1987 Field Report
(Part 4)
by
Jean-Luc Pilon
NOGAP Archaeologist
Canadian Museum of Civilization

Revisited Archaeological Sites

Field No: 85JLP-26
Borden No: MlTj-2
Location: lac à la Truite

The oval depression first noted in 1985 was selected for excavation on the basis of two important considerations. Firstly, a small chert flake, found in 1986 had been found in a small test unit placed inside the depression where buried and well-preserved wooden remains were also noted. Secondly, the depression was located on a well-drained sand terrace, and while increasing the likelihood that good preservation of organic remains could be expected, this also meant that permafrost was less likely to pose an impediment to excavation as it had at MlTk-2 (85JLP-38).

Two axes were established over the depression which essentially quartered the feature. In this way, profiles of both the long and short axes could bedrawn. Most of the excavation units measured 2 m x 2 m. Baulks measuring 20 cm in width were left between adjacent units in order to permit the drawing of the profiles. These were removed once the stratigraphies had been recorded.

Substantial pieces of wood were preserved. Their arrangement is without question structured and must reflect their original disposition. The most noteworthy characteristic of this arrangement seems to be the close spacing of long poles parallel to the long axis of the depression. The poles were not restricted by the limits of the depression, but were also found along the side of the feature. Some pieces were found which angled away from the predominant orientation of most of the poles.  These are especially evident in the eastern portion of Unit D.  Most of those, however, were found at a slightly lower stratigraphic level than the parallel poles in Unit C and so the association remains unclear.

The discussion of the site stratigraphy must distinguish between that found within the depression proper and that observed peripheral to this major disturbance.

Prior to beginning the excavation of the depression, distinct linear mounds were visible on either side of the feature. Excavations showed that these slight elevations consisted essentially of sand deposits (Level II) which buried an older humus level (Level III). As will be shown below, the sand deposits found along the sides of the depression are considered to be integral parts of the semi-subterranean structure, and are referred to as berms. There was also an apparently unrelated, circular elevation of sand (about 2 m indiameter) located about 2 m west of the depression. Its nature remains uncertain although the sand which composed this secondary accumulation was continuous and indistinguishable from the Level II sands which constituted the berm associated with the depression.

The berm sands (Level II), especially those in Unit D, contained a few quartzite flakes and some birch-bark fragments (apparently unworked). Also found within the berm sands were the numerous wooden poles.

The buried humus (Level III) was found to be continuous with the modern sod (Level I). It occurred everywhere berm sands were found except in the southeast portion of Unit C and the southwest portion of Unit B. There, a very substantial hearth feature was documented. It consisted of fire-reddened sand with pockets of ash, calcined bone and fire-cracked rocks. The buried humus level had clearly been truncated by the hearth feature.

In addition to the hearth feature identified to the south of the depression, a second feature, also identified as the remains of a hearth, was recorded adjacent to the southeast extremity of depression. It consisted of a grey sand layer containing calcined bone, and lay above fire-reddened sands. This second hearth feature was found barely 50 cm away from the first hearth feature, but appeared distinct from it. Unfortunately, it is virtually impossible to associate the artifacts which lay between the two features with one or the other.

Within the depression it is easiest to describe the stratigraphy by beginning with the bottom of the feature. The actual floor of the depression was easily distinguished from the in situ, bedded sand deposits. In fact, the builders of the structure had excavated through various sand layers to the top of a coarse sand which we found to be considerably more compact than the overlying strata. The depression bottom was fairly flat and level, rising somewhat along the lateral margins. Sticks or branches averaging 2-3 cm indiameter had been placed parallel to the length of the depression. A 5 cm layer of organic debris in a sand matrix, which included the sticks or branches, charred wood fragments, spruce bark fragments and a small number of faunal remains, indicates the careful preparation and use of the depression floor. The faunal material includes an articulated fish fin, various fish headbones, caribou bone including a broken phalange and longbone fragments, and small mammal bones.

A thick sand level overlay the organic layer which is considered to be the occupation level. This sand was continuous with the berm sand deposits described earlier. However, within the sand layer, very distinct organic strata (at least 2) were discerned, which may correspond to major periods of surface stability. At least one such stable surface existed prior to the collapse of the pole structure since it lies well below the neat line of poles. It is interesting to note that the reddened sand layer found along the slope of the southern portion of the north-south profile, is continuous with the hearth feature found in the southwest portion of Unit B an dthe southeast portion of Unit C. There is no doubt that this sand is derived from the hearth and has since slumped down the slope; the question remains whether this occurred during use of the semi-subterranean structure or following its abandonment. This point will be taken up later.

Outside the depression, the poles were found under the sod where the berm deposits were thin or non-existent, or near the top of the berm deposits where these were thicker. Within the depression, the poles were situated in the top portion of the depression fill. In plan view the poles appear arched toward the interior of the depression. This undoubtedly resulted from the pressure exerted on the dessicated poles by the substantial quantities of sand forming the berms on either side of the depression.

The sides of the depression, which sloped up to meet the top edge of the depression, were fairly steep and easily traced as the builders had cut through the distinctive bedded sand deposits.

Artifact Distribution

The collection of artifacts, gathered in the 22 square meters which were excavated in and around the semi-subterranean feature, consists of both faunal and lithic specimens. Of special note is a small tapering stem, bifacial projectile point and a scraper manufactured on a bifacial blank. The tip of a well-made biface, possibly similar to the complete specimen was found along with a modified beaver incisor and a slender bipointed bone tool. These were not restricted to any single stratigraphic level, but were found within the modern vegetation layer, in the buried humus level, where this strata existed, and also within the redeposited sands of the berm as well as in the Level II deposits making up the small rise in Unit D.

Clear associations exist in at least two instances. These involve clusters of lithic artifacts and faunal remains in and around the hearth features documented in Unit B. A very large scatter of fire-cracked rock in the southern portions of Units D, C and B is likely related to the cooking activities which took place around the combustion features. Likewise, the distribution of lithic debitage in Level III, although by no means restricted to these areas, is most concentrated in these same zones, on the south side of the semi-subterranean feature.

To the north of the pit feature, only 8 square meters were excavated, but disproportionately less material was recovered. This may be accounted for in one of two ways. Firstly, if the activities documented on the south side of the pit feature and associated with the buried humus, Level III, took place before the excavation of the semi-subterranean structure, it is merely coincidental that the pit feature was placed near the periphery of the material scatter left by the earlier occupation. On the other hand, if these artifacts were deposited during occupation, then they may indicate a preference for certain activities on the south side of the depression.

At present this question cannot be definitely resolved. The presence of lithic artifacts in Level II of similar raw materials as those found in Level III suggests that the debris of a previous occupation had been disturbed by the construction of the semi-subterranean feature. However, these items may have become incorporated into the redeposited sands during occupation of the structure, and also during its collapse following abandonment.

A second clear association was discovered in the very bottom of the depression. This consisted of fairly closely spaced branches or sticks laid down parallel to the long axis of the depression. The area covered was no larger than approximately 2 m x 1 m. Between and immediately above these branches, the disintegrating remains of spruce bark along with spruce needles were found. Also in close association were several faunal elements including fish, small mammal and caribou remains. Clearly the bottom layer of the depression fill represents the remains of a floor upon which some occupation debris had accumulated.

Discussion

Many of the remains described above can be attributed to the semi-subterranean feature and attest to a very considerable outlay of energy on the part of those who built the structure. A reconstruction of this structure must take into account at least three principal elements. Initially a fairly deep, rectangular depression with steep, sloping sides was dug in the sandy matrix of the terrace. In the bottom of this excavation, branches were arranged parallel to the length of the hole and covererd with spruce bark and boughs.

Another important task involved cutting numerous poles to serve in the construction of a roof over the depression. Many of these measured on the order of 4 meters in length or more with diameters often exceeding 10 cm. In the immediate site vicinity, spruce tree stumps, apparently cut with stone axes or adzes attest to the probable method of felling the trees for the roof of the structure. These closely spaced poles were placed, like the branches in th ebottom of the depression, parallel to the length of the excavation.

An integral part of the structure's roof was the sand excavated from the depression. Rather than simply scattering this material on either side of the hole, it was apparently placed on top of the tightly spaced poles. Following abandonment of the site, some of this sand slowly filtered down between the poles to partially fill the depression, while the remainder slumped to the base of the roof, forming the berms. Later, following the collapse of the roof, some of the sand which had slumped to the base of the roof on the lateral margins, gradually moved down into the depression through the combined action of wind and water.

The site was briefly revisited in 1991.