lable at ScienceDirect

Journal of Archaeological Science 35 (2008) 2523–2531

Contents lists avai

Journal of Archaeological Science

journal homepage: ht tp: / /www.elsevier .com/locate/ jas

Hunting with Howiesons Poort segments: pilot experimental studyand the functional interpretation of archaeological tools

Marlize Lombard a,*, Justin Pargeter b

a Institute for Human Evolution, University of the Witwatersrand, PO WITS 2050, Johannesburg, South Africab Department of Archaeology, University of the Witwatersrand, PO WITS 2050, Johannesburg, South Africa

a r t i c l e i n f o

Article history:Received 31 December 2007Received in revised form 8 April 2008Accepted 8 April 2008

Keywords:Howiesons PoortMiddle Stone AgeExperimental archaeologyHunting behaviourSibudu CaveUmhlatuzanaKlasies River Cave

* Corresponding author.E-mail address: marlize.lombard@wits.ac.za (M. Lo

0305-4403/$ – see front matter � 2008 Elsevier Ltd.doi:10.1016/j.jas.2008.04.004

a b s t r a c t

For many decades the use of backed pieces from the Howiesons Poort, between about 70 ka and 55 kaago, in South Africa has been a point of discussion. Recently direct evidence has been provided toassociate these tools with Middle Stone Age hunting strategies. Yet, whether they were used to tiphunting weapons or as barbs remained an open question. In this paper we introduce a set of pilotexperiments designed to test the effectiveness of Howiesons Poort segments, the type fossils of theindustry, hafted in four different configurations as tips for hunting weapons. It is shown that the mor-phological type can be used successfully in this way. We present the results of a macrofracture analysisconducted on the experimental tools and compare these to results obtained from three Howiesons Poortbacked tool samples. By correlating experimental outcomes, macrofracture data and the interpretation ofmicro-residue distribution patterns, we provide some insight into the functional variables that might beassociated with Howiesons Poort segments.

� 2008 Elsevier Ltd. All rights reserved.

1. Introduction

Few stone tool categories have received as much attention as thebacked tools from the Howiesons Poort Industry in South Africa.From early on they were considered unusual. Initially, the backedtools drew attention because they seemed so similar to Later StoneAge and Upper Palaeolithic artefacts, but appeared much earlier inthe archaeological sequence (e.g. Deacon, J., 1995). When debatesabout the origins of modern human behaviour became increasinglyrelevant, they were studied in terms of their symbolic or stylisticpotential (e.g. Deacon, 1989; Wurz, 1999). Recently, a series ofstudies was launched to generate evidence about the functions andhafting technologies associated with Howiesons Poort segments,the backed tool category that is considered the type specimen forthe industry (e.g. Delagnes et al., 2006; Gibson et al., 2004;Lombard, 2005a, 2006, 2007a, 2008; Pargeter, 2007; Wurz andLombard, 2007).

This paper continues along this line of investigation, combiningarchaeological use-trace evidence with new macrofracture datafrom experimental work. We build upon previous interpretations ofthese tools as insets in composite (hafted) hunting weapons. As ourmethods are chosen explicitly to explore and assess this particularfunction we do not exclude the possibility that segments were also

mbard).

All rights reserved.

used for other activities. In fact, it is most likely their functionalflexibility that contributes to the repeated appearances of similarblade-based technologies across the globe over the last 300 ka or so(Barham, 2002; Bar-Yosef and Kuhn, 1999; Bleed, 2002; Elston andBrantingham, 2002; Hiscock, 2002; Lombard and Parsons, in press;Marean et al., 2007; McDonald et al., 2007; Mitchell, 1988, 1995;Soriano et al,. 2007; Wadley, 1996, 2005).

Increasing interest in Middle Stone Age hunting technologiesand their role in modern or complex human behaviour stimulatediscussions around all tool types that could have been used inweapon systems. One of the debates centres on the origins andantiquity of projectile weaponry, where weapons are not hand-delivered (spears), but projected with the aid of spear throwersor bows (darts and arrows) (e.g. Brooks et al., 2006; McBreartyand Brooks, 2000; Shea, 2006; Villa and Lenoir, 2006). Because oftheir morphology and/or the uncertainty about whetherHowiesons Poort segments were used to tip hunting weapons ornot, these tools are often excluded from morphometric studiesconcerned with the evolution of hunting technologies (e.g.Brooks et al., 2006; Shea, 2006). Our methods are not yet suitableto generate convincing evidence for or against either in-terpretation. However, with this paper we aim to provide anddiscuss some evidence that suggests Howiesons Poort segmentscould have been used to tip hunting weapons, and that stone-tipped hunting technology during this phase was variable andadaptable.

mailto:marlize.lombard@wits.ac.zawww.sciencedirect.com/science/journal/03054403http://www.elsevier.com/locate/jas

M. Lombard, J. Pargeter / Journal of Archaeological Science 35 (2008) 2523–25312524

It has always been assumed that backed artefacts from theHolocene were hafted as parts of weaponry (e.g. Bocquentin andBar-Yosef, 2004; Clark, 1977; Clark et al., 1974; Crombé et al., 2001;Nuzhnyi, 2000; Phillipson, 1976; Wadley, 1987). Previous experi-ments with backed microliths similar to those of the EurasianMesolithic have shown them to be useful when used to tip and/orbarb hunting weapons (e.g. Barton and Bergman, 1982; Crombéet al., 2001). Conversely, an experiment by Binneman and Hall withreplicated San (ethnographic hunter-gatherers) bows and arrows,where segments were hafted transversely or end-on, showed themineffective for penetrating a calf carcass (Binneman, 1994). Seeingthat most ideas about the possible uses of Howiesons Poortsegments have been based on Later Stone Age/San ethnographicexamples, the question of whether these segments could have beenused successfully to tip hunting weapons remained open-ended.The Pargeter (2007) experiments are the first to focus specificallyon testing whether Howiesons Poort-like segments, hafted ina variety of positions can be used effectively to tip huntingweapons. The resulting macrofractures therefore represent the firstmodern reference sample that is directly relevant to the study ofHowiesons Poort segments that could have been used in this way.

2. A short background to the Howiesons Poort of South Africa

Lombard (2005b) provided a full review of research conductedsince the Howiesons Poort was first identified elsewhere. Here webriefly highlight the most relevant points concerning the industry,its evolutionary themes and dating. The Howiesons Poort ischaracterised by a blade-based technology and found primarily insub-Saharan Africa south of the Limpopo River (Fig. 1). The labelwas first used by Stapleton and Hewitt (1927, 1928) to describea stone artefact assemblage from a small rock shelter in the EasternCape. They gave the main characteristics for the ‘‘Howiesons PoortSeries’’ as the presence of burins, large segments, obliquely pointedblades, and trimmed points. However, the Stapleton and Hewittassemblage is probably from a mixed context as more recentexcavations at stratified sites show that points are absent or rareduring the Howiesons Poort, and that the technocomplex is rathercharacterised by its distinctive backed pieces such as segments(also referred to as crescents or lunates) and trapezes made onblades (Singer and Wymer, 1982; Deacon, H.J., 1995; Deacon, J.,1995; Deacon and Wurz, 1996; Wadley, 2005, in press; Wurz, 1999,2000, 2002).

Fig. 1. Archaeological sites with the Howiesons Poort mentioned in the text.

The industry has been recorded at more than 20 sites in SouthAfrica, a number of which have been comprehensively dated(Feathers, 2002; Grün et al., 1990; Grün and Beaumont, 2001;Miller et al., 1999; Parkington, 1999; Tribolo et al., 2005; Valladaset al., 2005; Vogel, 2001; Wadley and Jacobs, 2006). Age calcula-tions obtained from various dating methods applied to samplesfrom Border Cave, Klasies River, Rose Cottage Cave and Diepkloofpoint to the placement of Howiesons Poort assemblages firmlywithin the period of 70 ka to 60 ka ago. Recent thermolumines-cence data support this consensus, indicating that the HowiesonsPoort appeared about 70 ka ago and that it was still in use at 54 kato 62 ka ago at sites such as Rose Cottage Cave (Valladas et al.,2005).

The age of the industry, its similarities to Holocene backed toolindustries, and the use of unusual raw materials resulted in far-reaching behavioural hypotheses concerning the origins of humanmodernity (Lombard 2005b). Some of these models discuss aspectssuch as mobility patterns, gift exchange, symbolic behaviour,technical conventions and the origin of San languages (e.g.Ambrose, 2006; Ambrose and Lorenz, 1990; Deacon, 1989, 1992;Deacon, H.J., 1995; Deacon and Shuurman, 1992; Deacon and Wurz,1996, 2001, 2005; Minichillo, 2006; Wurz, 1999). Most recently,hunting technologies other than those based on stone tools aretentatively associated with the Howiesons Poort. Faunal data fromSibudu Cave indicate a predominance of the small blue duiker(Philantomba monticola) in the Howiesons Poort assemblage,suggesting the use of traps to capture these small antelope (seeClark and Plug, in press; Wadley, 2006), and an argument fora bone-tipped bow-and-arrow technology at the site during thesame period was presented recently (Backwell et al., 2008).

3. The experiments

It has long been acknowledged that not all variables can betested with a single set of experiments, and that controlledexperiments are most useful when variables are limited in order toisolate and address specific questions (e.g. Ingersoll and Macdon-ald, 1977; Mathieu, 2002; Odell and Cowan, 1986; Shea et al., 2002).Within this paradigm, Pargeter’s work (Pargeter, 2006, 2007)represents the pilot study in a series of experiments that will eachseek to address specific questions regarding the use and hafting ofbacked tools from the Howiesons Poort. Shea (2006: 841) indicatesthat most researchers who have discussed Middle Stone Agebacked pieces as possible weapon armatures have proposed theywere mounted as barbs on the sides of spears and other weapons,rather than on the tip. Nevertheless, use-trace studies indicate thatat least some pieces could have been used to tip hunting weaponsand that there could have been variability in the preferred haftingconfigurations of such weapons (Lombard, 2005a, 2008; Wurz andLombard, 2007).

Thus, for the purposes of this study we are not concerned withvariability in aspects such as raw materials, velocity, weight, angleof impact or the stopping power of targets. Rather, the experimentswere designed explicitly to investigate whether segments, hafted infour different configurations, could function effectively as tips forhunting weapons, and to provide some initial macrofracture data.Sylvain Soriano prepared 33 flint segments that were used toconstruct 27 hunting weapons. This is an average sample size forexperimentation with a single morphological type used for a singlefunction, and proved adequate for testing the effectiveness of thedifferent hafting configurations. Although the sample size is toosmall for the statistical analysis of macrofracture frequencies, theinitial macrofracture data raised interesting questions with regardto the interpretation of some archaeological pieces, and some ofthese aspects will be explored and assessed during futureexperimentation.

M. Lombard, J. Pargeter / Journal of Archaeological Science 35 (2008) 2523–2531 2525

The morphological characteristics of the replicated segmentsare comparable to Howiesons Poort backed tools excavated fromsites such as Sibudu Cave, Rose Cottage Cave and Klasies River Cave.They are thus suitable for establishing whether Howiesons Poortsegments, as a distinct morphological type, can be used effectivelyto tip hunting weapons. The segments were glued into slotted andweighted pinewood shafts in four different hafting positions. Thesepositions included longitudinal (n ¼ 7) (Fig. 2a), transverse (n ¼ 7)(Fig. 2b), diagonal (7 ¼ ) (Fig. 2c), and back-to-back hafting (n ¼ 6)(Fig. 2d). The weapons were launched at a suspended impala(Aepyceros melampus) carcass using a machine built for thispurpose. In this way variations in load, velocity and angles ofdelivery that can be expected during hand delivery was controlled(e.g. Shea et al., 2002). The weapons were each fired into the carcassten times or until they became unusable. This was done to assessthe robustness of the hafting configurations, and because it is likelythat hunters would have re-used undamaged weapons until theyneeded re-hafting or became too damaged for effective use. For thefull description and protocol of the experiments we refer toPargeter (2007).

The 27 weapons were shot into the carcass 167 times. Eighty-five percent of the weapons were considered successful as theypenetrated the target’s hide and flesh. Of these successful weapons,33% penetrated deep into the carcass (�30 mm), and a further 33%of these penetrations could have damaged vital organs. Theremaining successful weapons caused less severe damage to theanimal’s hide or flesh. Only 15% of the replicated weapons wereunable to penetrate the target. Thirty-seven percent of the weaponssurvived all ten shots. The various hafting configurations showedslight differences in success rates. Of the seven weapons haftedlongitudinally, six successfully penetrated the carcass. The seventransversely hafted weapons lacerated, rather than penetrated theanimal’s hide and flesh. Three of these penetrated somewhat, butthe other four deflected off the animal causing cuts to the hide andtissue. Although such wounds would not kill or cripple an animal,they are sufficient to create a blood trail for tracking downwounded prey or introducing poison into the bloodstream.

Five diagonally hafted weapons successfully pierced theanimal’s hide and flesh. The slicing action of the diagonal cuttingedge caused deep penetrations and relatively wide lacerations. Thedepth of the cuts in combination with the barbs lodged these

Fig. 2. Different hafting configurations of the replicated hunting weapons; (a) longi-tudinal, (b) transverse, (c) diagonal and (d) back-to-back.

weapons firmly in the carcass (Fig. 3). Of the six back-to-backhafted weapons five penetrated the carcass’s hide and flesh. Again,the slicing action caused by the cutting edges of the segmentsproved to be very effective. These weapons created large gapingwounds in the carcass (Fig. 4). Although we do not consider theseexperimental results as direct evidence for the interpretation ofarchaeological material, we believe that they clearly indicate thatHowiesons Poort-type segments, hafted in various positions, couldfunction successfully as tips for hunting weapons. In order to startbuilding a comparable database we conducted a macrofractureanalysis on this first experimental sample.

4. Macrofracture analysis

4.1. Background

Many fractures on archaeological tools no doubt resultedaccidentally, but some fractures are the result of use (Barton andBergman, 1982; Bergman and Newcomer, 1983; Moss andNewcomer, 1982; Shea, 1988). There is a wide degree of variation infracture patterns (combinations of fracture types and positions)associated with impact damage because hunting weapons aresubject to a wide range of targets, velocities, angles of impact andfracture initiations (e.g. Odell, 1981). Thus, even though ‘‘typicalpatterns’’ cannot always be reliably defined, there are certainaspects of fracture mechanics that can allow the elimination ofother causes (Odell, 1981; Dockall, 1997). For example, whenisolated, some fracture types produced by impact have certain traitsthat distinguish them from those caused under other circum-stances (Geneste and Plisson, 1993; but also see Shea et al., 2002).

In order to establish a tightly defined and replicable analyticalmethod, and to build a directly comparable database, we closelyfollow Fischer et al. (1984). They conducted experiments to isolateand define types of fractures that could be considered diagnostic forimpact use of stone-tipped hunting weapons. The term ‘‘diagnosticimpact fracture’’ (DIF) refers to evidence for impact use in the formof fracture types that could hardly have originated in any other waythan through forceful longitudinal collisions (Fischer et al., 1984).These wear traces are also diagnostic of the fracture mechanics ofbrittle solids subjected to both static and dynamic loadings(Cotterell and Kamminga, 1987).

Many other researchers have conducted hunting experimentswith stone tools and have discussed resulting fractures (seeDockall, 1997 and references therein), and although terminologyfrequently differs, it seems that the formation of the fracture typesis similar regardless of the type of hunting experiment (Dockall,

Fig. 3. Deep penetration of a weapon tipped with a diagonally hafted segment.

Fig. 4. Large wound caused by a weapon tipped with segments in the back-to-backposition.

M. Lombard, J. Pargeter / Journal of Archaeological Science 35 (2008) 2523–25312526

1997: 321), tool morphology (Fischer et al., 1984), or raw materials(Lombard et al., 2004). However, it is important to note that,although we consider macrofracture analysis a useful preliminaryindicator of whether an archaeological tool class could have beenused to tip hunting weapons, we do not necessarily consider suchresults conclusive (e.g. Shea et al., 2002). Functional inferences forarchaeological samples, including hunting interpretations, shouldideally be assessed with additional strands of evidence such asmorphometric analyses and more extensive functional analysessuch as microwear and micro-residues.

4.2. Method

The simplest DIFs are step terminating bending fractures. Theseresult from longitudinal pressure from the distal and proximal endsof the objects. In the case of occasional damaging processes, such astrampling, bending fractures will initiate by means of transversepressure more or less perpendicular to the dorsal and ventral sidesof the objects. Probably the most easily recognisable DIFs arebending fractures from which ‘‘spin-offs’’ initiate (see Fischer et al.,1984). Bending fractures that result from pressure perpendicular tothe dorsal and ventral sides of the objects will result in spin-offs ononly one broad side to a limited extent. But on artefacts used asimpact tools, where the forces run parallel to the broad sides, thespin-off fractures can have considerable dimensions (>6 mm), andlong spin-offs can occur on one, or even both, sides. Spin-offfractures on both sides, initiating from the same bending fracture,can in practice occur in hardly any other way than through use asa hafted impact implement. This type of fracture is thereforeconsidered diagnostic for impact use, irrespective of thedimensions of the fractures (for further fracture definitions andillustrations see Fischer et al., 1984; Ho Ho Committee, 1979).

Another fracture associated with impact use sometimesresembles the effect of a burin blow (pseudo burination or impactburination). These fractures usually occur along either one of thelateral edges (Barton and Bergman, 1982; Bergman and Newcomer,1983). They are difficult to distinguish from intentional burination,but usually lack the negative bulb of percussion at the burin initi-ation or Herzian ripples usually associated with intentionalremovals (Lombard et al., 2004; Paola Villa, pers. commun., 2004).Dockall (1997) describes these fractures as lateral macrofractures,and reports that it was documented on several experimentalsamples used to replicate hunting activities and many

archaeological samples associated with hunting (see Dockall, 1997:326). We are unaware of such fractures having been recorded inassociation with other use-related activities. Other fracture typesthat occur on broken stone tools such as snap terminating andhinge or feather terminating bending fractures are recorded andpresented in the data. These fracture types could result fromimpact-use, but have also been recorded as a result of trampling oraccidental breakage and are therefore not considered unambiguousevidence for stone-tipped hunting weapons.

We follow Fischer et al. (1984), using only the four DIF types(step terminating bending fractures, unifacial spin-off fractures�6 mm, bifacial spin-off fractures and impact burination) asanalytical criteria. Some adaptations have been made to the Fischeret al. (1984) method in an effort to further reduce the margin ofinterpretative error on archaeological samples. For example, flakesmanufactured using the bipolar technique are exposed to forcessimilar to those that tips of hunting tools are exposed to. Odell andCowan (1986) addressed this problem, and established the majordifferences between use-impact fractures and bipolar knappingfractures, the most important being that the location of bipolardamage coincides with other knapping characteristics. Normallythese traces are not located along the distal ends of the piece, andare not in association with a usable sharp tip or edge that couldhave functioned as the business end of a hunting tool (Odell andCowan, 1986). Therefore, DIFs occurring in close association withstriking platforms, ripples and bulbs of percussion are not includedin the data. During the analysis of archaeological samples, consid-eration is also given to the morphology of the pieces, and whetherthe tools could have been used for other impact functions such aschopping (Fischer et al., 1984).

Fischer et al. (1984) showed that the morphology of the stonepieces did not affect the fracture types that developed as a result ofhunting use. DIFs were recorded on 39% of their experimentalarrowheads, ranging from transverse to acutely angled in shape,and 55% of the points that were used as tips for throwing spears. Alarge number of prehistoric flint points from various sites, differingin age, size and shape, also displayed DIFs, suggesting that thedefined characteristics of impact function are universal for stonetips of flint and related raw materials (Fischer et al., 1984). Exper-iments conducted with convergent flakes made from South Africanraw materials, and with considerable variation in proportions,demonstrated that local stones of various grain sizes anddimensions develop similar frequencies of DIFs (57%) when all thepieces were used as stabbing or throwing spears (see Lombardet al., 2004). The seeming differences between DIF frequencies fromexperimental samples that have been hand-delivered (55–57%) andthose that were used as projectiles (30–40%) could possiblyindicate variation between the two weapon systems. While thismight be a future possibility we are, however, not confident thatthe current data can be used in this way. Continued experimenta-tion with this question in mind is needed.

4.3. Results

Of the replicated segments used in this experiment (n ¼ 33), 30tools were analysed for macrofractures. The remaining three wereunusable because of severe fracturing in the carcass. This observa-tion may account for the presence of many broken stone toolfragments excavated from archaeological sites associated with theprocessing and consumption of hunted prey. Also, it might be ofinterest to note that even though stone tips or insets are prone tosuch severe fracturing, some ethnographically-known hunters be-lieve that they are more lethal when they break up in the prey as thefragments increase tissue damage and bleeding (e.g. Rudner, 1979).

Forty percent of the analysed pieces developed DIFs and manymore had non-diagnostic fractures. While the non-diagnostic

Fig. 5. Diagnostic impact fractures and impact notches documented on the experi-mental tools; (a) impact notches, (b) impact burination, (c) unifacial spin-off fracture�6 mm with step termination, (d) step terminating fracture, (e) impact burination and(f) unifacial spin-off fracture �6 mm.

M. Lombard, J. Pargeter / Journal of Archaeological Science 35 (2008) 2523–2531 2527

fracture types are not indicative of hunting on their own, whencombined with the DIFs recorded in this analysis, they representa useful indication of fracture patterns that might occur on stonetipped hunting weapons. Table 1 provides the results of thedetailed macrofracture analysis according to hafting configurationssummarised below:

� One longitudinally hafted segment developed an impactburination.� Two transversely hafted pieces also developed impact burina-

tions, while two smooth, semi-circular notches developed onthe cutting edge on one of these tools (Fig. 5a).� Five segments hafted back-to-back developed DIFs in the form

of impact burinations (Fig. 5b), unifacial spin-off fractures>6 mm (Fig. 5c), and a step terminating fracture (Fig. 5d).� The diagonally hafted segments developed the highest DIF

frequencies (66%), three of which are impact burinations(Fig. 5e). One of the impact burinations has an associatedunifacial spin off fracture >6 mm (Fig. 5f).

These results verify previous conclusions that DIFs develop onstone tools used to tip hunting weapons regardless of raw material,tool morphology, dimension or mode of delivery. Macrofractureresults obtained from archaeological samples such as HowiesonsPoort segments can thus be compared successfully with resultsfrom other tool types and industries to generate comparable datafor hunting technologies and their associated behaviours acrosstime and space (e.g. Lombard, 2007b). The frequencies of impactburination on the experimental tools, and the development ofsmooth, semi-circular notches on one of the transversely haftedsegments have implications for the interpretation of archaeologicalassemblages and deserve some consideration.

5. Discussion

5.1. Impact burination and impact notching

The experimental sample shows a relatively high frequency ofimpact burination (26%). Impact burination were documented on4% of the backed tool sample from Sibudu Cave (n ¼ 132), 10% onthe sample from Umhlatuzana (n ¼ 101) and only 2% of the samplefrom Klasies River Cave 2 (n ¼ 85) (Table 2) (Fig. 6c, f). This could bedue to the fact that all of the experimental tools were shot into thecarcass multiple times, testing them towards the upper limits oftheir stress tolerance. Therefore, they could have developed suchfractures more frequently than the archaeological samples, some ofwhich may also have been used as barbs, cutting insets, functionsother than hunting or not used at all. It is further possible thatnuances in DIF type frequencies may reflect aspects of variablessuch as haft weight, velocity of delivery, angle of impact and

Table 1Macrofracture results according to the different hafting configurations

Fracture types Vertical (n ¼ 6) Horizontal(n ¼ 7)

n % n %

Step terminating 0 0 0 0Bifacial spin-off 0 0 0 0Unif. Spin-off >6 mm 0 0 0 0Impact burination 1 16 2 28Unif. Spin-off

Table 2Results of the experimental sample compared to the results obtained for three archaeological samples from Howiesons Poort contexts

Fracture types Pargeterexperiments(n ¼ 30)

Sibudu CaveHP (n ¼ 132)

UmhlatuzanaHP (n ¼ 101)

Klasies River Cave2 HP (n ¼ 85)

n % n % n % n %

Step terminating 1 3 13 10 15 14 12 14Bifacial spin-off 0 0 6 4.5 0 0 2 2Unif. spin-off >6 mm 3 10 9 7 0 0 2 2Impact burination 8 26 5 4 10 10 2 2Unretouched notches 2 7 5 4 4 4 No rec. No rec.Unif. spin-off

M. Lombard, J. Pargeter / Journal of Archaeological Science 35 (2008) 2523–2531 2529

(but not necessarily DIFs) develop in much lower frequencies ontools used as daggers, butchery knives or chisels, or on trampledtools (Shea et al., 2002).

It is also possible that not all archaeological pieces were used.Some may have been lost before use, or some may have beenprepared and stockpiled in anticipation of use. Stone-tippedweapons are composite tools in which the stone component isoften expendable (Bamforth and Bleed, 1997). Weapon systemsbased on blade technologies are especially well adapted to allowthe expedient replacement of a damaged tip or inset withouthaving to manufacture a new weapon. Large quantities of bladescan be manufactured in a short period of time, thus batches ofinsets can be produced in anticipation of need (e.g. Elston andBrantingham, 2002). It is consequently conceivable that archaeo-logical samples will often display relatively low frequencies of DIFscompared to some experimental samples where all the tools wereused in hunting activities.

Detailed micro-residue and use-wear analyses on 53 segmentsfrom Sibudu Cave supports the interpretation of Howiesons Poortsegments as adjustable insets used in hunting weapons. Almost2000 organic residue occurrences were documented on the samplethat was excavated and curated under ideal circumstances formicro-residue analysis. Most of the animal residues were docu-mented on the bladed portions of the segments. This is especiallytrue for animal tissue, collagen, blood residue and hair fragments.The distribution patterns of fatty and bone residues were lessdistinct and it is possible that some tools were hafted to bone(Lombard, 2008). In contrast, plant residues are concentrated onthe backed portions of the segments, with 67% of the plant residuecomponent comprising resin or gum, often associated with ochre(Lombard, 2007a). Additional use traces on the Sibudu sample, suchas striations, polish, micro-damage to tool edges and edge round-ing, show little indication of cutting, sawing or scraping. We do not,however, rule out multi-functionality for backed tools from theHowiesons Poort. Functions such as knives and implements forcutting plant material have previously been suggested (Deacon, J.,1995; Wadley and Binneman, 1995).

The distribution patterns of the micro-residues associated withhafting provided evidence for variability in hafting configurations.Two main clusters of hafting configurations could be identified. Thefirst cluster represents tools with adhesive residues recorded allalong the backed portions that could have been hafted transversely,back-to-back or longitudinally. The second cluster represents toolswith adhesive residues restricted to one end of a tool. These toolscould have been hafted end-on or in diagonal positions. During theoldest Howiesons Poort layer (PGS) at Sibudu, where segments weremostly made on hornfels and quartz and are generally smaller thanduring the younger layers, people seem to have preferred longitu-dinal, transverse or back-to-back hafting configurations. During theintermediate layers (GS and GS2) there seem to have been a shifttowards diagonal or end-on hafting configurations, with no seemingpreferences and a higher degree of variability during the youngestlayers (GR and GR 2) (Lombard, 2008; Wadley, in press).

Wadley and Mohapi (Wadley, in press; Wadley and Mohapi,2008) showed that the tip-cross-sectional area (TCSA) values (seeShea, 2006 for method) for Howiesons Poort segments from SibuduCave places them in the range of ethnographically known NorthAmerican arrowheads. However, if we assume that some werehafted transversely there are differences in the TCSA values for thedifferent morphometric groups that are correlated to raw materials.The short, deep and robust quartz segments, mostly present in theoldest Howiesons Poort layer at Sibudu, still have TCSA valuesfalling within the arrow range. Hornfels segments fall within therange of North American darts, and dolerite segments, mostly fromthe youngest Howiesons Poort layers at Sibudu, are closer to ex-perimental spears.

Single TCSA values for the experimental sample also fall in thearrow range; however, when the values are calculated for back-to-back hafted weapons they fall in the spear range (Pargeter, 2007).Even though these values represent hypothetical functional possi-bilities, they might be a further indication of change and variabilityin Howiesons Poort hunting technologies. Increasing evidence forsuch shifts and variability is of foremost interest for Middle StoneAge behavioural studies, and could have been caused by adapta-tions to aspects of economy, demography, social and symbolicbehaviour, procurement, maintainability, reliability, risk manage-ment, and the environment (e.g. Ambrose and Lorenz, 1990; Bleed,1986; Costa et al., 2005; Fitzhugh, 2001). While direct evidence foror against stone-tipped projectile weaponry during the HowiesonsPoort is still lacking, we suggest that tools such as segments couldhave been used successfully as tips or insets for both hand-delivered and projectile weapon systems in variable, adaptable andchanging hunting technologies.

In this context it might be useful to consider real-life huntingscenarios. We do not suggest that recent hunter-gatherer behaviourcan or should be indiscriminately mapped onto that of theHowiesons Poort. Yet, ethnographic case studies serve to illustratethat hunters operate in a real world of opportunity and risk,providing examples of the dynamics that could exist within a singlesocial/temporal context (Kusimba, 2002). Such examples providecontextual variables for choices of weapon systems, huntingstrategies and prey. More often than not, hunter-gatherer groupsuse a variety of weapon systems such as bow-and-arrow, spears,clubs, nets and snares. Although some variety in hunting behaviourand equipment may reflect social/group preferences (Bartram,1997; Hitchcock and Bleed, 1997), the !Kung of the Kalahari andother San groups also vary hunting strategies and the associatedweaponry according to season or prey type.

Schapera (1963) noted that the San would hunt certain antelopespecies during the rainy season by overtaking them and killingthem with wooden clubs or spears. During the dry season theyoften made pitfalls in game paths, and animals trapped in the pitsare killed with spears. Traps are also used mainly in the dry seasonto catch birds, hares, wild cats, foxes and small antelope. Manyvarieties of big game are run down until they are exhausted andkilled with spears. Some game species such as giraffe, eland,hartebeest and springbok are, however, generally hunted withbow-and-arrow (Schapera, 1963). More recent work with the Kuaand Tyua San of Botswana also shows that animal foods areobtained by methods that vary seasonally (Bartram, 1997;Hitchcock and Bleed, 1997). One of the noted differences betweenspear and bow hunting is that the former is an effective methodthroughout the year, whereas bow-and-arrow hunting is morerestricted to particular seasons. The effectiveness of poisonedarrows is also dependent on season, mainly because of variability inthe toxicity of the poison. Data collected amongst the Kua in theeastern Kalahari indicate that the quality of the poison declined inthe late dry season, and that during drought years and extremelywet years it was often not available at all (Hitchcock and Bleed,1997). If we accept that people in southern Africa behaved ina modern way at least since the Still Bay technocomplex or earlier,we predict that evidence for variability and complexity in theirhunting behaviours will become increasingly evident. Variation inthe hafting configurations of Howiesons Poort segments mightprovide one such example.

6. Conclusion

Replication experiments produce a range of variability withinparameters that can be controlled and understood. When suchvariability is compared with archaeological assemblages, we gainindirect insight into some aspects associated with the tool classes

M. Lombard, J. Pargeter / Journal of Archaeological Science 35 (2008) 2523–25312530

under investigation (Andrefsky, 2005). The Pargeter experimentsshowed that Howiesons Poort segments, as a morphological type,could function effectively as tips for hunting weapons hafted ina variety of positions. Although much larger experimental samplesare needed to generate reliable macrofracture data, the initialcomparisons between experimental and archaeological sampleshave highlighted interesting points with regard to impact burina-tion, impact notching and variability in DIF frequencies. Theseobservations can now be used to focus future research, and shouldlead to increased caution regarding some functional assumptions. Itis our opinion that through a reflexive process between replication,experimentation, testing and examination of archaeologicalmaterial, we are starting to gain insight into the intricacies ofhunting behaviour during the Howiesons Poort. Together with theresults from morphometric studies on segments (Wadley, in press;Wadley and Mohapi, 2008), faunal material (Clark and Plug, inpress) and worked bone (Backwell et al., in press) from Sibudu Cave,the evidence suggests that Howiesons Poort hunting technologiesand their associated behaviours were probably much more com-plex, varied and interchangeable than what has been accepted untilrecently.

Acknowledgements

We thank Sylvain Soriano for producing the segments used inthe experiments, Rodger Govender for providing the carcass, andLyn Wadley and Sarah Wurz for access to the material from SibuduCave and Klasies River Cave 2. The comments and suggestions ofthe reviewers strengthened the paper. J. P. received financial sup-port from the University of the Witwatersrand and the ACACIAResearch Unit at the University under the directorship of LynWadley. M. L. received support from the Palaeontological ScientificTrust and the Natal Museum. Opinions expressed in this article, orany oversights, are our own and do not necessarily reflect the viewsof our funding agencies or colleagues.

References

Ambrose, S.H., 2006. Howiesons Poort lithic raw material procurement patternsand the evolution of human behaviour. Journal of Human Evolution 50,365–369.

Ambrose, S.H., Lorenz, K.G., 1990. Social and ecological models for the Middle StoneAge in southern Africa. In: Mellars, P. (Ed.), The Emergence of Modern Humans.Edinburgh University Press, Edinburgh, pp. 3–33.

Andrefsky Jr., W., 2005. Lithics: Macroscopic Approaches to Analysis, second ed.Cambridge University Press, Cambridge.

Backwell, L., d’Errico, F., Wadley, I., 2008. Middle Stone Age bone tools from theHowiesons Poort layers, Sibudu Cave, South Africa. Journal of ArchaeologicalScience 35, 1566–1580.

Bamforth, D.B., Bleed, P., 1997. Technology, flaked stone technology, and risk. In:Barton, C.M., Clark, G.A. (Eds.), Rediscovering Darwin: Evolutionary Theory andArchaeological Explanation. Archaeological Papers of the American Anthropo-logical Association, Arlington, vol. 7, pp. 109–139.

Barham, L.S., 2002. Backed tools in the Middle Pleistocene central Africa and theirevolutionary significance. Journal of Human Evolution 43, 585–603.

Barton, R.N.E., Bergman, C.A., 1982. Hunters at Hengistbury: some evidence fromexperimental archaeology. World Archaeology 14, 237–248.

Bartram, L.E., 1997. A comparison of Kua (Botswana) and Hadza (Tanzania) bow andarrow hunting. In: Knecht, H. (Ed.), Projectile Technology. Plenum Press, NewYork, pp. 321–344.

Bar-Yosef, O., Kuhn, S.L., 1999. The big deal about blades: laminar technologies andhuman evolution. American Anthropologist 101, 322–338.

Bergman, C.A., Newcomer, M.H., 1983. Flint arrowhead breakage: examples fromKsar Akil, Lebanon. Journal of Field Archaeology 10, 921–947.

Binneman, J., 1994. A unique stone tipped arrowhead from Adam’s Kranz Cave,Eastern Cape. Southern African Field Archaeology 3, 60–62.

Bleed, P., 1986. The optimal design of hunting weapons: maintainability orreliability. American Antiquity 51, 737–747.

Bleed, P., 2002. Cheap, regular and reliable: implications of design variation in latePleistocene Japanese microblade technology. In: Elston, R.G., Kuhn, S.L. (Eds.),Thinking Small: Global Perspectives on Microlithization. Archaeological Papersof the American Anthropological Association 12, pp. 95–102. Arlington.

Bocquentin, F., Bar-Yosef, O., 2004. Early Natufian remains: evidence for physicalconflict from Mt. Carmel, Israel. Journal of Human Evolution 47, 19–23.

Brooks, A.S., Nevell, L., Yellen, J.E., Hartman, G., 2006. Projectile technologies of theMSA: implications for modern human origins. In: Hovers, E., Kuhn, S.L. (Eds.),Transitions Before the Transition: Evolution and Stability in the MiddlePalaeolithic and Middle Stone Age. Springer, New York, pp. 233–255.

Clark, J., Plug, I. Animal exploitation strategies during the South African MiddleStone Age: Howiesons Poort and post-Howiesons Poort fauna from SibuduCave. Journal of Human Evolution, in press, doi: 10.1016/j.jhevol.2007.12.004.

Clark, J.D., 1977. Interpretations of prehistoric technology from ancient Egyptian andother sources. Part II: Prehistoric arrow forms in Africa as shown by survivingexamples of traditional arrows of the San Bushmen. Paléorient 3, 127–150.

Clark, J.D., Phillips, J.L., Staley, P.S., 1974. Interpretations of prehistoric technologyfrom ancient Egyptian and other sources. Part I: Ancient Egyptian bows andarrows and their relevance for African prehistory. Paléorient 2, 323–388.

Costa, L.J., Sternke, F., Woodman, P.C., 2005. Microlith to macrolith: the reasons behindthe transformation of production in the Irish Mesolithic. Antiquity 79, 19–33.

Cotterell, B., Kamminga, J., 1987. The formation of flakes. American Antiquity 52,675–708.

Crombé, P., Perdaen, Y., Sergant, J., Caspar, J.-P., 2001. Wear analysis on earlyMesolithic microliths from Verrebroek Site, east Flanders, Belgium. Journal ofField Archaeology 28, 253–269.

Deacon, H.J., 1989. Late Pleistocene Palaeoecology and archaeology in the southernCape, South Africa. In: Mellars, P., Stringer, C. (Eds.), The Human Revolution:Behavioural and Biological Perspectives on the Origins of Modern Humans.Edinburgh University Press, Edinburgh, pp. 547–564.

Deacon, H.J., 1992. Southern Africa and modern human origins. PhilosophicalTransactions of the Royal Society B 337, 177–183.

Deacon, H.J., 1995. Two late Pleistocene/Holocene archaeological depositoriesfrom the southern Cape, South Africa. South African Archaeological Bulletin50, 121–131.

Deacon, H.J., Shuurman, R., 1992. The origins of modern people: the evidence fromKlasies River. In: Bräuer, G., Smith, S.H. (Eds.), Continuity or Replacement:Controversies in Homo sapiensEvolution. Balkema, Rotterdam, pp. 121–129.

Deacon, H.J., Wurz, S., 1996. Klasies River main site, cave 2: a Howiesons Poortoccurrence. In: Pwiti, G., Soper, R. (Eds.), Aspects of African Archaeology.University of Zimbabwe Publications, Harare, pp. 213–281.

Deacon, H.J., Wurz, S., 2001. Middle Pleistocene populations and the emergence ofmodern behaviour. In: Barham, L., Robson Brown, K. (Eds.), Human Roots: Africaand Asia in the Middle Pleistocene. Western Academic and Specialist Press,Bristol, pp. 55–63.

Deacon, H.J., Wurz, S., 2005. A Late Pleistocene archive of life at the coast, KlasiesRiver. In: Stahl, A.B. (Ed.), African Archaeology: A Critical Introduction. BlackwellPublishing, Oxford, pp. 130–149.

Deacon, J., 1995. An unsolved mystery at the Howiesons Poort name site. SouthAfrican Archaeological Bulletin 162, 110–120.

Delagnes, A., Wadley, L., Villa, P., Lombard, M., 2006. Crystal quartz backed toolsfrom the Howiesons Poort at Sibudu Cave. Southern African Humanities 18 (1),43–56.

Dockall, J.E., 1997. Wear traces and projectile impact: a review of the experimentaland archaeological evidence. Journal of Field Archaeology 24, 321–331.

Elston, R.G., Brantingham, P.J., 2002. Microlithic technology in Northern Asia: a risk-minimizing strategy of the late Paleolithic and early Holocene. In: Elston, R.G.,Kuhn, S.L. (Eds.), Thinking Small: Global Perspectives on Microlithization.Archaeological Papers of the American Anthropological Association 12, pp. 103–116. Arlington.

Feathers, J.K., 2002. Luminescence dating in less than ideal conditions: case studiesfrom Klasies River Main site and Duinefontein, South Africa. Journal ofArchaeological Science 29, 177–194.

Fischer, A., Vemming Hansen, P., Rasmussen, P., 1984. Macro and micro wear traceson lithic projectile points: experimental results and prehistoric examples.Journal of Danish Archaeology 3, 19–46.

Fitzhugh, B., 2001. Risk and invention in human technological evolution. Journal ofAnthropological Archaeology 20, 125–167.

Geneste, J.-M., Plisson, H., 1993. Hunting technologies and human behaviour: lithicanalysis of Solutrean shouldered points. In: Knecht, H., Pike-Tay, A., White, R.(Eds.), Before Lascaux: The Complex Record of the Early Upper Palaeolithic. CRCPress, Boca Raton, pp. 117–135.

Gibson, N.E., Wadley, L., Williamson, B.S., 2004. Microscopic residues as evidence ofhunting on backed tools from the 60,000 to 68,000 year-old Howiesons Poortlayers of Rose Cottage Cave, South Africa. Southern African Humanities 16, 1–11.

Grün, R., Beaumont, P., 2001. Border Cave revisited: a revised ESR chronology.Journal of Human Evolution 40, 467–482.

Grün, R., Schackelton, N.J., Deacon, H.J., 1990. Electron-spin resonance dating of toothenamel from Klasies River Mouth Cave. Current Anthropology 31, 427–432.

Hiscock, P., 2002. Pattern and context in the Holocene proliferation of backedartefacts in Australia. In: Elston, R.G., Kuhn, S.L. (Eds.), Thinking Small: GlobalPerspectives on Microlithization. Archaeological Papers of the AmericanAnthropological Association 12, pp. 163–178. Arlington.

Hiscock, P., Clarkson, C., 2007. Retouched notches at Combe Grenal (France) and thereduction hypothesis. American Antiquity 72, 176–190.

Hitchcock, R., Bleed, P., 1997. Each according to need and fashion: spear and arrowuse among San hunters of the Kalahari. In: Knecht, H. (Ed.), ProjectileTechnology: Interdisciplinary Contributions to Archaeology. Plenum Press, NewYork, pp. 345–368.

Ho Ho Committee, 1979. The Ho Ho classification and nomenclature committeereport. In: Hayden, B. (Ed.), Lithic Use-Wear Analysis. Academic Press, NewYork, pp. 134–158.

http://doi:10.1016/j.jhevol.2007.12.004

M. Lombard, J. Pargeter / Journal of Archaeological Science 35 (2008) 2523–2531 2531

Holdaway, S., McPherron, S., Roth, B., 1996. Notched tools and raw material avail-ability in French Middle Paleolithic sites. American Antiquity 61, 377–387.

Ingersoll, D., Macdonald, W.K., 1977. Introduction. In: Ingersoll, D., Yellen, J.E.,Macdonald, W.K. (Eds.), Experimental Archaeology. Columbia University Press,New York, pp. xi–xviii.

Kusimba, S.B., 2002. What is a Hunter-Gatherer? Variation in the ArchaeologicalRecord of Eastern and Southern Africa. University of Chicago Press, Chicago.

Lombard, M., 2005a. A method for identifying Stone Age hunting tools. SouthAfrican Archaeological Bulletin 60, 115–120.

Lombard, M., 2005b. The Howiesons Poort of South Africa: what we know, what wethink we know, what we need to know. Southern African Humanities 17, 33–55.

Lombard, M., 2006. Direct evidence for the use of ochre in the haftingtechnology of Middle Stone age tools from Sibudu Cave. Southern AfricanHumanities 18 (1), 57–67.

Lombard, M., 2007a. The gripping nature of ochre: the association of ochre withHowiesons Poort adhesives and Later Stone Age mastics from South Africa.Journal of Human Evolution 53, 406–419.

Lombard, M., 2007b. Evidence for change in Middle Stone Age hunting behaviour atBlombos Cave: results of a macrofracture analysis. South African ArchaeologicalBulletin 62, 62–67.

Lombard, M., 2008. Finding resolution for the Howiesons Poort through themicroscope: micro-residue analysis of segments from Sibudu Cave, SouthAfrica. Journal of Archaeological Science 35, 26–41.

Lombard, M., Parsons, I., in press. Blade and bladelet function and variability in riskmanagement during the last 2000 years in the Northern Cape. South AfricanArchaeological Bulletin 63.

Lombard, M., Parsons, I., Van der Ryst, M.M., 2004. Middle Stone Age lithic pointexperimentation for macro-fracture and residue analysis: the process andpreliminary results with reference to Sibudu Cave, South Africa. South AfricanJournal of Science 100, 159–166.

Marean, C.W., Bar-Matthews, M., Bernatchez, J., Fisher, E., Golberg, P., Herries, A.I.R.,Jacobs, Z., Jerardino, A., Karkanas, P., Minichillo, T., Nilssen, P.J., Thompson, E.,Watts, I., Williams, H.M., 2007. Early human use of marine resources andpigment in South Africa during the Middle Pleistocene. Nature 449, 905–908.

Mathieu, J.R., 2002. Introductiondexperimental archaeology: replicating pastobjects, behaviours, and processes. In: Mathieu, J.R. (Ed.), ExperimentalArchaeology: Replicating Past Objects, Behaviours, and Processes. BARInternational Series 1035, Oxford, pp. 1–11.

McBrearty, S., Brooks, A.S., 2000. The revolution that wasn’t: a new interpretation ofthe origin of modern human behaviour. Journal of Human Evolution 39, 453–563.

McDonald, J.J., Donlon, D., Field, J.H., Fullagar, R.L.K., Brenner Coltrain, J., Mitchell, P.,Rawson, M., 2007. The first archaeological evidence for death by spearing inAustralia. Antiquity 81, 877–885.

Miller, G.H., Beaumont, P.B., Brooks, A.S., Deacon, H.J., Hare, P.E., Jull, A.J.T., 1999.Earliest modern humans in South Africa dated by isoleucine epimerization inostrich eggshell. Quaternary Science Review 18, 1537–1548.

Minichillo, T., 2006. Raw material use and behavioural modernity: Howiesons Poortlithic foraging strategies. Journal of Human Evolution 50, 359–364.

Mitchell, P., 1988. Human adaptation in southern Africa during the last Glacialmaximum. In: Bower, J., Lubell, D. (Eds.), Prehistoric Cultures and Environmentsin the Late Quaternary of Africa. BAR International Series 405, pp. 163–196.

Mitchell, P., 1995. Revisiting Robberg: new results and revision of old ideas atSehonghong Rock Shelter, Lesotho. South African Archaeological Bulletin 50,28–38.

Moss, E.H., Newcomer, M.H., 1982. Reconstruction of tool use at Pincevanent:microwear and experiments. In: Cahen, D. (Ed.), Tailler! Pour Quoi Faire:Préhistoire et Technoloque. Studia Praehistorica Belgica 2, pp. 289–321.

Nuzhnyi, D., 2000. Development of microlithic projectile weapons in the Stone Age.Anthropologie et Préhistoire 111, 95–101.

Odell, G.H., 1981. The mechanics of use-breakage of stone tools: some testablehypotheses. Journal of Field Archaeology 13, 195–212.

Odell, H.G., Cowan, F., 1986. Experiments with spears and arrows on animal targets.Journal of Field Archaeology 13, 195–211.

Pargeter, J., 2006. To tip or not to tip: a multi-analytical approach to the hafting anduse of Howiesons Poort segments. Honours Dissertation, University of theWitwatersrand, Johannesburg.

Pargeter, J., 2007. Howiesons Poort segments as hunting weapons: experimentswith replicated projectiles. South African Archaeological Bulletin 62 (pp tofollow).

Parkington, J., 1999. Western Cape landscapes. Proceedings of the British Academy99, 25–35.

Parkington, J., Poggenpoel, C., Rigaud, J.-P., Texier, P.-J., 2005. From tool to symbol:context of intentionally marked ostrich eggshell from Diepkloof, Western Cape.

In: d’Errico, F., Backwell, L. (Eds.), From Tools to Symbols: From Early Hominidsto Modern Humans. University of the Witwatersrand Press, Johannesburg, pp.475–492.

Phillipson, D.W., 1976. The Prehistory of Eastern Zambia. British Institute in Zambia,Nairobi.

Rudner, J., 1979. The use of stone artefacts and pottery among the Khoisanpeoples in historic and protohistoric times. South African ArchaeologicalBulletin 34, 3–17.

Schapera, I., 1963. The Khoisan Peoples of South Africa. Routledge, London.Shea, J.J., 1988. Spear points from the Middle Palaeolithic of the Levant. Journal of

Field Archaeology 15, 441–450.Shea, J.J., 2006. The origins of lithic projectile technology: evidence from Africa, the

Levant, and Europe. Journal of Archaeological Science 33, 823–846.Shea, J.J., Brown, K.S., Davis, Z.J., 2002. Controlled experiments with Middle Palae-

olithic spear points: Levallois points. In: Mathieu, J.R. (Ed.), ExperimentalArchaeology: Replicating Past Objects, Behaviours, and Processes. BARInternational Series 1035, pp. 55–72. Oxford.

Shimelmitz, R., Barkai, R., Gopher, A., 2004. The geometric Kebaran microlithicassemblage of Ain Miri, northern Israel. Paléorient 30, 127–140.

Singer, R., Wymer, J., 1982. The Middle Stone Age at Klasies River Mouth in SouthAfrica. Chicago University Press, Chicago.

Soriano, S., Villa, P., Wadley, L., 2007. Blade technology and tool forms in the MiddleStone Age of South Africa: the Howiesons Poort and post-Howiesons Poort atRose Cottage Cave. Journal of Archaeological Science 34, 681–703.

Stapleton, P., Hewitt, J., 1927. Stone implements from a rock-shelter at Howieson’sPoort near Grahamstown. South African Journal of Science 24, 574–587.

Stapleton, P., Hewitt, J., 1928. Stone implements from Howieson’s Poort, nearGrahamstown. South African Journal of Science 101, 169–174.

Tribolo, C., Mercier, N., Valladas, H., 2005. Chronology of the Howiesons Poort andStill Bay techno-complexes: assessment and new data from luminescence. In:d’Errico, F., Backwell, L. (Eds.), From Tools to Symbols: From Early Hominids toModern Humans. University of the Witwatersrand Press, Johannesburg, pp.493–511.

Valladas, H., Wadley, L., Mercier, N., Tribolo, C., Reyss, J.L., Joron, J.L., 2005.Thermoluminescence dating on burnt lithics from Middle Stone Age layers atRose Cottage Cave. South African Journal of Science 101, 169–174.

Villa, P., Lenoir, M., 2006. Hunting weapons of the Middle Stone Age and the MiddlePalaeolithic: spear points from Sibudu, Rose Cottage and Bouheben. SouthernAfrican Humanities 18 (1), 89–122.

Vogel, J., 2001. Radiometric dates for the Middle Stone age in South Africa. In:Tobias, P.V., Raath, M.A., Moggi-Cecchi, J., Doyle, G.A. (Eds.), Humanity fromAfrican Naissance to Coming Millennia. University of the Witwatersrand Press,Johannesburg, pp. 261–268.

Wadley, L., 1987. Later Stone Age Hunters and Gatherers of the Southern Transvaal:Social and Ecological Interpretation. In: BAR International Series 380 Oxford.

Wadley, L., 1996. The Robberg Industry of Rose Cottage Cave, eastern Free State: thetechnology spatial patterns and environment. South African ArchaeologicalBulletin 51, 64–74.

Wadley, L., 2005. A typological study of the final Middle Stone Age stone tools fromSibudu Cave, KwaZulu-Natal. South African Archaeological Bulletin 60, 51–64.

Wadley, L., 2006. Partners in grime: results of multidisciplinary archaeology atSibudu cave. Southern African Humanities 18 (1), 315–341.

Wadley, L., in press. The Howiesons Poort Industry of Sibudu Cave. South AfricanArchaeological Society Goodwin Volume.

Wadley, L., Binneman, J., 1995. Arrowheads or pen knives? A micro-wear analysis ofmid-Holocene stone segments from Jubilee Shelter, Transvaal. South AfricanJournal of Science 91, 153–155.

Wadley, L., Jacobs, Z., 2006. Sibudu Cave: background to the excavations, stratig-raphy and dating. Southern African Humanities 18 (1), 1–26.

Wadley, L., Mohapi, M., 2008. A segment is not a monolith: evidence from the Ho-wiesons Poort of Sibudu, South Africa. Journal of Archaeological Science, 35 (9),2594–2605.

Wurz, S., 1999. The Howiesons Poort backed artefacts from Klasies River: anargument for symbolic behaviour. South African archaeological Bulletin 54,38–50.

Wurz, S., 2000. The Middle Stone Age sequence at Klasies River, South Africa. DPhilThesis, University of Stellenbosch, Stellenbosch.

Wurz, S., 2002. Variability in the Middle Stone Age lithic sequence, 115,000–60,000years ago at Klasies River, South Africa. Journal of Archaeological Science 29,1001–1015.

Wurz, S., Lombard, M., 2007. 70,000-year-old geometric backed tools from theHowiesons Poort at Klasies River, South Africa: were they used for hunting?Southern African Humanities 19, 1–16.

Hunting with Howiesons Poort segments: pilot experimental study and the functional interpretation of archaeological toolsIntroductionA short background to the Howiesons Poort of South AfricaThe experimentsMacrofracture analysisBackgroundMethodResults

DiscussionImpact burination and impact notchingVariability in Howiesons Poort hunting technology

ConclusionAcknowledgementsReferences