A recent paper by Ralph T. Pastore (1986) discusses the distribution of aboriginal sites on the Island of Newfoundland. One observation made by Pastore in reference to sites of Dorset Eskimo culture concerns the dichotomy between "base camps" and "exploitation camps". Large, extensive sites producing numerous artifacts and sometimes substantial structural remains are considered base camp locations; examples are the Cape Ray, Port au Choix-2, The Beaches, and Stock Cove sites. A variety of activities, including hunting, butchering, tool manufacturing, and general maintenance, is inferred for these sites from their toolkits. Smaller and artifactually poorer sites are equated with extractive or procurement stations where specific activities occurred, e.g., harp seal hunting, without the population concentration and range of activities suggested for the base camps.

Another system of site patterning is expressed in Louis Binford's paper "The Archaeology of Place" (1982). Binford describes a pattern of "economic zonation" (ibid:6-8) which he has applied to the Nunamuit, and which includes a "residential camp", a "foraging zone", a "logistical radius", and a "visiting zone". From the residential camp, groups would scour the surrounding territory, procuring resources which were transported back to the residential camp. The countryside exploited in such a manner constitutes the foraging zone. Once groups or individuals travelled further distances from the residential camp, overnight stays would become necessary. The logistical radius incorporates this larger territory exploited from the residential camp with the assistance of temporary encampments. The visiting zone lies beyond the boundaries of the logistical radius, and is the area contemporaneously occupied by related peoples. The pattern of residential camp, foraging zone, and logistical radius is repeated there.

On a more general level, Beardsley, et. al. (1956) have discussed seven primary types of community patterning, namely Free Wandering, Restricted Wandering, Central-based Wandering, Semi-permanent Sedentary, Simple Nuclear Centred, Advanced Nuclear Centred, and Supra-nuclear Integrated. The first three are applied by the authors to non-agrarian societies, and are of interest here.

Free Wandering involves unrestricted movement governed only by the availability of resources. The frequent relocation of camps resulted in dispersed and scanty archaeological remains (ibid:135-36). Restricted Wandering, a refinement of this pattern, denotes a degree of territoriality and possibly a planned seasonal round of activities. Restricted Wandering archaeological sites are not unlike those of free wanderers, yet may contain specialized tools indicative of a more intensive exploitation (ibid:136-37). A Central-based Wandering group established a central base, where a storable or abundant food source was available, for some of the year, and wandered for the remainder of the year. Central-base sites are distinguished from the preceding types by their deep midden deposits, indicative of repeated occupations (ibid:138-39).

The categories of these settlement classification schemes cannot be equated as the first was generated archaeologically, the second was derived from ethnology, and the last combines archaeological and ethnographical data. Although their points of similarity are notable, and suggest an underlying accuracy, such schemes should be applied with caution.

On the basis of its size and richness Stock Cove has been judged (by Pastore, myself, and others) to conform to the concept of a "base camp". This designation does not seem entirely appropriate however, as the definition of a "base camp" is unclear. Is it: (1) a centrally located camp, whose primary feature is its strategic location with respect to other camps which are periodically and alternately exploited (after Binford); (2) a settlement near a plentiful and dependable resource, where the grouping of a large number of people was possible (as seen amongst Beardsley's Central-based Wanderers); or (3) a location combining both these features in some manner? It seems there may well be several "kinds" of base camps, if base camps are only identified on the basis of site size and the abundance of artifacts.

Dorset Eskimo people inhabited much of insular Newfoundland. There is not a major coastal area (perhaps excepting the Burin and Avalon Peninsulas) where Dorset Eskimo sites have not been found. There are large sites with deep cultural deposits (e.g., Port au Choix-2, Stock Cove) and small sites with thin depositions (e.g., Broom Point, Shambler's Cove). Large and small sites are found in good harp sealing areas (northern and western Newfoundland) as well as in regions where harp seals were not available (southern Newfoundland). Harbour and gray seals were undoubtable more plentiful in some locations than others. Caribou were immediately accessible from some sites, but not from all. Some sites are located near salmon streams, while others are many miles from the nearest such river.

In other words, Dorset Eskimo sites throughout Newfoundland are not consistant with respect to relative size and richness, or resource orientation. Ecological factors may have determined whether or not a particular locale was suitable for Dorset Eskimo occupation, and if so, whether it would be the location of a large settlement (numerous inhabitants and/or repeated yearly occupations), or a small encampment (few inhabitants and/or short occupation). A locale near a plentiful resource (e.g., harp seals) might support numerous people for a season (in this case the spring), but would be abandoned unless other resources appeared at the same location during other seasons, or could be obtained within a reasonable distance. If either were true, then the locale had the potential to become a large site, with an artifact assemblage indicative of various hunting, butchering, and maintenance activities, and with substantial structural remains. If not, the site would likely exhibit a less cultural deposition, and its depth and extent would be dependent upon the number of years the locale was revisited. As well, the tool assemblage would be expected to reflect the site specialization (seal hunting, in the example being used) instead of being indicative of various hunting and maintenance activities.

Subsistence involves making efficient use of the resources which are available, and logic suggests that subsistence must differ in the face of different resources. The specific tactics necessary to exploit harp seals, harbour seals, caribou, salmon, etc., are quite different. In Newfoundland, Dorset people made use of these resources, yet it cannot be assumed that all bands participated in these different activities to the same degree; almost certainly they did not. If the balance of resources was such that each band could spend the same season and the same amount of time doing the same thing, then there would be a single Island-wide subsistence/settlement pattern discernable. In reality, however, regions of the Island differed with respect to the resource availability, therefore subsistence and settlement were necessarily adapted to local conditions. It seems logical therefore to examine each region of insular Newfoundland individually, with respect to its resource potential and archaeological sites, and suggest local systems of adaptation rather than, or at least prior to, attempting generalizations.

Published works by Elmer Harp Jr. and Urve Linnamae have implied or directly stated that the Dorset culture in Newfoundland may be "typified" by a few traits of artifact assemblages, and thus distinguished from Dorset Eskimo occupations in the eastern Arctic. This line of thinking has persisted in recent years. While some traits do distinguish Newfoundland Dorset from Arctic Dorset manifestations, these traits are not necessarily common to all assemblages on the Island, i.e., there is not a "typical Newfoundland Dorset" culture. Stone tools are the most obvious indicators of the extant variety, yet it is difficult at this point to elucidate differences without a detailed qualitative and quantatative examination of assemblages. Such an examination is greatly hindered by the inaccessibilty of some collections, and the incompatability of published reports. A thorough examination of Newfoundland Dorset Eskimo assemblages is beyond the scope of this work, as it would ultimately require the analysis and comparison of tens of thousands of lithic tools. For the present, there- fore, I will limit discussion of lithics to a brief look at a few tool types and the raw materials employed. This cursory examination is, I think, sufficient to demonstrate the diversity present among Newfoundland Dorset Eskimo artifact collections.

The west coast of Newfoundland has continued to interest researchers of Dorset culture since it was first systematic- ally surveyed during the decade of the 1920's. Elmer Harp Jr. performed excavations at the Phillip's Garden site at Port au Choix and tested several other sites in the area (Harp 1964). More recently, Urve Linnamae (1975) has conducted research at Cape Ray, at the southwestern corner of the Island. Since 1980, a Dorset site at Broom Point has been excavated (Tuck 1983), and a survey of the Port au Port Peninsula performed (Simpson 1984). In 1984 Dr. M.A.P. Renouf began a multi-season project at Phillip's Garden and environs. In sum, these investigations have provided a good look at the Dorset Eskimo occupation of western Newfoundland.

PORT AU CHOIX-2 (PHILLIPS GARDEN)

In discussing the Dorset artifacts from Port au Choix-2 and other northwest coast sites, Elmer Harp Jr. wrote that "...a strong cultural unity" was evident. More specifically: Many of the artifact types occur repeatedly; they are strikingly uniform in their small size; they show strong similarities in workmanship and technique; and they are made of the same materials (Harp 1964:89). These features of northwest coast Dorset artifacts -- relative frequencies, size, workmanship/technique, and raw materials -- gain new significance in light of Dorset assemblages from other regions of Newfoundland.

The two most popular artifact types at Phillip's Garden (as reported by Harp) are triangular projectile points and snub-nosed scrapers, each representing about 15% of the collection. The triangular projectile points, or harpoon end blades, have convex sides, slightly or deeply concave bases, and a width approximately 40-50% of their length. Most (86%) have surface flaking restricted to one surface, with the opposite face being either unworked or tip-fluted. The popular raw materials are chert and flint (Harp 1964:36). Snub-nosed scrapers, or end scrapers, are also fashioned from chert and flint, have convex working edges, and flaking on their "exterior" (dorsal) surfaces. Some examples are notched (ibid:54). Next in frequency are prismatic blades (7%), and "prismatic blade knives, blunt-edged" (4.5%), which are tip-flute flakes. The smaller prismatic blades are made from crystalline quartz, and the large ones from chert and flint. One-quarter show utilization or purposeful retouch (ibid:48-50). These artifact types constitute more than 40% of the Port au Choix-2 collection, with the remainder consisting of various knife forms, of which some are Groswater, but certain ones such as large, asymmetric, notched specimens can be identified as Dorset; ground and polished bevelled slate points with characteristic basal notches, and slate "chisels", also ground and polished; utilized chips; fragmentary artifacts; and a few rare types which in some cases may be recognized as Groswater rather than Dorset Eskimo. It should be noted that this collection represents only a portion of the total number of artifacts recovered by Harp from Port au Choix-2, as only his early work at Phillip's Garden is reported in his 1964 publication.

Other sites visited by Harp along the northwest coast were not worked as intensively as was Phillip's Garden, and the collections from them are of limited usefulness for comparative purposes. Mixing of Dorset Eskimo and Groswater deposits occurs. Some sites, such as Norris Point, are now known to have discrete Groswater components. Suffice it to say that other northwest coast Dorset sites examined by Harp probably have (or had) artifact assemblages similar to Phillip's Garden, with respect to artifact types, styles, and raw materials.

The west coast of Newfoundland borders the Gulf of St. Lawrence, and most locations along the coast are exposed to winds and seas from the gulf. As Harp notes (ibid:85), "Exposure to wind and climate... does not seem to have governed [site] location." He does say, however, that while some sites are open to the north, others, like Norris Point-1 and 2, are situated in more sheltered areas. Subsequent work at Norris Point-1 by Paul Bishop (n.d.) has shown it to be a Groswater encampment, and its sheltered location compares with the locations of other Groswater sites in Newfoundland, and contrasts with the "outer" locations common to Dorset Eskimo sites in western and northern Newfoundland.

These northwest coast Dorset sites were likely prime seal hunting locations. Analysis of food bone from House 4 at Port au Choix-2 showed almost 98% to be seal, mostly harp seal (Harp 1976:128). House 5 is thought to be a summer dwelling (ibid:130), as it is less substantial than other structures at the site and lacks an internal hearth. This would suggest that Port au Choix-2 was occupied at seasons - - perhaps summer and fall -- in addition to the spring harp sealing period.

BROOM POINT

Systematic archaeological research has been done at another Dorset Eskimo site on the northwest coast. Broom Point, in Gros Morne National Park, was investigated by a team from Memorial University under contract to Parks Canada in 1982. Additional excavation was done in 1984 by Carol Krol, a graduate student at Memorial University, again in cooperation with Parks Canada. Data from the 1984 excavation are not yet compiled, but a report of the results of the 1982 fieldwork is available (Tuck 1983).

Tool types from Broom Point closely agree with ones from Phillip's Garden, and the relative frequencies of types differ only slightly. The most frequent artifact type at Broom Point is the prismatic blade, accounting for 35.4% of the assemblage. End scrapers (11.5%) and harpoon end blades (10.4%) are next in order (Tuck 1983:66). In comparison to Phillip's Garden, prismatic blades are about five times as plentiful, while end blades and end scrapers are slightly less frequent. End blades and end scrapers occur with approximately equal frequency, as was the case at Phillip's Garden. End blades with convex sides, slightly or deeply concave bases, and which are unifacially flaked or tip-fluted, closely resemble Phillip's Garden specimens. Fine-grained coloured cherts were used to manufacture these and other artifacts.

There appears to be little significant difference between the assemblages from Phillips Garden and Broom Point with respect to the occurrence of artifact types and type frequencies -- except for the prismatic blade count-- as most chipped stone artifact types described or illustrated by Harp (1964) are represented at Broom Point. This would appear to argue against an hypothesis that Broom Point was a specialized procurement station. Instead it indicates that Broom Point was an independant and self-contained settlement, or "... a small encampment of Middle Dorset people whose equipment included all those chipped and ground stone tools and weapons necessary for maintenance and survival" (Tuck 1983:67). The primary subsistence focus was probably the exploitation of gulf seal herds during the winter and spring, although the possibility of summer and fall occupations at Broom Point cannot be excluded.

PORT AU PORT

The Port au Port Peninsula, on the southern part of the west coast, was first surveyed in 1975 by Paul Carignan (n.d.) and again by David N. Simpson in 1983 (Simpson 1984). The latter investigator records the presence of four prehistoric sites on the Peninsula. Only one, however, has yielded extensive evidence of Dorset Eskimo occupation. At this site (the Port au Port site, DdBq-1) harpoon end blades were relatively infrequent (6%), but closely resemble Phillip's Garden specimens with their convex sides, slightly or deeply concave bases and high frequency of tip-fluting (ibid: pl. 1). A variety of end scrapers, carefully manufactured from fine cherts, constitutes about 11% of the assemblage. An analysis of faunal remains from the Dorset component of the Port au Port site revealed a predominance of seal bones. Three species, harp, ringed, and harbour seals, are represented (Simpson pers. com.).

In addition to the habitation sites located during the Port au Port survey, three chert outcrops were discovered. Although no certain evidence linking the Dorset Eskimo population to these potential quarries was found, their existence may have provided a reason, other than seal hunting, for the Dorset occupation of the area.

CAPE RAY LIGHT

A large and rich Dorset Eskimo site is located at Cape Ray, at the southern extreme of the west coast, and was excavated by Urve Linnamae during 1967 and 1968 (Linnamae 1975). A total of 4,797 lithic artifacts were recovered, including all the major types recorded at Port au Choix-2, Broom Point, Port au Port, and elsewhere on the west coast. Prismatic blades, or microblades, constitute the single largest category (18.5%), followed by end blades (14.1%) and end scrapers (12.6%). The relative frequencies of the latter two types compare favourably with sites on the northwest coast. With respect to end blades, a greater variety of forms is present than at other west coast sites. Many specimens are similar to the Port au Choix type, but others present a different outline form -- longer and narrower (Linnamae 1975: fig. 15). Some of these end blades display basal grinding, a characteristic not seen in northwest coast assemblages.

DISCUSSION

These Dorset Eskimo sites, ranging along the west coast from Cape Riche to Cape Ray, share certain similarities of assemblages. Most notable is the style of harpoon end blades. Characteristically, end blades from west coast sites have convex sides, slightly or deeply concave bases, and a width equal to or slightly less than one-half their length. Most are carefully manufactured from fine-grained black, green, or blue chert. End scrapers are similarly well- made from the same materials. These two artifact types usually occur with approximately equal frequency, each constituting from 10% to 15% of the total artifact assemblage. Port au Port is the only site where the relative frequencies of these artifact types differ, as end blades constitute only 6% of that assemblage, but due to the small sampling of that site this difference may not be significant. Prismatic blades are numerically more variable, ranging from 7% of the Port au Choix-2 assemblage to 35.4% of the Broom Point collection. If these tools are, as is suspected, butchering knives, these variable frequencies may be indicative of on-site activity, with the relatively low frequency of blades recovered from Port au Choix indicating that butchering was not done in or close to the house structures excavated by Harp. Conversely, the excavated portion of the Broom Point site was a butchering area. All of the sites discussed above are located on the immediate coast, exposed to the Gulf of St. Lawrence. Exposed locations appear to have been preferred as they facilitated access to the gulf seal herds. The presence of harpoon equipment at each site, and the faunal remains from Phillip's Garden and Port au Port, indicate that the Dorset inhabitants of the west coast were seal hunters. The existence of at least one summer dwelling at Port au Choix-2, and the diversity of the seal faunal assemblage from Port au Port indicate that these sites were occupied during summer and fall months in addition to the spring -- perhaps even year-round.

It seems that the high biomass of seal herds, together with their availability for several seasons of the year, enabled Dorset people to congregate at several locations along the west coast. At each of these locations harp seals are obtainable during the spring as they moved north on pack ice. After the harp migration, harbour and grey seals return to the coast as the pack ice disappears, and can be taken on shore as they haul out to bask in the sun. These species abandon inshore locations in late fall/early winter, as landfast ice appears in coves and inlets. Such ice conditions are, in turn, preferred by the ringed seals, who maintain breathing holes and bask on landfast ice during winter months. In such a manner, seal resources may have been available to occupants of the west coast year-round, and permitted the establishment of large settlements maintained through most, perhaps all, of the year. "Budding" of communities may have led to the establishment of new, smaller encampments, (e.g., Broom Point) which functioned in the same ways as larger camps; i.e., were semi-permanent locations where several family groups lived for extended periods and participated in hunting, butchering, and maintenance activities. From each of these large and small villages, seals were hunted on the winter landfast ice, on the spring pack ice, and on or near shore during the summer and fall. Summer and fall subsistence could have been significantly augmented by salmon fishing. This may have necessitated the establishment of temporary camps by salmon rivers, yet while Dorset sites have been located near such rivers on the northwest coast (Wintemberg 1939), nothing of the nature of these sites is known.

This pattern of marine resource exploitation is in accordance with the available faunal remains and the geographic locations of known sites. It does not, however, include the possibility of terrestrial hunting. Caribou are not immediately accessible from the northwest coast, and no significant occurrence of caribou bone is noted by Harp in the Port au Choix-2 faunal assemblage. From this it might be inferred that either the Dorset people were not participating in caribou hunts, or they shifted to interior locations at some time of the year (possibly late fall/winter, when seal resources were at their lowest point and when caribou were migrating) to take caribou. On the southern part of the west coast caribou are perhaps more accessible from the coast, as the southern interior of Newfoundland supports a large herd. Faunal remains from the Dorset site at Port au Port includes a minimal representation of caribou (Simpson pers. com.). One of the few known interior Dorset Eskimo sites is located at Long Pond, about 30 km inland from the coastal community of St. Georges on the southern west coast (Penney n.d.b). Little is known about this site, as it is now submerged and probably destroyed as a result of hydroelectric projects in the area. End blades, microblades, an end scraper, and multi-notched bifaces were collected along the shore of Long Pond. The location may have been a Dorset Eskimo caribou hunting station.

Each of the west coast sites discussed above has been radiocarbon dated, and the results give a representation of the time-frame of the western Newfoundland Dorset Eskimo occupation. A total of twenty dates were run on Phillip's Garden samples, of which fourteen were derived from wood charcoal and are acceptable:

1736±48 B.P. (P-692)
1580±54 B.P. (P-727)
1712±40 (P-695)
1565±53 (P-733)
1683±49 (P-736)
1538±55 (P-729)
1659±48 (P-693)
1509±47 (P-696)
1623±47 (P-679)
1502±49 (P-676)
1602±49 (P-694)
1465±51 (P-734)
1593±49 (P-683)
1321±49 (P-737)

(Harp 1976:137)

Three dates are available for Broom Point:

1650±90 B.P. (Beta-4771) (Tuck 1983)
1420±70 (Beta-11375) (Krol pers. comm.)
1370±100 (Beta-11376) (Krol pers. comm.)

Two Dorset Eskimo radiocarbon dates were run at Port au Port:

1350±80 B.P. (Beta-7778) (Simpson 1984)
1300±60 (Beta-7777) (Simpson 1984)

A single reliable wood charcoal date is available from Cape Ray:

1565±95 B.P. (Gx-1198) (Linnamae 1975)

The Broom Point, Port au Port and Cape Ray dates all fall within the range established by Port au Choix-2, and this range of ca. 1300-1750 B.P., or A.D. 200-650, probably encompasses the major period of Dorset occupation on the west coast of Newfoundland.

In 1967-68 Urve Linnamae worked at the Pittman site at Sop's Island, White Bay. Dorset Eskimo sites at Englee and nearby Lane's Cove were excavated by Dr. James A. Tuck in 1969. An environmental impact study conducted by Gerald Penney for Newfoundland Hydro resulted in the discovered of a small Dorset Eskimo site at Devil's Cove. This, plus the excavation of the Norse site at l'Anse aux Meadows, where a few Palaeo-Eskimo artifacts were found, is the extent of Dorset archaeology along the east coast of the Northern Peninsula.

ENGLEE

The Englee site has produced a large collection of stone tools, most manufactured from black chert. As a typological analysis of the collection is not yet available, a quantatative comparison of the Englee assemblage with other Newfoundland assemblages is not possible at this time; however, some interesting points may be gleaned from a visual inspection of the artifacts. Chert end blades and end scrapers are common. The end blades bear a remarkable resemblance to specimens from Port au Choix-2 and Port au Port with respect to outline form (convex sides, concave base, length-width dimensions) and manufacture technique (fine workmanship, frequent tip-fluting). End scrapers are also visually similar to west coast ones, with, in several cases, the additional feature of ventral flaking.

LANE'S COVE, DEVIL'S COVE

Lane's Cove is located quite close to Englee. The assemblages from both sites are very similar, with no significant difference being immediately observable. The few end blades collected by Penney at Devil's Cove are also similar to specimens from the two preceding sites. They are slightly smaller than usual in overall size and are, in my view, the finest specimens recovered from any Newfoundland Dorset Eskimo site.

PITTMAN

The Pittman site on Sop's Island, White Bay, yielded a collection of 1,403 lithic tools. Mixing of Dorset Eskimo and Groswater artifacts occurs (see Linnamae 1975, fig. 28), and was not recognized at the time by the excavator, so the relative percentages of artifacts may be slightly skewed. Triangular end blades constitute 15.5%, end scrapers 12.2% (ibid:179). Microblades are the most frequent artifact type, making up 18.5% of the collection. Harpoon end blades similar to west coast and Englee/Lane's Cove specimens are present (ibid: fig. 28f), along with others whose longer, narrower outline contrasts with the style common along the west coast (ibid: fig. 28a-d).

DISCUSSION

There is no obvious difference between the Pittman site assemblage and west coast assemblages with respect to the relative frequencies of end blades and end scrapers. At the Pittman site each of these artifact types falls within the 10-15% frequency range established on the west coast. The frequency of blades at the Pittman site is most similar to Cape Ray Light, and is intermediate between Port au Choix-2 and Broom Point. The Englee and Lane's Cove assemblages are believed to conform to the Pittman site with respect to the relative occurrence of these artifact types, but this cannot be confirmed until a quantitative analysis of these assemblages is completed. The minor artifact classes present in the west are represented at these east coast Northern Peninsula sites.

This coast has not been explored to the extent the west coast has, and consequently any settlement/subsistence hypothesis is tentative. Resources are not unlike those of the west coast, in that numbers of migratory seals of the Front herd are found on the coast during the spring, along with non-migratory species at other seasons. On this basis alone one would expect a patterning of sites similar to that found on the west coast, and the little archaeological research that has been done appears to bear this out. Several large sites -- Englee, Lanes Cove, and Pittman site -- suggest large and/or repeated occupations. Toolkits from these sites indicate a balance of maintenance, hunting, and butchering activities, and are not indicative of site specialization. The impression is again of relatively large semi-permanent encampments centred around the intensive exploitation of sea-mammals. As was the case on the west coast, harp seals may have been taken during the spring, harbour and grey seals during summer and fall, and ringed seals during the winter.

It is possible that the Dorset occupants of this coast hunted caribou since the interior plateau of the Great Northern Peninsula supports a caribou herd. The relocation of settlement from the coast to the interior may have been necessary if this herd was to be exploited, as the east coast of the Northern Peninsula is very rugged and difficult to travel. Alternately, the nature of the topography may have funnelled herds into restricted areas during their migrations to the coast, making caribou hunting possible from coastal sites. No survey work has been done in the interior of the Great Northern Peninsula, and therefore the existance of inland caribou hunting sites remains hypothetical.

A single radiocarbon date is available from both Englee and the Pittman site. Both are acceptable, but they can only indicate a time of occupation at each site. At present we have no means of knowing how long each site was occupied, or at what stage in the occupation the radiocarbon date falls. Englee is dated 1585±95 B.P., and Pittman at 1340±100 B.P. These dates fall within the latter part of the range established for the west coast, perhaps indicating that the period after ca. 1600 B.P. (A.D. 350) was a time of widespread occupation throughout western Newfoundland. This agrees with the suggested time of Dorset occupation at l'Anse aux Meadows -- ca. A.D. 500 (Wallace n.d.).

The northeast coast of insular Newfoundland, including Notre Dame and Bonavista Bays, has not been explored to the extent the western region has with regards to Dorset Eskimo occupation. In 1972 and 1973 Paul Carignan excavated at the Beaches, Bonavista Bay (Carignan 1975). Shambler's Cove, located about 30 km from the Beaches, was investigated in 1982 by Reginald Auger, under contract to the Government of Newfoundland and Labrador. Dr. Ralph Pastore's 1981 survey of eastern Notre Dame Bay located several Dorset Eskimo sites, and is the only Dorset research reported for that bay (Pastore 1982).

THE BEACHES

The Beaches site incorporates at least four components: Maritime Archaic, Groswater, Dorset Eskimo, and Recent Indian. The latter component is probably more complex than previously thought, perhaps including several temporally and typologically discrete phases. These repeated occupations have resulted in mixed cultural deposits lacking clear stratigraphic separation, and many artifacts have been assigned cultural affiliation on the basis of typology rather than provenience, stratigraphy, and association. With respect to some artifacts this methodology is suspect and a degree of caution is necessary when dealing with the Beaches collection. Certain forms of side-notched end blades and asymmetric knives included by Carignan in the Dorset Eskimo assemblage are probably Groswater and are excluded here. Some of the scrapers and microblades are possibly Groswater, but as this is difficult to determine the figures are left unaltered in the following discussion.

An estimated breakdown shows that end blades constitute 18% of the assemblage, end scrapers 17% and microblades 15%. The latter two figures may be slightly high if they include some Groswater specimens. Compared to western sites, end blades and end scrapers are slightly more frequent, but again approximately equal in number. The outline form of end blades is, however, significantly different. A long and relatively narrow shape predominates, sides tend to be less convex, and bases are straight or only slightly concave. All specimens but one are manufactured from grey rhyolite -- this is the most popular material in the Dorset assemblage. The frequency of microblades is similar to Cape Ray and the Pittman site. Other artifact types represented in the Beaches assemblage are preforms, flake knives, tip-flute flakes, soapstone, and ground slate knives and projectiles, including most of the types present in western sites.

Carignan suggests that the Beaches area was occupied during the early spring to late fall period, at which time subsistence was based primarily on sea mammals. Seal bones were recovered from the Beaches, but were not clearly associated with cultural remains (Carignan 1975:39).

SHAMBLER'S COVE

The results of the Shambler's Cove excavation (Tuck 1983) have provided an interesting contrast with the Beaches, and indeed all Dorset Eskimo sites considered thus far. Chert and slate end blades constitute 38% of the Shambler's Cove collection. Chipped specimens are quite similar to Beaches end blades, with their elongated form and straight or slightly concave bases. The grey rhyolite popular at the Beaches is present. The relatively small Shambler's Cove assemblage also includes three ground slate projectiles, an unusually high representation of this tool type. Micro- blades account for another 35.2%, while five end scrapers represent only 3.5% of the collection. The frequency of microblades compares well with Broom Point, and contrasts with other sites, where this tool type was found to constitute from 7% to 18% of the assemblage. The occurrence of end scrapers is unusually low. Other artifact types present at Shambler's Cove include preforms, tip-flute flakes, and biface knives (ibid:62).

It has been suggested (ibid:62) that the large representation of hunting equipment (i.e., end blades) at Shambler's Cove indicates it was a specialized seal hunting encampment, and the concurrent popularity of blades suggest that Shambler's Cove was also a butchering station. The low frequency of end scrapers, and the complete absence of several artifact types, would appear to indicate that the full range of maintenance activities seen at other sites did not occur at Shambler's Cove. The visual similarities between Shambler's Cove and Beaches end blades with respect to style and raw material, and the close proximity of the two sites, suggest that both sites were utilized by the same people.

EASTERN NOTRE DAME BAY SITES

The 1981 survey of eastern Notre Dame Bay resulted in the discovery of three sites with large Dorset Eskimo components encompassing areas of 800 to 3000 m2, the largest being the Dark Hole site. At least six smaller ones were also found (Pastore 1982:164). Until the collections from these sites are analysed, however, their relationships to other Newfoundland sites will remain unclear. The locations of these sites, amongst the many small islands of Notre Dame Bay, stand in contrast to such exposed locations as the Beaches and Shambler's Cove, suggesting, perhaps, a slightly different subsistence routine.

DISCUSSION

The northeast coast of the island is populated by migratory seals during the spring, by harbour and grey seals in the summer and fall, and ringed seals during the winter. Caribou do not usually reach the coast but are available in the hinterland. Numerous streams support large populations of brook trout and salmon.

The presence of a varied sea-mammal population on the northeast coast would seem to make it similar to the regions considered above, yet the archaeological data suggest that a variant pattern of subsistence existed. In Bonavista Bay, the Beaches and Shambler's Cove sites exhibit remarkable similarities with respect to tool styles and lithic material utilization, but differ with regards to the representation of artifact types at each. The Beaches site has yielded a balanced assemblage with maintenance, hunting, and butchering tools represented, similar to assemblages from large sites in the west. Shambler's Cove, on the other hand, has produced an assemblage in which hunting and butchering equipment is over-represented, with harpoon end blades and microblades account for more than 73% of the assemblage. This would suggest that its primary use was as a seal hunting and butchering camp (Tuck 1983). Most artifact types associated with maintenance usually found at Dorset Eskimo sites (e.g., scrapers, various forms of knives, burin-like-tools, soapstone vessel fragments), are either absent or under-represented in the Shambler's Cove collection. From this it might be concluded that the sea-mammal resources of Bonavista Bay differed in some manner from those of western regions. Herds may have been dispersed along shore, requiring a more widespread and mobile hunt than was necessary in the west. Such a hunt necessitated the establishment of temporary hunting encampments, such as Shamblers Cove. The absence of large sites (even the Beaches does not compare with Port au Choix-2) may be seen to support the suggestion that concentrated hunts, which in turn allowed for the gathering of numerous family groups, were not possible in Bonavista Bay. Further exploration of the numerous small sites will shed more light on this possibility. It is expected that some small sites will show balanced assemblages where a few family groups were able to congregate for part of the year, while others will prove to be specialized hunting camps like Shambler's Cove.

Other specialized sites, besides sealing stations, may have been established for caribou hunting and salmon fishing, yet there are presently no data to support this possibility. This lack of data is primarily due to the incomplete nature of survey work in the hinterland of Bonavista Bay, a state of affairs shared by all interior Newfoundland regions.

Notre Dame Bay, the interlying coastal region between Bonavista Bay and the Great Northern Peninsula, was partially surveyed in 1981 and the results of this survey suggested that the Dorset population relied heavily on sea-mammals. Migratory seals of the Front herd are numerous during spring, along with other species during summer, fall, and winter. Large Dorset sites exist. One, Dark Hole, ranks among the largest and richest on the island. Numerous small sites are also present, yet little is known about them. Excavations at these small sites would reveal whether they were specialized sites, similar to Shambler's Cove, or "budding" communities after the Broom Point pattern on the west coast, and thus shed some light on Dorset Eskimo subsistence and settlement in Notre Dame Bay.

A radiocarbon date of 1650±95 B.P. (SI-1383) is available for the Beaches site, and a date of 1580 B.P. is reported for the Woodworth site in Notre Dame Bay (Carignan 1975). Shambler's Cove is dated 1890±100 B.P. (Beta-5369; Tuck 1983). The Beaches and Woodworth determinations fall within the early/middle part of the range established in the west, while the Shambler's Cove date is older than all west coast dates, and is presently the oldest Dorset Eskimo radiocarbon determination in Newfoundland. While it is tempting to hypothesis that the northeast coast was occupied earlier and perhaps for a shorter duration than the west, the evidence is scanty.

A number of sites are recorded in the vicinity of the Isthmus of Avalon. The Stock Cove site shows evidence of an extensive Dorset occupation, including structural remains. The Frenchmen's Island site (Evans 1981, 1982) has produced a smaller Dorset assemblage comprising stylistically similar artifacts. An archaeological survey of Placentia Bay (Linnamae n.d.) located several sites producing comparable artifacts.

STOCK COVE

The artifact assemblage from Stock Cove is detailed in a preceding chapter, and consequently only a few main points are reiterated here. The assemblage from this site is quite similar to the Frenchmen's Island collection with respect to stylistic forms of artifacts and raw material utilization. Harpoon end blades conform to the type observed in Bonavista Bay, with respect to outline form. The elongated shape, with a straight or slightly concave base is most popular. Unlike Bonavista Bay examples, many end blades from these sites are ground. Fifty-six percent of the Stock Cove end blades exhibit grinding. Most end blades, bifaces, and knives are manufactured from a chert that now exhibits a white or brown patina. Colourful, fine-grained cherts popular in western sites, and grey rhyolite as found at the Beaches and Shambler's Cove are very rare. End scrapers and microblades are commonly made from quartz crystal. The Stock Cove assemblage, comprising 1,367 Dorset tools, contains a large proportion of end blades (28.1%), numerous microblades (13.5%), 100 knives/bifaces (7.3%), and comparatively few end scrapers (5.3%).

FRENCHMEN'S ISLAND

Frenchmen's Island has produced a collection of 142 Dorset Eskimo artifacts, including harpoon end blades, bifaces, end scrapers, microblades, preforms, and tip-flute flakes. End blades, which constitute 16.9% of the assemblage, are visually similar to Stock Cove specimens, with respect to the type of raw material, outline, and occurrence of grinding. End scrapers make up another 10.5%, and the most popular artifact type, the microblade, represents 26.1% of the Frenchmen's Island Dorset Eskimo assemblage.

PLACENTIA BAY SITES

Sites located in Placentia Bay by Urve Linnamae (n.d.) have yielded artifacts closely resembling ones from Stock Cove and Frenchmen's Island. Notable is the presence of elongated, ground, patinated end blades, and quartz crystal end scrapers and microblades. The preliminary nature of the survey work in Placentia Bay makes further comparison of assemblages difficult, as sites were only briefly visited and intensive excavations not performed.

DISCUSSION

A variety of resources were available to prehistoric inhabitants of the Isthmus region, including harbour and grey seals from late spring to fall, caribou, salmon and trout, capelin, and sea birds and their eggs. Omitted from this list are harp seals, which were likely absent or only sporadically present in small numbers. As well, ringed seals do not usually range as far south as the Isthmus region, yet their presence during prehistoric times cannot be entirely discounted. The frequency of harpoon end blades at sites in Trinity and Placentia Bays and the specific locations of the sites are suggestive of sea mammal hunting, yet in the absence of harp seals hunting activities must have necessarily differed.

The presence of sites, including at least two large ones, in this region may be explained by the fact that two resources having large biomasses -- harbour seals and caribou -- were present within close proximity of one another. Coastal sites where harbour seals could be taken are only a few kilometres from interior locations where caribou were available. This situation is changed today, yet several sources of information suggest that such was the case in the past. The presence of harbour seals in Placentia and Trinity Bay is documented (Boulva and McLean 1979). Historical records show Beothuks were hunting caribou in this area during the early 1600's (Howley 1974:15), and a Fox Island, Placentia Bay, resident tells of caribou hunting in the inner Placentia Bay area within his lifetime. Unlike northern and western regions of the island, where coastal seal hunting and interior caribou hunting (if in fact the Dorset Eskimo occupants of these regions were hunting caribou at all) required a major geographic shift of camp locations, sealing and caribou hunting in the Isthmus region may have been done from the same coastal camps locations. Marine mammal resources and interior caribou herds are separated by only short distances, and the nature of the terrain makes a shift from coast to interior relatively easy.

At least two and possibly three large Dorset sites are located near the Isthmus, along with several small ones. All the small sites are located on or adjacent to sandbars in sheltered places, locations which appear highly favourable harbour seal hunting sites. The harbour seal prefers such locations during the summer and fall, when it hauls out on shore to bask in the sun. The large sites in the Isthmus region, Stock Cove, New Grove, and Dildo Island are located in small coves with extensive beachfronts, all in relatively sheltered inner bay locations. These locations may have also been favoured harbour seal habitats, as well as central places from which other sites could be reached. Hunting and butchering equipment (harpoon end blades, microblades) is present in collections from all Isthmus of Avalon sites. Maintenance tools are variably represented.

Frenchmen's Island has produced an assemblage which, with respect to the frequencies of major artifact classes, is similar to assemblages from most western and northern Newfoundland sites. Harpoon end blades account for 16.9%, end scrapers 10.5%, and microblades 26.1%. The number of end blades and microblades suggests an emphasis on hunting and butchering, yet the frequency of end scrapers, together with the presence of several types of biface and uniface knives, blade cores, preforms, tip-fluting flakes, adzes, and a burin-like-tool argue for maintenance activities as well. As such, Frenchmen's Island cannot be considered a specialized procurement station after the pattern of Shambler's Cove. Rather, it gives the impression of a small community, established for sealing and perhaps caribou hunting, that was occupied for a major part of the year by whole family groups. Its location is an ideal departure point for excursions into the interior and across the Isthmus to Placentia Bay.

Stock Cove has produced a much larger collection, but again there is a predominance of hunting and butchering equipment in the form of end blades and microblades. The extent of the Stock Cove site and the presence of substantial structural remains, however, argue against it being only a temporary sealing station. The occurrence of many tools and artifacts associated with maintenance (various biface and uniface knives, slate knives, end scrapers, preforms, hammerstones, burin-like-tools, soapstone vessel fragments, much lithic debitage) supports this conclusion, yet the relative frequencies of these artifact types are not as great as expected. Most notable is the infrequent occurrence of end scrapers. It should be recalled, however, that while numerous artifacts were recovered from the Stock Cove excavation, the excavation itself included only 1-2% of the site. It seems highly probable that this sample is not representative of all activities which occurred at Stock Cove, and other areas exist within the Stock Cove site where maintenance activities are more heavily represented. The 1981 work at Stock Cove may have, in effect explored a site within a site, i.e., a single summer/fall house structure occupied during the harbour sealing season. Unless the Dorset occupants performed all their activities inside the structure (where our excavations were concentrated), the full range of maintenance activities would not be represented. It appears unlikely that all activities would be done indoors during the season presenting the best weather conditions.

Along both shores of the Isthmus, it appears that Dorset people were pursuing harbour seals and possibly grey seals during the ice-free months of late spring to fall. Hunting may have shifted from one site to the next as local populations were temporarily depleted or dispersed, while large sites such as Stock Cove and Dildo Island were maintained by resident populations. In this manner Stock Cove was a central base, strategically located with respect to surrounding sealing stations, and at some times a hunting station itself. Stock Cove may have also been a base from which hunting bands travelled into the interior uplands of the Isthmus in pursuit of caribou during their fall migration, or during the winter when herds may have congregated in favourable locations on the Avalon uplands. Frenchmen's Island, through its location, may have also provided access to caribou herds on the uplands, as well as to Placentia Bay, which is visible from the high ground behind Frenchmen's Island.

The nature of food resources in the vicinity of the Isthmus apparently allowed the Dorset population to gather for at least part of the year in fairly large groups. Unlike the west and the northeast, however, this occurrence cannot be attributed to the presence of a single abundant food resource -- i.e., migratory seals. Rather, it seems that a combination of resources in a restricted geographic are -- notably non-migratory seals and caribou, and possibly salmon -- permitted the establishment of large semi-permanent sites, which were strategically located with respect to other sites, and were sometimes themselves the focus of subsistence activities. It is expected that an examination of the many small sites in the vicinity of the Isthmus will demonstrate that several were specialized procurement stations, primarily for harbour sealing and salmon fishing.

Four sites in the Isthmus region have been radiocarbon dated.

Frenchmen's Island:

1870±180 B.P. (Beta-2142) (Evans 1982)

New Grove:

1730±80 B.P. (GAK-1905) (Linnamae n.d.)

Bordeaux II:

1090±80 B.P. (GAK-1904) (Linnamae n.d.)

Stock Cove:

1560±60 B.P. (Beta-4064) (Robbins, this volume)
1280±60 B.P. (Beta-4062) (Robbins, this volume)
1280±60 B.P. (Beta-4065) (Robbins, this volume)

In absolute figures, these dates indicate that the Isthmus region witnessed the lengthiest occupation by Dorset Eskimos, from ca. A.D. 80 to A.D. 860. Occupation may or may not have been continuous throughout this period. Gaps in the chronology currently exist which might denote periods when the region was abandoned, yet these might be filled through additional radiocarbon dating. Urve Linnamae originally rejected the Bordeaux II date, on the basis it was too recent in comparison to the New Grove date, yet now the Stock Cove dates partly fill this 600-700 year hiatus. Consequently, the recent date from Bordeaux II might now be accepted.

The south coast of the Island -- Bay D'Espoir west to Cape Ray -- received little archaeological attention until quite recently. A series of surveys by Gerald Penney during the 1979-1981 seasons and an excavation of one site in 1980 have shown that southern Newfoundland was repeatedly occupied by prehistoric peoples, including Maritime Archaic Indians, Early and Late Palaeo-Eskimos, and, subsequently, an Recent Indian population termed the "Little Passage complex" (Penney n.d.c, 1981). The 1979 survey in Bay d'Espoir-Hermitage Bay identified Dorset Eskimo occupations at three coastal locations: Isle Galet, l'Anse … Flamme, and Eagle Head. A single harpoon end blade found in disturbed context at Branis Point, and three Dorset tools recovered from the eroded Copper Head site also indicate the presence of Dorset people (Penney n.d.b). Survey work to the west of Bay d'Espoir in 1980 and 1981 located several Dorset Eskimo sites in the Burgeo-Ramea area, at Brimball Store- house, Bay de Vieux, Island Cove, Sot's Hole, Sandbanks Island, Cuttail Island, Morgan Island and, farther to the east, Cape La Hune (Penney 1982).

L'ANSE A FLAMME

Excavation at the l'Anse … Flamme site in 1980 produced a collection of more than 700 Dorset artifacts in addition to cultural remains relating to Maritime Archaic, Early Palaeo-Eskimo and Little Passage occupations. Two seemingly discrete "components" are present in the Dorset Eskimo collection, distinguished according to the type of raw material used for making stone tools. One assemblage is manufactured from patinated chert -- as seen in the Isthmus of Avalon collections -- and contains tip-fluted and ground harpoon end blades, bifaces, ground slate projectiles, and crystal quartz microblades and end scrapers. The second assemblage is numerically larger and manufactured from colourful cherts, but comprises similar tool types (Penney 1981:99, pl.II). The relative frequencies of artifact types in both assemblages have yet to be determined.

ISLE GALET, EAGLE HEAD, COPPER HEAD

Approximately 80 Dorset Eskimo tools were recovered from Isle Galet. This assemblage includes harpoon end blades, some of which are tip-fluted and ground, microblades, tip- flute flakes, and bifaces, all made of patinated white chert (Penney n.d.b). Five harpoon end blades, two quartz end scrapers, tip-flute flakes, and a possible Dorset burin-like-tool were collected at Eagle Head (ibid). At Copper Head a ground slate tool and a quartz crystal micro- blade were found, along with a side-notched biface probably of Dorset origin (ibid).

BURGEO-RAMEA

Collections from sites in the Burgeo-Ramea area are very small, usually including only a few specimens, as these site were visited only briefly. An inventory of collected artifacts includes end blades, microblades, bifaces, tip- flute flakes, end scrapers, blade cores, and preforms. Patinated white chert occurs, along with green cherts, and black chert visually similar to Cow Head material.

DISCUSSION

Archaeological surveys along the south coast of the Island indicate the presence of numerous Dorset Eskimo sites, possibly extending along the entire coast as far west as Cape Ray. Most of these sites appear to encompass small areas and be relatively poor in cultural remains when compared to other habitation sites, especially those along the west coast. If the small samples available from these sites are representative, however, it appears that major artifact types are represented at most site -- including harpoon end blades, bifaces, microblades, and scrapers. Many of the end blades are similar in outline to those recovered from Bonavista Bay and the Isthmus of Avalon regions, and have the grinding popular in the latter location. The single most popular raw material type is patinated white chert, as found in the Isthmus region. Microblades and end scrapers fashioned from quartz crystal are also similar to those found around the Isthmus.

The preliminary nature of Dorset archaeology along Newfoundland's south coast impedes the formulation of settlement/subsistence hypotheses, and until further excavations are done any scheme remains tentative.

Migratory seals were and are now unavailable along the south coast. Harbour and grey seals are present, along with large herds of caribou in the hinterland. Salmon and trout are abundant in the many streams of Bay d'Espoir and Hermitage Bay, and at several location further to the west. In general, the resource base appears essentially similar to that of the Isthmus region. The recent surveys of the south coast have resulted in the discovery of many Dorset Eskimo sites. None, however, compares in size and richness to Port au Choix-2, Dark Hole, or Stock Cove, suggesting that the nature of food resources along the south coast did not permit the gathering together of large bands. A logical conclusion is, therefore, that there must have been a basic difference with respect to resources between the south coast and western, northeastern, and Isthmus of Avalon regions. In comparison with western and northeastern regions, this differences was likely the absence of migratory seals in the south. The difference between the south coast and the Isthmus of Avalon region might be tentatively identified as the concentration of resources, in that marine resources along the south coast were sufficiently distant from land resources (caribou) to make it impossible to exploit both from the same coastal location. Marine resources -- non-migratory sea mammals, and fish -- were in themselves insufficient to permit numerous family groups to congregate at large coastal camps. Instead it seems that a chain of relatively small coastal sites was established and populated by small bands during seasons when sea mammals were available. Caribou hunting probably required a settlement shift to inland locations, and a temporary abandonment of coastal sites. Excavations at south coast Dorset sites will in the future add to our understanding of subsistence on this coast, if research includes a comparative study of sites and assemblages. The interior plateau of southern Newfoundland, populated during winter by large caribou herds, seems to offer the greatest potential interior region of Newfoundland for Dorset Eskimo sites, yet much survey work remains to be done if sites are to be located, or their absence proved.

Two radiocarbon dates are available for sites in the Bay d'Espoir-Hermitage Bay area (Penney n.d.b):

Isle Galet:

1345±115 B.P. (Teledyne Isotopes 1-11, 076)

Eagle Head:

1660±85 B.P. (Teledyne Isotopes 1-11, 075)

These dates suggest that the south coast was occupied during the time of extensive western and northern Newfoundland occupations. The Isle Galet date may or may not indicate a recent (post-1300 B.P.) occupation similar to that seen in the Isthmus of Avalon region.

Since Dorset Eskimo sites were first recognized on the Island of Newfoundland in the 1920's, only a few years after the existence of that culture was first postulated (Jenness 1925), many sites have been discovered along the Island's coast. Researchers were for a time preoccupied with comparing the Newfoundland representation of Dorset Eskimo culture with the better-known Arctic occurrences. It was an unspoken assumption that Dorset sites on the Island constituted a "Newfoundland manifestation" of the eastern Arctic Dorset Eskimo culture; in other words, it was possible to conduct comparisons of Newfoundland Dorset with Arctic Dorset by examining one or two sites on the Island. These, it was assumed, would be representative of the whole.

In the last decade or so Dorset Eskimo research has expanded to northeastern, southeastern, and southern parts of Newfoundland and with a broader data base it has become apparent that this assumption is untenable. Sites vary with respect to such factors as area, depth of cultural deposition, assemblage composition, specific location, and resource potential. A single site cannot be taken as representative of the entire Dorset Eskimo occupation of the Island.

Based only on ecological data it seems reasonable to conclude that Dorset Eskimo subsistence, and concurrently settlement, in western and northern Newfoundland must have differed from southern regions due to the absence of migratory seals in the latter locale. Furthermore, the mutual occurrence of migratory and non-migratory species of sea mammals must have presented different -- and perhaps better -- regional opportunities than did the occurrence of only one or the other sea mammal resource. The joint availability of sea mammals and caribou likely worked in a similar fashion. Subsistence and settlement must have been necessarily adapted to these situations. An examination of Dorset Eskimo sites and assemblages in Newfoundland reveals regional differences which appear to co-vary, to some extent, with the resource base. The "regional expressions" of Dorset Eskimo culture proposed in the next chapter are, therefore, rooted in ecology. While "environmental determinism" may be unacceptable to some readers, it appears inescapable at present with respect to the Dorset Eskimo occupation of Newfoundland.