed from the earliest to the latest of these excavations, and each was performed with different objectives in mind, by individuals employing different methodologies (field and analytical), the data from these assemblages are not comparable in their present form. Therefore, the first step is to devise a standardized analytical methodology which may be applied to each assemblage.

Once this is done each assemblage may be examined in detail. All collections, with the exception of the one from Red Bay, southern Labrador, are presently stored in the Newfoundland Museum in St John's, Newfoundland. Arrangements have been made with the Museum's Curator of Archaeology to gain access to these collections. Unfortunately, the "political" situation is such that these collections cannot at this time be removed from the Province of Newfoundland, so I will have to perform basic data-collecting in St John's. Subsequent analysis can and will be performed at McGill.