1988 Field Report
(Part 3)
by
Jean-Luc Pilon
NOGAP Archaeologist
Canadian Museum of Civilization


Field No: 87JLP-21
Borden No: NbTj-3


Situated on a series of
benches cut into the southern extremity of the long esker forming the eastern end of Hyndman Lake, this site covered an area of more than 2000 square metres. Preliminary recoveries in 1987 suggested the possibility of a microblade component (Pilon 1988:24-25).

In order to obtain additional specimens which could aid in verifying the presence of the microblade component and provide additional insights into the activities carried out at the site, a total of 11 square metres were excavated. These units were located so as to encompass a particularly productive 1987 shovel test pit.

Initial testing in 1987 had indicated that artifacts would be found in the thin sod/humus and into the upper few centimetres of the underlying sand and gravels of the esker. Given the coarseness of these deposits, it was clear that very substantial downward movement of artifacts was to be expected and so vertical control was only maintained in 5 of the 11 square metres in order to verify this proposition. Otherwise, given the time constraints, provenance was limited to recording the square metre within which flake lots and implements were found.

Since the matrix within which the artifacts had been found was composed principally of coarse sand and gravel, all excavated matrix, with the exception of the surface vegetation, was screened through 1/8 inch mesh screening.

Vertical Artifact Distribution

Although the artifacts gathered in any one square metre unit were lumped together, regardless of the level within which they were found (with the exception of 5 square metres), excavation nonetheless proceeded according to each of the three visually distinct levels; 1-vegetation, 2-root mat/humus, 3-basal sand/gravel.

Artifacts were found in the lower two levels, but seemed most concentrated in the upper 10 cm or so of the basal sand/gravel. The upper portion of this level consisted mostly of sand while the proportion of gravel and cobbles increased with depth. There were a number of notable instance when specimens were found up to 15 cm or more into the sand/gravel. Rather than suggest a distinct occupation, I believe such depth is due to the nature of the deposits which could readily allow thin flakes to work their way down with very minimal trampling.

Whenever a lithic concentration was encountered in the root mat/humus, a corresponding one was found in the upper sand/gravel below it, thus suggesting that these were part of the same concentration. However, the highest densities occurred within the upper levels of the sand/gravel layer and not in the root mass/humus.

A very real link between the artifacts found in the root mat/humus and the upper sand/gravel is demonstrated by the numerous instances of refitting. At least 5 cases were documented in those units were materials were separated according to level. Additional refits occurred in the other units as well.

In spite of the above arguments which suggest that the artifacts found in the root mat/humus and the underlying sand/gravel should be considered as a single collection, there is some reason to be cautious in discussing the relation between the collection and the behaviour of a single group. The first caveat is the presence of numerous surface indications of a major forest fire in the site vicinity sometime on the order of 50-75 years ago which could have reduced the humus to ashes and thereby mixed previously separated assemblages. In fact, the greatest densities were found in thesand/gravel and not the root mass/humus. This observation corroborates the presumed effect of forest fires. Were this not the case, the opposite concentrations would have been found unless the people who used the area were in the habit of stripping the site of its humic layer before using it.

A second warning against a simple occupational model for the tested area is the discovery of a few charred bone fragments in association with fire-cracked rocks (2) in the humic layers (the fire-cracked rocks were resting in the root mat while a few charred bone fragments were found in the same vicinity, but in the humus). These items, along with a thin scattering of adze-cut stumps in the general site vicinity, suggest a late prehistoric or historic use of the locality.

In spite of the above warnings, the analysis which follows does not distinguish between separate occupations, if such is indeed the case, but must proceed by treating the collection as a unit. Hopefully the conclusions drawn from this study will reflect this problem.

Horizontal Artifact Distribution

The expansion of the excavated area was predicated upon the apparent concentration of material as the work proceeded. Indeed, the greatest concentrations were found in the southern squares of unit A and in the northern portions of the adjacent unit C. There was a marked reduction in the densities of artifacts in the adjacent units. Once again, cross-mending of linear flakes and microblades show at least some instances of horizontal movement between separated units.

Artifacts

The collection recovered from these 11 square metres is the largest yet found in the southest Anderson Plain. It also exhibits the greatest densities of artifacts. However, I am certain that should excavations be undertaken at the Thunder River quarry site, MiTi-1, that NbTj-3 would pale in comparison. Yet in relation to all other tested sites in the region, a remarkable amount of material was found with in a relatively small area. For this reason the lithic collection deserves careful consideration.

The most important artifact category of course is the debitage category (1823 pieces including 60 linear flakes and flake fragments). The next most numerically important artifact is the microblade and microblade fragment (62). The rest of the collection is made up of cores and core fragments (43) and variously retouched or shaped implements (39). The impression gained from the initial consideration of these figures is that the reduction of cores was a very important activity at this site.

Blades/Microblades

A great deal of difficulty was encountered in attempting to determine which pieces of debitage could legitimately be related to the systematic production of blades and/or microblades. Part of the problem is that a very significant activity which took place at this site was the acquisition and processing of lithic raw materials. Judging from the amount of cortical flakes and nodule fragments recovered, it would appear that the people were exploiting a local cobble source, quite possibly the esker itself.

Five items were found which can unequivocably be identified as microblades or fragments thereof (a-e). None represents a complete microblade. However it is possible to measure their widths and thicknesses. The values of the former are rather high (mean width 9.09 mm), approaching the practical limit usually employed to distinguish between blades and microblades (1 cm). All exhibit very regular parallel margins with 3 similarly oriented dorsal flake scars, giving them classic trapezoidal cross-sections. The three specimens which retain their proximal portions intact all have small, smooth punctiform striking platforms which were isolated by carefully trimming the dorsal edges of the platforms.

These microblades are accompanied by an additional 57 items which also share most of the essential characteristics of microblades. However, they exhibit a greater range of variation in many of the basic microblades characteristics (both morphological and metric) mentioned above. For example, although many of the dorsal surfaces of the striking areas were modified prior to microblade detachment by trimming, crushing, and grinding, the result is sometimes a narrow and wide platform remnant rather than a puntiform striking platform. In either case however, there is a universal trend to use smooth platforms. These microblades often exceed 10 mm in width and are usually thicker than the first five microblades.

The second group includes a number of complete or near-complete specimens (f-h). As a result it can be remarked that the outlines could stray from the presumed desired end-product of a thin parallel-sided microblade. In many cases the irregularity of the outline appears to reflect the stage at which the particular microblade was detached from the core. In a number of cases, vestiges of cortex or the weathered exterior of the core were observed on the dorsal surfaces of some microblades. This presence of cortex suggests that in certain cases there was no preliminary shaping of the core prior to microblade production. It is reasonable to expect these initial microblades to show a greater tendency to follow surface irregularities and internal flaws.

Following the successful removal of `ideal' microblades it is likely that a point would be reached at which the physical conditions necessary for microblade production (for example striking angle, core size, internal raw material homogeneity, etc.) ceased to exist within a certain range of tolerance. At this point, less than ideal microblades might be produced before changes were made to those conditions which could be modified (rejuvinating the platform and thus modifying the striking angle). Alternately, microblade production could simply cease. At this point, the core may still be useful for other purposes such as flake production or modification into an implement.

The is a third group of linear debitage which warrants specific attention. It consists of 25 specimens which were detached from the corners of blocky, angular cores. In some cases there had been at least one previous longitudinal removal prior to actually detaching the corner of the block. if these specimens were found in any other context they would not be accorded particular attention since it is reasonable to expect a lithic artisan to first removed these angularities when initial shaping a blocky, angular core. However, given the presence of intentionally produced microblades, these linear corner flakes take on a totally different meaning (i-m). Since they represent some of the first flakes removed from blocky, angular cores, they would have set up a pattern on the flaking face of these cores which would serve to orient later microblade removals. In their own right they might be considered blades/microblades, albeit crude ones.

The significant element in this whole discussion is that in the process of microblade production there will be a wide range of biproducts. It is difficult to exclude these in the examination of the present collection, but it is equally difficult to compare the microblades from this site with those found elsewhere because of the tendency to focus only on the refined and idealized products as most researchers conceive these to be.

One final point should be made in considering the linear debitage in the collection from NbTj-3. It has been discussed within the context of a single technological orientation, namely the production of microblades. It is entirely possible that the desired end-products could just as easily have been parallel-sided flakes of a range or sizes which included items usually identified as blades. Alternatively, the collection could very well represent the remains of different groups separated by centuries or millenia who shared a common technique but who put the products to markedly different uses.

In considering either alternate interpretation it is obvious that a detailed analysis of the existing collection might indicate a preference for one or the other. Additional excavations under more controlled conditions could also provide a better basis for interpretation.

Cores and Core Fragments

The collection from NbTj-3 consists to a large extent of the biproducts of the initial reduction of large pieces of cryptocrystalline raw materials. Although finished and evidently used lithic implements are present, the manufacture of such tools is not as well represented.

For convenience sake the cores and core fragments have been divided into 5 groups; 1-pebble nodules, 2-quartzite cores, 3-tabular cores, 4-bipolar cores and 5-exhausted cores/core fragments.

The first category, split pebbles, is exactly what one envisions upon hearing the term. They are black, cherty cobbles (6) which had undoubtedly been selected on the basis of the reticulate patterns of fractures on their smoothed exterior surfaces. My own experience on the cobble beaches nearby led me to systematically look for such surfaces in order to find nodules of chert. In any case, these particular specimens were totally unsuitable for further reduction since the removal of one or two flakes soon indicated an intricate network of internal flaws and fracture planes such that the flaked surface exhibited a very rough coarse granular surface. They were abandoned by the artisans.

Quartzite cores (3) have been separated from the chert and siliceous argillite cores and core fragments since they do not easily fit into either of the other groups which we have defined for those raw material types. Indeed, it might be argued that given the very significant microstructural differences between chert and siliceous argillite on the one hand, and quartzite on the other, basic approaches to working with quartzite would have differed. Such differences might be found in the general size of the nodules being used, the most efficient striking angles, the size of the flakes produced, etc. For the moment the basic descriptive data have been segragated to permit an initial evaluation of possible differences among the core remnants.

Tabular cores (14) constitute the most important primary raw material form if the mere frequency of such items in the collection is a reflection of their use by the site's inhabitants.

A core renmant is deemed to have originally been a tabular piece of stone by the fact that it retains cortex on two opposing parallel faces. These cortical surfaces usually indicate a relatively thick piece compared to the other dimensions. Moreover, the range of sizes indicates that such tabular cores also included items which could also be described as blocky or angular cores, the likely sources of the linear corner flakes described earlier.

In any event, the basic reason for the tabular character of these cores lies in the that they are all of siliceous argillite which has internal bedding planes, being a metamorphosed sedimentary rock.

An interesting feature of the tabular cores is that reduction often did not proceed beyond a few flake removals, usually from corners. No apparent internal flaws were discernable and no use was made of the cores as tools. Could some of these represent flaking not aimed at actually producing anything useful, something akin to doodling? Alternately, could these cores represent the remains of a novice, learning some of the principles of stone working?

Two cores, ultimately drawn from two of the above groups of cores (tabular and pebble) appear to have been rested on stone anvils while flakes were removed from them. In both cases, the extremity which was in contact with the hard surface consists of a ridge exhibiting crushing and microflaking. Neither was extensively flaked and both appear to be of relatively good quality raw material. It is difficult to entertain the idea that these items may have served as tools, however, it is also premature to put aside the suggestion.

Lastly, there are 7 items which have been grouped together as exhausted cores/core fragments. In all cases it seems clear that the intent was to remove flakes from these pieces rather than attempt to give them a particularly functional shape. The apparent reason for their abandonment was that they was been reduced to such a point as they could no longer be profitably worked or they represent fragments from a larger core. Some of these fragments were undoubtedly broken by the lithic artisans while others appear to have spalled subsequent to the abandonment of the core through the action of frost of heat. The following data are provided simply to impart an idea of the size range of these fragments. No further weight is attached to them.

Pièces Esquillées

Pièces esquillées are not an uncommon implement on northern archaeological sites although their ultimate function remains to be adequately demonstrated for the many different cultural contexts in which they have been recovered.

In this analysis, pièces esquillées are items which usually exhibit grinding, crushing and restricted flaking along a pair of opposing ridges. Some have had culumnar flakes removed from a lateral edge and may exhibit marked use at these corners, other do not. The blank can be just about anything. However, the artisans did not seem to chose thin or exceedingly thick pieces to modify. In one case (145a/270), the original pièces esquillées appears to have split into two thinner flat pieces which were subsequently used individually as pièces esquillées. In general these implement are significantly longer than they are wide.

There has always been some difficulty in determining exactly what could and should be described as a pièce esquillée. In some instances they can be confused with bipolar cores. However, not one of the items which have been grouped as pièces esquillées from NbTj-3 could be construed as cores. Most are made on flakes and many of these retain some cortex.

Gouges

For lack of a better understanding of the following group of implements, the term gouge has been selected in order to reflect a common trait. Like the pièces esquillées, these tools tend to be longer than they are wide, much longer. Some of them are simply made on long flakes (113,28) while others might be core fragments or core remnants (29,107). One (149) is simply a modified pebble. All of these items have exhibit crushing, grinding and/or flaking along one of the narrow extremities which form sharp angled edges. The edge modification can be bifacial or unifacial. When viewed from the side these working edges are quite flat while an end-on view reveals 2 of them as having marked curvatures.

Scrapers

A total of 8 implements have been identified as scrapers. The trait shared by 7 of these is a steeply retouched, convex, working edge. The unifacial marginal retouch is always located on the dorsal surface of the distal edges of the flake blanks. Four of these scrapers are made of siliceous argillite, two are of quartzite and one is made of a mottled grey chert. The eighth scraper, which might more properly be described as a concave scraper, is made of siliceous argillite.

There was no shaping of the siliceous argillite blanks. They all appear to have been made on blades (157, 92, 64) or elongated flakes with quite pronounced dorsal ridges (157, 133a). However the latter two siliceous argillite scrapers (147, 133a) split longitudinally such that comment on their original shapes or blanks is tenuous. The chert specimen (146) had two slightly invasive flakes removed from one lateral margin, but overall the general morphology of the tool, an oblong or rectangular shape, was not affected. The smaller quartzite scraper (187) was made on an unmodified short but thick flake retaining an especially pronounced bulb of percussion. Lastly, the longer quartzite scraper (267), although incomplete (the proximal portion and part of one lateral margin are missing), was retouch along its extant lateral margin possibly to give it an elongated shape. In contrast to the other scrapers, this specimen was much thicker. The lateral retouch may have also been intended to produce a working edge. Unfortunately the coarseness of this raw material makes microscopic determination of this fact virtually impossible.

The concave scraper exhibits steep dorsal retouch which forms a wide, shallow concavity at its distal extremity. The chord of this concavity is markedly skewed from the short axis of the tool. As such, there was a sharp tip left in the middle of the distal end. Although limited retouch is found on the ventral surface of this tip, the heaviest wear is located within the concavity and the apparent graver is just that.

Miscellaneous Implements and Tool Fragments

A number of tool fragments, both unificial and bifacial, were recovered. Since in most cases these are fairly small it is difficult to conjecture about the implements of which they were part. However, it should be pointed out that although many of these fragments appear to have been only marginally retouched, some were more than likely from more formal tools including large bifaces. This fact is corroborated by an initial assessment of the unmodified debitage which includes large biface thinning flakes.

Additional work was carried out this site in 1990.