The history of Dorset Eskimo research in the eastern Canadian Arctic has been chronicled by several scholars. In 1959 W.E. Taylor presented a summary of contributions to that time. More recently Moreau S. Maxwell reviewed the emergence and development of Dorset Eskimo archaeology and the current con­cerns in the eastern Arctic, in the introduction to "Eastern Arctic Prehistory: Palaeo-Eskimo Problems" (1976). Albert Dekin (1978) dealt not only with Dorset Eskimo research, but with the more general field of Arctic archaeology in a volume combining a narrative of participants, events, and theories with an extensive bibliography. The following review is intended to impart something of the growth and expansion of Dorset Eskimo archaeology, and does not attempt to detail regional work. It relies on the above volumes as well as the original publications.

It might be said that Dorset Eskimo archaeology began indirectly in 1925, when Therkel Mathiassen published an initial description of Thule culture. Later in the same year Diamond Jenness was able to separate "Cape Dorset" artifacts from Thule ones in a mixed collection, and thus identify the Cape Dorset culture (Jenness 1925). Diagnostic of this newly-recognized culture were stone implements and bone harpoon heads of hitherto unknown types. Although Jenness did not hesitate in postulating the existance of an entirely new population to account for these strange implements, Mathiassen was by no means as certain, and in the 1927 "Report of the Fifth Thule Expedition" he argued that Cape Dorset culture was in fact a local expression of Thule culture.

After these initial statements of the 1920's, Cape Dorset culture became an increasingly popular subject for investigation and debate. Mathiassen in 1930 and Jenness in 1933 both reaffirmed the essence of their differing opinions. W.D. Strong presented the results of two field seasons research in Labrador in a 1930 publication, and suggested that stone age sites there represented a "common substratum" from which Eskimo and Indian cultures arose. In 1935 Henry B. Collins accepted the Jenness stance on Dorset culture, i.e., that it was Eskimo, and an entity separate from and older than Thule. In 1936, nine years after his original comments on Cape Dorset culture, Mathiassen modified his position somewhat and agreed that Cape Dorset culture was distinct from Thule. At this point he stated that Cape Dorset was in fact an Indian culture, and implied that it was contemporaneous with the Thule culture. This suggestion brought a response from Collins in 1940; he disputed this Indian affiliation, and suggested that Cape Dorset culture held relations with the Alaskan Bering Sea cultures, the subject of his research in the 1930's.

The decade of the 1940's saw an increase in Dorset-related publications, and the appearance of new investigators and new ideas. By the decade's end considerable new infor­mation in the form of sites and assemblages had been amassed: in 1939 Wintemberg published a two-part report detailing Dorset sites discovered by Jenness and himself in Newfoundland during the years 1927 and 1929; Graham Rowley reported in 1940 on the Dorset occupation of the Abverdjar site in Foxe Basin; Eric Holtved (1944) investigated Thule and Dorset sites in west Greenland; Junius Bird (1945) excavated in the Hopedale region of Labrador; Leechman (1943) discovered two Dorset sites in Hudson Strait, and was the first to describe Dorset structures. These reports supplied supportive data for the separation of Dorset and Thule cultures, and for the greater age of the former. As well, the Alaskan affinities and the possibilities of a Bering Strait origin for Dorset culture received further attention from Birket-Smith (1948), Larson and Rainey (1948), and Giddings (1949, 1951).

Ideas which were born during the 1930's and 1940's were nurtured during the 1950's, and concurrently there developed a wealth of terms dealing with this increasingly specific knowledge. Some of these terms have become entrenched in the lexicon of present-day researchers, while others have slipped into disuse. With regards to culture origins, the "Denbigh Flint Complex" (Giddings 1951) became viewed as the ancestral expression of Eskimo culture. In northeastern Greenland Knuth (1954) identified the "Independence" occupation, which he later refined into two stages, Independence I and II -- the latter being the more recent (Knuth 1958). The Palaeo-Eskimo "Sarqaq" was recognized in west Greenland by Meldgaard (1952) and further described by Larsen and Meldgaard (1958).

The Independence and Sarqaq cultures demonstrated relationships with the Dorset culture as well as with the Denbigh Flint Complex of Alaska, and were thus held to be temporally intermediate between these two occupations. In Greenland there appeared to be a distinct lack of continuity between these occupations (Independence and Sarqaq), and investigators there assumed a number of migrations in accounting for their presence. Since their original defini­tion in Greenland, Independence I and II sites have been recognized in the Canadian Arctic (McGhee 1976), and are sufficiently different from Dorset culture to warrant the retention of the distinct name. American and Canadian archaeologists have placed Sarqaq in a Pre-Dorset stage of Dorset culture, thus implying a cultural continuity. Collins (1954) first employed the term "pre-Dorset" in describing the phase of Palaeo-Eskimo occupation in the eastern Arctic which appeared to be directly ancestral to classic Dorset. The Scandinavian archaeologists -- Meldgaard, Larsen and Mathiassen -- saw Sarqaq as a distinct entity, separate from Dorset culture and, on the basis of his work at Igloolik, Meldgaard (1960) disagreed with the idea of a Pre-Dorset-Dorset continuum. The North American viewpoint, which has become generally accepted, reconstructs an in situ development of Dorset culture from a Pre-Dorset base, which had developed from the older cultures of Alaska and the Bering Sea (Collins 1956, Taylor 1959).

The 1960's witnessed an expansion of fieldwork, espec­ially in the "core area" of Foxe Basin and neighbouring regions. Sites producing long series of occupations lent additional support to the theory of in situ development. The "core area concept", which held that the area of northern Hudson Bay-Foxe Basin supported a continuum of Pre-Dorset - Dorset occupation, and in turn supplied fluctuating popu­lations to surrounding fringe areas, gained acceptability. Ideas of "causative climatic change", which acted to control population expansion, were advanced.

Dorset archaeology of the 1970's has seen the expansion of previous ideas, and the rethinking of some old concepts. McGhee (1976) presented new information concerning the Independence and Pre-Dorset occupations of the central and high Arctic. In these areas Independence I and Pre-Dorset appear to have been sequential occupations by distinct populations, contrary to previous ideas. McGhee (1976:38-39) also presented convincing arguments against the concept of causative climatic change. McGhee and Tuck (1976) reviewed the use of radiocarbon dates, which had formed the basis of culture chronologies. A suggestion was made -- namely the elimination of all sea-mammal derived dates -- which reduced many of the previous inconsistancies in C-14-based chronologies.

The surveys conducted by Jenness and Wintemberg along the east and west coasts of the Great Northern Peninsula in 1927 and 1929 marked the beginning of Dorset Eskimo research in Newfoundland. Several sites of the Dorset culture were located, the largest and most prolific being the Phillip's Garden site at Port au Choix. In 1949 and 1950 Elmer Harp Jr. conducted excavations at Port au Choix. This research formed the basis of his Ph.D. thesis, "The Cultural Affinities of the Newfoundland Dorset Eskimo", later published by the National Museum of Canada (1964). Harp performed a comparison of Newfoundland Dorset with Dorset culture as it was then known (ca. 1950) from Hudson Bay, northern Labrador, Baffin Island, and Greenland. While he was able to demonstrate a high degree of correlation, a few discrepancies were noted. Certain types of chipped stone artifacts present in the Newfoundland assemblage were not recognized in the "parent complex", and many of the bone artifact types in the Arctic were not present in Newfoundland.

Since its completion Harp's work has been utilized by many writers and investigators, and the concept of a "New­foundland Dorset" or a "typical Newfoundland Dorset" -- a regional development of Arctic Dorset culture characterized by unique stylistic traits and tool forms -- is often attributed to him. This idea is not expressed as such in Harp's original work, although it is perhaps implied. Harp emphasized the "affinity", the similarity of character: "Thus I believe it has been demonstrated that there is a strong and widespread degree of cultural concurrence between the Newfoundland Dorset aspect and the parent complex." (Harp 1964:138).

In 1967 and 1968 Urve Linnamae excavated at the Pittman site in White Bay and the Cape Ray Light site at the south­western corner of the island. The concept of a unique "Newfoundland Dorset" is best expressed in her final report on these sites: "... it seems reasonable to postulate that the geographic isolation of Newfoundland, in conjunction with an adverse climatic period, was responsible for the development of the regional flavour of Newfoundland Dorset." (Linnamae 1975:93)

Linnamae's Newfoundland-Arctic comparison pinpointed several traits present in the Newfoundland assemblage but absent from Arctic assemblages. These traits include certain forms of side-notched end blades, "sickle-shaped" gravers, the presence of surface grinding on end blades, and variant forms of microblade cores, ground slate implements, and soapstone vessels. Such traits as these were used by Linnamae to distinguish the "Newfoundland Dorset".

In retrospect, the year 1972 marked the beginning of a significant change in Dorset archaeology in Newfoundland: in that year Paul Carignan began excavations at the Beaches site; the Norris Point site, first published by Wintemberg, was revisited by James A. Tuck and Paul Bishop; and William Fitzhugh's publication "Environmental Archaeology and Cultural Systems in Hamilton Inlet, Labrador" appeared. From the multi-component Beaches site Carignan was able to recognize at least two artifacts which he felt demonstrated "... an affinity to the early Groswater Dorset phase", as identified by Fitzhugh in Labrador (Carignan 1975:132). In his final report on the Norris Point site, Bishop dwelt on the simi­larity between the collection from that site and Fitzhugh's Groswater Dorset assemblages. At this time (the early 1970's) the realization that there were early Dorset traits -- as opposed to "typical" Dorset traits -- present in some New­foundland assemblages emerged. The status of Dorset culture research in Newfoundland at that time was expressed by Paul Bishop in his Norris Point report: "The nature and dating of the evolution or change -- from this early Dorset stage to a stage more similar to what is considered to be 'Typical Newfoundland Dorset' remains purely conjectural... Whether this change from an early form, as exhibited at Norris Point, was the result of centuries of contact and exchange with southern Labrador, or whether the later forms were established by a single new wave of migrating Dorset people from the north is a question only further research can hope to answer." (Bishop n.d.:23).

In the late 1970's a program of research was conducted by James A. Tuck of Memorial University at the Cow Head site in western Newfoundland. Excavations revealed a lengthy series of Indian and Palaeo-Eskimo occupations. The oldest Palaeo-Eskimo remains differed from any previously discovered on the island but did resemble some material from Labrador, and appeared to "... represent a step in the transition from Pre-Dorset to Dorset culture which took place about 3000 years ago" (Tuck 1978:139). Tuck stated that a con­tinuity between this material and the later Groswater or early Dorset was evident, yet the relationship with the "typical Newfoundland Dorset" still remained unclear.

These developments in the past decade revealed a deficiency, or a source of confusion, in the terminology being employed. The terms "Groswater Dorset" and "early Dorset" were used interchangably, but sometimes with different implications. Fitzhugh (1980) reorganized the terminology into two traditions, an "Early Palaeo-Eskimo" and a "Late Palaeo-Eskimo". The Early tradition included Pre-Dorset, 'transitional' Pre-Dorset-Dorset phases and, as its terminal expression, the Groswater phase. The Late Palaeo-Eskimo tradition incorporated the Dorset culture. "Early Dorset" and "Late Dorset" were used to refer to temporally distinct expressions of classic "Middle Dorset". Fitzhugh devised this scheme with respect to the Palaeo-Eskimo occupations of the Labrador coast. In reference to the island of Newfoundland he stated that: "... the luxuriant period of Dorset occupation on the island of Newfoundland is probably best seen as an amal­gamation of resident Early Palaeo-Eskimo traditions with new ideas introduced from the north about 2000 B.P." (Fitzhugh 1980:26).

The evidence which Fitzhugh offered for the "continuance of these early traditions" includes "tool types like the side-notched asymmetric or crescent type biface knives, high side-notched points, and perhaps the artwork" (1980:26).

Tuck reiterated this Early and Late Palaeo-Eskimo scheme in a paper presented to the 15th annual meeting of the Canadian Archaeological Association in 1982, and added: "... there clearly exists a 2-300 year gap, in both New­foundland and along the central Labrador coast, between the most recent "Early" or Groswater material of the early Palaeo-Eskimo tradition and the earliest Dorset cultures of the late Palaeo-Eskimo tradition" (Tuck 1982: 214).

It is difficult to reconcile this observation with Fitzhugh's expressed "amalgamation" of Early and Late Palaeo-Eskimo traits on the island. As very little evidence exists which might indicate the presence of Early or Late Palaeo-Eskimo populations on the island ca. 1900-2100 B.P., Fitzhugh's concept might be viewed with some reservation.

From the preceding review several points concerning New­foundland Dorset culture research become apparent. The first is that, with the exception of Paul Carignan's work, most research (prior to the 1980's) has been concentrated in western Newfoundland. Secondly, theses written by Urve Linnamae and Elmer Harp (based on western research) have openly stated or implied that Dorset culture in Newfoundland exhibits a distinctive character due, in part, to the insular nature of the region. This character is inherent in such terms as "Newfoundland Dorset" or "typical Newfound­land Dorset". As used here the term "Newfoundland Dorset" -- in quotations -- refers to this concept. Without quotations, the term simply refers to Dorset culture on the Island of Newfoundland. After six decades, then, we are left with a Newfoundland representation of Dorset culture which, despite the best efforts of several eminent researchers, is distinguished from Arctic manifestations or the "parent complex" by a very few specific artifact types, represented by very few individual specimens. The concept of a distinct "Newfoundland Dorset" culture has grown from the work of Wintemberg, Harp, and Linnamae in order to account for a number of artifact forms not previously seen in Arctic assemblages, and which were presumably common amongst the Newfoundland inhabitants. The presence of these artifacts lead to the postulation of a unique Newfoundland Dorset Eskimo culture.

Two concepts inherent in the idea of "Newfoundland Dorset" culture are uniqueness and homogeneity; i.e., that Newfoundland Dorset is different from Arctic Dorset, and that it is the same throughout the Island. With respect to the former, this author has little to add; the difficulties previous researchers have had in attempting to elucidate the distinct character of "Newfoundland Dorset" are obvious. As regards to the latter, the commonality of "Newfoundland Dorset" has been implied, largely accepted, but never proven -- and it is to this second assumption that the present work is addressed.

The existence of the Stock Cove site was first brought to my attention in the fall of 1980. As will be discussed below, the site had been previously explored, and found to be rich in Palaeo-Eskimo and Recent Indian remains. The site appeared to be an excellent one for a comparative Newfoundland Dorset Eskimo study, as it is located in a region of Newfoundland where little archaeological work has been done, and it would provide an excellent contrast with western Newfoundland's "Newfoundland Dorset" culture. Fieldwork at Stock Cove was, therefore, oriented towards the gathering of information pertaining to Dorset Eskimo culture that would allow a testing of the concept of "Newfoundland Dorset". This included aspects of material culture such as raw material utilization, functional and stylistic artifact types, frequencies of various artifact types, and structural remains.

Beyond the comparison of material culture, other objectives existed. Preliminary testing of the site had disclosed some distinctive traits, such as the frequent occurrence of grinding on chert end blades. It was suspected, on the basis of these traits, that Stock Cove might represent a very recent period of Dorset occupation in Newfoundland. A second objective was, therefore, to radiocarbon-date the site, for comparison with the chronology currently in existence for the Dorset occupation of Newfoundland. As Stock Cove is located in a region not usually reached by the migratory harp seal herds -- the mainstay of prehistoric hunters in northern and western Newfoundland -- the basis for the occupants' subsistence was not immediately obvious. A third objective therefore was to clarify this issue, by collecting identifiable faunal remains indicative of resource exploitation.

While good progress has been made towards the reali­zation of the first two objectives, the last still remains largely unresolved. Excavations resulted in the recovery of an extensive artifact collection, and the discovery of some interesting, though perplexing, structural features. Attempts at radiocarbon dating were also satisfactory, as three carbon samples from Dorset Eskimo features produced dates within the expected period. A reconstruction of subsistence on the basis of recovered faunal remains cannot be attempted at this time, however, and this stands as the single major failure of the Stock Cove project. Soil conditions at the site appear to inhibit organic preserva­tion. As excavations at Stock Cove were restricted to a single, small area, there exists the possibility that larger-scale work might locate areas where better preservation exists.

Conclusions drawn from the Stock Cove research have brought into question the concept of a homogeneous, Island-wide "Newfoundland Dorset" culture. A contrast between Stock Cove Dorset and western Newfoundland Dorset is apparent, and, with this recognition, it is becoming increasingly obvious that there was not a single and unique expression of the Dorset Eskimo culture on the Island. In the final chapter of this work some of the differences in the character of Dorset culture from various regions of Newfoundland are discussed and related in a hypothetical interpretation.

Situated on the eastern shore of the Isthmus of Avalon at a point where the Isthmus is about seven kilometres wide, Stock Cove is now uninhabited; it has, however, been a scene of intermittant human activity for perhaps 3,000 or 3,500 years. As this work is concerned exclusively with the Dorset Eskimo culture, the other prehistoric and historic inhabitants of this small cove are given only cursory mention. A fuller treatment will be presented in future papers. Archaeological evidence indicates that at least two prehistoric groups other than Dorset people resided there; ground stone celts and the stem of a slate "bayonet" indentify a Maritime Archaic Indian occupation, and many artifacts indicate the site was utilized by other Indians after the Dorset people had departed. Evidence of an eighteenth or nineteenth century historic occupation -- nails, glass, and clay pipe fragments -- is also present. The remnants of small houses and wharves, together with oral accounts of present-day occupants of neighbouring communities, give evidence for a not-long-past occupation by Newfoundland fishermen. These most recent occupants gave the cove its present-day name. "Stock Cove" is derived from the English word "stockfish", which refers to split, unsalted, sun dried codfish.

A small, steep and wooded headland divides Stock Cove into two parts, locally known as "Big" and "Little" Stock Cove. The past occupations of these two coves differ dramatically, and reflect the different requirements of aboriginal and white settlers. Little Stock Cove is fiord-like, bounded on both sides by steep and forested hills which rise to more than 90 m above sea level. The shelter which it offers from winds and seas encouraged temporary occupation by fishermen from the neighbouring communities of Sunnyside and Chance Cove in the past, especially in days before widespread use of the outboard motor speeded marine travel. Small houses or "shacks" were built, and spindly wharves erected against the rocky shore for this summertime fishery. This practice has now ceased, as technological advances permit fishermen to make the return trip between their permanent settlements and the fishing grounds in a day. Little Stock Cove shows no evidence of an earlier historic or a prehistoric occupation.

Little Stock Cove was attractive to Newfoundland fishermen in the past because of the shelter it offered, and for the opposite reason Big Stock Cove was unsuitable for their purposes. It is much more exposed to winds, and during a storm seas roll high up the sloping cobble beach, shifting and reshaping it. The prehistoric people of Stock Cove, unconcerned with safe anchorages for heavy plank boats, must have viewed Big Stock Cove very differ­ently. From its beach the inner reaches of Trinity Bay are clearly observable and accessible. The beach itself offers marine resources not available along a rugged and rocky shoreline.

The cobble beach extends along most of the shore of Big Stock Cove, interrupted at points by bedrock outcrops. The ground adjacent to the beach is level, until hills are encountered about 50 m inland. These hills rise steeply to a height of about 60 m and eventually reach an elevation of more than 100 m above sea level. A small stream discharges onto the beach in the middle of the cove. The shores of Big Stock Cove are, like Little Stock Cove, densely wooded; the prehistoric site, located on part of the level ground immediately adjacent to the beach, is covered with a growth of fir, birch, dogberry, and a variety of shrubs. A one or two metre high bank along the beachfront is presently undergoing erosion by the sea, and prehistoric cultural material, including lithic artifacts and flake debris, flagstones, and charcoal, occurs continuously for a distance of about 85 m. Inland, away from the beachfront, testing has revealed similar evidence extending, in some places, to the base of the bordering hills. The dense forest and brush cover make it difficult to ascertain the exact size and configuration of the once-occupied area, and estimates of site size range from 2000 to 4000 square metres.

Evidence of prehistoric occupation was first noted in 1976 when Gerald Penney collected Maritime Archaic and Dorset Eskimo artifacts from the eroding bank, during a survey of Trinity Bay (Penney n.d.a). In the fall of 1980 Dr. James A. Tuck and members of the Memorial University archaeology field school travelled to the site, and conducted further testing. Another occupation was recognized, referred to at the time as "Beothuk". Today this term is used in reference to the historic Indian occupants of insular New­foundland. The "Little Passage Complex" (Penney n.d.c) refers to their immediate (prehistoric) ancestors. Thus, the "Beothuk" occupation of Stock Cove is now termed Little Passage, or, more generally, Recent Indian.

As it was obviously impossible to excavate the entire Stock Cove site, some selectivity was necessary. There was, in truth, little reason to choose one area above another. It did not seem feasible to attempt to salvage the eroding bank as it was too extensive to deal with in the time and with the means available. Partial excavation of the bank, it was felt, would only hasten erosion. There was no other evidence of disturbance which might influence our selection of an area in which to work. Recent human activity had been concentrated in Little Stock Cove, and local residents were in fact unaware of the existence of the prehistoric site in Big Stock Cove. It was suspected that tree roots may have effected considerable disturbance, yet searching for an unwooded area proved fruitless. As the primary objective of the proposed research concerned the Dorset Eskimo culture, it was desireable to concentrate efforts in an area free from subsequent Recent Indian occupation. In this we were only partially successful, as some of the excavated area did produce evidence of Recent Indian activity. As an alternative to an intensive excavation of a portion of the site the possibility of conducting a random sample of the entire site existed. This approach was decided against on two counts. Firstly, the presence of structural remains encouraged exploration of at least one of these features; random sampling would not significantly advance our understanding of the Stock Cove structures. Secondly, it seemed unlikely than we could complete a representative random sample of the site, given our time and manpower limitations. A grid would have to be established over the entire site and, because of the dense forest cover, this alone would have consumed much of our short field season.

After some hours of exploring, an area in which we would concentrate our efforts was selected. Under a thin carpet of fir needles there was excellent evidence of a Dorset Eskimo occupation, including flagstones, diagnostic stone tools, and flakes of the material from which these tools were made. Additionally, no artifacts clearly of Recent Indian affiliation were located during this preliminary inspection. Initially it was our intention to complete an excavation in this area, exposing the feature, and then proceed to work in another area, slightly removed from the first. This optimistic scheme was short-lived as, in the first week, we fully realized the richness of the cultural deposits.

An area near the eroding bank of about 150 square metres was cleared of trees and brush. A large tree bordering the cleared area to the south was selected as datum, and a grid established. Work began with the clearing of four 50 cm wide trenches, forming a rectangle. The results from these preliminary trenches determined the direction of more intensive excavation.

Two natural strata were distinguished during excavation. While these levels are not clearly separated, some distinction appears obvious with respect to the cultural debris contained in each, and thus they are thought to have some temporal and occupational significance. Stratum I, the upper level, was distinquished from the lower Stratum II on the basis of soil colour and density. Generally, the upper stratum was lighter and less compacted than the lower, containing a greater amount of cobble rock and fewer flagstones. Most of the Recent Indian material was recovered from this level, along with artifacts of Dorset origin. Stratum II yielded artifactual and structural evidence of Dorset Eskimo occupation. Stratum IIa also relates to the Dorset Eskimo culture, and differs from Stratum II in containing greater amounts of decayed organics. Stratum IIa is re­stricted to a portion of the excavated area, whereas Stratum II is continuous throughout.

In all, an area of 55 square metres was excavated at the Stock Cove site. Only in isolated locations, however, was the entire Dorset stratum excavated to a sterile sublevel. The features described below were located in Stratum II, and the stratum continued below them. Thus, excavating the Dorset deposit to its bottom required the removal of flagstones, and the destruction of these features. Nine weeks (and five people) were required to accomplish the work summarized here -- including the collecting of some 1,900 artifacts. Removing the flagstones and fully excavating the exposed area required time -- and finances -- that were unavailable. It was felt that a "rush job" in the last few days would do little but add to the artifact count, and be essentially an act of destruction. Features were therefore left intact, with a thought towards future work.

Before concluding fieldwork, however, it seemed advisable to probe at least a small part of the site to a greater depth to see what (if anything) lay underneath the tremendously rich Dorset deposit. Accordingly, one square metre unit was selected, simply on the criterion that it was one of the very few places we could dig without disturbing the exposed features. Stratum II was excavated to its bottom, where it merged with an apparently sterile gravel layer (about 30 cm below surface). Continuing excavation into the gravel uncovered a thin black level some 38 cm below surface. This level -- labelled Stratum III -- produced flakes of patinated white and green cherts, along with several artifacts, including a bifacial, side-notched triangular end blade, a crystalline quartz end scraper, a miniature crystalline quartz end blade, a microblade core of brown-green chert, and a small side-notched end blade reminiscent of Early Palaeo-Eskimo culture. Scattered charcoal was also encountered amongst these artifacts in Stratum III, and collected. Radiocarbon analysis produced a date of 2140 +/- 60 B.P. (Beta-4063), near the end of the known range of the Groswater occupation of insular Newfoundland. We can therefore add yet another period of occupation to the culture history of Stock Cove.

Feature 1

Two flagstones standing on edge and protruding from the forest carpet were initially thought to mark the location of a hearth. These flagstones were located on the periphery of an oval-shaped raised or "humped" area measuring approxi­mately 2-3 m in diameter through which a tree, aged about 30 years, had grown. Preliminary examination disclosed flakes of patinated white chert. Clearing the carpet of fir needles exposed Stratum I, a dark brown, loosely packed soil containing fist-sized cobbles and larger rocks, some of which were fire-cracked, together with occasional small flagstones. An admixture of lithic raw materials was present, including patinated white chert, crystalline quartz, purple rhyolite, green- and grey-banded chert, and occasional small fragments of coloured cherts. This stratum continued to a depth of about 8-12 cm below surface, and contained, predominantly, artifacts of Recent Indian origin. A few artifacts of Dorset affiliation were mixed in this deposit, along with an occasional iron nail. None of the nails showed aboriginal modification.

At a depth of 8-12 cm below surface a soil change was noted. The previous dark brown deposit gave way to a blacker soil mixed with gravel and containing fewer cobbles. This level, designated Stratum II, proved to contain only material of Dorset Eskimo origin. Patinated white chert and crystalline quartz were the predominant lithic materials found. Dorset Eskimo artifacts recovered include harpoon end blades, knives/bifaces, grindstone fragments, end scrapers, microblades, and various retouched and utilized flakes. Charcoal deposits were encountered, and samples taken. A radiocarbon analysis (Beta-4064) yielded a date of 1560 +/- 60 B.P. Stratum II contained greater amounts of gravel at lower depths, and eventually merged with a sterile gravel layer at about 25-30 cm below surface.

Charcoal deposits and fat-stained flagstones indicate that Feature 1 was a hearth. There is, however, no clearly delineated structure. It seems likely that the once-present structure has been disturbed and literally "uprooted" by tree growth. The flagstones first noted, those standing on edge, were found to be supported by tree roots, and therefore are probably not in their original position. The Dorset hearth may also have been disturbed to some extent by Recent Indian activity, possibly in a search for chert cobbles left by the Dorset inhabitants, as there is some evidence indicating that Indians were making use of this material. The profusion of rocks and cobbles in Stratum I, some of which evidenced fire-cracking, suggests the presence of an Indian hearth overlying the Dorset Eskimo hearth. Similar cobble features have been noted at other Recent Indian sites, notably Frenchmen's Island and Boyd's Cove.

Feature 2

Feature 2 is another probable hearth, lying to the west of Feature 1 and seemingly connected to it by two irregular rows of flagstones. These flagstone rows are slightly less than one metre wide, approximately parallel and about one metre apart, and extend for a distance of six metres. The stones are not neatly arranged; some are angled slightly and partially resting on top of others. The edges of several stones appeared freshly broken, and larger ones could sometimes be reconstructed from smaller fragments. None of the flagstones was set vertically.

Feature 2, the western extremity of these rows of flagstones, consists of scattered flagstones and other rocks. Charcoal and fat-stained rocks suggest that it was a hearth; again, however, there is no well-defined structure discernable. Stratigraphy in the area of Feature 2 was similar to that of Feature 1. The dark brown Stratum I encountered immediately below the surface overburden contained an admixture of Recent Indian and Dorset Eskimo artifacts. Compared to Feature 1, there were fewer Recent Indian artifacts in this upper stratum, and more Dorset Eskimo ones. Stratum II was a pure Dorset Eskimo deposit, containing an assortment of stone tools very similar to Feature 1. Charcoal collected in Feature 2 produced a radiocarbon date of 1280 +/- 60 B.P. (Beta-4062).

Feature 3

The rows of flagstones mentioned above lie roughly parallel to the beachfront (ESE-WNW). Towards their longitudinal centre, and on the side closer to the beach, a change in the Stratum II deposit is noticeable. Here, the Dorset deposit is very black, greasy, and compacted, compared to elsewhere. Few flagstones are present; instead the deposit is mixed with rocks of irregular shapes, some football-sized and larger. The artifact frequency is quite high,and the occurrence of many broken specimens is notable. This area, Feature 3, is thought to be a midden. The greasy black deposit is referred to as Stratum IIa.

A one by three metre section of this midden was excavated to a depth of 35 cm below surface, and testing in one location continued to a depth of 55 cm below surface without encountering a sterile sublevel. As this midden was situated about three metres from Feature 1 and four metres from Feature 2, it was not obviously associated with one hearth rather than the other. Indeed, if Features 1 and 2 proved contemporaneous, the midden might relate to both. A charcoal sample was recovered from Feature 3 at a depth of 25-30 cm below surface, and analysis gave a date of 1280 +/- 60 B.P. (Beta-4065).

Two general interpretations appeared feasible while excavation was ongoing. The rows of flagstones, linking two hearths, might constitute a single house structure. Alternately, the hearths might pertain to different occupations, with the flagstone being remnants of structures occupied at different times. If the former was correct, then charcoal collected from the two hearths would be expected to produce contemporaneous radiocarbon dates. Analysis of a sample from Feature 2 (Beta-4062) yielded a date of 1280 +/- 60 B.P., which is at least 160 years later than the Feature 1 date (1560 +/- 60 B.P.). Features 1 and 2, then, represent temporally distinct occupations. As Feature 2 is the more recent of the two hearths, most of the flagstone arrangements may relate to it. It seems a reasonable presumption that, in an area of repeated occupation, the most intact structure pertains to the most recent occupation.

The date from the Feature 3 midden is identical to the one from the Feature 2 hearth, suggesting that the midden is associated with the Feature 2 hearth and the flagstone rows. A tentative interpretation, then, would see the linear arrangements of flagstones as being the remnants of a structure built and occupied ca. 1280 B.P., incorporating the Feature 2 hearth and the Feature 3 midden. The discovery of a cross-mend between a tool preform fragment from the Feature 2 hearth and another fragment from the midden lends additional support to the suggestion that these two features were utilized contemporaneously.

The internal arrangement of this Dorset structure still remains unclear (see Excavation Plan). The flagstones appear to mark the limits of the structure. Bordering the flagstones (towards the outside of the structure) Stratum II is quite loose-packed and gravelly, with a relatively low artifact content. This may have resulted from gravel being heaped against the outside periphery of the structure, to seal and hold down the edges of a skin covering. Between the flagstone rows (in the centre of the structure) the ground is very hard-packed, and artifacts are again scarce. Artifacts were most frequently found amongst the flagstones and around the hearth feature. Initially we thought that the flagstone rows were sleeping platforms, but this idea does not agree with the artifact distribution. It now appears more plausible that the central unpaved area was used for sleeping, and the surrounding paved areas were work places. The structure's hearth (Feature 2) was placed at the western end of the structure. The entranceway was presumably located towards the eastern end of the structure, away from the hearth and facing the ocean. This would agree with the location of the midden -- immediately adjacent to the structure on the seaward side.

Feature 1, in the eastern part of the excavated area and bordering the structure described above, likely pertains to an older Dorset occupation (ca. 1560 B.P.). Excavation was not extended to the area around Feature I as time did not permit. A little testing was done, yet this did not reveal a flagstone structure, and therefore we presumed that the flagstones had been removed, perhaps to be used in later constructions. The Feature I date, compared with the Features 2 and 3 dates, gives a minimal time range for the Dorset occupation of Stock Cove, ca. 1300-1500 B.P.

Stock Cove has made a contribution to the prehistory of southeastern Newfoundland and is promising for future work as excavations to date have, literally and figuratively, only scratched the surface. Following is a brief summary of the phases of prehistoric occupation discovered to date at Stock Cove.

Recent Indian

The work described above, conducted during May, June and September of 1981, produced evidence of repeated occu­pation by Palaeo-Eskimo and Indian people. The most recent inhabitants -- Indians of the Little Passage Complex -- were widespread throughout Newfoundland, and are thought to have been the immediate ancestors of the historic Beothuks.

Despite our efforts at Stock Cove to avoid areas of Recent Indian activity (for reasons mentioned earlier), some 300 artifacts relating to that culture were collected. The single most common diagnostic tool type encountered was the corner-notched/expanding stem projectile point. The range of other tools recovered, such as scrapers, knives/ bifaces, linear flakes and bipolar cores, along with consid­erable amounts of flaking debris, indicate that the Recent Indians were performing a variety of activities at this site. Two activity areas were identified by concentrations of fire-cracked rock, flaking debris and artifacts. Stock Cove may prove to be one of the larger known sites of Recent Indian occupation on the Island.

In one sense, however, Stock Cove is not ideal for a Recent Indian study because of the mixing of cultural deposits which occurs. Recent Indian artifacts are restricted to the upper stratum, yet are often intermixed with objects of Dorset origin. In order to separate the two clearly, a comprehensive knowledge of Recent Indian and Dorset Eskimo assemblages is required. In the spring of 1981, Little Passage archaeology was in its infancy, and my very incomplete knowledge of what might be included in such an assemblage presented some on-site difficulties in separating and interpretating a mixed collection. This problem has been somewhat alleviated in recent years, with the excavation of unmixed Recent Indian sites (Pastore 1983). Some confusion still exists; for example, the distinction between Dorset and Recent Indian end scrapers is not entirely clear. With further work on single component Recent Indian sites, the Stock Cove site -- carefully approached -- can undoubtable yield a great deal of information pertaining to the Recent Indian period in southeastern Newfoundland.

Maritime Archaic

Maritime Archaic artifacts were rarely encountered during the excavation. The few diagnostic Archaic artifacts recovered include the basal portion of a stemmed, ground slate lance, and two celt fragments. Two leaf-shaped bifaces are possibly of Archaic origin, yet their provenience is more suggestive of a Recent Indian affiliation. The lance point and the celt fragments were found in Stratum II, and are not believed to be in situ. A few biface fragments collected from the eroding bank are thought to relate to a Maritime Archaic occupation, and suggest that much of the Archaic component may now be eroded from the site.

Groswater

The single radiocarbon date and the small, side-notched end blade from Stratun III hint at a Groswater occupation at Stock Cove prior to the arrival of Dorset people. Only in one restricted location was the excavation pursued to the depth of Stratum III, so very little can be said at present about this occupation. The presence of several centimetres of sterile gravel between Stratum II and Stratum III suggests, however, that there may have been a considerable time lapse between the Groswater and Dorset occupations. This is in agreement with the 200-300 year hiatus between the Dorset and Groswater occupations of Newfoundland proposed by Tuck (1982).

With regards to future excavation of the Groswater com­ponent, any excavator proposing to do so would first have to deal with the artifact-rich Dorset and Recent Indian deposits. This would certainly prove a major hinderance to an investigator concerned solely with Groswater phase Palaeo-Eskimo culture.

Dorset Eskimo

The most extensive and intensive occupation of Stock Cove was undoubtably by the Dorset Eskimos. Radiocarbon dates suggest that the Dorset people were present in Stock Cove over a 200 year period, and the actual span of occupation was probably at least somewhat greater. From the size of the occupied area and the structural evidence it might be inferred that the site was a semi-permanent residence, or at least something more than a casual camp. It is obvious that the Dorset inhabitants of Stock Cove were exerting considerable effort in moving and arranging flagstones, and it should be emphasized here that the area excavated was but one of many locations where flagstone features were evident. The depth and richness of the Dorset cultural deposit also suggest repeated occupations, and perhaps extended yearly stays. If, as was previously suggested, future excavations are successful in recovering preserved faunal remains, a better understanding of seasonality might be achieved.

Only 55 square metres were excavated during the 1981 season, out of some 2000-4000 square metres of occupied area (1.4-2.8%). This work has supplied a good look at one phase, at least, of Dorset Eskimo history at Stock Cove. It is impossible to say at this time whether this is representative of the entire period of Dorset occupation at the site. Undoubtedly the site is worthy of further attention, yet the direction of future research must be carefully considered. Two primary objectives which should be addressed are (a) a reconstruction of resource exploitation and (b) a clarification of the form of structures.

Difficulties encountered with organic preservation have already been related. Hopefully more extensive work will provide analyzable faunal material. The reconstruction of subsistance may also be aided through the exploration of neighbouring sites in the Isthmus region, as the Stock Cove inhabitants may have utilized other nearby locations and their various resources in a seasonal exploitation round. These other sites may supply preserved faunal material, or through their specific locations imply a resource-related activity. In this sense, the Isthmus of Avalon constitutes a prime location for a settlement-subsis­tence study.

Flagstone structures are well-known in the Arctic and on the Labrador coast, yet are quite rare on the Island. With the exception of Phillip's Garden at Port au Choix, few Dorset sites in Newfoundland have produced structural evidence of any kind, and therefore Stock Cove presents a unique opportunity. A clearer understanding of Dorset structures at Stock Cove can only be achieved through more extensive excavation. The flagstone structure described above, which included a hearth and a midden, occupied about 27 square metres. Approximately 130 man-days were required to expose it, as excavation was slowed by the high artifact frequency. Obviously, productive work cannot proceed on a piecemeal basis. A comprehensive interpretation of the internal organization of Stock Cove structures and rela­tionships between structures must begin with large areal excavations. Such an undertaking will require a considerable manpower and time commitment -- e.g., an eight or ten person crew engaged in a multi-season project -- and backup facil­ities to deal with the thousands (perhaps tens of thousands) of retrieved artifacts. There are many secondary objectives that might be addressed by future researchers, for example: (a) additional radiocarbon dates would augment the rather meagre chronology; (b) a distinctive chert type dominates the lithic collection, suggesting that the source of this material is close to Stock Cove. The locating of this source would be of assistance in reconstructing the Dorset cultural system in the region of the Isthmus of Avalon.

The Dorset Eskimo people who lived at Stock Cove 1300-1500 years ago undoubtably inhabited other sites in the regions, to hunt, fish, or collect stone material for their tools. As such, the Stock Cove site cannot be viewed in isolation. While such data as hearth features, house structures, and tool assemblages derived from a single site are of fundamental importance, this information must be incorporated into a regional framework -- i.e., an inter-site comparative study -- in order to interpret past lifeways of a population. The next two chapters present a descriptive analysis of Dorset Eskimo stone tools from Stock Cove, and a summary of the natural and cultural history of southeastern Newfoundland. Chapter four attempts to intergrate this data and the Stock Cove site information presented in the preceding pages, in a hypothetical reconstruction of the Dorset Eskimo occupation of insular Newfoundland.