Across Atlantic Ice
Page 18
LITHIC TECHNOLOGY
To better understand its similarities to early assemblages in North America through Clovis, we offer this examination of the details of Solutrean flaked stone technology.
Material Selection As one might expect, the Solutreans mainly used stone from their immediate area, yet their focus on exotic raw materials has often been noted. Even in northern Spain, where access to exotic stones appears to be limited, several Solutrean occupations at La Riera Cave contained many artifacts made of non-local raw materials. At Laugerie Haute in the Vézère Valley in France, where high-quality black and brown flint is abundant, a relatively large proportion of the Solutrean bifaces were made from nonlocal chalcedony and multicolored banded flints from the Bergerac area, as well as from jasper, quartz crystal, and even obsidian. The combination, unusual for the Paleolithic assemblages in the region, of these exotic materials is also present in southern Spain and Portugal. We have been told more than once that when archaeologists come across exotic chalcedony or quartz crystal Paleolithic artifacts, they can be sure that the tools will turn out to be Solutrean.
There are known sources of high-quality flint, such as the Upper Turonian flint at Les Maitreaux, that Solutrean knappers used to produce biface blanks. A Solutrean biface manufacturing site has also been discovered near Bergerac, which may be the source of the material for the many flint bifaces at Laugerie Haute. Otherwise knappers mainly used a high-quality local black flint. Although both of these locations fall within the known distribution of Solutrean sites, no major occupations have been found in their vicinity. It is clear that Solutrean knappers traveled significant distances to obtain preferred materials, especially for making laurel leafs.
Heat Treatment Another aspect of Solutrean technology was the intentional heat treatment of stone to improve its flaking quality. François Bordes and later Mike Collins presented convincing examples of this on some of the Laugerie Haute artifacts, and Marc Tiffagom has shown that it was a common practice in the eastern Spanish Mediterranean and Portuguese Solutrean.31 Using the strict standards applied by Collins, we thought the frequency of heat treatment was fairly low in southwestern France.32 However, we recently re-examined a large sample of laurel leaf fragments from the sites of Laugerie Haute and Fourneau-du-Diable, which indicated that intentional heat treatment was indeed fairly common, perhaps in as many as 30 percent of cases, at least in southwestern France. The only way to determine the precise proportion of heat treatment would be to make a detailed analysis of large numbers of biface flakes and preforms broken in the early and middle phases of manufacture from completely collected deposits. Unfortunately, such pieces were not usually saved from the majority of Solutrean excavations. Yet all flaked stone items have been recovered from the Solutrean manufacturing site of Les Maitreaux, and there is no evidence of heat treatment there. This doesn’t necessarily indicate that it wasn’t used in the Solutrean—just that it may not have been part of the technology at that source location. This could have been the case because the flint there was fine enough without treatment or because the unfinished bifaces from that site were treated elsewhere.
Blades Like other Upper Paleolithic people, Solutrean flintknappers made large blades. We have identified at lease three different technologies, resulting in three different core types. It is unclear if all three methods were in use at the same time or if they form a succession through time. At Les Maitreaux the dominant approach was similar to that of the earlier Gravettian: two opposed platforms with blades struck from both ends toward the middle, resulting in what are known as bidirectional cores (figure 5.8a). The main product of this technology seems to have been straight blade blanks to be made into shouldered points. The second approach used a single primary platform with a single blade-making surface (figure 5.8b). Flakes and occasionally blades were struck from the opposite end, mainly to maintain the core surface or correct flaking errors rather than for the production of usable blades. The main objective of this technology was the production of not necessarily straight blade blanks for a wide range of tools. This approach resulted in wedge-shaped cores. The third approach had a single platform from which most blades were struck, but they were removed from the entire circumference of the core, giving it a conical shape (figure 5.8c).
In all three technologies great attention was paid to the production of each blade. Frequently the whole core platform was removed, creating core tablet flakes. Individual platforms were carefully adjusted by flaking and grinding on the striking surface and the adjacent ridges. The resulting blade platforms were usually bifacial and nipple-shaped. Although the archaeologist John Whittaker has suggested that indirect percussion was employed, we think the technique was direct percussion, with either a soft hammerstone or a billet of antler or wood.33
Unlike other Upper Paleolithic blade-making technologies, which began with fully bifacial precores, Solutrean technology started with knappers either taking advantage of natural ridges or making precores from medium to large pieces of flint. Frequently they bifacially flaked a single ridge at slightly less than a right angle to a natural or prepared platform surface. For single-direction cores they often bifacially flaked the end opposite the platform, commonly perpendicular to the intended flaking surface, to produce a sharp angle. After striking an initial, crested blade (a series of alternating flake scars that form a dorsal ridge), removing much of or the entire bifacial ridge, they readjusted the platform, re-established the flaking surface projection (by removing corner blades, flaking laterally, or both), and removed another center blade. This process continued until either the core became too small or a major flaking mistake was made. Occasionally blades and flakes were struck from the bottom of the core, but this was done to correct mistakes rather than as a standard blade-making strategy. The same basic procedure was followed for the opposed platform cores, except that blade making occurred from both ends, and cores tended to be abandoned sooner, when they became unsuitable for the production of shouldered point blanks. On these cores it is usually evident that one platform was preferred over the other, and most of the desired blade blanks were obtained from it.
FIGURE 5.8.
Solutrean blade core types: (a) obverse, reverse, and section of a bidirectional core; (b) obverse, reverse, and top view of a wedge-shaped core; (c) face and top view of a conical core.
Although this bidirectional, opposed platform process was generally similar to the earlier and contemporaneous Gravettian and later Magdalenian blade-making technologies, there is one significant difference: exhausted Gravettian and Magdalenian blade cores usually retain a portion of a bifacially flaked ridge on the back, opposite the primary flaking face; Solutrean cores, by contrast, usually have flat backs, either a remnant of the natural nodule surface or the result of flake scars originating from the sides. This distinction may seem minor and may have made little difference in production efficiency, but it does indicate a significantly different approach to core surface maintenance. The devil is in the details, and even such a small variation may have had cultural or traditional significance.
Noninvasive unifacial retouch of flakes and blades produced scrapers, gravers, backed tools, and other implements, as we see in other Paleolithic times. In addition to the usual types of percussion retouch, “flat retouch” was used to form some tools. In this process, retouch flakes invaded the surface of the blank to the extent that they shaped the entire contour of the piece, not just the edge. Normally, we would consider this shaping or thinning rather than retouch, as it was done to shape the piece rather than produce a tool edge.
Bifacial Flaking In Solutrean times there was much less emphasis on blade blanks and more inclusion of flakes as blanks than is seen in antecedent technologies. Flakes were frequently used unmodified as tools and as blanks for such implements as backed knives, side scrapers, and gravers. Most flake blanks probably came from laurel leaf manufacture, but there is an indication that they were also produced from bifacial disc-shaped cores.
&nbs
p; Recent research on the Maitreaux assemblage has clearly shown how Solutrean laurel leafs were produced.34 Our examination of hundreds of laurel leaf fragments from all phases of manufacture has not shown any significant departure from the steps taken there. Solutrean knappers selected large pieces of high-quality stone, then removed big flakes to create blanks or reduced entire pieces to bifaces. Blanks were usually flaked all the way around their circumference to produce a bifacial edge before thinning and regularizing began. Bifacing proceeded through to the finished product. The best evidence for intentional staging—planned intervals where flaking was either stopped and then re-initiated later or re-initiated in a different location—comes from Les Maitreaux, a knapping site to which initially edge-trimmed pieces were brought and from which nearly finished bifaces were taken. We have used this evidence plus some arbitrarily identified phases to see whether flaking decisions changed during the bifacing process.
We recognize four production phases after initial trimming (figure 5.9a–d): early, middle, late, and finished. These arbitrary phases do not have distinct boundaries, and we have not used width-to-thickness ratios to allocate pieces to them. To see if such characteristics as thinning method, pressure flaking, heat treatment, and width-to-thickness ratios changed during the sequence, we recorded a series of attributes of a large group of bifaces and biface fragments in the collections from Laugerie Haute and Fourneau-du-Diable, Solutrean sites in southwestern France. The results of this analysis allow us to describe the general biface flaking procedure.
FIGURE 5.9.
Laurel leaf production phases: (a) early; (b) middle; (c) late; (d) finished.
Overshot flaking was applied during the early (26.5 percent of the pieces that exhibit overshot flake scars) and middle phases (28.1 percent) until a regular, flat, thin biface was made.35 Flaking in the early phase focused on regularity of the edges and outline, with thinning beginning in earnest in the middle phase. Overshot flaking was applied in the late phase (23.9 percent) as the main thinning method. For finishing, Solutrean knappers switched to the diving flake method, although overshot was used all the way to the end (15.8 percent). Pressure flaking was used only for finishing, mainly for tip and base regularization, although some of the smaller pieces exhibit all-over pressure shaping and retouch. It is possible that some of these were flaked all over during reworking rather than initial production.
The following discussion details what effects each thinning choice had on bifaces. This is important, as we think decisions at this level can be used to demonstrate historical connections between technologies.
Overshot and diving flaking are both effective ways to thin a biface, and the preparation of platforms and the spacing of flake removals are critical aspects of both of these techniques. Platform preparation for overshot flaking is especially tricky because even a small variation in the angle or isolation of the platform or the place of fracture initiation will result in failure, frequently in the form of a diagonal (perverse) break. The angle of application and the amount of force applied are also crucial, as is the location of flake removal in relation to the contours of the biface’s surface. There must be enough mass immediately below the platform so that it does not shatter when struck, and the surface must have contours so that the flake tends to form with nearly parallel sides. A miscalculation of any of these features will result in either too much of the distal edge being removed (called a plunging fracture) or a deep hinge fracture, in which the flake terminates in a deep scoop taken out of the surface rather than an even taper (called feather termination). Interestingly, when possible, Solutrean knappers alternated sides on the same face during thinning. When successful, this is an effective way to achieve full facial coverage with the removal of only a few large flakes.
Diving flaking, by contrast, requires slightly less control and can be accomplished with less preparation. This is not to say that it is easy, but success may be achieved with a greater range of edge angles, platform placement, and angle and force of impact. The goal is to remove a flake that terminates in a shallow hinge fracture near the midline of the biface, and then to remove the hinge termination with flakes struck from the opposite margin. This technique is highly effective, barring major mistakes, but produces many small, relatively thin flakes. It also results in a high proportion of manufacture errors, especially when the flake struck to remove a hinge fracture itself terminates in a hinge short of the goal, leaving a stack near the middle of the biface. Once a stack is established, it is extremely difficult to remove from an otherwise thin biface.
Since most Solutrean laurel leafs are bipointed, basal thinning was accomplished from the sides rather than the base. Finishing was frequently done with pressure flaking, either by marginal retouch or, on the narrower pieces, by thinning, including overshot flaking. The last modification on many laurel leafs was a light to heavy grinding of the lower edges from the point of greatest width to the base.
The Solutrean biface thinning process was far and away more complex and difficult than any other method of flaking during the Upper Paleolithic. It also resulted in an overall greater use of raw material, related to the proportion of flakes used as tools (an exception being the large overshot flakes). Generally the technology may be thought wasteful if one considers only efficiency of tool manufacture. With limited supplies of exotic stones, it is curious that Solutrean people used them to make bifaces, unless they were for symbolic as well as practical purposes.36
Blades were used as blanks for various types of tools, including backed knives, denticulates (tools with a jagged-toothed edge), end scrapers, burins, and especially the distinctive single-shouldered points that are as diagnostic of Solutrean as laurel leafs. Shouldered points were made in different ways, some with indentations formed by abrupt retouch and minor ventral surface retouch at the tip and base. This style was most common in Spain. Alternatively, especially in southwestern France, shouldered points were frequently flaked by invasive pressure, leaving only small remnants of the original blank surfaces. Many shouldered points exhibit diagonal serial pressure flaking, which is especially effective for removing flake or blade ridges while only minimally narrowing the piece. It also keeps pressure flakes from breaking off prematurely, especially on some of the lower-quality stones, such as quartzite.
A COMPLEX COINCIDENCE?
Fresh out of university, I spent the summer working in southwestern France with the noted French prehistorian and flintknapper François Bordes. This was where I first encountered beautiful and technically complex Upper Paleolithic artifacts. The highly sophisticated thinned bifaces, called laurel leafs because of their shape, were the greatest challenge for me to reproduce. Surprisingly, I found that I used the same approach to make laurel leafs as I used to make Clovis points. I thought this was curious but just a coincidence. It wasn’t until years later that I began to think otherwise.
I am now involved in a project that studies the technology of the manufacture of laurel leafs. This project, directed by Thierry Aubry, has excavated a manufacturing site, Les Maitreaux, where the artifacts were preserved in their original discard locations. It is possible to see nearly complete flaking sequences in this assemblage, revealing specific decisions made by ancient knappers. The project has sponsored several experimental knapping sessions, each with a specific research question. A range of knappers with different skill levels have participated, including Thierry Aubry, Michel Lenoir, Serge Maury, Jacques Pelegrin, Marc Tiffagom, Bertrand Walter, and me. These experiments have been highly instructive and have opened up new levels of understanding of the complexity and difficulty of laurel leaf and blade manufacture. They have also led to various areas of interpretation in relation to site formation and the social implications of the archaeological record. It has been a great learning experience and a privilege to be part of this project. Bruce
Another style of projectile point occurs in many Vasco-Cantabrian assemblages. These are symmetrical with an indented base, and many exhibit diagonal p
ressure flaking on one face and little to no flaking on the other, except at the tip and in the basal indentation (figure 5.10a–b and f). These points are most frequently made of local quartzite and were probably made from straight, elongated flakes. Fully bifacially flaked indented base points are present in slightly smaller numbers (figure 5.10c–e and g–h). These bifacial points tend to be plano-convex (flat on one face and convex on the other) in cross section and occasionally have light basal lateral margin grinding and significant basal thinning (figure 5.10c and f), some of which would be identified as fluted if found in the Americas (figure 5.10g–h). Although indented base points were probably used in varying hunting situations, they are especially well designed for tipping harpoons and fishing spears.
FIGURE 5.10.
Spanish Solutrean indented base points: (a–b) obverse and reverse of unifacially flaked points; (c–d) obverse, reverse, and side views of bifacially flaked points; (e) obverse and reverse of bifacially flaked points; (f) obverse, reverse, and cross-section of a partially flaked biface; (g–h) obverse, reverse, and cross-section of end-thinned (fluted) points.
FIGURE 5.11.
Solutrean rock art figures of riverine and marine animals: (a–b) great auks; (c) salmon or tuna; (d–f) deepwater flatfish; (g) walrus or bearded seal caught in net; (h–i) walrus or bearded seals pierced by spears; (j–l) seals.