lable at ScienceDirect

Journal of Archaeological Science 49 (2014) 302e316

Contents lists avai

Journal of Archaeological Science

journal homepage: http: / /www.elsevier .com/locate/ jas

Technological behaviors in Paleolithic foragers. Testing the role ofresharpening in the assemblage organization

Juan I. Morales a, b, *, Josep M. Verg�es a, b

a IPHES; Institut Catal�a de Paleoecologia Humana i Evoluci�o Social, C/ Marcel·lí Domingo s/n, Campus Sescelades URV (Edifici W3), 43007 Tarragona, Spainb �Area de Prehist�oria, Universitat Rovira i Virgili (URV), Avinguda de Catalunya, 35, 43002 Tarragona, Spain

a r t i c l e i n f o

Article history:Received 14 March 2014Received in revised form19 May 2014Accepted 20 May 2014Available online 8 June 2014

Keywords:Technological organizationUse-wear & residue analysesResharpeningReductionForager mobilityUpper PaleolithicLa Cativera

* Corresponding author. IPHES; Institut Catal�aEvoluci�o Social, C/ Marcel$lí Domingo s/n, Campus43007 Tarragona, Spain. Tel.: þ34 654329942.

E-mail addresses: jignacio.morales@gmail(J.I. Morales).

http://dx.doi.org/10.1016/j.jas.2014.05.0250305-4403/© 2014 Elsevier Ltd. All rights reserved.

a b s t r a c t

This paper describes the evaluation, based on archaeological materials, of the role that resharpeningplays in the continuum of stone-tool reduction. We define a multi-evidence-based approach that com-bines use-wear intensity and location, traces that could be related to hafting and the distribution ofmineral residue. By combining these methods, we have observed a minimum resharpening ratio of 52%in the selected end-scraper sample. If one takes into account ethnographically obtained informationabout end-scraper management, this result is an unexpectedly low value. Dynamics of mobility, tech-nological organization and raw material availability causes high variability in the archaeological visibilityand characteristics of lithic remains. Our results are in line with a technology organized wholly orpartially on the basis of expediency, in which tools are not curated for more than the time taken tocomplete the activity or the length of the occupation. Tools that were not exhausted were abandoned atthe site, leading to recycling behaviors in periodic site reoccupations.

© 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Measuring the degree of reduction in lithic tools has sparked theinterest of many archaeologists during recent decades. Here,reduction is understood to be the continuous process of toolresharpening caused by the edge exhaustion while a given task isperformed. The more reduced a tool is, the longer the usage timededuced, or the greater the intensity of the work. Thus, if we areable to measure reduction objectively and precisely, this will openthe door to intra-site/multi-site and diachronic/synchronic com-parisons between prehistoric groups. This will result in a robustapproach to human behavior as related to several factors, includingtool management, site occupation intensity, territorial mobilitydynamics and raw material administration as a function of itsabundance or scarcity.

Several researchers have focused their work on investigatingways to quantify reduction intensity and its implications. Theirresearch takes different conceptual approaches to dealing with the

de Paleoecologia Humana iSescelades URV (Edifici W3),

.com, jmorales@iphes.cat

question of reduction. Technological analyses (Dibble, 1984, 1987,1995; Marwick, 2008), experimentally tested indexes (Hiscock andAttenbrow, 2002, 2003; Hiscock and Clarkson, 2005; Clarkson,2002; Eren et al., 2005; Kuhn, 1990), and allometric relationships(Blades, 2003; Clarkson and Hiscock, 2011; Shott et al., 2000; Iovit�a,2010; Eren, 2013; Eren et al., 2013) are the principal approaches toestimating reduction in archaeological and experimental samples.

All of these approaches have emphasized the importance ofreduction intensity as one of the characteristics that can provide themost useful information in the fields of lithic analysis and humanbehavior. All of them have the same strong initial hypothesis e thattool reduction is a function of time and usage. In order to be able toaccept the measurements of reduction, it is necessary to confirmempirically that tools are reduced or resharpened during theirfunctional life. It seems logical to think that maybe not all types oftool are sequentially reduced. Some kinds of highly standardizedtools, or individual parts of composite implements (such as arrowarmatures) can undergo only one shaping phase in which theirfunctional morphology is defined. The sequence of the life cycle ofthese tools would be defined as (1) blank selection, (2) shaping, (3)use and (4) abandonment. In contrast, the sequence for tools thathave suffered reductionmust be (1) blank selection, (2) shaping, (3)uset1, (4) resharpeningt1, (5) uset2, (6) resharpeningt2 …, (7) aban-donment, (8) (re)-selection, (9) (re)-use. In these specific cases, a tool

J.I. Morales, J.M. Verg�es / Journal of Archaeological Science 49 (2014) 302e316 303

is configured again and again, probably for the same use, soresharpening is an indicator of maintenance, not of retooling or alateral change to a different cycle as proposed by Schiffer (1972).

We can derive from this that the basic cause of tool reduction isedge resharpening by retouching. An accumulation of resharpeningevents results in a progressive loss of mass and the morphologicalevolution of the tool. Identifying resharpening events in the lithicassemblage should be used as an initial test for deducing howreductionwas present, and howmuch was present, in the technicalbehavior of prehistoric groups, or, at least, in the physical remainsof the technology of the occupation we are studying.

The physical action of resharpening causes a wide range ofmodifications to lithic tools. It must be possible to observe thesuccession of working phases and resharpening events empiricallyin various ways e this cannot be just assumed a priori. Inmorphologically stable tool types, such as end-scrapers, the changein form after retouch is gradual, not abrupt. In these cases oneexpects that the functional retouched zone would be more or lessthe same after each resharpening event. Nonetheless, minimalmorphological variations are accumulated after each modification,and this could create a final morphology that is slightly differentfrom the initial one (e.g. Eren, 2013). These subtle modifications ofthe working edge must imply that functional evidence such as use-wear traces or residue preservation will not be completely super-imposed after each resharpening phase. If this hypothesis isconfirmed, then some kind of use-resharpening “stratigraphy”could be established through microscopic analysis. Initial use-weartraces should be partially removed by successive retouching, anddifferent degrees of use-wear should be visible. Several functionalanalysts (Loebel, 2013; Jard�on and Sacchi, 1994; Jard�on, 2000) haveused similar reasoning, but the authors have found no extensivedocumentation. The same reasoning could be applied to residuepreservation and distribution. Organic residue is not commonlypreserved at sites, but mineral residues such as iron oxides or thegums used in hide tanning or tool hafting are frequently docu-mented. When these residues coat the tool, their distribution willbe modified if resharpening occurs.

Resharpening and usage are not necessarily continuous overtime, so tools did not have to be used immediately after retouching.It is well known in studies of lithic technology and mobility thatstone tools were curated, or maintained and transported, to meetfuture necessities (Binford, 1977, 1978, 1979). Resharpening eventsand the moment of use could therefore be discontinuous in time,because some tools could be resharpened after one working phaseto leave them ready for future use. When these tools are found inthe archaeological record and no traces of the working event can beobserved, they could be misinterpreted as unused tools. In thesecases it is necessary to find indirect evidence of use in order todemonstrate that those tools have also been resharpened.

Recently, a modification of Eren's (Eren et al., 2005) ERP meth-odology for quantifying the mass lost in reduction has been pro-posed (Morales et al., in press). This new approach focused on thestudy of distally retouched tools such as end-scrapers. In order toapply this methodology systematically to the archaeological record,it is necessary to verify that this morphotype is systematicallyresharpened during its use-life. Ethnographic studies have demon-strated that end-scrapers are perhaps some of the most frequentlyresharpened tools in the lithic toolkit. We are therefore not dealingwith an archaeological demonstration of rejuvenating end-scrapers,but with incidents of this technical behavior in specific archaeo-logical samples. Taking into account the above working hypothesis,we have analyzed a verywell preserved Late Upper Paleolithic stonetool assemblage in order to observe whether there is evidence ofresharpening, and to establish the inferences that can bemade fromthe reduction sequence with respect to site occupation.

2. Ethnographical remarks on the resharpening of scrapers

Even if ethnographical studies can't be directly transferred toPaleolithic studies they give an insight of concrete behavior of stonetool management. In this case, lithic tools used for scraping are welldocumented in living societies all over the world, and these areoften related to woodworking, (Gould et al., 1971) or processinghides (e.g. Gallagher, 1977;Weedman, 2006; Mansur, 1986; Beyries,2002; Beyries and Rots, 2008).

Our research into the archaeological evidence of functionalrejuvenation is well supported by ethnographical studies of severalof the abovementionedmodern hidescrapers. Ethiopian people uselithic scrapers as formal tools for hide processing. Resharpening hasbeen reported as a technical behavior commonly used to rejuvenateedges when they become dull. Weedman reports a mean of eightresharpenings for each scraper during its use-life among the Gamoand a mean of 281 scrapes between resharpenings (Shott andWeedman, 2007). Among the Konso, hide-scrapers are resharp-ened regularly while processing a whole hide (Brandt andWeedman, 1997; Brandt, 1996). Among the Gurage people,resharpening behavior varies depending on the source, but it is alsocommon. Gallagher (1977) reports a mean of 100 scrapes betweenresharpenings, while Clark and Kurashina (1981) note thatresharpening occurs every 15/20 scrapes. All of them agree on theintensive use of scrapers, reporting that a mean of four to fivescrapers are used during the processing of a cattle-sized hide.

Similar scraper use and management has been reported for theAthapaska people. Edge resharpening is performed regularly whileprocessing a hide, especially during the hair removal phase. Incontrast, during the softening process, resharpening rarely occurs(Beyries, 2002; Beyries and Rots, 2008). The same researchersreport a case of unresharpened tools among the Chukchi people(Beyries and Rots, 2008). In this case, a broader toolkit of scrapingtools is documented, and specialized scrapers are used during thedifferent phases of hide processing. The Chukchi rarely resharpenthe scrapers used.

Among the Tehuelche scrapers, periodic resharpening is alsodocumented.Both the rawmaterialsusedeglassandchalcedonyeareregularly resharpenedwhenedgesbecomedull (Casamiquela,1978). Ingeneral, there is evidence of resharpening in all ethnographicallyobserved scraping activities, not only in hidescraping but also inwoodscraping, as in the case of the Ngaanyatjarra tulas (Gould, 1971).

All of these ethnographical references are valuable in order tounderstand how scrapers are used, maintained and discarded in itseconomic context. Nevertheless, in order to understand theimportance of tool curation and technological organization inPaleolithic foragers there is a great obstacle to using these ethno-graphic examples. All of the mentioned groups are sedentaryfarmers e they are no longer mobile hunter-gatherers (McCall,2012). In their cases, the distance from and accessibility of, placeswhere they can obtain raw materials is constant, provisioning iseasily schedulable, and the storage of both unretouched andretouched tools is easy and unlimited. This gives them constantcertainty of having tools ready for work when needed, and theprevision efforts are clearly reduced. For highly mobile Paleolithicforagers, their mobility patterns, the constraints on raw materialavailability and the resulting technological organization mustchange the specific ways in which tools are managed. This mustconsequently be reflected in the tools' versatility, suitability forcarrying, and use-life prolongation. We therefore cannot takeethnographical resharpening patterns to be a direct analogy to thepatterns of Paleolithic foragers. The stress agents suffered by thelatter were different [i.e. (Torrence, 1983)], so it is reasonable tothink that their technical and technological responses weredifferent, too.

J.I. Morales, J.M. Verg�es / Journal of Archaeological Science 49 (2014) 302e316304

3. Material & methods

3.1. Materials

In order to evaluate the initial hypothesis from the archaeologicalrecord, a sample that fulfills some rigorous requirements in terms ofpreservation is required. We therefore looked for a sample from amodern and rigorous excavation in which the incidence of tapho-nomical alterations, such as burning, patina, concretion, edgerounding and trampling seems to have been very low. By beingrestrictive in the selection of the sample we can ensure that thepreservation andobservationuse-wear isa priorigood, and also avoidhydrochloric or ultrasonic washes and mechanical cleaning, whichcould affect residue preservation.With respect to the sample sizewefavored one that was significant but affordable, in order to accuratelyapply the multi-evidence approachmethodology consisting in ochredistribution characterization and use-wear/hafting-wear analyses.

Taking into account these criteria, we selected the Late UpperPaleolithic lithic record recovered at level B of La Cativera rock-shelter, a PleistoceneeHolocene transition site in northeasternIberia. La Cativera's level B is dated by AMS 14C to the early Boreal,between 10.3 and 9.6 kyrs BP in the 2-sigma calibration (Moraleset al., 2013). Micromorphological studies (Angelucci, 2003) haverevealed that the archaeological remains were quickly sealed byfine-grained slope sediments (see inline Supplementary Figure 1).The excavation of this level has provided a lithic record consistingexclusively of regional flint and made up of 2406 pieces, 203 ofwhich are retouched tools. Patina formation is almost inexistent,affecting only 7% of the assemblage and where present, is always invery incipient stages. Alterations due to burning are even less sig-nificant, affecting 3.3% of the assemblage. In most cases, fire dam-age is present on the pieces affected by patina formation. No icedamage or weathering has been documented. In preliminaryfunctional approximations no water rounding has been observed,and very good edge preservation has been recorded. The conser-vation of mineral residues was also an important criterion inselecting the sample. The presence of ochre in the assemblage hasbeen documented since the first studies (Fontanals, 2001) of thematerial from the site. Ochre remains have been recovered in formof burned centimetric fragments and also embedded in the surfaces

Fig. 1. m-XRD plot showing chert composition in

of lithic tools. The presence of ochre at archaeological sites is wellknown, has been well studied and is usually related to preparingpigments, hide processing and the preparation of hafting mastic.No evidence of art has been found at La Cativera, but intense hideprocessing activities and systematic tool hafting are being docu-mented. The relationship between ochre and lithic tools is there-fore not casual, because there is a clear connection between toolsand ochre-requiring activities such as hide processing (Philibert,1993). The presence of ochre on lithic tools, and specifically, onend-scrapers has also been documented (Keeley, 1980), and espe-cially in the European late Upper Paleolithic (Seronie-Vivien, 1986;de Beaune, 1989; Philibert, 1993).

Inline Supplementary Figure S1 can be found online at http://dx.doi.org/10.1016/j.jas.2014.05.025.

In northeastern Iberia, the late Upper Paleolithic is characterizedby a reduced typological variability. The retouching patternobserved indicates that end-scrapers and backed elements are al-ways the dominant groups, and together they usually amount to ca.70e80% of the complete retouched assemblage. Taking this ratiointo account, we selected the end-scraper assemblage from LaCativera's level B for analysis. This is made up of 129 pieces, 63% ofthe retouched tools.

3.2. Equipment

Fourdifferentmicroscopeswereused for traceological and residueanalysis: two Optical Light Microscopes and two Scanning ElectronMicroscopes. A m X-Ray diffractometer was also used. Technicalcharacteristics of the equipment, sample preparation and imagecapture procedure are detailed in supplementary information.

4. Residue analyses and use-wear observation

4.1. Residue analyses

The macro- and microscopic analyses of the collection allowedus to identify evidence of reddish residues on 82 end-scrapers,approximately 64% of the total sample. A m-XRD analysis gave thecomposition of the residue as iron oxide (Fe2O3) confirming that iscan be attributed to hematite (Fig. 1). Aside from two imprints of

form of SiO2 and hematite of ochre as Fe2O3.

Fig. 2. Ochre distribution patterns archaeological examples and schematic view. Patterns 1 and 2 are macroscopically visible while Pattern 3 is not. Pattern 3 right image shows themicroscopically visible ochre remains associated to a hide-working polish (OLM).

Fig. 3. Distribution of the different ochre distributions observed in La Cativera end-scraper assemblage.

J.I. Morales, J.M. Verg�es / Journal of Archaeological Science 49 (2014) 302e316 305

vegetal fibers preserved in the concretion no other organic residuehas been documented.

Ochre distribution and preservation is not homogeneousthroughout the entire sample. A high degree of variability wasobserved in the amount of residue present on each piece. Thisvariability ranges from pieces which are clearly completely coveredwith ochre, to pieces on which there are isolated spots or smallconcentrations that are only visible with some magnification(Fig. 2). We are not currently able to assess whether this variabilityderives from functional differences or is just due to differences inpreservation. On the basis of the distribution of the ochreembedded in the tool's surface, three different categories or pat-terns have been established:

Pattern 1. Elements with residue concentrations on one surfaceor partially distributed on several surfaces.Pattern 2. Pieces with ochre distributed over circa the whole ofthe surface.Pattern 3. Pieces with a residual presence of ochre (isolatedspots or particles) randomly distributed on one or moresurfaces.

Of the end-scrapers with ochre, 70% fit the Pattern 2 distributiongroup (Fig. 3), conserving ochre more or less homogeneouslydistributed on the surface. Within this group, dorsal e ventral(z38%) and dorsal e ventral e distal (z25%) distributions areclearly dominant. It is therefore possible to say that there are twodominant groups that characterize the distribution patterns. Onecontains those pieces that are fully-covered (Pattern 2a in Fig. 2),and the other those pieces that are fully-coveredwith the exceptionof the functional (or active/retouched) surface (Pattern 2b in Fig. 2).

Fig. 4. Example of hide-working use-wear traces. Top left, SEM secondary electron micrograph of scraping polish. Top right, enlarged detail of the same polish. Bottom left andright, equivalent OLM micrographs of the same zones observed at SEM. Magnification bar is equivalent for both SEM and OLM.

J.I. Morales, J.M. Verg�es / Journal of Archaeological Science 49 (2014) 302e316306

4.2. Use-wear observation

Firstly, we should comment on two aspects that affect theidentification and analysis of use-wear formation in the La Cativeraassemblage. On the one hand, the variety of chert from which theend-scrapers were made has differing responses in terms of lightreflection, and as a result the quality of optical observations candiffer from one piece to the next. On the other hand, the highermagnification power of the SEM has allowed us to identify lighteruse-wear traces that were not observed with the OLM. As a resultthe level of documentation of tools observed with both SEM andOLM is higher than for those observed only with the OLM. Thecombination of different and complementary observation methodsresults in a higher level of information (Borel et al., 2014).

The use-wear analyses allowed us to identify wear traces on 105end-scrapers, 82% of the total assemblage.

In this work wewill focus on the activities performed using onlythe retouched part of the end-scrapers, dismissing those onesattributed to previous work carried out with the unretouchedflakes. The type of work most documented was hide processing(95% of the cases), while woodworking (3 cases) and bone or antlerworking (2 cases) are very poorly represented, and need someclarification. Two of the tools had been used for woodworkingwhen they were fully configured as end-scrapers. Another toolshows that it was first used on wood as an unretouched flake, andwas later configured as an end-scraper used in hide processing. Oneof the two tools on which bone/antler wear was identified under-went a similar process. In this case, the end-scraper shapingretouch was superimposed on the traces of bone/antler working.The second tool in which bone/antler polish has been observed

shows an alternation between bone microwear on one edge(pushing), and wear caused by hide working on the other, but theretouched end of the tool seems to be fresh.

Use-wear traces resulting from hide-working show a generalappearance of abrasion without striation formation. Despite thedifferent degrees of development, these surfaces are characterizedby a process of attritional homogenization of the microrelief. Edgestend to be rounded and the contact surfaces smoothed. The textureobserved with the OLM is slightly greasy and bright, while with theSEM it is coarse, and the micro- or cryptocrystalline texture of therock is visible. In some cases it is possible to observe brighter,smoothened and isolated spots with the OLM. However, SEMobservation of the same areas shows that these surfaces are notcompletely smooth (Fig. 4).

Wear characteristics indicates that the contact between the tooland the surface worked were marked by a moderately viscousinterfacial layer that restricts the abrasion caused by the particlesdetached. We have documented these contact conditions when thehide is not completely dry through experiments (Oll�e, 2003;Verg�es, 2003). Although the inner surface of the hide seems to bedry, the subcutaneous tissue may contain adipose cells and mois-ture that act as lubricants and prevent striation formation. Innormal weather conditions (no rain and moderate humidity) thedrying process starts just after skinning, being perceptible within afew minutes. However, if no other action is taken, hides can takeseveral days to dry completely. This process varies depending onhide size and animal species. Taking these into account, the end-scrapers studied are used before the hide is completely dry.Despite the influence of the raw material, most of the variabilityobserved in use-wear traces must be attributed to the different

J.I. Morales, J.M. Verg�es / Journal of Archaeological Science 49 (2014) 302e316 307

degrees of dryness of the hides as well as to the variation in thegrease present in the interfacial layer.

Woodworking use-wear traces (Fig. 5) have rounded edges andrippled surfaces. The more highly polished zones show clear con-vexity. Their appearance is dense, bright and smooth and theydisplay striations oriented in the direction of use. In the lessaffected zones, wear is only visible on the higher parts of the relief,while the lower parts remain unmodified.

In the bone/antler use-wear case (Fig. 6), the microrelief of therock appears highly modified with little of the surface showingripples, and the formation of flat plains. A characteristic is the

Fig. 5. Example of wood-working use-wear traces. Top, OLM panoramic image showing woramig image from SEM secondary electron micrographs. Bottom left, detail of the OLM p

formation of sub-parallel linear depressions oriented in the usedirection. This association sometimes creates comet morphologies.Wear aspect is dense, bright, smooth and without striation. Theboundaries are clear with abrupt ends.

4.3. Working cinematic

In all cases the edge was used transversally to the worked ma-terial. We have observed a pulling movement using the ventral faceof the flake as a contact surface. Only in the 3 cases of wood andbone/antler working have pushing movements been observed.

ood-scraping polish on the functional edge of an endscraper. Middle, equivalent pan-anorama. Bottom right, detail of the SEM panorama.

Fig. 6. Example of bone/antler-working use-wear traces. Top, OLM micrograph-basedpanoramic image showing bone/antler scraping polish. Middle, equivalent SEM sec-ondary electrons micrograph-based panoramic image. Bottom, SEM secondary elec-trons micrograph-based panoramic detail of the middle image.

J.I. Morales, J.M. Verg�es / Journal of Archaeological Science 49 (2014) 302e316308

4.4. Working angle

Taking into account the extension and penetration of polish inthe ventral face of the tool, compared with our own experimental

data (Verg�es, 2003; Oll�e, 2003) and similar data already published(Guti�errez, 1996; Jard�on, 2000), the analyzed end-scrapers used inhide processing shows high working angles of over 60�. A directconsequence of these working angles is the fact that the wear doesnot penetrate to the tool's contact surface. In the most markedcases, the combination of high working angles and abrupt or obtuseedge delineations affects the ridges of the end-scraper retouchscars. Tools used for processing wood or bone/antler, however,show lower working angles.

4.5. Wear location and distribution

Use-wear traces caused by hide processing are alwayslocated in the retouched front edge and, sometimes, in thedistal parts of the lateral edges. Given the high number of hideprocessing cases we have observed a high degree of variabilitywithin this pattern. In contrast, the wear distribution caused bywood and bone/antler working is focused exclusively on themiddle part of the retouched edge, with a clearly delimiteddistribution. In wood processing cases, the width of wear is1 mm and 3 mm respectively, with degraded penetration of100 mm in one case and 500 mm in the other. In the only bone/antler case, the width is 1.5 mm with a very homogeneouspenetration of 700 mm.

4.6. Use-wear intensity and working time

The process of developing polish is not regular or progressive, dueto small fractures occurring as a brittle response to the strain (Oll�eand Verg�es, 2008, 2014), and also due to the total or partialresharpening of the working edge. An assessment of the intensity ordegree of deformation cannot be used as an unbiased method forcomparing the duration of the work developed by different tools.Nevertheless, categorizing intensity could be useful for observingdifferences in development between different zones of the same tool.

In order to characterize the degree or intensity of deforma-tion in the La Cativera's end-scrapers, five different categorieshave been established and tools have been assigned to one ofthem ( see distribution in Table 1; chi-square ¼ 9.714, df ¼ 4p ¼ 0.046).

1) Slight and discontinuous use-wear traces along the workingedge. It provides weak criterions to establish which variableshave taken part in the formation process. In some cases there issome doubt as to whether they can be identified as use-weartraces or just technical traces related to the retouch.

2) Light use-wear traces showing continuity along the workingedge.

3) Well developed and continuous use-wear traces.4) Well developed and continuous use-wear traces with clear edge

rounding.5) Well developed and continuous use-wear traces with dulled

edges.

4.7. Hafting

The criterions used for identifying hafting traces have beenbased on those published by Rots (Rots, 2008, 2010) and on ourown experimental series (Oll�e, 2003; Verg�es, 2003).

A total amount of 97 tools (75%) show some kind of evidenceof hafting in the form of micro-chips or polishes that we haveconsidered to be haft-related. We have defined three differentdegrees of development to classify hafting wears for each of thelateral edges of every single tool (see distribution in Table 2 and

Table 1Distribution of use-wear intensity into the different categories.

Intensity Absolute Relative

Slight and discontinuoususe-wear traces alongthe working edge

21 18.58

Light use-wear tracesshowing continuityalong the workingedge

28 24.78

Well developed andcontinuous use-weartraces

28 24.78

Well developed and continuoususe-wear traces with a clearedge rounding

16 14.16

Well developed and continuoususe-wear traces with dulled edges

12 10.62

Ø Observable 8 7.08Total 113 100

J.I. Morales, J.M. Verg�es / Journal of Archaeological Science 49 (2014) 302e316 309

correspondence plot in Fig. 7; chi-square ¼ 1.391, df ¼ 3,p ¼ 0.71):

1) Presence of micro-chip(s).2) Presence of micro-chip(s) and slight polish.3) Presence of micro-chip(s), highly polished, and dulled edge.

Other marks identified as resulting from hafting are abrasionsand polishes on the dorsal ridges and elevated zones of the dorsalsurface. Abrasions are characterized by matt appearance and acoarse texture, generated by an accumulation of microfractures.Polishes appear in form of flat plains with clear boundaries andabrupt ends. Close to the lateral edges, these deformations may beassociated and/or superimposed (Fig. 8).

5. Evidence of resharpening

The distribution pattern of ochre embedded in tools, the use-wear distribution, and the hafting-wear identification hasallowed us to observe and describe a set of criterions fromwhich atechnical process of resharpening is deduced. In some cases onlyone of these proofs is observed in tools, but in others, differentproofs converge for the same tool.

5.1. Evidence from ochre

Of the ochre distributions, the totally-covered tools with theexception of the distal end, are characterized by a very specificdiagnostic pattern stand out. It has been observed that, in mostcases, retouch scars cuts the ochre-covered surface, resulting in anew surface without any remaining ochre (Fig. 9). The end of theochre distribution is abrupt, and not transitional. In some of these

Table 2Cross-table of hafting wears intensity correlation between right and left lateral edges.

Right edge hafting intensity Left edge hafting intensity

Micro-chip Micro-chip & slight polish Mi

N % N % N

Micro-chip 13 59 2 9.09 0Micro-chip & slight polish 3 13 10 43.48 6Micro-chip & polish & dulled 1 3 2 5.71 24Ø present 7 16 3 6.82 2Ø Observable 2 40 0 0.00 3Total 26 20 17 13.18 35

end-scrapers, it is possible to observe an even more complexpattern of ochre distribution. The dorsal and ventral surfaces andthe more lateral and marginalized retouch scars, coinciding withthe delineation change between the distal end and the edge of thetool, are still covered with ochre, but the central part of theretouched area has been uncovered. This fresh uncovered surface isclear evidence of resharpening.

5.2. Use-wear evidence

In comparison with experimental series (Oll�e, 2003; Verg�es,2003; Jard�on, 2000; Keeley, 1980; Gonz�alez and Ib�a~nez, 1994)“anomalous” use-wear distributions have been documented. Theseare characterized by discontinuous wear distributions or distribu-tions within which the intensity varies (Fig. 10). Three categories ofanomalous use-wear distributions have been established:

A) Use-wear discontinuity along the working edge withdeformed and non deformed sections alternating. In somecases, this distribution is observed as small polished surfacesisolated along an unmodified edge.

B) The presence of use-wear in secondary zones of the workingedge, while the most probable contact zones appear fresh.

C) Use-wear continuity along the working edge, but with dif-ferential intensities that cannot be explained by the delin-eation of the edge, its morphology or working movement.

The fact that these use-wear patterns can be identified impliesthe existence of partial or incomplete resharpening stages duringuse.

5.3. Evidence from hafting traces

Hafting traces have been compared with use-wear traces on theworking edge in order to assess how they correspond in terms ofintensity. When the intensities of the hafting-traces and the use-wear traces do not correspond, we then consider that this is dueto the effect of resharpening. Clear differences between the activeedge and hafting zone polish development can be observed inFig. 11. Active edge shows none or minor polish but hafting zonedisplays a well developed and strong polish along the completeedge. These differences are reinforced by the fact that hide-workingis an activity that generates strong polishes faster, implying thatuse-wear must be generated earlier and more intensively thanhafting-wears.

Two different cases have been identified:

A) Presence of hafting-wear and absence of use-wear.B) Lack of correlation between use-wear and hafting-wear. We

can cautiously state that this has been only noted in cases ofstrong polish, dulled edge and chip presence in the haftsection and (1) slight and discontinuous use-wear traces

cro-chip & polish & dulled Ø present Ø observable Total

% N % N % N %

0 6 27.3 1 4.55 22 10026.09 4 17.4 0 0 23 10068.57 6 17.1 2 5.71 35 1004.55 32 72.7 0 0 44 10060.00 0 0 0 0 5 10027.13 48 37.2 3 2.33 129 100

Fig. 7. Correspondence plot showing the co-occurrence relationship between different hafting wear developments in both edges of the tool.

J.I. Morales, J.M. Verg�es / Journal of Archaeological Science 49 (2014) 302e316310

along theworking edge or (2) slight use-wear traces showingcontinuity along the working edge. When it has not beenpossible to clearly establish differences between tracescaused by retouching and/or intentional abrasion and haftingtraces, caremust be taken inmaking resharpening inferencesfrom hafting traces. These modifications may have beenmade intentionally, in order to regularize the edge and pre-vent damage to the bonding elements of the haft.

6. The archaeological visibility of resharpening

The lithic assemblage studied presents a level of preservation ofedges and mineral residues that has allowed us to follow a multi-evidence approach to identifying resharpening (identificationsare shown in Table 3).

Through the functional analyses of active-edges, it has beenpossible to identify which of the end-scrapers have been resharp-ened (21.8%). Only those with several clearly observable use-weargenerations were identified in this way.

Complementing functional analyses with observations of haft-ing wear increased the percentage of resharpened tools identifiedto 41.3% of the total. This method made it possible to identifyresharpening in 19.5% of the tools that show no other characteristicfeature (Table 4). Tools had to display wear deriving directly fromhafting, which had to be cross-compared with functional evidence.If there is an absence of use-wear traces, or the degree to which

these traces had developed does not correlate with the intensity ofthe hafting-wear traces, it is then very reasonable to assume that aresharpening phase occurred prior to the tool being abandoned(see Table 5).

Residue conservation introduces another type of direct evidenceof resharpening from the ochre distribution. Thirty end-scrapersshow a distribution pattern that fits with resharpening (Table 6)and in 10 of these (7.8%) ochre appears to be the only evidence ofresharpening.

The total number of resharpened end-scrapers comprises 49.1%of the sample. This means that it has been possible to establish thatat least half of the tools were used, maintained and used again. Wehave documented use-wear traces in 106 end-scrapers, 82% of thetotal. Taking this as a reference for the ratio of used to unused tools,we can say that the 45% of tools showing evidence of use had beenmaintained through resharpening. In addition, resharpening hasbeen observed in 16 tools that show no evidence of use. There aretherefore a total of 121 end-scrapers that may have been used, and63 of those have been resharpened. This gives a final resharpenedratio of 52%. The ratio could be understood to be the ResharpeningIndex for the assemblage, in the knowing that it always representsthe minimum percentage of tools resharpened. It is reasonable tosuppose that some of the tools showing no evidence of use-wearmay have been used and then retouched (or their use may havenot developed any traces of use) and also that other hafting tech-niques didn't result in observable hafting wear.

Fig. 8. Example of haft-derived traces. Top, SEM secondary electrons micrograph-based panoramic image view of a hafted edge section showing hafting traces. Bottom, scheme ofthe hafting distribution. Long arrows and continuous line shows the bifacial micro-chips. Small arrows show the flat plains presence along the edge. Middle enlarged details of thedashed polygons: left, strong polish and dulled edge caused by the bonding elements (probably leather cords), right, flat plains caused by the tool and wooden haft friction.

J.I. Morales, J.M. Verg�es / Journal of Archaeological Science 49 (2014) 302e316 311

7. From resharpening to reduction and mobility patterns

The La Cativera's end-scraper assemblage has yielded evidenceof systematic resharpening in nearly half of the total number oftools studied. Although this might seem a high percentage, the factis that ethnographical studies have shown that scrapers are

continuously used and resharpened until they have been exhaus-ted, and they are then discarded (Shott and Weedman, 2007;Brandt, 1996; Brandt and Weedman, 1997; Gallagher, 1977).

Although there is not a lot of quantitative information aboutend-scraper reduction, there are some comparable references thatcould be used to create a framework for the reduction carried out

Fig. 9. End-scraper fully embedded in ochre except for the distal and retouched zones. Bottom left, SEM secondary electron micrograph. Bottom right: SEM image of back-scattered electron showing the ochre distribution (white particles) and the abrupt transition between embedded and non-embedded surfaces caused by a micro-flake scar.

J.I. Morales, J.M. Verg�es / Journal of Archaeological Science 49 (2014) 302e316312

on the La Cativera tools. In Morales et al. (in press), the 3D-ERPmethod was developed to measure reduction in end-scrapers. Inthis study, the reduction carried out after the shaping phase wascalculated, and this is equivalent to the reduction after the first useof a tool that had not been resharpened. The mean value obtainedfor the reduction intensity was 0.05 ± 0.03. La Cativera end-scrapers have a 3D-ERP value of 0.26 ± 0.17. The difference be-tween these measurements clearly fit with the resharpening evi-dence documented by analyzing use-wear. That is to say, the toolshave been reduced to a greater extent than the initial shapingperformed on the 3D-ERP experimental series. A more extensivelyused measure of reduction e Kuhn's geometric index or GIUR(Kuhn, 1990) e gives a very similar result. The value of GIUR for theexperimental assemblage is 0.47 ± 0.13, while for the assemblagestudied is 0.62 ± 0.17. Highly resharpened and reduced Gamo end-scrapers showamean reduction value of 0.68 (Shott andWeedman,2007). If we try to know which is the lowest GIUR sensibility toend-scraper reduction (how much of the reduction could beattributed to resharpening events and how much to the shapingphase), GIUR value after shaping from Gamo end-scrapers is 0.53,being statistically similar to that provided by the only shapedexperimental material (0.47 ± 0.13). If we construct a relative GIURscale for the end-scrapers reduction ranging from 0.47 to 0.68, theLa Cativera end-scrapers are almost 30% less reduced than theGamo ones. This scale has been constructed based on the meanvalues, and a significant number of individual measurements ofreduction are higher or lower than these two extremes, but this isstill a good illustration of the difference in the amounts of reduction(Fig. 12).

There are several possible ways of interpreting this apparentexhaustion of tools after non-intensive use. Firstly, it is possible thateven more intensive SEM observations of all the specimens will

yield a higher number of resharpening events due to the identifi-cation of use-wear traces that are lighter or purely testimonial.Additionally, very intensive resharpenings before discarding couldhave erased all the features of functional wear, and hafting tech-niques that did not result inwearmay have been employed. Despitethese known limitations, it is also obvious that there must be dif-ferences between the hide-scraper curating behavior of Paleolithicforagers and the known behavior of historic and sedentary hide-scrapers. Radically different settlement dynamics and mobilitypatterns should be reflected in the people's technological organi-zation (Bamforth, 1986; Andrefsky, 1991), and at the same time,should condition the formation processes that resulted in thearchaeological record of lithic technology.

Since they are dependent on settlement and mobility, reductionintensity and tool use-time should vary from site to site, so onewould expect to find varying values for reduction. Using Binford's(1980) extremes of mobility patterns intuitively, reduction carriedout at the base camp of a group with some level of logisticalmobility should be high due to the re-use of a tool that is main-tained for as long as is possible. On the contrary, reduction carriedout at the base camp of a residential mobility group will dependexclusively on thework's intensity and on the degree towhich toolsare curated. If a tool is used to its maximum potential, and isconsequently transported from site to site, reduction values shouldbe high by the time it is discarded. Nevertheless, when tools areused and discarded on the site and not incorporated into the toolkitand transported from site to site, reduction values should vary, andshould probably be lower. In this case, the archaeological recordwould contain both exhausted tools and also tools that are stillusable.

Our study case shows a degree of technological organization inwhich end-scrapers are used and reused while the group occupies

Fig. 10. Example of anomalous use-wear trace distribution. SEM secondary electron mosaics showing: Top, complete edge of the tool. Black line indicates use-wear presence, whiteline indicates resharpened edge without evidence of use. Middle, partial detail showing clearly the alternation between previously used and resharpened surfaces. Bottom left,enlarged detail of traces of hide-working. Bottom right, contact point between used (left) and resharpened (right) surfaces. The change is marked by the change in delineationcaused by the resharpening retouch.

J.I. Morales, J.M. Verg�es / Journal of Archaeological Science 49 (2014) 302e316 313

the site and are probably abandoned when either the work or theoccupation ends, independently of its remaining potential ormaximum utility (Shott, 1996). In this way, end-scraper manage-ment could be defined as expedient, and this expediency allows usto make several inferences.

During the Late Upper Paleolithic, in the region in question therewas no type of standardization in the blanks selected for shapinginto end-scrapers. There is a great degree of variety in the flakesselected, and we found that cortical flakes, chunks, regularizationflakes, regular blades and flakes produced during full-productionhad been used. This heterogeneity is also enforced by the highpercentage of recycled pieces selected for shaping into end-scrapers. At other neighboring sites a Minimum Recycling Indexof 20% has been observed (Vaquero et al., 2012a). In La Cativera wecompute MRI using the double patina identification in both burnedtools and non-fire patinated tools, and, despite the low sample, MRIis even higher, 30.77% (Table 7). This behavior must be related towell-established mobility routes that imply periodic reoccupationof the sites. Thus the certainty of the immediate availability of lithic

rawmaterial in form of discarded pieces from previous occupationsreinforces the expedient behavior and the lack of curation of certaintool types.

The geological context in the region is characterized by theabundance of raw materials in primary outcrops drained by severalrivers and channels which create an infinity of secondary catch-ment areas (Vaquero et al., 2012b; Soto et al., 2014). Thus theabundance of raw materials in the territory, the establishedmobility routes, and the periodic reoccupation of the sites all in-fluence the technological organization (Andrefsky, 1994), making itunnecessary to transport tools (Bamforth, 1986), at least for certainspecific functional typologies.

The presence of end-scrapers resharpened before being aban-doned indicates contradictory behavior in maintaining tools thatare going to be discarded. Assuming cyclical reoccupation of thesites and the high indexes of recycling, it can be inferred that theusers anticipated returning from the fact that these pieces aremaintained and then not abandoned, but cached or kept at the siteto meet future needs (Stevenson, 1982). The result is a mixed

Fig. 11. SEM secondary electron mosaic images showing that the intensity at the end-scraper functional edge and the hafted left edge is clearly different. Top, panoramic view ofboth functional edge and left hafted edge. Dashed line indicates the active edge, and continuous line the hafted edge. Bottom left, generally fresh appearance of the functional edgeshowing some very low intensity use-wear traces between two resharpening retouches. Bottom middle, magnified detail of the low intensity use-wear traces. Bottom right, veryhigh intensity hafting traces showing bifacial micro-chips, strong polish and dulled edge.

Table 4Absolute and relative values of use-wear distribution.

Ø Use-wear Normal distribution A&B distribution C distribution Total

N 23 78 19 9 129% 17.83 60.47 14.73 6.98 100

J.I. Morales, J.M. Verg�es / Journal of Archaeological Science 49 (2014) 302e316314

picture in which tools are produced expediently for that situationand then cached as an “insurance-gear” for future use. Althoughtechnological organization is strongly related to mobility, a widevariety of scenarios is possible for the same mobility patterns(Sellet, 2013) depending on the resource management, functionalactivities and planned behavior in any given case.

Table 5Cross-table of hafting-wear intensity against use-wear intensity.

8. Conclusions

Use-wear studies of lithic tools constitute a valid approach notonly to reconstructing prehistoric work processes but also whenthe approach is behavioral or organizational. At many sites, lithictools are the principal archaeological remains, if not the only ones,and must be studied from a wide range of approaches. The classictechno-typological approach usually provides very useful infor-mation about the periodicity and geographical distribution ofprehistoric cultures, but technology clearly has the potential toprovide more information. This potential is often untapped in favorof more simplistic explanations of variability.

Technological variability is cross-correlated with behavioralvariability, but not always with cultural differences. Depending on

Table 3Absolute value, relative value and cumulative percentages of the identification ofresharpening depending on the kind of evidence.

Resharpening case Absolute Relative Cumulative

Use-wear 28 21.8 21.8Hafting 25 19.5 41.3Ochre exclusive 10 7.81 49.1

combined 31 e e

the mobility patterns, raw material availability, resource exploita-tion and other factors, prehistoric groups could generate a variety ofdifferent archaeological records within the same cultural scenario.In La Cativera, tool abandonment in the early stages of reduction is afactor that conditioned the end-scrapers’ abundance in thearchaeological record, and also its morphology. This implies largertools, less extensive retouching and greater morphological het-erogeneity. A change in this pattern, linked to raw material scarcityfor instance, would transform the lithic record recovered. Abun-dance would decrease, tools would be transported, reduction

Use-wear intensity

1 2 3 4 5 ØObservable

P

Hafting-wearintensity

N % N % N % N % N % N % N %

1 4 17.4 8 34.8 4 17.4 4 17.4 1 4.3 2 8.7 23 1002 4 19.0 6 28.6 4 19.0 1 4.8 4 19.0 2 9.5 21 1003 6 13.6 8 18.2 12 27.3 9 20.5 5 11.4 4 9.1 44 100Ø present 7 28.0 6 24.0 8 32.0 2 8.0 2 8.0 0 0.0 25 100P

21 18.6 28 24.8 28 24.8 16 14.2 12 10.6 8 7.1 113 100

Table 6Cross-table of ochre presence, functional use-wear distribution and hafting traces.

Ochre ØUse-wear

Normaldistribution

A&Bdistribution

Cdistribution

Total Haftingtraces

No 9 30 6 1 46 33Yes A 7 35 8 2 52 26

B 7 13 5 6 31 38Total 23 78 19 9 129 97

Fig. 12. Left, boxplot showing the GIUR values of the La Cativera sample comparedwith experimental and Gamo values. Right, histogram of the distribution of the LaCativera GIUR values, with the hypothetical distribution of a highly reduced assem-blage shown shaded.

J.I. Morales, J.M. Verg�es / Journal of Archaeological Science 49 (2014) 302e316 315

would be greater and therefore size would be smaller. Classificatoryapproaches sometimes attribute these variations to different lithictraditions e the microlithic “nail-like” end-scrapers versus theflake or laminar end-scrapers. Technological structure is aniso-tropic and changes depending on the conditions under which it isobserved, this anisotropy is caused by dynamism in the economicalstructure of foragers. Static approaches can thus transform humanresponses into unreal cultural variations.

Table 7Raw data for the Minimum Reduction Index calculation based on burned andpatinated tools as proxy for the minimum estimation.

N %

Endscrapers 129 100Patinated 12 9.30Burned 10 7.75Patinated & Burned 4 3.10P

26 100Recycled 8 6.20MRI 30.77

Acknowledgments

We are grateful to the staff of the Scientific and Technical Re-sources Service of Rovira i Virgili University for their help in theSEM observation process. This work has been developed within theframework of the Spanish MICINN projects CGL2012-38434-C03-03 and HAR2012-32548, and the Catalan AGAUR projects2009SGR-188. J.I. Morales is a beneficiary of a predoctoral researchfellowship (FI) from the AGAUR of Generalitat de Catalunya (FI-B2-2013). Authors also wants to strongly thank Michael Shott andKathryn Weedman Arthur for their clarifications and commentsabout the Gamo endscraper, to Andreu Oll�e, Maria Soto and PalmiraSaladi�e for their helpful comments, and to Dr. Richard Klein fortheir advices. The three anonymous referees clearly detect theweak points of the work and really help unraveling the previousversion of the manuscript. Any existing mistake is author re-sponsibility only.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.jas.2014.05.025.

References

Andrefsky, W., 1991. Inferring trends in prehistoric settlement behavior from lithicproductions technology in the southern plains. North Am. Archaeol. 12 (2),129e144.

Andrefsky, W., 1994. Raw-material availability and the organization of technology.Am. Antiq. 59 (1), 21e34.

Angelucci, D.E., 2003. Geoarchaeology and micromorphology of Abric de la Cativera(Catalonia, Spain). Catena 54, 573e601.

Bamforth, D.B., 1986. Technological efficiency and tool curation. Am. Antiq. 51 (1),38e50.

Beyries, S., 2002. Le travail du cuir chez les Tchouktches et les Athapaskans: im-plications ethno-arch�eologiques. In: Le travail du cuir de la pr�ehistoire �a nosjours. XXIIe rencontres internationales d'arch�eologie et d'histoire d'Antibes.APDCA, Antibes.

Beyries, S., Rots, V., 2008. The contribution of ethno-archaeological macro- andmicroscopical wear traces to the understanding of archaeological hide-workingprocesses. In: Longo, L., Skakun, N. (Eds.), Prehistoric Technology 40 Years after:Functional Studies and the Russian Legacy. Proceedings of the InternationalCongress Verona (Italy), 20e23 April 2005, BAR International Series. Archaeo-press, Oxford.

Binford, L.R., 1977. Forty-seven trips. In: Wright, R.S. (Ed.), Stone Tools as CulturalMarkers. Australian Institute of Aboriginal Studies, Canberra.

Binford, L.R., 1978. Dimensional analysis behavior and site structure: learning froman Eskimo hunting stand. Am. Antiq. 43 (3), 330e361.

Binford, L.R., 1979. Organization and formation processes: looking at curatedtechnologies. J. Anthropol. Res. 35 (3), 225e273.

Binford, L.R., 1980. Willow smoke and dog's tails: hunter-gatherer settlement sys-tems and archaeological site formation. Am. Antiq. 45 (1), 4e20.

Blades, B.S., 2003. End scraper reduction and hunter-gatherer mobility. Am. Antiq.68 (1), 141e156.

Borel, A., Oll�e, A., Verg�es, J.M., Sala, R., 2014. Scanning electron and optical lightmicroscopy: two complementary approaches for the understanding and inter-pretation of usewear and residues on stone tools. J. Archaeol. Sci. 48, 46e59.

Brandt, S.A., 1996. The ethnoarchaeology of flaked stone tools use in southernEthiopia. In: Pwitit, G., Soper, R. (Eds.), Aspect of African archaeology. 10thCongress of the PanAfrican Association for Prehistory and Related Studies.University of Zimbabwe Publications, Harare.

Brandt, S.A., Weedman, K.J., 1997. The ethnoarchaeology of hideworking and flakedstone-tool use in southern Ethiopia. In: Fukui, K., Kuimoto, E., Shigeta, M. (Eds.),Ethiopia in Broader Perspective: Papers of the XIIth International Conference ofEthiopian Studies. Shokado Book Sellers, Kyoto.

Casamiquela, R.M., 1978. Temas patag�onicos de inter�es arqueol�ogico. III. La t�ecnicade la talla del vidrio. Relac. Soc. Argent. Antropol. XII, 213e223.

Clark, J.D., Kurashina, H., 1981. A study of the work of a modern tanner in Ethiopiaand its relevance for archaeological interpretations. In: Gould, R.A., Schiffer, M.(Eds.), Modern Material Culture: the Archaeology of Us. Academic Press, NewYork.

Clarkson, C., 2002. An index of invasiveness for the measurement of unifacial andbifacial retouch: a theoretical, experimental and archaeological verification.J. Archaeol. Sci. 29, 65e75.

Clarkson, C., Hiscock, P., 2011. Estimating original flake mass from 3D scans ofplatform area. J. Archaeol. Sci. 38 (5), 1062e1068.

J.I. Morales, J.M. Verg�es / Journal of Archaeological Science 49 (2014) 302e316316

de Beaune, S.A., 1989. Fonction et d�ecor de certains utensiles pal�eolithiques enpierre. L'Anthropologie 93 (2), 547e584.

Dibble, H.L., 1984. Interpreting typological variation of Middle Palaeolithic scrapers:function, style or sequence of reduction. J. Field Archaeol. 11 (4), 431e436.

Dibble, H.L., 1987. Reduction sequences in the manufacture of Mousterian imple-ments of France. In: Soffer, O. (Ed.), The Pleistocene Old World. Regional Per-spectives. Plenum.

Dibble, H.L., 1995. Middle Palaeolithic scraper reduction: background, clarification,and review of evidence to date. J. Archaeol. Method Theory 2, 299e368.

Eren, M.I., 2013. The technology of Stone Age colonization: an empirical, regional-scale examination of Clovis unifacial stone tool reduction, allometry, and edgeangle from the North American Lower Great Lakes region. J. Archaeol. Sci. 40(4), 2101e2112.

Eren, M.I., Dominguez-Rodrigo, M., Kuhn, S.L., Adler, D.S., Le, I., Bar-Yosef, O., 2005.Defining and measuring reduction in unifacial stone tools. J. Archaeol. Sci. 32(8), 1190e1201.

Eren, M.I., Jennings, T.A., Smallwood, A.M., 2013. Paleoindian unifacial stone tool‘spurs’: intended accessories or incidental accidents? PLoS ONE 8 (11), e78419.

Fontanals, M., 2001. Noves aportacions a la intervenci�o del límit pleistoc�e-holoc�e alsud de catalunya: l'estudi de la indústria lítica del jaciment de la Cativera (ElCatllar, Tarragon�es). Butll. Arqueol. 23, 73e100.

Gallagher, J.P., 1977. Contemporary stone tools in Ethiopia: implications forArchaeology. J. Field Archaeol. 4 (4), 407e414.

Gonz�alez, J.E., Ib�a~nez, J.J., 1994. Metodología de an�alisis funcional de instrumentostallados en sílex. Universidad de Deusto, Bilbao.

Gould, R.A., 1971. The archaeologist as ethnographer: a case from the WesternDesert of Australia. World Archaeol. 3 (2), 143e177.

Gould, R.A., Koster, D.A., Sontz, A.H.L., 1971. The lithic assemblage of the WesternDesert aborigines of Australia. Am. Antiq. 36 (2), 146e149.

Guti�errez, C., 1996. Traceología. Pautas de an�alisis experimental, Foro. Arqueología,Proyectos y Publicaciones S.L., Madrid.

Hiscock, P., Attenbrow, V., 2002. Morphological and reduction continuums ineastern Australia: measurements and implications at Capertee 3. Tempus 7,pp. 167e174.

Hiscock, P., Attenbrow, V., 2003. Early Australian implement variation: a reductionmodel. J. Archaeol. Sci. 30, 239e249.

Hiscock, P., Clarkson, C., 2005. Experimental evaluation of Kuhn's geometric index ofreduction and the flat-flake problem. J. Archaeol. Sci. 32 (7), 1015e1022.

Iovit�a, R., 2010. Comparing stone tool resharpening trajectories with the aid ofelliptical Fourier analysis. In: Lycett, S.J., Chauhan, P.R. (Eds.), New Perspectiveson Old Stones. Analytical Approaches to Paleolithic Technologies. Springer.

Jard�on, P., 2000. Los raspadores en el Paleolítico Superior. Tipología, tecnología yfunci�on en la Cova del Parpall�o (Gandía, Espa~na) y en la Grotte Gazel (Sall�eles-Cabard�es, Francia). In: S.I.P. (Ed.), Serie de Trabajos Varios. Diputaci�on Provincialde Valencia, Valencia.

Jard�on, P., Sacchi, D., 1994. Traces d'usage et indices de r�eaffûtages etd'emmanchements sur des grattoirs magdal�eniens de la Grotte Gazel �a Sall�eles-Cabardes (Aude e France). L'Anthropologie 98 (2e3), 427e446.

Keeley, L.H., 1980. Experimental determination of stone tool uses. A microwearanalyses. In: Butzer, K.W., Freeman, L.G. (Eds.), Prehistoric Archaeology andEcology Series. University of Chicago Press, Chicago.

Kuhn, S.L., 1990. A geometric index of reduction for unifacial stone tools. J. Archaeol.Sci. 17 (5), 583e593.

Loebel, T.J., 2013. Endscrapers, use-wear, and Early Paleoindians in Eastern NorthAmerica. In: Gingerich, A.M. (Ed.), In the Eastern Fluted Point Tradition. TheUniversity of Utah Press.

Mansur, M.E., 1986. Microscopie du mat�eriel lithique pr�ehistorique. Traces d'uti-lisation, alt�erations naturelles, accidentelles et technologiques. Exemples dePatagonie. CNRS, Paris.

Marwick, B., 2008. What attributes are important for the measurement of assem-blage reduction intensity? Results from an experimental stone artefactassemblage with relevance to the Hoabinhian of mainland Southeast Asia.J. Archaeol. Sci. 35 (5), 1189e1200.

McCall, G.S., 2012. Ethnoarchaeology and the organization of lithic technology.J. Archaeol. Res. 20 (2), 157e203.

Morales, J.I., Verg�es, J.M., Fontanals, M., Oll�e, A., Allu�e, E., Angelucci, D.E., 2013. Losniveles B y Bb de La Cativera (El Catllar, Tarragona). Procesos t�ecnicos y cul-turales durante el Holoceno inicial en el noreste de la Península Ib�erica. Trab.Prehist. 70 (1), 54e75.

Morales, J.I., Lorenzo, C., Verg�es, J.M., 2014.Measuring retouch intensity in lithic tools:a new proposal using 3D scan data. J. Archaeol. Method Theory, 1e16 (in press).

Oll�e, A., 2003. Variabilitat i patrons funcional en els sistemes t�ecnics de Mode 2.An�alisi de les deformacions d'ús en els conjunts lítics del Riparo Esterno deGrotta Paglicci (Rignano Garganico, Foggia), �Aridos (Argando, Madrid) i Galería-TN (Sierra de Atapuerca, Burgos). Departament de Geografía i Hist�oria, Uni-versitat Rovira i Virgili, Tarragona.

Oll�e, A., Verg�es, J.M., 2008. SEM functional analysis and the mechanism of micro-wear formation. In: Longo, L., Skakun, N. (Eds.), “Prehistoric Technology” 40years later: Functional Studies and the Russian Legacy, BAR International Series,pp. 39e49.

Oll�e, A., Verg�es, J.M., 2014. The use of sequential experiments and SEM in doc-umenting stone tool microwear. J. Archaeol. Sci. 48, 60e72.

Philibert, S., 1993. Quelle interpr�etation fonctionelle pour les grattoirs ocr�es de laBalma Margineda. Traces et focntion: les gestes retrouv�es. In: Anderson, P.C.,Beyries, S., Otte, M., Plisson, H. (Eds.), Actes du colloque international de Li�ege.ERAUL, Liege, pp. 131e137.

Rots, V., 2008. Hafting traces on flint tools. Prehistoric technology 40 years alter:functional studies and the Russian legacy. In: Longo, L., Skakun, N. (Eds.), Pro-ceedings of the International Congress Verona (Italy), 20e23 April 2005, BARInternational Series. Archaeopress, Oxford, pp. 75e84.

Rots, V., 2010. Prehension and Hafting Traces on Flint Tools. Leuven UniversityPress, Leuven.

Schiffer, M.B., 1972. Archaeological context and systemic context. Am. Antiq. 37 (2),156e165.

Sellet, F., 2013. Anticipated mobility and its archaeological signature: a case study ofFolsom retooling strategies. J. Anthropol. Archaeol. 32 (4), 383e396.

Seronie-Vivien, M.R., 1986. L'utilisation de la gravure et de la peinture �a l'Azilien:l'apport du gisement de P�egouri�e (Caniac-du-Cause, Lot). BSPF 83 (11e12),416e422.

Shott, M., 1996. An exegesis of the curation concept. J. Anthropol. Res. 52 (3),259e280.

Shott, M.J., Weedman, K.J., 2007. Measuring reduction in stone tools: an ethno-archaeological study of Gamo hidescrapers from Ethiopia. J. Archaeol. Sci. 34(7), 1016e1035.

Shott, M.J., Bradbury, A.P., Carr, Ph. J., Odell, G.H., 2000. Flake size from platformattributes: predictive and empirical approaches. J. Archaeol. Sci. 27, 877e894.

Soto, M., G�omez de Soler, B., Vallverdú, J., Vaquero, M., 2014. Potential siliceoussources during Prehistory: results of prospecting in the East margin of the EbroBasin (NE Iberian Peninsula). J. Lithic Stud. 1, 293e318.

Stevenson, Marc G., 1982. Toward an understanding of site abandonment behavior:evidence from historic mining camps in the southwest Yukon. J. Anthropol.Archaeol. 1 (3), 237e265.

Torrence, R., 1983. Time budgeting and hunter-gatherer technology. In: Bailey, G.(Ed.), Hunter-Gatherer Economy in Prehistory: a European Perspective. Cam-bridge University Press, Cambridge.

Vaquero, M., Alonso, S., García-Catal�an, S., García-Hern�andez, A., G�omez de Soler, B.,Rettig, D., Soto, M., 2012a. Temporal nature and recycling of Upper Paleolithicartifacts: the burned tools from the Molí del Salt site (Vimbodí i Poblet,northeastern Spain). J. Archaeol. Sci. 39 (8), 2785e2796.

Vaquero, M., Chac�on, M.G., Cuartero, F., García-Ant�on, M.D., G�omez de Soler, B.,Martínez, K., 2012b. The Lithic Assemblage of Level J. In: Carbonell, E. (Ed.), HighResolution Archaeology and Neanderthal Behavior. Time and Space in Level J ofAbric Romaní (Capellades, Spain). Springer, New York.

Verg�es, J.M., 2003. Caracteritzaci�o dels models d'instrumental lític del Mode 1 apartir de les dades de l'an�alisi funcional dels conjunts litot�ecnics d'Aïn Hanech iEl-Kherba (Alg�eria), Monte Poggiolo i Isernia la Pineta (It�alia). Departament deGeografía i Hist�oria, Universitat Rovira i Virgili, Tarragona.

Weedman, K.J., 2006. An ethnoarchaeological study of hafting and stone tool di-versity among the Gamo of Ethiopia. J. Archaeol. Method Theory 13 (3),188e237.