Text Size

The 1996 Excavation of a Dorset Palaeoeskimo Dwelling in Cape Ray, Newfoundland: Interim Report

Lisa Mae Fog

Getting Started: Surveying the Site

Cape Ray is located on the southwest coast of Newfoundland about 10 kilometres west of Port aux Basques, one of the two gateways into the island. I left to begin my first season of fieldwork at Cape Ray in June of 1996, having received funding from the Institute of Social and Economic Research (ISER) and from the Culture and Heritage Division of the provincial government. My research assistant and I arrived on the southwest coast just in time for caplin weather. We were greeted with terrible winds, rain, fog, and the constant blow of the nearby fog horn.

1996 Field Objectives

We immediately set out to accomplish our three fieldwork objectives:

  • To find Linnamae's (1975) primary datum and excavation areas.
  • To begin a systematic survey of the site to determine the limits and nature of the prehistoric occupation.
  • To locate and excavate a house structure.

The first thing I wanted to establish was where Linnamae had dug almost thirty years ago. Finding her datum point would be ideal, but we soon realized that the maps which were provided to us by Linnamae were difficult to interpret. Eventually, I was able to pinpoint the area where her datum point could be found from comparisons between an aerial photograph taken in the early 1990s, and her survey and excavation map. We found the datum point three days after we began our search. This helped to clarify where her three excavation areas had been.

We were now ready to survey the site. Initially, I had planned a systematic survey of test-pitting at five and ten metre intervals. However, this type of survey is impossible at Cape Ray due to the extensive growth of tuckamore on the site. A considerable amount of time was spent clearing the tuckamore for test pits. This made a systematic survey inappropriate given time and labour constraints. Instead we surveyed the site using random methods, which usually meant that we placed test pits in areas free of tuckamore in addition to placing some test pits in the tuckamore.


It took us three weeks to survey the site, which was accomplished by digging test pits measuring 50 cm2 each, and by general surface reconnaissance. By the end of these three weeks we had dug 72 test pits, with 37 of them yielding a cultural layer.

Test-pitting enabled me to get a good idea of the limits of occupation at the site. There were some surprises. For instance, I did not expect to find a cultural layer in the tuckamore, and yet five test pits came out positive. This means that in some places, the tuckamore has extended onto the areas of occupation of palaeoeskimo groups. Also, the site was not as large as I had expected it to be. In her thesis, Linnamae had generalized about the evidence for occupation at the site based on the limited testing of seven areas. Though never explicitly stated, it seemed from her 1975 monograph that the site encompassed a region with an area of approximately 30 000 m2. However, our survey has significantly modified the known extent of the occupation at Cape Ray. Based on this past summer's survey the site is contained within an area of approximately 7500 m2.

Test pits indicated that the greatest concentration of artifacts, and the most consistent presence of an occupation layer was found along the ridges between 20 and 30 m above sea level. This region extends approximately 168 m along the coastline and encompasses Areas F, E, and C, excavated by Linnamae, and Area C excavated by our crew. Small traces of occupation in the form of a few artifacts were found on the Northwest boundary of the site just west of the existing road towards the shoreline, and north of the road. The last traces of occupation at the site are found to the east in the low lying tuckamore, and to the south about 20 m away from Area C. This area (south of Area C excavation) is now mostly a bog mixed with tuckamore. The area immediately to the south of Area C had, at some time before 1960, been traversed by a road leading towards an old fog horn. It is fortunate that this road did not traverse Area C where there is a heavy occupation layer, and a Dorset dwelling. To the west along the shoreline the site continues on indefinitely. Artifacts and/or an occupation layer can be found in areas which in some cases extend all the way to the rough coastline edge. Lithic debris was even found in the rugged edges of the coastline and perhaps was produced by someone sitting over the edge of the rock making tools while looking out towards the Gulf of the St. Lawrence.

Surface reconnaissance, as opposed to test-pitting, resulted in the discovery of a second site about two kilometres Southeast of Area C. This site was situated on high ground, about 20 m inland on a peninsula where a perfect view of the southern and western coastlines could be obtained. Six test pits were placed in this area but to no avail. Instead, the only indication of a site came in the presence of a handful of flakes seen resting on a dark layer in three erosion cuts. Directly beneath the dark layer was a clay subsoil. Either this site was used on a temporary basis, for instance in the case of tool manufacturing or procurement of nearby resources such as quartz, or a more substantial site has eroded away.

Following three weeks of survey we had a clear idea of the limits and the nature of occupation at the site. However, the test pits had also been dug to search for evidence in the form of features which included a dwelling, an activity area, and a midden. Logistically, with the use of small 50 cm2 test pits, identifying these features was a potentially difficult thing to do, but not impossible as we soon found out.

In Search of a Midden

Areas E and F, excavated by Linnamae, were located on a plateau of raised land surrounded by lower trench-like areas. Linnamae had informed me (personal communication, 1996) that these trenches had not been tested and that, if the raised areas had been lived on by prehistoric groups, then it seemed likely that these trenches could have been middens, made as a result of the inhabitants throwing their waste over the edges and into them. I decided to test her theory, which proved to be correct, as two test pits dug in the trench surrounding Area F produced evidence for a midden. The occupational layer in these pits was exceptionally dark, greasy, and deep. The artifacts were numerous and the majority of them were broken and of tertiary manufacture. The impression was that most of the lithics were the swept away debris from the living area above.

Testing of the trenches was only accomplished in Area F. However, the trench in Area E, which is identical to Area F in that it surrounds the higher living area plateau, could also be tested. For Area C the occupation is not found on a raised plateau and as such, there are no trenches surrounding it. However, as will be discussed later, a midden was found in a low trench-like area which was only discernible after excavation and which was adjacent to the dwelling.

In Search of a Dwelling

While I was searching for any indication of a feature such as an activity area or midden, what I really wanted to find was a dwelling. Finding evidence for a dwelling meant that I was looking in the test pits for signs of structural evidence. Since structural evidence is mostly found in the form of paved stones, I was looking specifically for this or for the presence of an organized array of stones. I found this in six test pits: two in Area F (one of which was within the tuckamore), two in Area E (one being within the tuckamore), and two in Area C. All of this evidence was very promising, for in some cases the rocks were flat and butted against each other, forming the appearance of paved stones. The evidence was particularly strong in Area C, where Test Pit 2 came down upon a series of stones, including two large flat stones pressed up to each other in a clear and undisturbed grey-black cultural layer interspersed with artifacts. There was some evidence for burning on one of the stones. The stones themselves were of different varieties and included granite, sandstone, and schist—all of which rested directly above the clay dirt subsoil. Altogether, this feature was almost certainly cultural, as the organization of the rocks was not seen in any naturally occurring state in the area.

It was time to decide on where to lay out a grid for excavation. Area C looked to be the most promising, not only because the test pits produced structural evidence, but also because a large part of it appeared to be undisturbed by potting or by Linnamae's prior investigations.

Therefore, by the end of the third week, we laid out an excavation grid in Area C in search of a palaeoeskimo dwelling.

The Excavation of Area C—The Discovery of a Dorset Palaeoeskimo Dwelling

We began excavating Area C on July 17, 1996. The caplin weather was finally receding and clear skies were becoming a common thing. In fact, when we finished excavating after four weeks, only two days had been washed out due to the rains.


Area C excavation is located immediately south of where Linnamae had excavated in the 1960s. This excavation grid, oriented towards magnetic north, was made up of one metre square units, and divided by baulks into four quadrants representing the southwest, southeast, northwest and northeast.

Units were excavated according to naturally occurring stratigraphic levels. As such, artifacts and flakes were collected from two layers: Layer 1 being the peat and sod layer, and also in some cases the disturbed cultural layer representing backdirt from Linnamae's excavation; and Layer 2 being the cultural layer. Trowels were used to excavate both layers and soil from both layers was screened on one quarter inch mesh. Artifacts, flakes, and charcoal were collected separately from both layers. Artifacts from the cultural layer were given north, east and depth co-ordinates, and charcoal samples from the cultural layer were given depth coordinates. Soil samples were taken from each layer. Other samples taken included ash and sand deposits, a bark-like layer from the midden, a leather-smelling layer from the south-east quadrant, and burned fat. Flake concentrations were water-screened and seed concentrations were floated.

The unit as a whole was described in terms of its stratigraphy, inclusions (artifacts and stones) and connections or dissociations with adjacent units. The cultural layer was described in terms of colour, texture (smooth, gritty), level of greasiness and compactness, and microtopography. Features, concentrations (flakes, artifacts, charcoal, or other) and stones were mapped by unit. And finally, the height above sea level was measured for each rock in case a three-dimensional reconstruction of this house is ever attempted.

Midden versus Dwelling

Two distinct areas were found in Area C excavation: a dwelling located on a plateau of raised ground, and a midden located in a lower wet and boggy area adjacent to the dwelling. These areas were differentiated due to differences in the following attributes: elevation, thickness and texture of cultural layer, artifact yield, organization of rocks, and overall appearance. The main difference between the two areas is that the raised area is where the dwelling was located and where the inhabitants lived, while the lower trench area is where a midden was formed when the inhabitants deposited their waste into it.

The midden was found in units that were located throughout a lower wet and trench-like region in the excavation. These units produced a thick cultural layer of loose to medium compactness which was black, greasy, and yielded a large number of artifacts. The rocks found within these units were disorganized and appeared throughout various levels of the cultural layer.

In contrast, the dwelling was located on raised dry ground. In this area, the cultural layer was thin, compact, moderately greasy, grey-black, and produced a small to medium number of artifacts. For the most part, the rocks in the dwelling were organized, rested directly above the subsoil, and appeared to be part of a feature. Relatively large open spaces that were free of rocks were also found in the dwelling. Table 1 summarizes the differences between the midden and dwelling areas:

Table 1. A Comparison of the Midden and Dwelling Areas in Area C.

Lower Area: Midden Upper Area: Dwelling
Elevation low - in bog high - dry land
Cultural Layer thick thin
very greasy moderately greasy
wet dry
black grey-black
loose to medium compactness medium to heavy compactness
Rocks disorganized organized
resting throughout cultural layer resting directly on subsoil
Artifacts large number small to medium number
Overall appearance cluttered: area is covered with rocks organized: area is divided into places that have rock features and open areas free of rocks

The nature of the occupation of Area C can be inferred using this information

Dorset groups built their dwelling on the raised plateau, or knoll, because it was a flat and dry area suitable for living on. The rocks in the dwelling comprise features directly associated with the dwelling, while the midden rocks are indirectly associated with it. Large open areas free of rocks were living areas within the dwelling.

The midden has a thick, black and greasy cultural layer with a large number of artifacts because this is where the inhabitants continually deposited their waste (bones, tools, etc), forming organic- and artifact-rich layers. The soil in this area is not as compact as in the raised area where the inhabitants lived because there is not as much compression.

As indicated, the rocks in the midden were found vertically throughout the cultural layer rather than just resting above the subsoil. I think this is because many of these rocks have been discarded over the years from continual re-occupation of the dwelling. According to Linnamae (1975:48), Palaeoeskimo groups lived at Cape Ray for a period extending over six hundred years. The large amount of artifacts retrieved from this dwelling indicates an intense occupation that could be the result of repeated occupations over many years. This particular dwelling was most likely occupied seasonally, perhaps for a maximum of twenty years (Renouf 1997, personal communication). Over the years the Dorset may have modified the construction of the dwelling, removing rocks or replacing them and throwing the broken or less desirable ones in the midden. It is also possible that hold-down rocks from the ridge above had rolled down into the midden.

Evidence for the repeated occupation of this dwelling may exist in the layout of some of the rocks in the both the midden and the raised area. These rocks have a coherent structure and appear to be associated with the dwelling. However, they are situated on a pre-existing cultural layer. Therefore, these rocks may have been part of a structure that was built after the dwelling had been occupied for a period of time. While this structure could have functioned independently, it could also have been associated with the dwelling, perhaps extending its boundary.

By the third week of excavation, almost the entire southwest quadrant was exposed and units in the southeast, northwest, and northeast quadrants were slowly being excavated. A total of six units in the southeast quadrant were excavated. These units were located directly beside what appears to have been Linnamae's earlier trench excavation of Area C. A fairly thick disturbance in the upper stratigraphy of the cultural layer was noted in these units, and was in all likelihood caused by this previous investigation.


On the whole, the stratigraphic sequence in Area C excavation was simple, as was expected from the test pit evidence. From top to bottom, the sequence was as follows: an upper peat and sod layer that varied in thickness from 1 to 30 cm, occasionally mixed in with or followed by a disturbed cultural layer which was really the displaced dirt from Linnamae's excavation close by; a cultural layer which reached its maximum thickness in the trench areas or midden; and lastly, the bottom layer which was a yellow-brown clay subsoil. While most units had a cultural layer which conformed to this sequence, there were some exceptions.

A layer resembling bark was found in the lower strata of the cultural layer in nine units throughout the midden. This bark-like layer was dark brown or black to orange-brown in colour, was approximately 5 to 10 cm thick, and had the consistency of wet bark under the trowel. I am still unsure of what this layer represents. Could it indicate that the prehistoric group used bark or wood to cover the wet mucky terrain of the midden so that they could walk along the edge of the dwelling more easily? Could this be waste of an organic nature that had been thrown into the midden? Or, could it be the remains of trees which once stood there in prehistoric times? While all of these are possibilities, I think this layer most likely represents an old peat and sod layer which, having been covered by so much organic material in the cultural layer, has now taken on a woody or bark-like appearance. In this case, the dwelling was situated on what was formerly a relatively large and flat knoll of clay, dirt, and pebbles, surrounded by peat and sod. Although these knolls, surrounded by peat, can even now be found at the site, they are uncommon. Furthermore, the clay, pebbly knolls that I came across this past summer are not nearly as large as the one where the dwelling is located. Therefore, if this situation was similar prehistorically, then it is possible that Dorset groups were limited by the availability of large flat and dry knolls which were suitable for a dwelling.

There were a few other units in the midden with an interesting stratigraphy. These units had a complex layering of soils which was either caused by disturbance, or by repeated dumping episodes. Isolated pockets of charcoal, sand, clay, and unidentified organic material (possibly fur) were found.

Two Axial Features

By the fourth week we had exposed almost an entire Dorset dwelling. Two axial features, a main and secondary one, could be seen intersecting each other in the middle of the dwelling. Axial features are typical in Dorset dwellings, and usually consist of a line of rocks upon which various activities, such as cooking and tool manufacturing, are carried out. An axial feature also delineates space, so that the inhabitants of a dwelling live in open spaces on either side of the axial feature.

The dwelling in Cape Ray is unique because it appears to contain two axial features instead of the usual individual axial feature found in Dorset dwellings. It is difficult to interpret the contemporaneity (if both were constructed at the same time as opposed serially) of the two features because both are situated directly above the subsoil, with no identifiable superposition between them. However, one axial feature, running on a northeast angle along the dwelling, is larger, contains more activity areas, and connects with a rear sleeping platform (which will be discussed later). I call this the main axial feature. The smaller axial feature will be referred to as the secondary axial feature.

Evidence for activity areas was found on the main axial feature. Soapstone shatters were found in two areas: in the centre and on the end of the feature. The centre of the axial feature also appears to be the focal point or "centrepiece" for activity within the dwelling. I refer to this area as the centrepiece because it contains so much evidence for activity: three soapstone shatters, two flake and charcoal concentrations, and a concentration of burnt rocks. Therefore, this centrepiece appears to be an area where they cooked and made tools. It is also located in the area of highest elevation within the dwelling.

The other soapstone shatter was found on the western end of the axial feature, right before the drop-off in elevation where the midden begins. Here again the shatter is in close proximity to a number of burnt rocks. The majority of the burnt rocks recorded for the entire dwelling make up this main axial feature.

The secondary axial feature is located perpendicular to the main one, and the two intersect at the centrepiece. This smaller axial feature also traverses the dwelling so that in combination both axial features appear to divide the dwelling into quadrants of relatively open space.

Open Spaces

Four open spaces, relatively free of rocks, were found within the dwelling. The two open spaces on either side of the main axial feature are fairly large and flat, and would have been suitable living areas within the dwelling. A third open space on the northwestern boundary of the dwelling could have been a living area, or used for some other purpose, like storage. The fourth open space was most likely the entrance-passage to the dwelling. It naturally slants upward and thereby acts as a perfect cold-trap. Furthermore, it faces the ocean, which is common in Dorset entrance-passages. Finally, it is marked by a large boulder in the inside of the dwelling, which is also characteristic of the entrance-passage (LeBlanc 1997, personal communication).

Rear Sleeping Platform

The rear sleeping platform is made up of over one hundred rocks. All of these rocks are lying flat on the ground rather than on their edge so that none of them protrude upwards obtrusively, or more appropriately for a sleeping platform, uncomfortably. It is a fairly level platform, and the rocks have the appearance of being paved or slightly embedded into the ground. The overall impression is that the ground beneath the platform has been hollowed out, creating a trough for these rocks to be placed on. Whether this feature was created naturally or culturally is unknown. But it does seem that the rocks were intentionally placed within this trough. If this trough was cultural, perhaps their intent was to level the floor upon which the sleeping platform of rocks would be laid.

Dwelling Boundary

One of the most difficult things to determine about this dwelling is where its boundary lies, and what the superstructure consisted of. We know that other palaeoeskimo dwellings can have a superstructure of whale ribs and other bones, animal skins and sod. Often the perimeter of a cold-climate dwelling is delineated by hold-down rocks which were used to keep the skins in place, or by a low stone wall upon which whale ribs might have rested. There is no clear indication that either of these existed in this dwelling.

Natural Bedrock Outcrops

Another note on the presence of the naturally occurring bedrock: it does not seem to be a coincidence that the dwelling, in its entirety, is situated right in between the two large slabs of bedrock. I think the decision to locate the dwelling in this area was partially influenced by the availability of this bedrock. The bedrock could have been incorporated into the construction of the dwelling, both practically and symbolically.

On a practical level, the inhabitants could have used the bedrock as a foundation or wall. The symbolic use of the bedrock is also noted because, in addition to marking the edges of the dwelling, the main axial feature runs right through two smaller outcrops of bedrock coming through the living floor.

Therefore, the location of this dwelling was probably chosen because (a) it is a raised, flat and dry area, and thus suitable for living on, and (b) the natural bedrock could be incorporated into the construction of the dwelling, both practically and symbolically. In my opinion, both points could have equal weight in the selection of this location. But, it is interesting that in the entire span of coastline along this site and beyond—about two kilometres—this is the only spot where we noticed the bedrock jutting out of the land so obviously and invitingly.

Other Features

Other features within the dwelling include one large irregular pit located in unit 513N 508 directly beside the large bedrock and on the edge of the dwelling, six smaller pits in the living floor located mainly near the sleeping platform, and a circular ring of stones directly beside the sleeping platform.

The large irregular pit on the edge of the dwelling was filled with charcoal and burnt fat, and could have been used as a storage pit. A silty greyish-green soil was also found within the pit, which could be discoloured ash.

I question why the inhabitants would have dug a pit and filled it with waste if they could have thrown it in the midden. Perhaps the pit was naturally formed and this was a way of filling it in. Or perhaps it was just an easier way of depositing the waste at the time.

The six smaller irregular pits are located near the sleeping platform. All of these pits were filled in with the regular dark cultural layer, and their function is unknown. Inhabitants could have used these pits for storage of some kind, keeping tools or other personal items in compartments that were less likely to be stepped on or disturbed as they might have been by remaining on level ground. Another possibility is that some or all of these were post-molds created by support beams, or by a tripod used to suspend soapstone while cooking.

The circular ring of stones is located directly beside the sleeping platform, and has a diameter of 70 cm. According to Renouf (1996, personal communication), circular rings like this one have been observed in Palaeoeskimo dwellings in Greenland, and were used for storage.

By the end of the fourth week we had uncovered more than enough material for analysis. The next two weeks were spent continuing the artifact processing in the basement of the interpretation centre (cleaning, labelling, cataloguing), conducting resource interviews in the nearby village of Cape Ray, mapping features, drawing profiles, and excavating the north-south baulk. Depths were taken from 62 points along the living floor so that we could later reconstruct the dwelling's topography. This served as a backup and counterpart to the unit measurements for the microtopography of the cultural layer.

761 rocks were classified in Area C excavation. For each rock we recorded its type (granite, sandstone, schist, conglomerate, slate, bedrock), width, and indicated if it had been burnt, pedestalled, stacked upon other rocks, and if it rested on an angle.

The Artifacts

The usual repertoire of the Dorset lithic assemblage was found in the Area C excavation. Almost two thousand artifacts were recovered, the majority of which were lithics. The overwhelming choice of material was chert and then quartz crystal. Chert was probably obtained (either through trade or direct recovery) from the Port au Port and Cow Head regions. Quartz crystal was probably recovered in the immediate area of Cape Ray where large veins of quartz can be found.

As is common on Dorset sites, microblades were the most prevalent artifact. This was followed by endblades, endscrapers and sidescrapers, bifaces, retouched and utilized flakes, and tip-flute spalls. A substantial amount of soapstone was also recovered. This had been anticipated given that Linnamae had also come upon large amounts of soapstone. By comparison to other Dorset sites in the High Arctic and in Newfoundland and Labrador, Cape Ray has an unusually high amount of soapstone in its assemblage. This is curious given that Fleur de Lys, the large Dorset soapstone quarry, is located over 940 km away on the northern shore of the island. Other artifacts include whetstones, abraders, incisors, and hammerstones.

Small patches of fur and hide were recovered, most of them from the wet boggy midden area. Also recovered was a fragment of a small toy soapstone vessel, and a ground slate implement, probably a needle, with binding material still lodged in a side-notch. Scanning and transmission electron microscopy was performed on this material by Carolyn Emerson from the Biology Department at Memorial University of Newfoundland. The analysis of the internal and surface structures of this material provided good evidence that it is sinew from animal tendon fibres.

A DNA sequencing analysis was attempted by Sylvia Bartlett from the Biochemistry Department at Memorial University of Newfoundland, to determine what species the tendon fibres belonged to. Unfortunately, this proved to be unsuccessful due to the inability to amplify the DNA for the complete analysis.

Organic material such as wood, fur, hide, and sinew was preserved due to the acidity of the peat which rested on top of the cultural layer. The acid would have prevented bacterial growth, which is what destroys soft tissues. On the other hand, the acid decomposed any bone remains so that we do not have a record of the rich bone technology associated with land and sea hunting (shafts, foreshafts, harpoon heads, arrows, spears and prongs, etc.), or with food processing and personal items (wicks, ulus, awls, pendants, etc.).


After nine weeks of fieldwork we had determined the spatial limits and concentration of the prehistoric occupation at Cape Ray, discovered a complete Dorset dwelling, and recovered almost two thousand artifacts from it.

The dwelling was situated on the plateau of a knoll and surrounded by a midden located in a lower trench area. Dorset groups probably selected the site of the dwelling for two reasons: it is a raised, flat and dry area which is suitable for living on; and the inhabitants were interested in the natural bedrock outcrops which could have been used, both practically and symbolically, in the construction of the dwelling.

There is evidence that the dwelling was re-occupied over many years, indicated by the large number of artifacts, the discard rocks in the midden, and the presence of rocks which have been placed on a pre-existing cultural layer.

While a lot of information concerning the dwelling can already be determined from the features within it, it now becomes a matter of expanding our interpretations with other evidence.


Linnamae, U.

1975 - "The Dorset Culture: A Comparative Study in Newfoundland and the Arctic." Technical Papers of the Newfoundland Museum, No. 1. Culture and Heritage Division, Department of Tourism, Culture and Recreation, Government of Newfoundland and Labrador, St. John's.

Last Updated:
This page and all contents are copyright, Government of Newfoundland and Labrador, all rights reserved.