Five types of scrapers are symmetrical end, combination, discoidal and bifacial scrapers, and asymmetrical sidescrapers. Except for discoidal scrapers and rare scrapers with scraping edges at both ends, scrapers have two parts: an abruptly retouched scraping bit at the tip and/or sides; and a retouched or unretouched supporting base and midsection, sometimes called a handle or haft. Symmetrically retouched or unretouched tanged or stemmed scrapers were inserted in antler or bone handles, while unretouched sidescrapers were often hand-hafted. Tanged bases are narrower than stemmed or tapered bases, and may or may not be parallel-sided. Combination side/end scrapers have paired scraping bits meeting at a corner or along an arc, the latter having imperceptibly merging edges. As combination scrapers are resharpened, they shorten to resemble endscrapers, distinguishable only by their wear striae. Discoidal scrapers were hand hafted and rotated with use. The final category of bifacial scrapers are confined to ululike ASTt skin flexers. Mainly chert, they dull quickly on wet gritty caribou hides, producing extensive striae over their edges. End and combination scrapers have flat or upturned ventral bits, their dorsal surface flat, moderately convex, keeled or domed tortoise-back. If a prominent juncture interrupts an otherwise tortoise-back, it is called keeled. A base may be retouched on one or both sides.
Plan and section of broken scrapers were approximated by comparing fragments with full scrapers. Estimates of nonmetric (e.g.s, tearshape plan & biconvex section) and metric (e.g.s, length & width) traits were used. Care was taken to differentiate rhomboid from square plans in endscrapers, and tearshape from triangular. Triangular scrapers have bases tapered almost to a point, while rhomboids or keystones end in a flat base at the striking platform. Square scrapers may have rounded corners. Tearshaped scrapers resemble triangular ones, except the bit is semi-circular and the base is more tanged than stemmed; even concave sided and meeting at a point. Ovoid plans occur when bases with very convex plans have an overall circular plan. Most broken scrapers have missing bases. Discoidal scrapers are circular, with complete peripheral retouch and missing striking platform. Some combination scrapers approach a circle, especially in MT, except that the striking platform interrupts the circle. Discoidal scrapers are confined to ASTt and SA. ASTt discoids may be finely bifacially retouched; SA discoids are unifacial and rougher. A single stratified ASTt discoid, KjNb-7:18-47, serves as the example for typing fine bifacial discoids in surface sites. A single stratified SA discoid, KjNb-6:45-49, serves as type tool for rougher unifacial discoids. ASTt and SA discoidal scrapers are easily separable, the SA discoid thicker and cruder, with a less regular periphery.
In sections, keeled means thick and prominently ridged. Tabular means flat and usually thin, with the possibility of 1-2 low ridges that are less than keeled. Tortoise-backed means arcuate or 2-dimensional convex retouch across the dorsal face. Spurs or projections often occur where front and sides meet, and scrapers may be single or double spurred. Single spurs may be large, making the endscraper asymmetric.
Most scrapers are either quartzite or chert, with rare basalt. All were used for scraping hide, but may have been used on organic materials of wood, bone and antler. A few scrapers are striated, allowing the analysis of application forces and angles in functional interpretation. Many have thinned hafts. Striae occur most frequently on the bit rather than ventral surface like pushplanes, indicating oblique angles of application. Some are so rounded that retouch is obliterated. Scraper application is >30 deg. to the work, often 45 deg. Some rare huge endscrapers resemble some abruptly flaked flat cores and core tablets, but are separable using telltale wear.
Stone scrapers were used by Déné Indians well into the historic period because they are less apt to pierce and ruin fur and hide than steel scrapers. Their Taltheilei ancestors and predecessors of the ASTt, Shield Archaic and Northern Plano used a variety of scrapers. But end scrapers are ubiquitous and command a common shape. They were simple tools-of-the-moment, requiring little shaping and edge grinding except for the more elegant bifacial ASTt and tanged ET scrapers which may have been carried across the range by herd-followers.
795 stratified forest and tundra scrapers were compared to test trends and estimate culturally meaningful types for assigning surface scrapers. By number and culture (including fragments), scrapers from earliest to latest include 14 NP, 40 SA, 101 ASTt, 204 ET, 398 MT & 37 LT, plus 1 Chipewyan. Bifacial scrapers are confined to ASTt tundra sites and may reflect warm season use only. Ethnographically, most caribou skins are processed in late summer and autumn, but summer was when wind-blown sand produced striae, like they did on burins and spalls, when sand grit was caught between tools and object worked. A lone Chipewyan endscraper came from a forest site, while another came from the forest (KcNf-4), a crude cortex spall.
Scraper Metric Analysis
(1) Regarding length, SA, ASTt, ET, MT & LT endscrapers are all shorter in the forest. There are no forest NP or tundra Chipewyan ones for comparison. The shorter forest endscrapers may mean they were resharpened rather than replaced, probably due to limited access to raw material under snowcover in the forest or winter range. Little can be said regarding sidescrapers because they are few; e.g.s, all NP, ASTt, Mt & LT sidesrapers are tundra. ASTt combination scraper lengths are very similar and may reflect adequate forest chert supplies or curation of tundra carryovers.
(2) Regarding width, SA, ET (big sample), MT (huge sample but small difference) & LT (small but different) endscrapers are narrower in the forest. Tundra & forest ASTt width similarity, like length, may reflect curated supplies or a rectangular plan of constant width. Sidescrapers are too few for comparison, except for MT scrapers which are much shorter in forest. Tundra and forest SA combination scrapers are too unequal for comparison, but forest ASTt combination scrapers are wider, reinforcing my earlier suggestion of local supplies. ET and MT combination scrapers are narrower in the forest.
(3) Regarding thickness, tundra & forest SA endscrapers are equally thick. ASTt, ET, MT & LT endscrapers are thinner in forest. Again, sidescrapers are to few, except MT scrapers which are thinner in the forest. SA, ET & MT combination scrapers are thinner in forest, while forest ASTt are thicker, reiterating possible local chert supplies.
(4) Regarding weight, SA endscrapers weigh less in forest, with ASTt & ET endscrapers ca. half. Forest MT endscrapers are a little more than half the weight of tundra scrapers; LT endscrapers less than half. ASTt combination scrapers weigh more in forest, while ET combination scrapers weigh less, with MT scrapers less than half the weight of their tundra equivalents. The longest endscrapers are SA, followed by NP, then LT, then ASTt & MT, with the smallest being ET. The longest sidescrapers are also SA, followed by ASTt, then MT and ET. The longest combination scrapers are again SA, followed by MT, then LT and ET, and finally ASTt. Endscraper length, width, thickness & weight do not vary significantly by culture, except possibly ET which is smallest. Of interest are ASTt endscrapers which are from a supposed microlithic tradition but are larger than ET endscrapers. Sidescrapers vary greatly, but again with ET smallest. ET miniaturization continues into combination scrapers, but ASTt are even smaller. This is interesting considering the cold-adapted ASTt were quickly followed by warmer-adapted ET Indians. Could miniaturization of tools reflect colder environments, a landscape that was warming slowly in ET times?
Plans decrease in order: triangular, rhomboidal, tearshape, ovoid, rectangular & ululike, with discoidal uncommon and only one parallelogram. From earliest to latest, NP is equally represented as rhomboid, triangular & ovoid; SA mainly ovoid, but high in rhomboid & tearshape; ASTt as very ululike, followed by tearshaped, triangular & ovoid. ET is mainly rhomboid & triangular (>50%), while MT has this reversed, indicating increasing popularity of reduced striking platforms & narrower stems, possibly for narrower haft insertion. LT is mainly triangular, followed by ovoid and rectangular. Rhomboid plans become reduced throughout Taltheilei, ending in a return to wider but non-triangular hafts, possibly for hand clenching. This is reinforced by the single square Chipewyan scraper. Plans are alike in tundra & forest, but forest scrapers have higher rhomboid and square plan frequencies (32 & 20%; tundra at 18 & 7%), but fewer ululike (1; tundra at 5%) due to the absence of bifacial forest ones. Other differences are the large number of forest ET rhomboid & MT square scrapers.
Sections decrease in order: tabular, tortoise-back & keeled, with minor biconvex and rare biconvex cortex spall-like scrapers. From earliest to latest, NP is almost entirely tearshaped; SA, mainly tearshaped with rising tortoise-backed. Taltheilei tabular sections drop from ET to MT to LT at 80, 70 & 50%, with LT having increasing keeled & tortoise-backed. The single Chipewyan scraper is also tortoise-backed. It appears that Taltheilei scrapers change from tabular rhomboid to keeled/tortoise ovoid/rectangular scrapers; i.e., thicker rounder scrapers. Tundra and forest section differences are immediately apparent. Tundra scrapers are more tabular and tortoise-backed (73 & 20%; forest at 64 & 14%) and less keeled & biconvex (6 & 0.5%; forest at 15 & 5%). All cortex spall planoconvex scrapers are tundra. The forest tabular & keeled plan frequencies of ET tabular (high) and MT (low) are responsible. Combined with plan (above), forest scrapers are mainly flat tabular rhomboids, while tundra ET are mainly tabular triangles; i.e., tundra ET scrapers have thinner bases, possibly for easier haft insertion during more transient hide preparation times accompanying migration. Tortoise-backing makes scrapers thicker, stronger and longer to make and is more plentiful in the semi-sedentery forest. Keel-backing makes scrapers thicker & stronger, but requires less time and is frequent in the tundra. Ululike bifacial scrapers were found only in the tundra and raise the frequency of biconvex sections.
Serrated scraper edges are in all prehistoric Beverly cultures at about 8% frequency, except for NP which doubles in a small sample. Serrated scraper edges are similar in both ranges. Type of scraper base is mainly ground unretouched & dorsally retouched taper, totalling 80%. Ventrally retouched stems & bifacially retouched tangs lag at 6% each, while unground unretouched tapered bases are few. Base types are similar overall, but tundra scrapers have a higher frequency of ground taper (40, compared to forest at 33%) & a lower frequency of discoidal (<1, compared to forest at 6%). So far, the data suggest that tundra scrapers are tabular with ground triangular bases. Cortex occurs in all cultures, with decreasing frequency in the order NP (28%), SA & LT (22%), ASTt & MT (9-10%), and ET only 6%. Since ET was the earliest Taltheilei culture following the cold ASTt period, ET peoples may have had more raw material or a dislike of cortex on finished tools. The frequency of scrapers with cortex is higher in the tundra and probably reflects unneeded retouching & resharpening due to more raw material.
Striae occurs in 12% of scrapers, most cultures having 8-14%. Striae are very high in ASTt (34%) due to ease of formation & identification. The incidence of striae is similar in both ranges. Moderate basal grinding occurs in 90% of Beverly scrapers and is >85% in all cultures except ASTt, where it is 74%. Heavy basal grinding is present in 1% of scrapers, and is seen rarely in SA, ASTt (more common), ET & MT. Basal grinding frequency is similar in forest and tundra. 77% of scrapers have modestly worn bits, while 9% are very worn. Taltheilei ET, MT & LT all have >80% worn bits, a possible cultural marker. SA has 67%, NP 50% and ASTt lowest at 47%. Very worn bit frequencies are higher in the tundra (11 compared to 5%), while worn & unworn frequencies are similar. Many endscrapers have ground multifacetted bifacial bases with edge angles like those of bifacial knives. Many may have been made from knife sharpening flakes and are called thinning flake scrapers. Scrapers on biface thinning flakes or which have knife-edged striking platforms have similar frequencies in both ranges.
Spurs are lateral projections from endscraper bits. They may be sharp in Taltheilei and dull in ASTt. 13% of Beverly scrapers are single spurred, while 6% are double spurred. Single spurs are common in ET at 20%, less so in MT at 13% and minor in LT at 6%. Spurring decreases throughout Taltheilei and may relate to the trend towards thicker rounder scrapers. However, double spurs in Taltheilei are low at 2-6%. Single spurring is uncommon in ASTt at 8%, but double spurring is common at 34%. Spurring frequency is generally similar in tundra & forest, except for double spurring which is common in tundra bifacial scrapers.
84% of Beverly scrapers are quartzite, 13% are chert, 1% are quartz, and <1% for the remainder. Tundra material frequencies are similar generally, while no forest sandstone, basalt or copper scrapers are reported. Tundra & forest frequencies of quartzite are 84 & 83%, while chert are 13 & 12%, showing little difference in choice of scraper material in the tundra and forest. Cultural phase-pertinent traits and variables are as follows:
Plan: 0=unknown; 1=rhomboid/keystone; 2=tearshape; 3=triangular; 4=ovoid; 5=rectangular/lanceolate/bladelike; 6=square; 7=ululike; 8=discoidal; 9=parallelogram.
Section: 0=unknown; 1=tabular; 2=keeled; 3=tortoise-back; 4=biconvex; 5=planoconvex. Many tabular sections could be called planoconvex, but that is used here to show pronounced unretouched round dorsal & flat ventral surfaces, as in cortex spall scrapers.
Edge type: 0=unknown; 1=serrated; 2=unserrated (usually abruptly retouched).
Base: 0=unknown/missing; 1=natural unground tapered base of common flake; 2=1 but ground; 3=dorsally retouched base; 4=ventrally retouched base; 5=bifacially retouched tang (common in ASTt ululike scrapers); 6=discoidal; 7=double-bitted with no base; 8=alternately retouched basal edge; 8=where unretouched base is broken & grinding undetermined. In broken specimens where bit retouch ends in unretouched midsections, bases are also called unretouched. If bits are unworn or worn, bases are assumed to be unground (1) or ground (2).
Striae: 0=unknown/absent; 1=present and over the edge. Cortex: 0=absent; 1=present (dorsal face). Grndbase (ground base): 0=absent/unknown; 1=ground; 2=very ground; 3=unground. Wornbit (worn bit): 0=absent/unknown; 1=worn; 2=very worn 3=unworn. Thinflake (from biface thinning): 0=unknown/absent; 1=present. Spur: 0=absent; 1=single spur; 2=double spur.