Flexibility of stone tool manufacturing methods on the Georgina River, Camooweal, Queensland.
The Australian Aboriginal approach to stone technology is often characterised as highly flexible, a phenomenon well-documented by ethnographic observation. In the Australian context, it would appear that a stone's function was only loosely related to its form. Nevertheless, many ethnographic studies recognise that artefact manufacture was "aimed at" producing specific forms (Home and Aiston 1924:92). This study examines the extent of rigidity in artefact manufacture through an archaeological analysis of a large stone assemblage from Camooweal, northwestern Queensland, Australia. The reduction sequence which created the assemblage is modeled and the rigidity of the various trajectories comprising the reduction sequence is assessed by the degree to which blanks for "aimed at" forms crossed between trajectories. While the ethnographic literature indicates that various artefact categories tended to be used in an ad hoc fashion, the results of the technological analysis indicate that blank production for "aimed at" forms was, in fact, relatively rigid. This stands at odds with sweeping generalisations about the flexibility of Aboriginal lithic technology.
The "flexibility" of a lithic technology can be seen as the degree to which a stone might be used for more than one purpose. As such, flexibility is a continuum with a hypothetical "inflexible" structure at one end, reflecting the manufacture of a specific type of stone for a specific use, and a "hyperoflexible" structure at the other end, reflecting the use of any stone for many different tasks. Historical changes in the Australian scholars' perspective of the Aboriginal toolkit tracks a shift along this continuum. In the formative years of Australian archaeology, models borrowed from Old World paradigms were applied that assumed a degree of inflexibility in stone technology; archaeologically defined artefact types were thought to reflect specific functions or cultural groupings (Veth et al. 1998). The paradigm has shifted in recent decades towards a consensus that Aborigines used stone tools in a highly flexible manner and rarely linked tool form and function in any systematic way (Hiscock 1998).
This intellectual shift is due largely to enhanced appreciation of the ethnographic literature. For example, Daisy Bates' observations of stone tool use among Aborigines in the 1920s inspired her to rail against the assumptions of Australian antiquarians. When confronted by the position that artefact types reflect specific functions, she noted "no stone--except the initiation flint--can be said to be made for a definite purpose ... [they] use their little knives and flakes for any purpose" (Bates 1922 in Wright 1977:2; emphasis in original). George Aiston made a similar observation:
In describing these tools it must always be remembered that the casual nature of the black does not allow him to keep any tool for the one purpose. He is just as likely to use his best stone knife to scrape a weapon as he is to use any flake he may pick up. At the same time he may get an affection for a certain tool and only keep it for the purpose for which it is most suitable ... This casualness is what makes it so hard to say specifically that a tool is used for any one purpose ... (Horne and Aiston 1924:91-2).
Nevertheless, Aiston concludes "... but in describing them I have carefully asked [the Aborigines] until I could arrive at what was aimed at in each particular tool, and so have classed them" (Home and Aiston 1924:92). Other researchers have taken a similar tack, determining through observation and questioning what a stone tool was "aimed at," while discovering, often to their surprise, that the tasks to which these tools were applied epitomised "casualness and opportunism" (Gould et al. 1971:154). Hence, Aborigines sometimes used large blade "fighting knives" as woodworking adzes (Cane 1992:25), "points" as wood engraving tools (Davidson 1935:162; Kamminga 1985), "woodworking adzes" as butchering tools (Thomson 1964:418) and throwing weapons (Davidson 1935:160), and "axes" as knapping hammers (Smythe 1878:379). Observers have also seen Aborigines scavenge byproducts from the manufacture of "aimed at" tools for use in various tasks. For example, Home and Aiston (1924:87, 101) comment that flakes produced in making "ideal stones" were scavenged for use as "casual tools." Basedow (1925:363-4) describes how flakes struck in manufacturing axe blanks were used as-is or retouched into scrapers, and how flakes apparently struck in retouching a scraper were mounted as barbs on wooden spear heads (367). Elkin (1948:111) describes how percussion flakes struck in making Kimberley points were used for "cutting flesh," and Tindale (1985:9) implies that pressure flakes were used in initiation ceremonies. Roth (1904:16) notes that the detritus of large blade manufacture might be retouched into scrapers, and Binford (1989:181-2) indicates that large blades might sometimes be reduced as cores for the production of small unmodified cutting tools. "Hyperflexibility" is suggested in some parts of Australia by the manufacture of complex wooden implements using, to a large degree, minimally modified or even unmodified pieces of stone (Mountford 1941; Thomson 1964:412-4; Hayden 1979; see also Gould et al. 1971:163, Gould 1978:819). Indeed, even carefully hafted stone tools sometimes consisted of naturally-occurring stones with little or no modification (e.g. Tindale 1965:133, 135, 160; Mountford 1941: 316).
A tension exists in Australian lithic studies because of an apparent contradiction in these observations. On the one hand, ethnographic accounts clearly indicate that the functions of stone tools was unstructured and highly flexible, while, on the other hand, it would appear from both archaeological analysis (e.g., Akerman 1976; Akerman et al. 2002:18-20; Hiscock 1993; Moore 2003a, b) and ethnographic evidence (Roth 1904; Spencer and Gillen 1904; Elkin 1948; Baines 1866) that certain stone reduction sequences were quite structured, or, to use Aiston's phrase, were "aimed at" specific artefact forms. So, in light of overwhelming evidence for unstructured tool use by Australian Aborigines, just how structured were the Aboriginal approaches to tool manufacture? One way to examine this question is to explore the source of blanks for "aimed at" retouched forms through reduction sequence modelling. A reduction sequence model is a way of describing the manipulations a stone knapper applied to a block of stone. The model can be presented in the form of a flow chart with technological choices shown as pathways or "trajectories." In a rigid, inflexible technological structure, flake blanks for "aimed at" forms derive from the culmination of the technological steps within a single reduction trajectory. Technological structure in this case consists of a set of distinct trajectories with minimum movement of flake blanks laterally between them. These lateral movements are referred to here as "crossovers." In a flexible technological structure, on the other hand, flake blanks for "aimed at" forms derive from any of a number of highly interconnected trajectories.
This study examines the structural rigidity of the lithic technology reflected in a large surface assemblage recovered on the upper Georgina River in northwest Queensland (Figure 1). Prior studies into this assemblage are summarised and unpublished elements of the technology are described. The various technological reconstructions are drawn together to provide a reduction sequence model of lithic technology on the Georgina River (Figure 2). The results suggest a relatively rigid structure to stone artefact manufacture. A comparison is then made between the reduction sequence models for the upper Georgina River, the Hunter Valley of New South Wales, and Tasmania (Moore 2000a, b), and the implications for Australian lithic studies are discussed.
[FIGURES 1-2 OMITTED]
Project setting and methodology
A recent bridge construction project on the Georgina River at Camooweal resulted in the mitigation of impacts to chert outcrops used by Aboriginal people to manufacture a variety of stone tools. The recovery of 16,645 stone artefacts (Table 1) provided an excellent opportunity to explore the ways in which the Aborigines on the upper Georgina River approached their stone technology. Decomposition of the local dolomite has created abundant residual deposits of fine-grained to relatively coarse-grained nodular and tabular marine chert. The larger nodules at Camooweal occur in the 100 to 150 mm size range and are sub-spherical in shape, consisting of sections of spherical or ovate nodules broken and rounded through weathering. The finer-grained pieces of chert generally occur as nodule sections measuring 100 mm and smaller. Tabular chert can occur in flat slabs up to 500 mm across and 60 to 70 mm thick, although internal cracks limit the effective size of tabular pieces to a maximum of about 200 mm across. The chert is available as a continuous lag deposit in some places forming pavements of gravel and cobbles. Concentrations in other areas drop to 1 or 2 pieces per square metre.
The impact area for the bridge replacement project consisted of a corridor measuring 5.5 km long and, on average, 70 m wide. The undisturbed portion of the corridor measured 23.63 hectares. Aboriginal stone artefacts were scattered in a continuous distribution across this area. The Indjilandji traditional owners required the salvage of all surface artefacts and a sample of subsurface artefacts. The lithic analysis was conducted hand-in-hand with collection cataloguing. The debitage collection was sorted into various flake types, including early reduction flakes, blades, uniface retouching flakes, and tula adze flakes. Formed objects included early reduction cores, blade cores, tula cores, tula adzes, bifaces, retouched blades, and retouched early reduction flakes (after Moore 2000a, b; 2003a, b, c). "Early reduction" refers to debris that is morphologically similar to relatively non-diagnostic debris produced in the very earliest stage of biface reduction. The label is retained for consistency's sake although only one of the Camooweal trajectories involves biface production.
The analytical methods used in this study are described in detail in Moore (2003c). Briefly, the analytical approach began with the archaeological assemblage and traced the steps of manufacture backwards to the initial blank. Analysis of the stone assemblage initially focused on formed objects. The order of flake removals was inferred through examination of arrises for tell-tale signs of flake overlap, such as truncated hackles on intruded scars and minute step fracturing on complete scar margins. Recurrent patterns of flake overlap became apparent and allowed the identification of several technological "groupings." The order in which flakes were removed from formed objects has logical implications regarding the scar patterning to be expected on the dorsal surfaces of flakes, and, to verify these expectations, flake scar analysis was also applied to the dorsal surfaces of flakes and blades. Knapping experiments were conducted throughout the study to examine the feasibility of the reconstructions and to provide insight into interpretations of the knapping products and by-products.
Five reduction trajectories have been identified, including a "Levallois" core trajectory, a prismatic core trajectory, a tula adze trajectory, an "early reduction" core trajectory, and a biface trajectory (Figure 2). The model covers the range of variation seen in the archaeological assemblage from Camooweal, although it should be noted that not all of these trajectories were necessarily part of the organisation at any single point in time. Further research on dated assemblages from the region may allow chronological sorting of various elements of the model. Although a "use" category is included, this is portrayed in a highly simplistic fashion, as will be discussed below.
Trajectory 1: "Levallois" cores and blades
The "Levallois" trajectory (after Dortch 1972, 1977; Dortch and Bordes 1977) involved producing "target blades" with subparallel edges that contracted to a pointed tip (Figure 3d-h) (Moore 2003a). Target blades are the so-called "aimed at" products of the reduction trajectory (see Roth 1904). Three reduction methods were employed to produce blades of this morphology directly from the core by controlling the blades' termination, although the shapes of target blades were sometimes adjusted by retouching. The Camooweal Standard Method involved creating a single straight arris on the blade core face by removing two corner blades and then striking the target blade down this arris (Figure 3a, b). The blade core face was cleared by removing two new corner blades, thereby setting up a new arris for striking an additional target blade. The process was repeated until the core was exhausted. The Camooweal Multiple Scar Method involved striking one or two target blades down the lateral arrises created by the first target blade, then clearing the blade face by corner blade removals (Figure 3c). Again, the process was repeated until the core was exhausted. The Camooweal Shark's Tooth Method involved creating a straight arris on the blade core face by removing two corner blades, and then isolating a platform at the apex of the arris by removing two corner blades/flakes across the platform surface (Figure 4a, b). The isolated platform was then struck, producing a target blade with a distinctive "shark's tooth" shape (Figure 4c-e). Target blades made from Camooweal chert were used ethnographically as hafted knives and perhaps spearpoints (Roth 1904) and most measured between 30 and 90 mm long.
[FIGURES 3-4 OMITTED]
Trajectory 2: prismatic cores and blades
The second blade production trajectory involved producing small blades from prismatic cores (Moore 2003a) (Figure 5a, b). These small blades are less than 50 mm long, 5 to 25 mm wide, roughly parallel-sided, and with squared-off, rather than pointed, distal ends (Figure 5c-g). The latter feature resulted from blades carrying through to the bottom of the core. A comparison of prismatic core scar and target blade widths demonstrates some morphological overlap although small blades struck from prismatic cores are generally narrower and thinner than target blades struck by the Levallois methods. Prismatic cores are typically single-platform. A variant seen in the Camooweal region involved striking blades down the faces of chert tablets, resulting in a relatively flat core. The platform for this was the natural edge of the tablet or a unifacially prepared platform. Some tabular cores were reduced from two directly opposed platforms.
[FIGURE 5 OMITTED]
Trajectory 3: tulas
The manufacture of flake blanks for tula adzes at Camooweal demonstrates a technical mastery of the physics controlling bulb formation (Moore 2003b). Ethnographic accounts note that the goal of this trajectory was to create a flake blank with a pronounced bulb of percussion (Roth 1904:17). The Aboriginal knapper then produced a gouge-shaped woodworking edge on the blank by bisecting this bulb of percussion through unifacial retouch. The prominence of tula bulbs--and hence the degree of curvature of the gouge-shaped edge--was enhanced by directing the percussion blow directly into the concavity created by the previous flake removal, creating a flake with a "gull wing" cross section (Figure 6b). The width and length of the gull wing flake was controlled by the width and thickness of the stone cobble (Figure 6a). The tula flake was retouched around its perimeter and dorsal convexity was sometimes decreased by thinning the dorsal surface from the dorsal edge of the platform (Figure 6c). The tula was hafted at the end of a handle or spearthrower and progressively resharpened until it was too small to be used, at which point is was discarded as a worn-out slug (Figure 6d).
[FIGURE 6 OMITTED]
Trajectory 4: early reduction cores and flakes
Unifacial cobble cores
Five cores were recovered in the project area consisting of relatively flat chert cobbles (or, in one case, a large flake blank) that have been unifacially flaked around 1/2 to 2/3 of the perimeter (Figure 7) (Table 2). The major flake scars are relatively large although smaller peripheral trimming scars are present on several. The latter scars sometimes dramatically undercut the core face in a similar manner to that seen on horsehoof cores. This raises the possibility that these artefacts served both as cores for flake blanks and as woodworking tools (Akerman 1993). Three of the five chert cores are heavily patinated by iron oxide staining. Another specimen, discovered on Nowranie Creek, 10 km east of the Camooweal project area, is also heavily patinated. Patination of this nature can be a poor indicator of age due to the myriad of microenvironmental factors which lead to patina development (Luedtke 1992:108-11). Indeed, similar staining was observed on tula adze flakes in the Camooweal assemblage, a technology dating to the late Holocene. However, these unifacial cobble cores are the only artefact category where the majority of the artefacts are patinated and this might indicate a relatively greater antiquity for the type.
[FIGURE 7 OMITTED]
Bifacial cobble cores
Sixteen bifacially flaked chert cobbles were recovered from the Camooweal project area (Moore 2003c) (Figure 8). About one-half of the periphery of each of these artefacts is bifacially flaked with the remaining portion consisting of the cobble's cortical surface. Most of these cores measure smaller than 70 mm in maximum dimension and between 30 and 45 mm thick. The cores were manufactured by hard-hammer alternate flaking delivered to well-spaced unprepared platform surfaces formed by previous flake removals. This resulted in relatively large, deep percussion scars and a sinuous core edge. No attempt was made by the Aboriginal craftsmen to trim or straighten the edges of these cores.
[FIGURE 8 OMITTED]
Multiplatform early reduction cores
These cores are small--all are less than 55 mm in largest dimension--and are marked by two or more platform surfaces. Nine multiplatform early reduction cores were recovered. Core rotation presumably maximised the number of flakes struck from the core. Flakes were struck along areas of high mass rather than down straight arrises.
Flake blank early reduction cores
Flake blank cores are large flakes which have themselves been reduced into relatively large flakes. The latter flakes were presumably for retouching and use. The flake scars on these artefacts are larger and less systematically arranged than the scars seen on retouched flake tools although it can be difficult to differentiate between the two. It is possible that some of these cores functioned both as the source for flake blanks and as tools.
Six flake blank cores in the Camooweal assemblage were reduced on their dorsal face, three were reduced on their ventral face, and six were reduced on both faces. As with multiplatform cores, flakes were struck along areas of high mass rather than down straight arrises. In some cases a flake might be struck down the edge of the blank but in most cases the flakes were struck directly into the blank's mass.
Certain flakes, referred to here as "contact removal flakes," were produced by a percussion blow that propagated onto the flake blank's ventral surface and removed the ring crack and part of the bulb of percussion. The latter features are preserved on the dorsal surface of the contact removal flake (Figure 9). Five contact removal flakes were recovered in the Camooweal assemblage. Most of these flakes probably derived from the reduction of the ventral surface of flake blank cores, although one is interpreted as resulting from retouching a small blade. Contact removal flakes are technologically similar to certain "Kombewa" flakes (e.g., Owen 1938; Araho et al. 2002; Dag and Goren-Inbar 2001) but at Camooweal they appear to be a byproduct of reduction rather than a deliberately-produced flake.
[FIGURE 9 OMITTED]
Other early reduction cores
In total, 438 early reduction cores were recovered from the Camooweal project area that fall within the "other" category. These cores were reduced in an opportunistic fashion and most are marked by one to five medium- to large-sized flake scars originating from one platform. Many are cores rejected due to knapping error, such as hinge or step fracturing, or due to flaws in the raw material. It is suspected that some are failed cores from the initial stages of blade manufacture or from tula blank manufacture. These cores are the by-products of on-source stone extraction and reduction and are typical of those seen on many Aboriginal quarries throughout the region.
Retouched early reduction flakes
In total, 13 retouched early reduction flakes were identified in the Camooweal assemblage. Due to the difficulties in separating deliberately retouched flakes from those produced by trampling (McBrearty et al. 1998)--cattle damage is a serious post-depositional factor in the collection's preservation--a conservative approach was taken to identifying retouched flakes. A flake was considered deliberately retouched if the flake scars were distributed regularly along a sizable portion of the flake's margin. Hence, the total number of deliberately retouched flakes in the assemblage is probably underestimated. All of the identified retouched flakes were trimmed towards the dorsal surface using the ventral surface as the platform.
Comparison of early reduction core products
The various core types produced flakes of different morphologies. The flakes struck from bifacial cobble cores tended to be wider than long. This contrasts with, for example, small blades struck from prismatic cores (Figure 10). The shapes of scars on other core types fall between these two extremes. The dimensions of flakes struck from flake blank cores and multiplatform cores are similar and most are no greater than 50 mm in maximum size. The unifacial cobble cores produced the largest flakes, with a sizable proportion measuring between 50 and 55 mm in greatest dimension.
[FIGURE 10 OMITTED]
Trajectory 5: bifaces
Three biface variants are represented in the Camooweal area (Moore 2003c). One variant includes large hand axe-like bifaces made by hard-hammer percussion using a variety of edging, edge-turning, and face contouring techniques. These bifaces average 143 mm long, 85 mm wide, and 33 mm wide (Figure 11). The second variant includes small bifaces made by hard hammer and/or soft hammer percussion. These are absent from the project assemblage although one repatriated museum specimen from Camooweal is curated by the Indjilandji. Well provenanced specimens have been discovered northeast of Camooweal (Moore and Sachs 1999; Brumm 2001). The third variant includes an edge-ground metabasalt axe reworked by bifacial flaking. One flake from this process is present in the Camooweal assemblage (Figure 12). A hammer-dressed dacite axe was recovered (Figure 13), and, although percussion scars are not apparent, the tool may have been initially manufactured by bifacial flaking followed by pecking, as described by Spencer and Gillen (1904:656-7).
[FIGURES 11-13 OMITTED]
Bifacial reduction techniques were used in a number of reduction trajectories. Most notable of these is the alternate flaking technique used to produce the bifacial cobble cores. In addition, small bifacial early reduction cores were recovered which were reduced by similar edge turning and contouring techniques used in manufacturing large bifaces, combined with the face reduction strategy used in striking small blades from chert tablets. They differ from the latter by having early reduction flakes struck across the face from various points around the margin (Figure 14). Bifacial flaking was used to prepare the basal end of some target blades for hafting, and the edges of target blades sometimes show sporadic bifacial flaking presumably to repair or resharpen worn edges.
[FIGURE 14 OMITTED]
On-source and off-source reduction
The spatial distribution of stone reduction is an important factor to consider in reconstructing a reduction sequence. Central to this is determining what sort of reduction occurred at the stone source and the form in which stone was carried away. Stone knapping that occurred at the source is "on-source" reduction and that which occurred away from it is "off-source" reduction. Defining this in the Camooweal assemblage is in a sense moot as the entire project area and the surrounding landscape is a chert source. However, by tracking various flake blanks through the technological system and assessing the spatial distribution of these artefacts across the Camooweal region, it is possible to postulate the forms in which chert was carried from the source. By extension, the procurement debris abandoned on-source can be identified.
The sizes of the flakes struck from several types of early reduction cores are compared in Figure 15. Figure 16 shows that many of the flake blanks for retouched flakes and flake blank cores are larger than most of flakes produced from the core types plotted in Figure 15. (Since attrition to length and/or width occurs during reduction, the artefact dimensions in Figure 16 are minimum values.) The flake blanks for these forms were probably struck from the large early reduction cores in the "other" category. Most of the latter cores are quite large and hundreds are present in the project area and surrounding region. A proportion of these were evidently discarded on-source after the production of large blanks for reduction as cores away from the procurement locality, or for retouching into tools. This form of stone procurement--the reduction and discarding of early reduction cores at the stone source and transport of flake blanks away from the area--has been observed ethnographically (Aiston 1928:123; Tindale 1965:140-1; Gould et al. 1971:160-161).
[FIGURE 15 OMITTED]
In a similar fashion, gull wing flakes for tula adze manufacture appear to have been struck from cores which were discarded at the stone source. The ubiquity of tula-related debris in the Camooweal assemblage--76 tula cores and 752 tula flakes were recovered--is characteristic of on-source reduction. Tula cores are absent from caches of gull wing tula adzes (Hewitt 1976; Hiscock 1988a). It also appears that tula flakes were often retouched at the source prior to export. Nevertheless, the presence in caches of unretouched gull wing flakes with fully retouched tula adzes indicates that quantities of both retouched and unretouched blanks were carried from the source area.
In total, 564 blade cores and 6765 flakes struck in reducing these cores were recovered from the Camooweal project area. The majority of this material derived from the various Levallois methods of producing target blades. As with the tula manufacturing trajectory, finished target blades are regularly encountered at off source sites although blade cores resulting from the Camooweal Standard, Multiple Scar, and Shark's Tooth methods are rare in off-source settings. This is strong evidence that target blades were the result of on-source Levallois core reduction. On-source blade manufacture is amply represented in the ethnographic literature (Roth 1904:17-8; Spencer and Gillen 1904:641-2; Aiston 1928:128; Thomson 1983; Binford and O'Connell 1989; Jones and White 1988; Paton 1994:174). Both Aiston (1928:123) and Binford (Binford and O'Connell 1989; Binford 1989) report the transporting of target blade cores. The transported cores observed by Aiston were apparently reduced into small blades, although he implies that this was not a common practice. The core observed by Binford was transported by a vehicle.
Sites in off-source areas in the Camooweal region contain three types of cores: prismatic blade cores, small early reduction cores, and flake blank early reduction cores. These items were struck to provide small blades and flakes for various purposes and were carried across the landscape to produce tools on an as-needed basis. Again, this aspect of Aboriginal stone technologies was observed ethnographically (Roth 1904:22; Aiston 1928:123; Tindale 1985:4-5). Small early reduction cores and prismatic blade cores were manufactured on high-quality stone nodules or tablets, and reduction at the source may have been minimal. Conversely, large flakes for flake blank cores were struck from early reduction cores abandoned at the quarry, as discussed previously. Flake blanks may have been deliberately produced for reduction as early reduction cores due to the ideal platform angles naturally present around a rake's perimeter. Also, flake production essentially provided a window through the stone, ensuring that it was unflawed and of good quality.
Both flakes and rejected early stage items from the large biface trajectory were recovered from the Camooweal project area. This indicates that the manufacture of these items occurred at the raw material source. Completed examples were probably exported from the source area for use elsewhere on the landscape. Small bifaces may have been manufactured at the stone source area, as described ethnographically by Love (1936:74), although little direct evidence for this was encountered in the Camooweal assemblage.
Bifacial and unifacial cobble cores may have served both as producer cores and as woodworking tools. Flakes struck from the unifacial varieties were generally somewhat larger than those struck from the other types of offsource early reduction cores.
As modeled in Figure 2, the lithic technology of the Camooweal Aborigines involved five reduction trajectories: the Levallois core trajectory (Trajectory 1), the prismatic core trajectory (Trajectory 2), the tula trajectory (Trajectory 3), the early reduction core trajectory (Trajectory 4), and the biface trajectory (Trajectory 5). As such, these categories may reflect, to reiterate Aiston's phrase, the "aimed at" products of reduction (Home and Aiston 1924:92). Most of these "aimed at" categories are also indicated in Roth's (1904) historical description of the Camooweal knappers and/or other ethnographic studies in northern Australia (discussed in Moore 2003a, b, c).
So just how "structured" was the Camooweal knappers' technological system? The model presented in Figure 2 suggests that there were only two crossovers between trajectories and, indeed, one of these is conjectural.
The conjectured crossover involved the production of blades from prismatic cores for retouching into hafted spearpoints/knives. In Moore and Sachs' (1999) study northeast of Camooweal, small unifacially retouched spearpoints/knives were made on materials procured nearby, and these were the same materials comprising the various prismatic cores (usually fine-grained quartzite). Unlike blades from the various Levallois reduction methods, these blades were not preformed on the core, but were usually retouched into a triangular shape. Clear evidence for this crossover in the Camooweal assemblage is not present, but, since the low-end of the size range for target blades overlaps the sizes of scars on prismatic cores, such a crossover is conceivable. A second, less ambiguous crossover involved striking non-gull wing flakes from early reduction cores and retouching them into tula adzes. Eight early stage tulas were recovered which were produced from the early reduction core trajectory rather than from the typical gull wing trajectory. These compose 9% of the retouched tula assemblage (Moore 2003b).
Variation exists in the internal character of the reduction trajectories. While the Levallois core, prismatic core, tula adze, and biface reduction trajectories are linear and rather inflexible in approach, the early reduction trajectory reflects a more fluid and complex strategy for creating blanks for tools. In the latter trajectory, five distinct core reduction approaches appear to have provided early reduction flake blanks, including bifacial cobble cores, unifacial cobble cores, tabular bifacial cores, multiplatform cores, and flake blank cores. It is possible that certain cores were themselves used as tools. Hence, this trajectory reflects a high degree of internal flexibility.
Variations in technological structure occur in nearby areas. The reduction sequence at Lawn Hill Station, some 130 km to the north of Camooweal, involved a crossover between the tula adze trajectory and a small blade-making trajectory. At Lawn Hill, tula cores produced both blanks for adzes and blanks for heat-treating and reducing into small blades. The flake blanks for this were not gull wing in morphology (Hiscock 1988b:148, 273). This type of crossover has not been observed in the Camooweal assemblage and implies a somewhat less rigid technological structure.
Few reduction sequence models are available in Australia. Exceptions include examples of single trajectories or segments of trajectories (e.g., Akerman 1976; Akerman et al. 2002; Hiscock 1993; Mulvaney 1998:814; Webb et al. 1994:114-24; Witter 1988), or a general example encompassing all Australian technologies (Flenniken and White 1985). An approach falling somewhere in the middle includes the author's work in the Hunter Valley of New South Wales and Tasmania.
The Hunter Valley technological structure is linked by a myriad of crossovers (Moore 2000a). Flake blanks struck from early reduction cores might be reduced into early reduction flakes and flakes from both might become blanks for retouched forms such as microliths, eloueras, and other retouched flakes. Early reduction flakes struck from these cores, or from small cobbles, might become blanks for reduction as blade cores, which in turn provided blanks for retouched artefacts. It would appear that early reduction cores carried in the toolkit were capable of providing virtually the entire range of blanks for the various retouched forms, although getting from core to the presumed "aimed at" product, such as backed microliths, might sometimes involve rather rigid reduction trajectories (e.g., Hiscock 1993). These formal aspects of the technology are perhaps best conceived of as "subtrajectories" linked by a high degree of flexibility in the earlier stages of core reduction.
Tasmanian technology involved the production of early reduction flake blanks for retouching into "scrapers," the presumed "aimed at" forms. Blank production was often done by freehand percussion but exploitation of small pebbles often involved bipolar percussion with retouching applied to the resulting bipolar flakes (Moore 2000b). Stones of virtually any size might be reduced into the retouched artefacts within this strategy, and in this sense the technological structure falls even further towards the flexible end of the gradient. However, an entirely distinct trajectory involving the production of ballywinne stones is indicated for a small area in the north part of the island (Webb et al. 1994:114-24). It would appear that neither the debris or "aimed at" product of this trajectory crosses over with other elements of the lithic technology.
This study has examined the flexibility of lithic technology from the standpoint of the production of blanks for "aimed at" forms. A high degree of technological rigidity was encountered at Camooweal which is a bit surprising given the general characterisation of Aboriginal technology as casual, opportunistic, and ad hoc (e.g., Gould et al. 1971). In the upper Georgina River region, this characterisation best applies to the realm of artefact function rather than to artefact manufacture. Ongoing residue analysis of the Camooweal assemblage suggests that waste flakes produced in the different trajectories were sometimes used for a number of different tasks involving the processing of both animal and vegetable materials (T. Loy pers. comm. 2001; Loy and Nugent 2002). This accords well with the ethnographic observations discussed earlier. The reduction model in Figure 2, focused on stone artefact manufacture rather than function, presents tool use in an overly simplistic fashion. Residue analysis suggests that incorporating a more realistic portrayal of use would involve overlaying an intricate web of arrows which link the flakes produced in various stages of "aimed at" artefact manufacture, as well as the "aimed at" forms themselves, with a myriad of functions. This would reflect the high degree of functional flexibility that appears to be inherent in the upper Georgina River lithic technology.
Given the lack of functional rigidity, why were "aimed at" forms produced in the first place? Indeed, ethnographic accounts indicate that many of the functions for these "aimed at" stone artefacts were also accomplished by ad hoc flakes or non-stone elements of the material culture. For example, hardwoods were successfully worked with unmodified or minimally modified stones, ad hoc flakes, and informal retouched flakes, rather than tula adzes or hafted axes (Mountford 1941; Thomson 1964:412-4; Hayden 1979); spears were tipped with wood, bone, or teeth rather than stone (Davidson 1934; Kamminga 1985:8); yams were dug and processed with digging sticks and large blades rather than bifaces (O'Connell 1974; see discussion in Moore 2003c); and ceremonial fighting was conducted using small flakes and blades rather than large leilira blades ostensibly made for the purpose (Aiston 1928:129; Horne and Aiston 1924:96-7; see discussion in Moore 2003a). White's conclusion that "... the majority of stone tool forms were not necessary, in a utilitarian sense, at all" (1977:26), resonates in this context.
Several authors have suggested that the linkage between technological and social domains of human existence is a defining characteristic in the emergence of modern human behaviour some 40,000 years ago in Eurasia (Kuhn and Stiner 1998; Mithen 1996a, b). According to this view, the explosion of stone artefact forms and technologies at the beginning of the Upper Palaeolithic reflects an expansion and embedding of stone technology from the economic domain into the social and symbolic realms of culture. Abundant Australian ethnographic evidence exists to indicate that stone artefacts did, in fact, perform social and symbolic roles; hence, it is conceivable that the discordance discussed here between ad hoc stone function and rigid technological structure is related to this phenomenon. Given the possibility that Australia was colonised by 50,000 BP, Australasian studies of stone artefacts assemblages offer an important, and largely untapped, source of data on the emergence and global spread of modern human behaviour (Foley and Lahr 1997). An understanding of the technological structure behind stone artefact assemblages is an important prerequisite for tapping this potential.
Insights by Klm Akerman, Adam Brumm, Iain Davidson, and an anonymous reviewer improved this research enormously, but errors in fact or logic are my own. Fieldwork for the Georgina River Bridge Project was carried out jointly by the Dugalundji Aboriginal Corporation, representing the Indjilandji people, and ARCHAEO Cultural Heritage Services, Brisbane, on behalf of the Queensland Department of Main Roads. Special thanks to Ruby and Colin Saltmere for articulating the needs and wishes of the Aboriginal community, facilitating in-depth discussions regarding traditional culture and technology, and ensuring that the archaeology was treated properly from the outset. Ann Wallin, Michael Strong (ARCHAEO Cultural Heritage Services), and Queensland Department of Main Roads personnel provided crucial support for the archaeological mitigation and lithic analysis. Figure 13 was prepared by Michael Strong.
Table 1. Stone Artefacts Recovered in the Camooweal Project Area. Artefact Type Number Unifacial Cobble Core 5 Bifacial Cobble Core 16 Flake Blank Core 15 Small Bifacial Tablet Core 2 Multiplatform Early Reduction Core 9 Other Early Reduction Core 438 Standard Blade Core 528 Multiple Scar Blade Core 11 Shark's Tooth Blade Core 11 Prismatic Core 14 Tula Core 76 Large Biface 8 Early Reduction Flake 7789 Contact Removal Flake 5 Uniface Retouching Flake 19 Biface Thinning Flake 14 Tula Flake 752 Blade 6765 Eraillure 4 Reflex Flake 2 Janus Flake 1 Axe Reworking Flake (Biface Thinning) 1 Retouched Target Blade 34 Other Retouched Blade 22 Retouched Early Reduction Flake 13 Tula Adze 89 Dacite Edge-Ground Axe 1 Quartzite Hammerstone 1 Total 16,645 Table 2. Dimensions (mm), Patination (P) and Blank Type (T) of Unifacial Cobble Cores, Camooweal Assemblage. Ref L W Th P T 10666-9 93 79 40 Heavy Flake 777-3 107 93 47 Heavy Cobble 87500-2 78 84 39 Minimal Cobble 124-1 92 101 63 Heavy Cobble 23666-1 91 80 48 Minimal Cobble
Aiston, G 1928 Chipped stone tools of the Aboriginal tribes east and north-east of Lake Eyre, South Australia. Papers and Proceedings of the Royal Society of Tasmania, pp. 123-131.
Akerman, K. 1976 Notes on the experimental manufacture of long blades and points by percussion flaking. University of Queensland Occasional Papers in Anthropology 6:117-128.
Akerman, K 1993 The status of the horsehoof core. Records of the Australian Museum, Supplement 17:125-127.
Akerman, K., Fullagar, R. and van Gijn, A. 2002 Weapons and wunan: Production, function and exchange of Kimberley points. Australian Aboriginal Studies 1:14-42.
Araho, N., Torrence, R. and White, J. P. 2002 Valuable and useful: Mid-Holocene stemmed obsidian artefacts from West New Britain, Papua New Guinea. Proceedings of the Prehistoric Society 68:in press.
Baines, T. 1866 On the flint-flakes in the drift and the manufacture of stone implements by the Australians. Mackies Geological and Natural History Repository 1865-67:258-262.
Basedow, H. 1925 The Australian Aboriginal. F. W. Preece and Sons, Adelaide.
Binford, L. R. 1989 An Alyawara day: Making men's knives and beyond. In L. R. Binford Debating Archaeology, pp. 172-189. Academic Press: New York.
Binford, L. R. and O'Connell, J. F. 1989 An Alyawara day: The stone quarry. In L. R. Binford Debating Archaeology, pp. 121-146. Academic Press: New York.
Brumm, A. 2001 Cultural heritage survey and assessment of the proposed Gulf Communities fibre optic cable route, near Camooweal, northwest Queensland. Unpublished report to Network Design and Construction, Brisbane. Ms. on file, ARCHAEO Cultural Heritage Services: Brisbane.
Cane, S. 1992 Aboriginal perceptions of their stone technology. Australian Archaeology 35:11-31.
Dag, D. and Goren-Inbar, N. 2001 An actualistic study of dorsally plain flakes: a technological note. Lithic Technology 26:105-117.
Davidson, D. S. 1934 Australian spear-traits and their derivations. Journal of the Polynesian Society 43:41-162.
Davidson, D. S. 1935 Archaeological problems of Northern Australia. Journal of the Royal Anthropological Institute 65:145-183.
Dortch, C. E. 1972 An Archaeological Site in the Chichester Range, Western Australia: Preliminary Account. Journal of the Royal Society of Western Australia 55:65-72.
Dortch, C. E. 1977 Early and Late Stone Industrial Phases in Western Australia. In R. V. S. Wright (ed) Stone Tools as Cultural Markers: Change, Evolution and Complexity, pp. 104-132. Prehistory and Material Culture Series 12. Australian Institute of Aboriginal Studies, Canberra.
Dortch, C. and Bordes, F. 1977 Blade and Levallois technology in Western Australian prehistory. Quartar 27-28:1-19.
Elkin, A. P. 1948 Pressure flaking in the Northern Kimberley, Australia. Man 130:110-113.
Flenniken, J. J. and White, J. P. 1985 Australian flaked stone tools: A technological perspective. Records of the Australian Museum 36:131-151.
Foley, R. and Lahr, M. 1997 Mode 3 technologies and the evolution of modern humans. Cambridge Archaeological Journal 7:3-36.
Gould, R. A., Koster, D. A. and Sontz, A. H. L. 1971 The lithic assemblage of the Western Desert Aborigines of Australia. American Antiquity 36:149-169.
Gould, R. A. 1978 The anthropology of human residues. American Anthropologist 80:815-835.
Hayden, B. 1979 Palaeolithic Reflections: Lithic Technology and Ethnographic Excavations Among Australian Aborigines. Australian Institute of Aboriginal Studies, Canberra.
Hewitt, R. 1976 Aboriginal adze stone hoards found on the Arcoona Plateau near Woomera, South Australia. Australian Archaeology 4:14-52.
Hiscock, P. 1988a A cache of tulas from Boulia District, westem Queensland. Archaeology in Oceania 23:60-70.
Hiscock, P. 1988b Prehistoric settlement patterns and artefact manufacture at Lawn Hill, Northwest Queensland. Unpublished PhD thesis, University of Queensland.
Hiscock, P. 1993 Bondaian Technology in the Hunter Valley, New South Wales. Archaeology in Oceania 28:65-76.
Hiscock, P. 1998 Revitalising artefact analysis. In Archaeology of Aboriginal Australia, A Reader, edited by Tim Murray, pp. 257-265. Allen and Unwin, St Leonards.
Horne, G. and G. Aiston 1924 Savage Life in Central Australia. Macmillan, London.
Jones, R. and White, N. 1988 Point blank: Stone tool manufacture at the Ngilipitji Quarry, Arnhem Land, 1981. In B. Meehan and R. Jones (eds) Archaeology with ethnography: An Australian perspective, pp. 51-93. Department of Prehistory, Australian National University.
Kamminga, J. 1985 The pirri graver. Australian Aboriginal Studies 2:2-25.
Kuhn, S. L. and Stiner, M. C. 1998 Middle Palaeolithic `creativity': reflections on an oxymoron? In S. Mithen (ed) Creativity in Human Evolution and Prehistory, pp. 143-164. Routledge, London.
Love, J. R. B. 1936 Stone-Age Bushmen of To-Day: Life and adventure among a tribe of savages in north-western Australia. Blackie and Son, London.
Loy, T. H., and Nugent, S. 2002 Residue Analysis of a Portion of the Georgina River Bridge Collection, Camooweal, Queensland. Molecular and Residue Analysis Service, MARS Report 200. University of Queensland, St. Lucia.
Luedtke, B.E. 1992 An archaeologist's guide to chert and flint. University of California: Los Angeles, Institute of Archaeology, Archaeological Research Tools 7.
McBrearty, S., Bishop, L., Plummer, T., Dewar, R. and Conard, N. 1998 Tools underfoot: Human trampling as an agent of lithic artifact edge modification. American Antiquity 63:108-129.
Mithen, S. 1996a The Prehistory of the Mind: The Cognitive Origins of Art, Religion and Science. Thames and Hudson, London.
Mithen, S. 1996b Social learning and cultural traditions: interpreting Early Palaeolithic technology. In J. Steele and S. Shennan (eds) The Archaeology of Human Ancestry: Power, Sex, and Tradition, pp. 207-229. Routledge, London.
Moore, M. W. 2000a Technology of Hunter Valley microlith assemblages, New South Wales. Australian Archaeology 51:28-39.
Moore, M. W. 2000b Lithic technology in Tasmania. Archaeology in Oceania 35:57-73.
Moore, M. W. 2003a Australian Aboriginal blade production methods on the Georgina River, Camooweal, Queensland. In press, Lithic Technology.
Moore, M. W. 2003b Australian Aboriginal tula adzes on the Georgina River, Camooweal, Queensland. Submitted.
Moore, M. W. 2003c Australian Aboriginal biface reduction techniques on the Georgina River, Camooweal, Queensland. Submitted.
Moore, M. W. and Sachs, K. 1999 Indjilandji cultural heritage survey, NORQEB transmission Line, Gunpowder to Century Zinc. Unpublished report to Sinclair Knight Merz Pty Ltd and North Queensland Electricity Bureau. Ms. on file, Gordon Grimwade and Associates: Yungaburra, Queensland.
Mountford, C. P. 1941 An unrecorded method of manufacturing wooden implements by simple stone tools. Transactions of the Royal Society of South Australia 65:312-316.
Mulvaney, K. 1998 The technology and Aboriginal association of a sandstone quarry near Helen Springs, Northern Territory. In R. Fullagar (ed.) A Closer Look: Recent Australian Studies of Stone Tools, edited by Richard Fullagar, pp. 74-94. Archaeological Methods Series 6. Sydney University.
O'Connell, J. F. 1974 Spoons, knives and scrapers: the function of Yilugwa in Central Australia. Mankind 9:189-194.
Owen, W. E. 1938 The Kombewa Culture, Kenya Colony. Man 38(218):203-205.
Paton, R. 1994 Speaking through stones: A study from northem Australia. Worm Archaeology 26:172-184.
Roth, W. E. 1904 Domestic implements, arts, and manufactures. North Queensland Ethnography Bulletin No. 7. The Home Secretary's Department, Department of Public Lands, Brisbane.
Smythe, R. B. 1878 The Aborigines of Victoria: with notes relating to the habits of the natives of other parts of Australia and Tasmania. Government Printer, Melbourne.
Spencer, B. and Gillen, F. J 1904 The northern tribes of Central Australia. Macmillan: London.
Thomson, D. F. 1964 Some wood and stone implements of the Bindibu Tribe of Central Western Australia. Proceedings of the Prehistoric Society 17:400-422.
Thomson, D. F. 1983 Donald Thomson in Arnhem Land. Compiled and introduced by N. Peterson. Currey O'Neil: Melbourne.
Tindale, N. B. 1965 Stone implement making among the Nakako, Ngadadjara, and Pitjandjara of the Great Western Desert. Records of the South Australian Museum 15:131-164.
Tindale, N. B. 1985 Australian Aboriginal techniques of pressure-flaking stone implements: Some personal observations. In M. G. Plew, J. C. Woods and M. G. Pavesic (eds) Stone Tool Analysis: Essays in Honor of Don E. Crabtree, pp. 1-33. University of New Mexico Press, Albuquerque.
Veth, P., Hiscock, P., O'Connor, S. and Spriggs, M. 1998 Gallus on the crossroads: diffusionist models for Asian artefact `traditions' in Australia and the last hurrah for cultural evolutionism. The Artefact 21:14-18.
Webb, J. A., McKay, M. D. A. and Sagona, A. G. 1994 Analysis of the finds. In A. Sagona (ed) Bruising the Red Earth: Ochre Mining and Ritual in Aboriginal Tasmania, pp. 79-132. Melbourne University Press.
White, J. P. 1977 Crude, colourless and unenterprising? Prehistorians and their views on the stone age of Sunda and Sahul. In J. Allen, J. Golson and R. Jones (eds) Sunda and Sahul: Prehistoric Studies in Southeast Asia, Melanesia, and Australia, pp. 13-30. Academic Press, London.
Witter, D. 1988 From butchering caribou to butchering stone. In B. Meehan and R. Jones (eds) Archaeology with Ethnography: An Australian Perspective, pp. 33-41. Australian National University Press, Canberra.
Wright, R. V. S. 1977 Introduction and two studies. In R. V. S. Wright (ed) Stone Tools as Cultural Markers: Change, Evolution and Complexity, pp. 1-3. Australian Institute of Aboriginal Studies, Canberra.
Department of Human and Environmental Studies, University of New England, Armidale NSW 2351
|Printer friendly Cite/link Email Feedback|
|Author:||Moore, Mark W.|
|Publication:||Archaeology in Oceania|
|Date:||Apr 1, 2003|
|Previous Article:||Productivity, production and settlement in precontact Rarotonga, Cook Islands.|
|Next Article:||Hammer dressed stone hatchets in the Lake Eyre Basin. (Research Reports).|