Ochre in hafting in Middle Stone Age southern Africa: a practical role.

Wadley, Lyn ; Williamson, Bonny ; Lombard, Marlize 等

Introduction

There appears to have been a demand for colouring material during the Middle Stone Age (MSA) and large quantities have been found at some African sites. At the enigmatic site of Lion Cavern, in the eastern part of South Africa, tons of specular haematite were mined from iron pods perhaps earlier than 40 000 BP. Mining hammers in the form of grooved heavy duty stones were found alongside MSA tools in a Lion Cavern adit (Beaumont 1973: 140). Colouring material was, however, used systematically even earlier than this at c. 200 000 BP at Kapthurin, Kenya (McBrearty 2001), and possibly earlier still at Twin Rivers in Zambia where pigments were found associated with a Lupemban Industry (Barham 2002). Barham (2002: 186-7) estimates that at Twin Rivers about 60 kg of colouring material were recovered from the 1950s excavations, while about 16 kg were recovered from his own sampling in the 1990s.

There is considerable speculation amongst archaeologists about the potential uses of the early pigment. Some archaeologists, for example Knight et al. (1995), Deacon (1995), McBrearty & Brooks (2000) and Watts (2002), claim that the presence of ochre in MSA occupations, and particularly ochre crayons, implies ritual, such as body-painting, a hypothesis that has stemmed from analogy with modern hunter-gatherers living tens of thousands of years later. For example, !Kung girls observed in north-western Botswana in historic times used ochre body-paint for puberty and marriage rituals (Marshall 1976: 277). There is, however, no way of testing whether ancient people in Africa practised body painting. Recently another use for ochre was discovered at Blombos Cave in the Western Cape where two pieces of engraved ochre were found (Henshilwood et al. 2002). The larger Blombos ochre tablet has a crosshatched design engraved inside several broken boundary lines and the tablet was found in a layer believed to date to c. 77 000 BP. The decorated ochre is interpreted by Henshilwood et al. (2002) as evidence for the presence of symbolic behaviour and therefore cultural modernity by c. 77 000 BP.

In addition to these more symbolic uses, there is some evidence that ochre had practical functions. For example, ochre has been shown to have medicinal purposes as an antibacterial agent (Mandl 1961: 196; Velo 1984) and it therefore inhibits collagenase, making it ideal for tanning, softening and colouring leather (Audouin & Plisson 1982). Watts (2002) contests the use of ochre as a hide preservative, argues for its primarily decorative use on hides and claims that there is no functional reason for the use of red rather than other colours of ferruginous material. It is true that the extinct/Xam from Bushmanland, South Africa, used it to colour leather bags (Bleek & Lloyd 1911) and the San practice of colouring some leather bags continues even today in parts of Botswana and Namibia (LW personal observation). It is also true that there are many ways of tanning leather and that ochre is not an essential component of the process, for example, southern Ethiopian hide workers use only water on their skins (Kimura et al. 2001). However, ochre is a very successful tanning agent. In 1906 Steinman observed Tehuelches in Argentina tanning guanaco hides with ochre and fat and Sollas described the same process for hide tanning in Tasmania (Audouin & Plisson 1982:57). Furthermore, Audonin and Plisson (1982) demonstrate that ochre tanning of hides is beneficial, particularly for preventing or reversing the process of decay. One of their experiments involved treatment with ochre of a three day old moose skin that was already beginning to putrefy. The rotting skin was scraped with simple flakes and ochre was then applied everywhere, except the tail. Notwithstanding the disadvantageous start to the experiment the ochred skin dried quickly and became thinner and softer. The tail section which had not been treated with ochre became green and malodorous. In a further experiment they treated two pieces of skin cut from the same ox hide, one was rubbed with yellow ochre and the second with red ochre. The yellow ochred piece remained stiff, thick and rough whereas the piece treated with red ochre dried rapidly, lost 1mm in thickness and started to get soft. Audouin and Plisson therefore concluded that ochre stops hide from rotting and helps to dry it rapidly. Additionally, they concluded that the higher iron content of red ochre produces better quality hide than yellow ochre. The use of ochre seems to be particularly appropriate where some delay in the processing of the hide threatened decay.

We do not yet know whether people living in the MSA used ochre in the tanning of their hides, but it seems likely that ochre was part of the processing of skins. The evidence comes from eight MSA bone tools from Blombos Cave that were stained with a dark red ochre that had been incorporated into the tool polish (Henshilwood et al. 2001:661). Experimental piercing of ochred hides by d'Errico showed that ochre permanently stains bone awls used in this way and the authors accordingly suggest that people were processing hides with ochre and bone awls (Henshilwood et al. 2001:661-2).

Our own research suggests that there may be yet another use for pigment. Preliminary residue analyses of stone tools from Rose Cottage Cave, South Africa (Williamson 1997), revealed the presence of red ochre on many classes of tools. A subsequent investigation of the distribution of residues on backed tools from the Howiesons Poort Industry of the cave showed that ochre and plant material residues were often concentrated on or near the backed edge (Tomlinson 2001). The coincidence suggested that ochre might be part of the hafting process, but the sample was small and it was therefore decided to attempt a more detailed study of ochre residues remaining on stone tools.

Tools from Sibudu Cave were chosen for the new hafting study. Sibudu Cave is approximately 12 km inland of the Indian Ocean and about 40 km north of Durban in KwaZulu-Natal. The surface of the site contains debris from Iron Age occupations and multiple MSA layers are immediately below the Iron Age horizons. The tools that comprise the residue study come from a variety of MSA layers that date between about 26 000 BP and c. 60 000 BP (Wadley 2001). One hundred and fifty-four points, 83 scrapers and 217 flakes make up most of the collection analysed here, but 77 retouched tools from several other tool classes were also included. The tools are made predominantly of hornfels and dolerite, but there are a few quartz pieces.

Methodology

Most of the tools were nor touched prior to the study; they were removed from the excavation using sterilised plastic tweezers and were bagged individually in plastic. This careful removal means that the tools for residue analysis were not examined in detail prior to the microscopic analysis. Consequently we did not know in advance of the study whether a tool would have ochre on it. Soil samples were removed from the locality where tools were collected and the soil samples were analysed to check that residues on the tools were not emanating from the cave deposit. The few specimens that were handled during excavation were lightly rinsed with water in the laboratory and, after drying, were then separately bagged in plastic. The surfaces of the tools were microscopically examined at magnifications ranging from 50x to 800x under incident light. Cross-polarised light was used to identify starch grains and to distinguish between cellulose and collagen. The Hemastix[R] test (Williamson 2000) was used to identify blood films and to distinguish blood residues from resin deposits. The resin deposits are clearly of plant origin, but they cannot at present be more exactly identified. Animal residues are not as common as plant residues, and they will not be discussed in detail here. While some of the ochre deposits were visible to the naked eye, others required microscopy for their recognition, but relatively low magnification (50-100x) was sufficient to enable the reliable plotting of red ochre residues.

Sketches of dorsal and ventral surfaces were made for each tool and residue and damage positions were noted, using a grid that divides the tool into proximal (near or on the tool platform or bulb of percussion), medial (in the middle of the tool) and distal (the tip) sections. in Microscopy Study 1 there is also a spatial category for residues on the 'working edge', that is on the retouched or utilised edge of the tool, which may be a lateral in the case of scrapers. Other spatial categories combine some of the elements, for example, proximal and medial sections, medial and distal sections and all over the tool surface.

In Microscopy Study 2 a formula was devised to equalise the surface areas represented by the proximal, medial and distal portions of the points. This was done to allow for the skewing of results that could be caused by the varying surface areas that make up the three portions within the triangular or lanceolate shape of a point. Each 5[mm.sup.2] block covered by the tool was counted as one whole block. Where the block was not entirely filled, but contained more than 50 per cent of tool surface, it was counted as one whole block. Where the block contained less than 50 per cent of tool surface, it was ignored. The following formula was then used to establish comparable surface area values (c values) for the residues in each of the proximal, medial and distal portions of each point:

All the c values and statistical tests are available elsewhere (Lombard 2003) because space does not permit them to be included here.

c value = f of 5[mm.sup.2] blocks covered with residue x 100/ f of 5[mm.sup.2] blocks covered by the tool

Microscopy Study 1

The studied artefacts comprise retouched points, scrapers, a variety of other retouched tool types and unretouched flakes (Table 1). Among points examined, 29 out of 104 had ochre residues on them, 30 out of 83 scrapers had ochre on them, 23 out of 77 pieces of "other" retouch had ochre and 26 out of 113 flakes had ochre on them. With only a few exceptions there are more plant tissue, starch and white starchy residues on tools that have ochre on them than on tools without ochre. The microscopic analysis of the 108 Sibudu flake tools with ochre residues (Table 2) shows that the majority of the ochre residues occur on proximal, medial or a combination of these two positions. When tool classes are individually examined, 69% of points have their ochre residues occurring on the proximal portions, medial portions or a combination of these two positions. Scrapers have 80 per cent of ochre residues on proximal, medial or combinations of these two positions and flakes have 47 per cent of their ochre residues in these positions. Flakes have the highest proportions of ochre residues on their working edges (27 per cent) and they also have the highest proportions of ochre residues all over their surfaces. It appears that flakes were used for processing ochre.

Encouraged by these results, which strongly suggest that ochre residues are most likely to be found on the bases of retouched tools like points, it was decided to explore this conclusion further.

Microscopy Study 2

Multiple lines of evidence are always more convincing than single lines so, in addition to the residue analysis, ML decided to study macro-fractures and other physical damage as well as the residues of a new sample of 50 retouched points. Experimental studies were not conducted before examining the archaeological material because analyses of fracture patterns have been conducted elsewhere by Ahler (1977), Barton & Bergman (1982), Bergman & Newcomer (1983), Fisher et al. (1984), Odell & Cowan (1986), Shea (1988), Holdaway (1989) and Geneste & Plisson (1993). Research on the fracture mechanics of brittle solids was conducted by Cortell (1972) and Cortell and Kamminga (1987).

Where experimental points have been lashed to hafts, the sudden loading of weight from impact onto edges in contact with the haft bindings often produces clusters of bending fractures on the lateral margins (Shea 1988: 443). Small clusters of bending fractures tend to be feather-terminated and of relatively small size. As might be expected, the use of resins or mastic for the attachment of a point to its shaft tends to shield the tool laterals from this type of fracturing. Prolonged tool use and repeated weight loading can also produce worn, polished areas where the artefact abrades against the handle (Shea 1988: 443). Crushing on the proximal end of the tool (distinct from the crushing that takes place when flakes are produced by the bipolar flaking technique) can be a further manifestation of hafting (Abler 1977; Odell & Cowan 1986; Holdaway 1989: 80). Crushing is defined as severe subsurface fracturing and the severity of fracturing associated with crushing requires force through impact or percussion (Ahler 1977: 309). In the case of crushing on the proximal end of tools, fracture direction indicates that force was primarily directed through the edge directly into the centre of the body of the tool. Finally, hafting can be indicated by the deliberate modification through thinning of the proximal end of a tool by flaking of the bulb of percussion and/or part of the dorsal ridges of the platform.

Fifty Sibudu retouched points (whole and broken) were analysed for macro-fractures and only three of the 24 whole points were found to be without polish, fractures or crushing. Clusters of bending fractures were macroscopically observed on 88 per cent of the whole points and this damage was to the laterals on the proximal and/or medial portions (Table 3; Figure 1). Crushing of the proximal ends was observed on 83 per cent of whole points. However, all the crushed proximal ends had also been modified through thinning and it therefore could not be established whether the crushing resulted from hafting or knapping. Removal of part of the proximal dorsal ridges or other parts of the proximal end (Figure 2) to thin the base of the tool is an attribute present on 84 per cent of the whole points.

[FIGURES 1-2 OMITTED]

Ochre traces were then examined, and were documented as concentrated on the proximal and medial portion, on the distal portion or all over. Ninety per cent of the points and point fragments revealed ochre traces. Sixty-eight per cent of the whole points (n=24) have ochre concentrated on their proximal and medial portions (as in Figure 3, Table 4) whereas only 13 per cent of whole points have ochre concentrations exclusively on their distal portions (Table 4). Furthermore, of the whole points with ochre concentrated on proximal and/or medial portions, 68 per cent also display damage in the form of clusters of bending fractures (Figure 1), removal of part of the proximal dorsal ridges (Figure 2), proximal crushing and polish on the proximal surfaces or dorsal ridges at the proximal end of the tool (Figure 4) (Table 3). The experimental work cited previously suggests that all of these damage types are likely to have been caused by hafting.

[FIGURES 3-4 OMITTED]

When the c values (explained in the methodology section) are calculated to make allowance for different surface areas of the three tool portions (Table 5) the situation changes. The c values make it clear that the most significant concentrations of ochre (58 per cent) are, in fact, on the medial portions of the whole points. [chi square] tests using the grid analysis data recorded in Table 5 (for the 24 whole points) are informative, too, for they suggest that the scarcity of ochre on distal portions and the concentration of ochre on combined proximal and medial portions is unlikely to have occurred by chance. The frequency of ochre residues on medial and proximal portions combined was compared with the frequency of ochre on distal portions of the whole points and also with 5mm grid blocks in which no residues were found. At one degree of freedom the probability of there being no significant difference between the observed and expected frequencies is less than 1 per cent for proximal/ medial portions and less than 0.1 per cent for distal portions of whole points.

Discussion

Microscopy study 1 showed that ochre is often associated with plant derived residues such as plant tissue, starch, white starchy deposit which appears to be cooked starch, resin and plant exudate and often these plant-derived residues are coincident (Table 1). Although multiple, serial use of tools is likely, it is also possible that the combined residues are providing evidence for an activity involving the joint use of ochre and plant material. This interpretation might be supported by the microscopic examination by BW of three crayons of colouring material from Sibudu. All three had starch grains on them and two were extensively worn and polished. One crayon in particular had copious amounts of plant residue in the form of resins, plant tissue, white starchy deposit and starch grains (1-3 [micro]m in diameter). This suggests that the crayons were used in conjunction with plant materials, but it is also possible that ochre pieces were rubbed with sticks in order to extract powder. In cases where plant residues and ochre are combined at the base of tools it is possible that both were ingredients of the glue that had once attached the stone tool to a handle or shaft. The cooked starch would then be the result of heating the glue to make it pliable. Replication work that will be commented on shortly suggests that plant residues may also been accidentally incorporated in the resins.

More plant tissue and starch was found on the tools that have ochre than on those that do not, but, with the exception of the scraper and flake categories, there is a tendency for more resin to occur on tools that do not have ochre on them. The origin of the plant residues is unknown, but their presence on the tools is certainly the result of deliberate action by prehistoric people. We know this because soil samples taken from where the tools were lying in the cave contained only small traces of starch grains and plant residues compared to those on the tools. Plant resins are used in glues for tool hafting; even today in southern Ethiopia the hide workers use stone scrapers mounted in mastic that is largely constituted of hardwood resins (Kimura et al. 2001). Resin is obtained from many southern African woody plants, such as Ozoroa spp, Arctopus echinatus, Widdringtonia cedarbergensis, Commiphora pyracanthoides, Combretum erythrophyllum, Protorhus longifolia and Acacia spp (Grant & Thomas 1998:236, 318, 350; Van Wyk & Gericke 2000: 140, 278, 230, 250, 284). Both Protorhus longifolia and Acacia spp occur near Sibudu Cave today and they also occurred in the area in the MSA (Wadley in prep.). Gum from Protorhus longifolia was used in KwaZulu-Natal in historic times for attaching blades into assegai handles (Grant & Thomas 1998: 236). It could be argued that it is possible to accumulate resin residues on tools in the process of cutting or scraping resinous wood, but this would not account for the presence of the white starchy material which seems to be cooked starch.

The position of ochre on the tools is considered extremely important for the study at hand. The ochre that is on or near the working edges of tools may have been processed by those tools, but we suggest that where ochre is found on proximal and/or medial parts of tools, this ochre was used as part of the hafting process of those tools. Hafting with ochre is inferred, first, from the position of ochre on the proximal and medial parts of tools and, secondly, from its association with damage to the tools in positions where handles or shafts may have been attached. The most convincing evidence for the use of ochre in the hafting process thus comes from Microscopy Study 2 where, in addition to residues, 50 retouched points were examined for the presence of fractures, crushing and polish. Of the 24 whole points with ochre collected on their proximal and/or medial sections, 68% exhibit compelling physical evidence for hafting, that is, clusters of bending fractures on proximal laterals, proximal crushing, polish on the proximal surfaces or dorsal ridges at the proximal end of the tool and deliberate removal of all or part of the proximal dorsal ridges. In the Near East a wide range of tools dating from the Mousterian to the Bronze Age are ochre stained, but they appear (based on micro-wear studies) to have been wrapped in ochred hide and then hailed into bone handles (Buller 1988). The Sibudu tools do not have the mat polish and capillar striations that characterise the tools examined by Buller and we conclude that the ochre stained Sibudu tools obtained their ochre from a different form of processing.

An explanation for why ochre would have been used in the process of hafting tools seems to come from experimental work conducted in France. Replication experiments by Allain and Rigaud (1986:715) show that an adhesive recipe using mastic, wax or resin requires an inert powder such as ochre for at least two reasons. First, the ochre acts as an emulsifier because wax and resin would not otherwise mix well and, secondly, ochre encourages the hardening of the mastic when it dries. Allain and Rigaud (1986) made glue in their experimental study by heating one part beeswax, four parts resin and one part yellow ochre. At about 120[degrees]C the products liquidise and then solidify on cooling. Heating of ferruginous material to between 230 and 250[degrees]C during the making of the glue would automatically transform the yellow goethite to a red iron oxide. There is, however, another reason for using filler in adhesive recipes: pure resins are too brittle when they are heated alone and they would not resist high impact pressure (Rots 2002:57-59). Thus Australian aborigines use a mixture of vegetal fibre, ochreous dust and sand with their resins (Rots 2002:60).

Recent replication studies by one of us (LW) confirm that the heating of an ochre and plant resin paste is appropriate when attempting to set a stone tool into a wooden shaft. The heated paste, made of Acacia karoo gum and ochre by LW, begins to set within minutes of the stone being inserted into it (Figure 5). Left-over paste also hardens quickly and experiments suggest that left-over paste cannot be used. Reheating makes the paste a gritty mess and although the addition of a few drops of water reconstitutes the paste, the water appears to destroy the emulsion and prevent the paste from setting. More replication work needs to be undertaken because Acacia is the only species worked with thus far. When collecting gum from trees, by scraping it from wounds in the trunk, plant fibres are automatically included in the resin. Thus plant fibres may be unintentionally incorporated into the adhesive.

[FIGURE 5 OMITTED]

Some of the Sibudu stone tools do appear to have resin or mastic mixed with their ochre residues, but not all do. The ochre stain may survive better than the organic components of the glue. An alternative possibility that should be considered is that the ochre came from binding the tool to a handle using a wet leather thong that had been treated with ochre during the tanning process, but this idea has less support than that of ochre mixed into the hafting paste. Twine may well have been used, but it may have been of vegetable rather than animal origin, a hypothesis supported by the presence of plant material on many tools. Many southern African woody plants produce excellent twine, for example, Acacia tortillis, Acacia robusta (Grant & Thomas 1998:322,360), Acacia karoo, Pouzolzia mixta, Adonsonia digitata, Hibiscus tiliaceus, Grewia flava, Terminalia sericea and a variety of palm trees (Van Wyk & Gericke 2000: 284, 232, 304, 308). Acacia, Grewia and palm trees were in the Sibudu area in the MSA (Wadley 2004). The twine may have been used to attach the stone tool to a handle by means of slot hafting, that is, wedging the stone into a cleft stick or bone and binding the composite tool. However, the notch technique of hafting (Shea et al. 2001: 811) would work equally well using twine. In southern Ethiopia, Gamo hide workers insert a scraper into the split end of a straight piece of wood and secure it with tightly wrapped cord only (Kimura et al. 2001: 47). It is possible that some MSA tools were also halted with twine alone and that no mastic residues would then be present on the stone tools. Anderson-Gerfaud (1990: 407-10) believes that some Mousterian tools were bound without mastic into their split hafts using vegetable string or animal materials. She believes, further, that hafting abrasion to the tool surfaces would not occur if the tool was cushioned by mastic.

This Sibudu Cave residue study suggests that ochre found on MSA tools resulted from at least two activities. The first involved the use of stone tools to process ochre, perhaps together with plant material, and the second entailed the use of ochre in the process of hafting tools. Sibudu Cave is not alone in having ochre used in the mastic for hafting purposes. Some Palaeolithic tools found in France have been similarly interpreted (Audouin & Plisson 1982; Beyries & Inizan 1982). A backed blade with ochre on its backing was recovered from the Magdalenian Gouy in northern France and, at Lascaux in south-west France, mastic mixed with red ochre was found on backed bladelets (Audouin & Plisson 1982: 52). At the Magdalenian site of la Garenne in central France ochre was found at the base of a bone point and on the internal faces of 13 out of 68 navettes, which are grooved bone rods that are thought to have been handles for the mounting of end-scrapers and possibly burins (Allain & Rigaud 1986: 715, 724). Ochre has also been found on lithics from a further eight French sites, but the position of the ochre is not documented (Allain & Rigaud 1986). Based on the French archaeological data and their own replication experiments Allain and Rigaud (1986: 715) also suggest that red ochre was part of a paste used to fix stone tools to shafts or handles. Their microscopic studies suggest that the ochre and mastic mixture was liquid when it was applied and that it dried after the stone insert was put in place. The mixture would have been heated in order to inch the components. In the Sibudu context, this would explain the presence of cooked starch on some of the tools.

At Twin Rivers, Zambia, the shift from the Acheulean industry to the Lupemban is accompanied by a range of hafted tools and pigment use (Barham 2002: 188). It is perhaps no coincidence that hafted tools and extensive pigment use appear simultaneously. In Kenya, backed blades from Enkapune Ya Muto, dated to between 50 000 and 40 000 BP, have red ochre residues on their backed portions (Ambrose 1998). The position of the ochre on at least one of the tools suggests that it had been hafted parallel to the long axis, with the sharp edge exposed. Closer to home, in South Africa, ochre was found on the retouched laterals of backed tools from the Howiesons Poort Industry at Rose Cottage Cave (Tomlinson 2001). Ochre was also noted on some Die Kelders MSA stone tools (Thackeray 2000: 157) and mastic was found on an MSA tool at Apollo 11, Namibia (Wendt 1976), but most other hafted tools or tools with traces of mastic have been found in Later Stone Age sites dating to the Holocene (Phillipson 1976; Schweitzer 1979; Binneman 1983; Binneman & Deacon 1986; Deacon 1979; Deacon & Deacon 1980; Mitchell 1995; Jerardino 2001). Jerardino notes that the mastic used for the mounting of a stone adze from Steenbokfontein Cave contained plant fibre (Jerardino 2001:862), but no ochre was observed in the mastic and it is possible that a different inert powder was used in this particular recipe. Interestingly, Jerardino (2001: 863) also recovered a cigar-shaped mastic object with folds and fingerprints that had probably been formed while it was soft and warm from heating. Jerardino interprets the mastic object as a glue-stick that would have been repeatedly heated when small quantities of glue were required. The surface of the object, excluding the tip, was covered with ochre, which Jerardino (2001: 864) interpreted as an incidental residue. In the light of the information provided by the French material and the stone tools from Sibudu, it is possible that the ochre on the mastic object was the loading agent that was added in the last stages of manufacturing the adhesive, after the glue-stick was heated in the fire.

Some archaeologists may venture the suggestion that the use of ochre in hafting was a symbolic rather than functional act. It is impossible to know whether symbolism was involved in the placing of ochre in mastic, but this study has been able to show that there are sound functional reasons for doing so. We do not disclaim the potential for the ritual use of the ochre; we merely point out here that ochre can play an important role in some utilitarian activities. Archaeologists can no longer assume that the presence of ochre in a site is automatically and exclusively equated with ritual and symbolic behaviour. Another important issue signalled by the discovery of ochre as part of the hafting process is that it appears to signal sophisticated technical knowledge. An understanding of the properties of the individual ingredients is implied both by the combination of several items for the hafting paste and the subsequent application of heat to the paste. Furthermore, the creation of the hafting pastes would appear to have involved planning to ensure that all the necessary ingredients were simultaneously available.

Table 1: Microscopy Study 1: Residues on Sibudu tools that
have ochre versus those without ochre

 white
 plant plant starchy
 tissue fibre starch deposit resin

 f % f % f % f % f %

Points with ochre 13 45 3 10 20 69 8 28 1 3
Points without ochre 20 27 6 8 24 32 22 29 12 16
Scrapers with ochre 15 50 8 27 19 63 11 37 8 27
Scrapers without ochre 13 25 5 9 20 38 13 25 12 23
"Other" retouch with
 ochre 4 17 1 4 14 61 8 35 3 13
"Other" retouch
 without ochre 9 17 3 6 21 39 9 17 9 17
Flakes with ochre 12 46 2 8 12 46 7 27 10 38
Flakes without ochre 8 9 6 7 21 24 9 10 15 17

 plant char-
 exudate coal ash Total

 f % f % f % f

Points with ochre 5 17 1 3 -- -- 29
Points without ochre 11 15 6 8 3 4 75
Scrapers with ochre 11 37 -- -- -- -- 30
Scrapers without ochre 7 13 2 4 -- -- 53
"Other" retouch with
 ochre -- -- 1 4 1 4 23
"Other" retouch
 without ochre 6 11 6 11 1 2 54
Flakes with ochre 1 4 2 8 -- -- 26
Flakes without ochre 4 5 3 3 -- -- 87

Table 2: Microscopy Study 1: Position of ochre on a sample
of Sibudu points, scrapers and flakes

 position of ochre on tool

 re-
 touched
 proxi- proxi- or uti-
tool mal medial mal & lized all
type end portion medial edge over Total

 f % f % f % f % f % f

Points 8 27 11 38 1 3 5 17 4 14 29
Scrapers 5 17 11 36 8 27 1 3 5 17 30
Flakes 2 8 8 31 2 8 7 27 7 27 26

Table 3. Microscopy Study 2: Position of polish, bending fractures
and crushing on 50 Sibudu points

 whole points (n = 24) f %

Polish on proximal ridges and/or surfaces 21 88
Clusters of lateral bending fractures on proximal and/or
 medial portions 21 88
Crushing and/or alteration of the proximal portion
 through trimming 20 83
No polish, fractures or crushing 3 13
Ochre in combination with polish, crushing or fractures
 on proximal and/or medial portions 16 67

 broken points: proximal and medial portions (n = 1)

Polish on proximal ridges and/or surfaces 1 100
Clusters of lateral bending fractures on proximal and/or
 medial portions -- --
Crushing and/or alteration of the proximal portion
 through trimming 1 100
No polish, fractures or crushing -- --
Ochre in combination with polish, crushing or fractures
 on proximal and/or medial portions 1 100

 broken points: distal and medial portions (n = 14)

Polish on proximal ridges and/or surfaces -- --
Clusters of lateral bending fractures on proximal and/or
 medial portions -- --
Crushing and/or alteration of the proximal portion
 through trimming -- --
No polish, fractures or crushing 14 100
Ochre in combination with polish, crushing or fractures
 on proximal and/or medial portions -- --

 broken points: distal portions (n = 11)

Polish on proximal ridges and/or surfaces -- --
Clusters of lateral bending fractures on proximal and/or
 medial portions -- --
Crushing and/or alteration of the proximal portion
 through trimming -- --
No polish, fractures or crushing -- --
Ochre in combination with polish, crushing or fractures
 on proximal and/or medial portions 11 100

Table 4. Microscopy Study 2: Position of ochre on 45 Sibudu points

 whole points (n = 24) f %

Ochre on proximal & medial portions 16 67
Ochre on distal (retouched) portions 3 13
Ochre all over 5 21

broken points: proximal and medial portions (n = 1)

Ochre on proximal & medial portions 1 100

broken points: distal and medial portions (n = 13)

Ochre on medial portions 9 69
Ochre on distal (retouched) portions 2 15
Ochre all over 2 15

 broken points: distal portions (n = 7)

Ochre on distal (retouched) portions 7 100

Table 5. The grid system analysis of 24 whole points.

 Tools with Results of grid system
 ochre analysis on 24 points

 frequency frequency
Tool portions of tools % of ochre c/value %

Proximal portions 23 of 27 85 68 6 25
Medial portions 32 of 39 82 63 14 58
Distal portions 35 of 49 71 33 4 17
 164 24 100

Acknowledgements

We should like to extend warm thanks to Dr Paola Villa for the translation of the French references and for commenting on this paper. Grant Cochrane kindly pointed out the need to consider surface area differences on the points. We thank the two anonymous referees whose comments have improved this paper. Financial assistance for the Sibudu Cave project comes from the University of the Witwatersrand and the NRF. Opinions expressed here are, however, the sole responsibility of the authors and cannot necessarily be attributed to either of the funding institutions.

References

AHLER, S.A. 1977. Functional analysis of nonobsidion chipped stone artefacts: terms, variables and quantification, in B. Hayden (ed.), Lithic usewear analysis: 301-28. New York: Academic Press.

ALLAIN, J. & A. RIGAUD. 1986. Decor et fonction quelques exemples tires du Magdalenien, L'Anthropologie 90:713-38.

AMBROSE, S.H. 1998. Chronology of the Later Stone Age and food production in East Africa. Journal of Archaeological Science 25: 377-92.

ANDERSON-GERFAUD, P. 1990. Aspects of behaviour in the Middle Palaeolithic: functional analysis of stone tools from southwest France, in P. Mellars (ed.) The emergence of modern humans: 389-418. Edinburgh: Edinburgh University Press.

AUDOUIN, F. & H. PLISSON. 1982. Les ochres et leurs temoins au Paleolithique en France: Enquete et experiences sur leur validite archeologique, Cahiers du Centre de Recherches Prehistoriques 8:33-80.

BARHAM, L.S. 2002. Systematic pigment use in the Middle Pleistocene of South-Central Africa, Current Anthropology 31:181-90.

BARTON. R.N.E. & C.A. BERGMAN. 1982. Hunters at Hengistbury: some evidence from experimental archaeology, World Archaeology 14:237-47.

BEAUMONT, P.B. 1973. The ancient pigment mines of southern Africa. South African Journal of Science 69: 140-6.

BERGMAN, C.A. & M.H. NEWCOMER. 1983. Flint arrowhead breakage: examples from Ksar Akil, Lebanon, Journal of Field Archaeology 10:238-43.

BEYRIES, S. & M-L. INZAN. 1982. Typologie, ochre, function, in D. Cahen (ed.) Tailler! Pour quoi faire: 313-22. Tervuren: Museroyal de l'Afrique centrale.

BINNEMAN, J. 1983. Microscopic examination of a halted tool, South African Archaeological Bulletin 38:93-5.

BINNEMAN, J. & J. DEACON. 1986. Experimental determination of use wear on stone adzes from Boomplaas Cave, South Africa, Journal of Archaeological Science 13:219-28.

BLEEK, W.H.I. & LLOYD, L.C. 1911. Specimens of Bushman folklore. London: George Allen & Company.

BULLER, J. 1988. Handling, hafting and ochre stains, in S. Beyries (ed.) Industries Lithiques Traceologie et Technologie Vol. 1: Aspects archeologiques: 5-32. Oxford: BAR International Series 411 (i).

CORTELL, B. 1972. Brittle fraction in compression, International Journal of Fracture Mechanics 8:195-208.

CORTELL, B. & J. KAMMINGA. 1987. The formation of flakes, American Antiquity 52:675-708.

DEACON, H.J. 1979. Excavations at Boomplaas Cave: a sequence through the Upper Pleistocene and Holocene in South Africa, World Archaeology 10:241-57.

--1995. Two late Pleistocene-Holocene archaeological depositories from the southern Cape, South Africa, South African Archaeological Bulletin 50:121-31.

DEACON, H.J. & J. DEACON. 1980. The hafting, function and distribution of small convex scrapers with an example from Boomplaas Cave, South African Archaeological Bulletin 35:31-7.

FISHER, A., P, VEMMING-HANSEN, & P. RASMUSSEN, 1984. Macro and microwear traces on lithic projectile points: experimental results and prehistoric examples, Journal of Danish Archaeology 3:19-46.

GENESTE, J-M. & H. PLISSON. 1993. Hunting technologies and human behaviour: lithic analysis of Solutrian shouldered points, in H. Knecht, A. Pike-Tay & R. White [TEXT UNREADABLE IN ORIGINAL SOURCE.] Lascaux: the complex record of the early Upper Palaeolithic: 117-35. Boca Raton: CRC Press.

GRANT, R. & V. THOMAS. 1998. Sappi tree sporting: KwaZulu-Natal coast and Midlands. Johannesburg: Jacana.

HENSHILWOOD, C.S., F. D'ERRICO, C.W. MAREAN, R. MILO & R. YATES. 2001. An early bone industry from the Middle Stone Age at Blombos Cave, South Africa: implications for the origins of modern human behaviour, symbolism and language, Journal of Human Evolution 41:631-78.

HENSHILWOOD, C.S., F. D'ERRICO, R. YATES, Z. JACOBS, C. TRIBOLO, G.A.T. DULLER, N. MERCIER, J.C. SEALY, H. VALLADAS, I. WATTS & A.G. WINTLE. 2002. Emergence of modern human behaviour: Middle Stone Age engravings from South Africa, Science 295:1278-80.

HOLDAWAY, S. 1989. Were there hafted projectile points in the Mousterian. Journal of Field Archaeology 16:79-85.

JERARDINO, A. 2001. Diversity in mastic-mounted stone adzes and the use of mastic in precolonial South Africa: evidence from Steenbokfontein Cave, Antiquity 75:859-66.

KIMURA, B., S.A. BRANDT, B.L. HARDY & W.W. HAUSWIRTH. 2001. Analysis of DNA from ethnoarchaeological stone scrapers, Journal of Archaeological Science 28:45-53.

KNIGHT, C., C. POWERS, & I. WATT. 1995. The human symbolic revolution: a Darwinian account, Cambridge Archaeological Journal 5(1): 75-114.

LOMBARD, M. 2003. Closer to the point: a residue and damage study of Middle Stone Age points from Sibudu Cave, South Africa. MSc dissertation, University of the Witwatersrand.

MANDL, I. 1961. Collagenasis and elastases. Advances in Enzymology 23: 164-264.

MARSHALL, L. 1976. The !Kung of Nyae Nyae. Cambridge: Harvard University Press.

MCBREARTY, S. 2001. The Middle Pleistocene of East Africa, in L. Barham & K. Robson-Brown (eds), Human roots: Africa and Asia in the Middle Pleistocene: 81-92. Bristol: Western Academic and Specialist Press.

MCBREARTY, S. & A.S. BROOKS. 2000. The revolution that wasn't: a new interpretation of the origin of modern human behaviour, Journal of Human Evolution 39:453-563.

MITCHELL, P. 1995. Revisiting the Robberg: new results and a revision of old ideas at Sehonghong Rock Shelter, Lesotho, South African Archaeological Bulletin 50:28-38.

ODELL, G.H. & F. COWAN. 1986. Experiments with spears and arrows on animal targets, Journal of Field Archaeology 13:195-211.

PHILLIPSON, D.W. 1976. The prehistory of Eastern Africa. Nairobi: British Institute in Eastern Africa. Memoir 6.

ROTS, V. 2002. Hafting traces on flint tools: possibilities and limitations of macro- and microscopic approaches. PhD dissertation, Katholieke Universiteit, Leeuven.

SCHWEITZER, F. 1979. Excavations at Die Kelders, Cape Province, South Africa: the Holocene deposits, Annals of the South African Museum 78:101-232.

SHEA, J.J. 1988. Spear points from the Middle Palaeolithic of the Levant, Journal of Field Archaeology 15:441-50.

SHEA, J.J., Z. DAVIS & K. BROWN. 2001. Experimental tests of Middle Palaeolithic spear points using a calibrated crossbow, Journal of Archaeological Science 28:807-16.

THACKERAY, A.I. 2000. Middle Stone Age artefacts from the 1993 and 1995 excavations of Die Kelders Cave 1, South Africa, Journal of Human Evolution 38:147-68.

TOMLINSON, N.E. 2001. Residue analysis of segments, backed and obliquely backed blades from the Howiesons Poort layers of Rose Cottage Cave, South Africa. MA project, University of the Witwatersrand.

VAN WYK, B-E. & N. GERICKE 2000. People's Plants: A guide to useful plants of southern Africa. Pretoria: Briza Publications.

VELO, J. 1984. Ochre as medicine: a suggestion for the interpretation of the archaeological record, Current Anthropology 25:674.

WADLEY, L. 2001. Preliminary report on excavations at Sibudu Cave, KwaZulu-Natal, Southern African Humanities 13:1-17.

--2004. Seeds from the Middle Stone Age of Sibudu Cave: the environmental implications, South African Journal of Science 100:167-173.

WATTS, I. 2002. Ochre in the Middle Stone Age of southern Africa: ritualized display or hide preservative? South African Archaeological Bulletin 57:1-14.

WENDT, W.E. 1976. 'Art Mobilier' from the Apollo 11 Cave, South West Africa: Africa's oldest dated works of art, South African Archaeological Bulletin 31:5-11.

WILLIAMSON, B.S. 1997. Down the microscope and beyond: microscopy and molecular studies of stone tool residues and bone samples from Rose Cottage Cave, South African Journal of Science 93:458-64.

WILLIAMSON, B.S. 2000. Prehistoric stone tool residue analysis from Rose Cottage Cave and other southern African sites. Ph.D. dissertation, University of the Witwatersrand.

Lyn Wadley * (1), Bonny Williamson * & Marlize Lombard *

* Archaeology, School of Geography, Archaeology and Environmental Studies, University of the Witwatersrand, PO Wits 2050, South Africa.

(1) (Email: wadleyl@geoarc.wits.ac.za)

Received: 15 October 2002; Revised: 1 October 2003; Accepted: 29 October 2003