In 1977, a Dorset Eskimo site on Shuldham Island in northern Labrador was found by a biologist intent on examining the flora of the richly-vegetated terrace on which the site stood. Upon noting several housepit-like depressions, he alerted his colleagues, archaeologists with the Smithsonian Institution/Bryn Mawr College Torngat Archaeological Project, and the site was tested. Evidence was found of an occupation that extended the known Dorset culture history of the region by several hundred years and suggested that this outer part of the coast was an important exploitation zone during the Middle and Late Dorset periods. In 1978, a short return visit to the site resulted in the excavation of three miniature soapstone figurines (see Fig. 4) associated with Late Dorset stone tools. During the summers of 1980 to 1982 and with the generous support of members of the Torngat Archaeological Project, I investigated two of the winter houses at this site, “Shuldham Island 9.”
At times during the past 30 hours I had wondered whether we would be here on Shuldham Island tonight. Late yesterday afternoon a representative from Petro-Canada, the company conducting exploratory work from drill rigs on the Labrador Shelf, called to inform us that there was room for our small archaeology crew on this morning’s flight from St. John’s, Newfoundland, to Saglek Bay in northern Labrador. After a hectic night of final purchasing and packing, we hauled all our gear out to the airport and by 6:30 a. m. the Convair was climbing out of St. John’s en route to Goose Bay, where we stopped briefly to pick up the rest of our five-person crew.
As we flew over the thick boreal forests of interior central Labrador, we could see rapid-strewn rivers tumbling down from the interior plateau. Around Hopedale, a small settlement on the central coast, the forests thinned into a transitional forest-tundra zone which graded into open tundra where isolated stands of stunted spruce grew in sheltered valleys. Passing Nain, the northernmost settlement, we could see that summer was coming slowly to northern Labrador. Streams of drift ice were flowing down on the Labrador Current, and several giant icebergs could be seen off to the east. To the north the coastal terrain became more mountainous. Peaks in the Kiglapait and Kaumajet ranges were still snow-covered and the Torngat mountains, dominating the extreme northern coast, could, be seen looming in the distance. Steep-sided fjords cut deeply into the interior. As we approached Saglek, we could see that the bay was full of floating ice (Fig. 3). However, there were navigable lanes among the icepans, the beach below the runway was clear and, looking across to Shuldham Island, it seemed that the cove where we were to work and camp was fairly free of ice. The plane landed and rolled smoothly up the runway first built for a DEW (Distant Early Warning) line station back in the ’50s.
We were anxious to set up our camp on Shuldham Island while the weather held and were able to talk the Petro-Canada station manager into supplying us with some outboard motor gas. He then kindly ran our gear down to the beach in the base truck. As we worked to get our inflatable boat set up, two helicopters droned off with crew changes to the rigs, stationed beyond the horizon in the Labrador Sea. When we had the boat ready, we put a small first load of gear and all the crew aboard and zig-zagged off through the ice to Shuldham Island. By late evening we had transferred all our supplies, the sleeping tents and cooktent were erected near a stream in the lee of a low rise, and food, kitchen gear, and digging equipment were stored away. Sleep came easily that first night, to the soft sounds of a calm sea, the stream burbling away down to the beach, and a pair of loons on the pond indignantly complaining about the intruders.
This morning, the first order of business after breakfast was to make sure that everyone was familiar with the handling of our rifle and shotgun in case we were troubled by polar bears. We were well aware of a denning area on the north side of Saglek Bay and, although we thought it unlikely that any would remain this far south through the summer, experience in previous years with bears coming uncomfortably close to camp made us properly respectful.
We began work on Shuldham Island 9 after a tour of the site. The Middle and Late Dorset Eskimo groups who had lived in Saglek Bay between 1500 and 5(X) years ago had chosen this site well. A circle of small hills behind the terrace ensured shelter from the north, the arms of the cove deflected the worst of the storms coming in off the Labrador Sea, nearby prominences afforded good views for spotting game, the pond at the back of the terrace had a steady supply of fresh water, and the terrace and beach offered ample building materials in the form of deep sod and rocks (Fig. 1).
Our first glimpse of Shuldham Island 9 was much as the members of the Torngat Archaeological Project had described it. Seven shallow depressions in the moss and lichen covered terrace were easy to spot by the darker, more luxuriant vegetation growing inside them. These features were the remains of semi-subterranean houses dug 1 to 2 feet into the terrace, with walls of stone and sod raised by a similar amount. Each structure would have been roofed with a framework of bone and wood covered with skins, sod, and flat stone slabs. These houses might have been built in fall, before the ground froze, to be occupied during the winter.
Stone tent rings on the beach below the terrace indicated that this had remained a popular occupation site into the succeeding Thule and/ or Labrador Inuit periods, although during a different season. The tent rings were the remains of summer dwellings, simple skin tents over wood or bone frames with the lower edges and guys held clown by a circle of substantial boulders. Like the sod houses, tent rings were occupied by one or two small families living and hunting together. A number of large stone meat caches around the site suggested that subsistence activities had resulted in surpluses, which were stored for future use beneath piles of heavy stones to discourage pilfering by polar bears or foxes.
Once we had isolated the houses I wanted to investigate, we set up a grid system of string over the structures to control measurement of objects and features found below the surface. The contours of the houses were mapped, and then we proceeded to strip sod from a test trench across each house that would give us an understanding of strati-graphic sequences within the structures. Beneath the upper 3 to 6 inches of recent sod and roots we came to a black, greasy soil containing stone tools, soapstone vessel fragments, manufacturing debris, animal bones, charcoal and burned fat, and discarded or in situ structural rocks, all of which we carefully plotted on maps as we trowelled.
Based on knowledge of what had been found previously in test pits at this site and from a study of excavation results from elsewhere in Labrador, I expected to find that the depressions were the remains of winter houses occupied sequentially by Middle Dorset and Late Dorset Eskimos during the past two millennia. The present state of knowledge indicates that Dorset Eskimos had initially entered Labrador from the Eastern Arctic about 2600 years ago, gradually spreading down the Labrador coast and into the island of Newfoundland. In doing so, they replaced their predecessors, the Pre-Dorset Eskimos, who had also entered from the north about 4000 years ago, after migrating east from Alaska. In Labrador, the Pre-Dorset Eskimos had themselves replaced the incumbent Maritime Archaic Indians who had been resident since about 9000 years ago, when glacial retreat first made the coast habitable.
During the Dorset period, Point Revenge Indians, the ancestors of the Innu population which today occupies parts of the central coast and interior, moved onto and took over the central coast, separating the Newfoundland and Labrador Dorset groups. The final native cultural group to enter Labrador, the Thule Eskimos, arrived about 500 years ago and replaced the remnant Dorset population. The Inuit, descendants of the Thule, today occupy several communities as far north as Nain. Each of the cultural replacements was accompanied by and probably partially caused by climatic change. The degree and nature of interaction between overlapping cultural groups is the subject of considerable interest and one of the problems we planned to investigate.
I had come up to Saglek Bay, and specifically to the site of Shuldham Island 9, with a research design aimed at seeking evidence, currently lacking on this coast, of a stage transitional between Middle and Late Dorset which would indicate that there had been an in situ evolution of Dorset culture in Labrador, not a series of repopulations from the eastern Arctic, which was the alternative hypothesis. Because of the time frame indicated by the earlier testing of the site, we also hoped to find traces of contact between Late Dorset Eskimos and other cultural groups who had interests in the north coast: the Greenland Norse, who certainly passed right by the bay in their voyages, the Thule Eskimos, and the Point Revenge Indians, all of whom were present in northern Labrador within the first few centuries of the present millenium. A third objective was to recover more of the curious soapstone figurines which had been found during the earlier testing of the site (Fig. 4). The figurines bear some resemblance to Late Dorset ivory, antler, bone, and wood carvings found elsewhere in the Arctic, but differ in style and degree of realism as well as medium. Only Shuldham Island 9 has produced large numbers of soapstone carvings. We also planned to study the Dorset use of the outer part of the bay, which previously had not received any thorough archaeological or ecological investigation. This would be done through a program of boat and foot surveys, interpretation of site function based on ethnographic analogy and knowledge of animal seasonality, comparison with data from other areas, and analysis of faunal and artifactual material.
Our hopes of finding more of the soapstone carvings were realized within the first few hours of work at the site with the recovery of an exquisite miniature bear (this and other carvings are described together below). We stopped work early on the first day, well satisfied with our beginnings, and went off for a walk around some of the other archaeological sites at the south end of the island. The south, east, and north coasts of the three-mile long island are indented with small coves, most of which have seen occupation by at least one cultural group. The spine of the 400- to 500-foot hills running down the center of the island drops off steeply to the sea on the west side. Much of the island’s surface is composed of barren gneiss, but where soil has accumulated the ground is covered with short grasses, sedges, and moss, with berry plants and tiny flowers adding a brief splash of color during summer. In sheltered wet gullies a few stunted willow and birch shrubs have caught hold, but Saglek Bay is some 70-miles north of the tree line, so human residents of the area have to depend upon driftwood and imported timber for their needs. We found a large number of Inuit tent rings on the shore of the channel that separates Shuldham Island from Big Island, and concluded that these had most likely been occupied in spring or fall when the great harp seal migrations take place and where, ice permitting, the seals could be most easily harvested. We also noted evidence of an extensive Middle Dorset occupation, and several Early Dorset sites identified by their generally small tools, the high percentage of tiny microblades or cutting tools, and a distinctive type of harpoon-head blade. Late that night, after another cool, calm, cloudless day, we were treated to a vivid display of northern lights. Sometimes great bands of white light shimmered across the sky; at other times the bands would become infused with delicate greens, blues, and yellows, and flares would veer sinuously about the sky.
Despite some stormy weather, we have been making progress and have expanded the excavations to take in most of the apparent limits of the two house structures. House 1 appears to have been most recently occupied by a Thule family who cleared out the Late Dorset house, throwing sods, artifacts, waste material, and paving rocks onto the midden and beyond the house walls, then constructed their own semi-subterranean house on the same site. The interior of this more recent house shows certain traits that are unmistakably Thule, such as a depressed entrance passage which acts to deflect cold air away from the raised entrance step, an elevated rear sleeping platform, a lamp stand inside the door over which a soapstone cooking pot would be suspended, a small paved alcove off the entrance passage, and a few scraps of baleen (the plates which hang from a baleen whale’s upper jaw and act to sift out krill and plankton). It is possible that this small Thule house, measuring about 10 feet in interior diameter, was roofed with rafters of whalebone, although little of this material was preserved.
We discussed the possibility that this might have been a Dorset copy of a Thule house, or an occupation by a Thule male and a Dorset female, as we found a couple of Thule drilled slate endblades and two Dorset lamp support slabs on the floor, and the architecture indicated an occupation date of around A. D. 1400-1500, at which time both of these unrelated cultural groups were vying for control of the coast and its resources. The evidence, however, was not strong enough to positively support our contentions. In front of the house, towards the edge of the terrace which sloped down a 12 foot drop to the active beach, we found a few other Thule tool and utensil fragments as well as the remains of a paved floor which apparently is all that is left of the Late Dorset structure.
House 2, 30 feet to the north, was producing a clear, paved Late Dorset house floor, with a mid-passage bisecting two side platforms. The mid-passage, about 15 feet long, was set perpendicular to the shore and bordered with heavy rectangular rocks to demarcate clearly the working and sleeping side platforms, and the central area in which cooking activities were confined. The walls were composed of piled rocks and organic material. At the back, away from the sea, we found rock slabs that had fallen into the house from the walls where they had evidently been set to elevate rafters of bone or wood.
This past week we had our first good storm. During the day the wind had gradually risen, coming out of the northwest. While our site remained quite sheltered, the strait between Shuldham Island and the mainland was filled with whitecaps and blowing spume, and surf was climbing high up the cliffs on the south side of the bay. As we returned to camp at the end of the day and reached the top of the rise above the site, the full force of the wind hit us and bowled us down the slope into camp. The tents were standing up well, but we replaced and tightened several guy ropes and piled additional rocks on the tent pegs. Early in the morning, the wind shifted around to the northeast and brought a cold miserable rain as well as more severe gusts. I dragged myself reluctantly out of a warm sleeping bag, wondering why these gales seemed always to reach their screaming crescendo in the middle of the night, and trudged off down to the site to check on the boat. It was riding easily, well sheltered in its little cove below the Shuldham Island 9 site. As I got back to camp, I found one of the crew out fighting with the fly sheet of the cooktent which had come adrift. We struggled to put it back in place but the wind ripped ropes out of our numb hands, sent pegs spinning off into the darkness, and tied impossible knots in the guys. Eventually, we gave in, removed the tattered fly, shifted the perishable contents of food boxes away from the walls of the tent, which were now beginning to leak, and headed back to our warm, dry beds.
By daylight, the wind had eased, but the rain remained steady and cold. We spent the morning removing sod from some new squares, then set off after lunch to survey the north shore of the island. Here we found the surf crashing up on the gravel beaches, sending spray far in land. This side of the island had also been well occupied prehistorically. We found a great many tent rings, hearths and caches, and three sod houses. The sites were well situated for access to winter and spring sealing on the sea ice and at the ice edge.
The following day was calm and sunny, although a swell continued to push great bursts of spray high up the cliffs. About midday a regular visitor to the site, a large male caribou, came trotting by. Evidently the flies were tormenting him terribly, as every now and then he would stop, quiver violently, jerk his legs, then go trotting blindly on. At one point he went berserk, charged madly along the beach below the site, plunged into the cove, and swam the quarter mile across to the opposite side. We had seen about 30 caribou on the outer islands, which they cross over to on the late spring ice in their annual effort to evade the worst of the flies and larger predators, and had counted a herd of about 500 one day on the mainland. As these animals were vital to all Labrador residents for their meat, skins for clothing and tent covers, antler and bone for tools and utensils, and back sinew for thread, we wondered whether the numbers that came out to the coast regularly provided sufficient for the needs of a local Indian or Eskimo band, or whether in some years a trek had to be made into the interior to intercept the migrating herds.
This morning was cold and foggy and all of us were huddled down low in our excavation squares when we heard a swooshing sigh in the cove and turned in time to see the finned back of a minke whale disappearing beneath the waves. It rose twice more, expelling a light jet of moisture-laden breath and inhaling, then cruised of out of the cove in search of krill or a shoal of char. At one point it had been no more than 150 feet from shore, and we pictured the excitement such a sight would have caused in a Dorset or Thule community. Despite the usual characterization of them as a whaling people, it seemed to us unlikely that the early Thule residents of Saglek Bay did much sitting around waiting for a whale to appear. They would probably locate themselves in areas where they could actively hunt more abundant, predictable resources, and take advantage of fortuitous catches such as a whale when such an opportunity presented itself. The Dorset did not possess the technology or the necessary manpower for whale hunting, so may not have been able to do much more than watch them pass by, as we did, and scavenge bones, baleen, and perhaps meat from beached whales.
As excavation proceeded, we began to recover a good sample of faunal material. Preliminary field analysis of these bones indicated the presence of young as well as adult seals and walrus. This kind of population is present in late winter and spring at the ice edge. Our analysis supported the belief that, like other northern Labrador Dorest winter sites on the outer coast, Shuldham Island 9’s location had been chosen for its proximity to these resources. Although the ice edge is a dangerous place from which to bunt because of the tendency for sudden storms or a rising swell to break up the ice, sea mammals are most highly concentrated at this location and can be taken with harpoons and lines there.
Today, we recovered artifacts related to the question of Dorset Eskimo and Point Revenge Indian contact. Most of the chipped stone tools that we were finding were made of Ramesh chert, a translucent, coarse-grained material which is found only in a narrow bed some 30 miles long just to the north of Saglek Bay. We were aware that the central Labrador coast Point Revenge Indians favored this material, but we were not sure whether they came north to obtain their own supplies or traded for it with Dorset Eskimos. Therefore we were greatly excited when we found two small corner-notched points quite clearly of Point Revenge Indian manufacture (Fig. 5). The finds suggested that the Indians came north to the quarry zone where they extracted the stone themselves, which would have entailed making some economic arrangements with the. resident Dorset population.
By the end of July we had completely revealed the outlines of the main structures in Houses 1 and 2 and decided to leave them intact for future investigation of the levels below. Most of the diagnostic stone tools that we had recovered from the two excavations were from the Middle Dorset phase and seemed to date to around 1600 to 1700 years ago, but no structural remains had been assigned to this period. We had also found a few specimens from the Early Dorset phase, and several triangular points that demonstrated a combination of technological and typological attributes from both Middle and Late Dorset, thus providing at least a hint of the in situ transitional phase we were seeking. Our small diagnostic Late Dorset collection suggests a fairly short occupation, although many of the undiagnostic tools would also have been of Late Dorset manufacture. It seemed evident that there had been about 2000 years of discontinuous Dorset reuse of this site. The Thule and Inuit structures extended the span of occupation by another several hundred years.
Many evenings after work at the main site had been taken up with short boat or foot surveys of other parts of the outer bay, and we had found evidence of Maritime Archaic Indian inhabitations dated to about 600() to 5000 years ago, and a large Pre-Dorset Eskimo site that appeared to overlap in time with other Maritime Archaic sites dated to about 4500 to 3800 years ago. These discoveries raised again the question of contact between these two distinct ethnic groups who seemingly shared the north coast and its resources for several hundred years. We also located a number of large Inuit tent ring sites and some more Dorset remains. The resources in this exposed outer section of Saglek Bay have been of great importance to residents over several thousand years.
During the last two weeks, we completed our excavations, and photographed and drew the two structures revealed. We also dug some test pits on the terrace between the houses and the pond and found that the cultural material extended some 100 feet back from the shore and reached a depth of up to 3 feet. A sketch map was drawn of the whole site, including surface features, elevations, and contours. When this was finished, we got out the long-handled shovels and began the most unpleasant job of the season—replacing all the dirt we had removed over the past seven weeks. This was done to tidy up the site and leave it as much as possible as we had found it, and also to provide a layer of insulation for the organic material which remained below the deepest levels that we had excavated. When this backbreaking job was complete, we packed up the camp, loaded our boat for the trip back across the bay, and returned to the Petro-Canada base to catch our flight home. It had been a good summer. We accomplished most of what we set out to do and, through our survey work, had posed many more questions that would require answers through future field projects.
In both structures that we excavated at Shuldham Island 9, we found on the exposed house floors and more particularly in the wall and midden areas a curious assemblage of soapstone carvings. They were distributed amongst the food bone, stone tools, soapstone pot and lamp fragments and manufacturing remains, discarded structural elements, and chipping debris from tool-making and re-sharpening. The fact that the carvings had been discarded, often in a complete, finished state, has in the past been interpreted to mean that their spiritual qualities had been exhausted and that they no longer functioned efficiently in their intended capacity. While this may have been true at Shuldham Island 9, we were also finding many complete stone tools in the same contexts. Unless their users thought that these too had outlived their usefulness, it seemed as likely to us that either (1) these tools and carvings were for use at this particular location and season, (2) were left for retrieval at a future date and then thrown out by new occupants, (3) were unintentionally lost while still functional and had ended up in the midden or wall and roof sods, or (4) had been abandoned in a sudden departure.
On our very first day of excavation, one of our crew members suddenly sat back from his square, cupping a small object in his hands. He turned it over and over, grinning delightedly, then walked slowly down to the water’s edge and cleaned the find before sharing it with the rest of us, who were by now quite intrigued. It proved to be an exquisite rendering of a seated adult polar bear about 2 inches tall, carved from a piece of soapstone that must have been deliberately chosen by the artist for its colors (Fig. 6). The midline up the back, over the head and down through the chest divided a dark green side from a lighter shade. Polar bears, universally revered for their stealth, strength, and prowess as hunters, are the only predators of man in the Arctic and as a consequence are feared and yet admired, a dualism perhaps intended in the colors of this piece. Because of its unusual realism, we had initial doubts about the carving’s Dorset origin. Dorset bear figurines found elsewhere generally tend to be either conventionalized abstractions featuring skeletal motifs or lifelike renderings often in a pose suggesting swimming or flight, with a gouged slot in the throat or abdomen connoting ritual killing or a life line. The striking resemblance between a skinned bear, or a standing polar bear, and a human acted to promote a concept of man-bear transformation among contemporary Inuit groups. This belief perhaps was also extant in the Dorset period. Our bear, however, lacks these features and is instead a superbly sculptured rendition of an animal that was held in high regard and carefully studied by an artist. An additional unusual aspect of this carving is that it stands, or rather sits, perfectly on its four legs and rear. It was not designed for suspension around the neck or affixment to a person’s apparel. This miniature bear radiates a sense of knowledge of its own pre-eminence. It may have sat in a special location in the carver’s house or been carried in a pouch, where its power and monu mentality could he seen and felt.
Later we found two more tiny soapstone carvings that seemed to answer our questions as to whether or not shamanism was being practiced at this site. There has always been debate among students of Dorset art as to who carved the miniature figures found in many Dorset Eskimo sites, and what purpose(s) they served. Some scholars feel that any hunter had the technical ability to make himself amulets for protection, for increased hunting luck, or from which to derive some of the qualities of the being represented. Others maintain that carvings of human and animal figures and many of the other art objects found are so foreign to the standard tool-kit that they were almost certainly made by specialists, the shaman/artists who possessed both special skills and a latitude for departure from mundane thinking, action, and activities. A more balanced interpretation may be that the regulations and requirements concerning communication with helping spirits changed through time. In periods of cultural or environmental stress a shaman might be much more important and more jealous of his or her position and scope of activities that when life was proceeding smoothly. These two carvings we found clearly indicated that a shaman had been at work, and speculation ran riot as to the meaning of the objects.
One carving was half of a maskette, portraying a triangular human face with incised tattoo marks on the cheek and beside the eye, and a wide open mouth. In the historic period, such tattoos were applied only to women. While the maskette seemed too small to have had much direct visual impact upon an audience, it may have been used in a ritual performance in which light was directed through the orifices and a larger shadow cast upon a wall or screen. The second piece was a vividly realistic carving of a human skull (Fig. 7), with savagely incised lines on the back, and deeply cut features including eye sockets, nasal cavity, and teeth on the front. This object held great menace. A shaman may have carved it to represent a person or a group held in fear or loathing; ritual killing would help to rid the community of the real or perceived threat. The threat might have been a new group such as the Thule Eskimos about whom the Dorset would certainly have heard before the Thule actually arrived in the area. Another arctic resident revered for its hunting abilities and eagerly sought as a spirit helper, the snowy owl, was also depicted by our Dorset artist. One specimen has a pose emphasizing its strength, keen eyesight, fierceness of beak and talons, and awesome invulnerability (Fig. 8).
Our carving collection grew to include four more pieces with shamanic properties, as well as two more bear depictions, an apparent cross between a bird and a fish, and a miniature egg. The egg provoked two theories: one, that it represented the rebirth of life in spring or an appeal for increased fertility, the other that the carver was demonstrating a light, humorous view of what we, from the comfort of our southern homes and laboratories, regard as a harsh, unforgiving environment. The egg, like most of the other carvings, is minute and perfectly shaped (Fig. 9).
The bird/fish was more commonly identified as a fish by crew members, in which case a char or salmon would be what was intended, but it also resembles a short-billed duck such as an eider (Fig. 10). It is carved of a soft, almost translucent green soapstone, and may have been meant to represent fish and water fowl, both of which would have been sought in spring and summer by the Dorset. Like the egg, this carving might have been made to hasten the advent of spring.
Two of the human figurines bring to mind a wide range of theories about the people represented and the purposes behind their depiction. One is in the shape of a Late Dorset triangular point hafted onto an incomplete arrow shaft or harpoon head. It is also a sculpture of a person wearing a long, hooded parka similar to those worn by Thule people (Fig. 14). An X incised into the abdomen, and the weapon portrayed, may represent an attempt to kill this person or his/her group. This piece can also be seen as a caribou hoof, a common image that would bring swiftness to the hunter and plentiful supplies of the animal, and also as a swimming seal. A second figurine is also most shamanic in content, having a deep slot in the back into which a killing sliver of some material could be inserted (Fig. 15). The attitude suggests flight or swimming, both being methods adopted by shamans during certain of their rituals in which visits were made to communicate directly with spirits such as Sedna, who resides at the bottom of the ocean and controls the release of game. The high collar at the back of this figure’s head is distinctive of Dorset garments portrayed elsewhere in the Arctic.
Three more bears were recovered from House 1, one a seated polar bear cub in a child-like pose with its forepaws stretched out to touch its toes (Fig. 16). The most puzzling discoveries were of one complete and several fragments of miniature soapstone vessels with faces incised into the base or side. We had previously found some finely carved, unadorned tiny vessels which we reasoned were toys, or hunting lamps taken out onto the sea ice and held lit below the parka, or perhaps blueprints for larger vessels (Fig. 17). Interestingly, while most were plainly Dorset, some of them resembled Thule cooking vessels in shape although lacking distinctive Thule traits such as drilled holes, handles, and inward flaring walls, as though they were poorly remem bered copies. But the face-pots had us baffled until it was suggested that they may have been intended as soul catchers beneath which a parent would hide the soul of a newborn child so that an evil spirit could not capture it. Alternatively, the face may represent the soul inherent in all animate and inanimate objects.
In House 2, we found perhaps the most extraordinary carving of them all, a perfect copy of a seashell, a whelk (Fig. 1.8). We had found in our excavation a number of the small plates that seal the mouths of these whelks, so assumed that the contents were occasionally eaten, but we were astounded by the discovery of a full-size soapstone carving of one. As with the egg, the seated polar bear cub, and the caribou-like bear, this seemed to show that the carver exercised a sense of humor and a taste for whimsy in his/ her work.
While preservation of organic material at Shuldham Island 9 permitted the recovery of some 500 animal bones, many of these were large or dense and less susceptible to deterioration than smaller, finer elements. Wood, hide, ivory, antler, and other materials were not preserved, due to the pervasive influence of the salt air, and no Dorset artifacts of organic material were found. The paucity of soapstone carvings representing sea mammals at this site may indicate that carvings of these vital food resources and the embodiments of their spirits used by the hunter to invoke sympathetic hunting magic were rendered in sea mammal bone, ivory, or some other material no longer preserved. Two soapstone ca wings of walrus-like figures and a possible seal were found, however, and can be added to the four owls and three of the bears which might have been carved for enhancement of the relationship between a hunter and his helping spirit.
The two bears caused a great deal of discussion among the crew. One was identified as a polar bear or a caribou (Fig. 11). The confusion arose over the configuration of thefront and rear legs, which are shaped like those of an ungulate. The stubby tail and abdomen are also more reminiscent of a .grazing animal than a bear. The outstretched neck, prominent ears and the head are quite caribou-like, too, but the overall bulk seems to be more suggestive of a bear. A provocative addition is a small incision and sewing marks on the flank, as though one of the animals was wearing the hide of the other. The second bear (Fig. 12) is most evocative: it portrays an adult that is old, wounded, ill or pregnant, or perhaps no longer able to withstand tile hunter’s superior power and technology and therefore representative of a human foe.
Three soapstone carvings of people and some clusters of human faces were also recovered from the midden (Fig. 13). The face cluster is a motif which is found in Dorset and on into Thule and recent Inuit contexts, usually in media such as ivory and antler, and is the subject of considerable debate. Hypotheses as to its functions include casual portraits of family or community members, a tally of people helped or cured or harmed by a shaman, and the product of idle whittling; however, the grotesque, open-mouthed nature of the faces perhaps better indicates that these are connected with the many souls controlled by the carver. The deeply incised features of the Shuldham pieces are characteristically Dorset in execution.
Other carvings, such as the maskette, skull, the six human figurines, six vessel fragments with incised faces, the four face clusters, and the other three bears were surely used by a shaman in activities designed to protect the community from harm. The nature of this harm is open to speculation, but could include a cycle of bad or unseasonable weather or a climatic change to which the people were proving unable to adapt, the impending or actual threat of a competing cultural group, most likely the Thule, or perhaps a disease to which the Dorset had no immunity being transmitted from the Thule, Norse, or Indians from the south.
So far, the tally of soapstone carvings from Shuldham Island 9 seems to reflect an even distribution of hunting magic amulets between the two houses and a marked concentration of shamanic objects in House 1. If the bears are all treated as hunting amulets, then House 2 in fact has, only one article of shamanic significance. This distribution, the present uniqueness of the collection, the care with which most of the pieces were carved and treated, the use of soapstone, rarely found in other sites, and the seeming homogeneity of the soapstone used suggests that one person, perhaps a shaman/artist living in House 1, was the producer of most of these objects, dispensing amulets and other objects to the family in House 2. It remains for future investigations to determine whether there was a wider distribution of this art form in Labrador.
It has been suggested that the florescence in artistic production and general homogeneity in Dorset artwork across the eastern and central Arctic might have resulted from a common cause: stress created by environmental change and competition from a new population. About 1000 years ago the Thule Eskimos began their spread across the High Arctic on the heels of a climatic warming trend. They would have encountered Late Dorset Eskimos who were not as well adapted to cope with the changes in animal distributions and necessary technological adjustments. Some exchange of adaptive strategies between the two groups undoubtedly took place, but both the archaeological record and the oral tradition that were passed down to today’s Inuit suggest that within a few generations the Thule had replaced the Dorset population in the Arctic Islands.
By the middle of the present millennium, the climate had begun a dramatic cooling trend, forcing the Thule evacuation of the High Arctic and a consequent spread into Labrador, where they encountered a remnant Late Dorset population. If the Shuldham Island 9 soapstone carvings can be seen as part of a Dorset shaman’s paraphernalia used to deal with the environmental and cultural threat, the gross similarities in content with other Late Dorset collections can be viewed as an example of cultural continuity, while the uniqueness of the realism, the subject matter of some of the carvings, and the use of soapstone can be seen as functions of Labrador’s isolation from High Arctic Late Dorset and perhaps a higher degree of interaction between Thule and Dorset than had occurred elsewhere. The Dorset in Labrador had been quite successful in their methods of adapting to environmental pressures for some 2000 years, but, ultimately, despite what appears to have been a strong attempt in Saglek Bay to combat new climatic, ecological, and cultural pressures through shamanic devices, they succumbed, leaving their house ruins, tools, utensils, and soapstone carvings with which we attempt to piece together the pattern of their lives and deaths.