neMrePessis Anne-Marie Ni hitleLahaye Christelle rcDouville Eric rcBo Eric (Piau da Furada Vale Pedra the America: South in sequence archaeological Pleistocene late new A 5 4 3 2 1 upeetr aeili rvddoln thttp://antiquity.ac.uk/projgall/boeda341 at online provided is material Supplementary methods dating tools, quartz tools, cobble technology, Keywords: ANTIQUITY C niut ulctosLtd. Publications Antiquity dad oae iad,Cmu saTj,Vlii,Chile Valdivia, Teja, Isla Campus Miranda, Morales Eduardo asnd l’Arch de Maison l’Universit nttt eCeca minae vltvs nvria uta eCie dfii mloPug Emilio Edificio Chile, de Austral Universidad Evolutivas, y Ambientales Ciencias de Instituto du Domaine CNRS-CEA-UVSQ, France 8212 Universit Yvette, UMR sur l’Environnement, Gif de 91198 et CNRS, Climat du Sciences des Laboratoire Arqueolog de Departamento Universit ArScAn–AnTET, CNRS 7041 UMR d Guidon ede ` Vale daPedraFurada eda ¨ 8(04:9795http://antiquity.ac.uk/ant/088/ant0880927.htm 927–955 (2014): 88 ´ odaxMnage M 00IAA-R2,Isiu eRcecesrlsArch les sur Recherche de Institut IRAMAT-CRP2A, 5060 UMR Bordeaux-Montaigne, e er uaa er eCpvr,stlmn fteAeia,tpooy lithic taphonomy, Americas, the of settlement Capivara, de Serra Furada, Pedra ´ ,903Nner,France Nanterre, 92023 e, 1 Brasilia gai Clemente-Conte Ignacio , ´ ooi,EpaaedsAtle,367Psa,France Pessac, 33607 Antilles, des Esplanade eologie, 3 6 0 iliHoeltz Sirlei , 4 6 ,MarioPino , ´ ayAntropolog y ıa 9 ieiViana Sibeli , m2000 km ´ a M-SC al Egipc Calle IMF-CSIC, ıa, 8 non Lourdeau Antoine , 5 ´ ai ,Mio el’Arch de Maison X, Paris e N ieeDlrn Felice Daltrini Gisele , ´ ,Brazil) ı, 10 927 ,Am one f yarpyfo h authors. the from reply a by comments, off re-thought. of rounded series be a by followed to is article have The in currently will settlement America the human South of that narrative conclude accepted authors assemblage. The tool stone the of analysis technical yasre of series a supported by is argument The more years. 000 back 20 than extending occupation human Pedra for evidence da new Vale produced of however have site Furada open-air nearby the at okhle fBoqueir of rockshelter (see accepted Furada universally Meltzer Pedra not were at Brazil occupation in early very for claims Previous subject. Americas contentious the a of remains settlement first the of date The 2 ihlFontugne Michel , leD Costa Da elie ´ tal. et ´ aus1,Breoa001 Spain 08001, Barcelona 15, ıaques ´ ooi td ’tnlge 1all 21 l’Ethnologie, de et eologie 14 1 94.Nwwr tthe at work New 1994). n S ae,adby and dates, OSL and C , 9 aiaPagli Marina , od er uaaand Furada Pedra da ao ˜ 1 6 , 7 & , 3 , ´ n Avenida ın, ´ eomat ´ eriaux, 1 ´ ede ee ,
Debate A new late Pleistocene archaeological sequence in South America
6 Fundac¸ao˜ Museu do Homem Americano (FUMDHAM), Centro Cultural S´ergio Motta s/n, Bairro Campestre, Sao˜ Raimundo Nonato – PI, 64770-000, Brazil 7 Universidade Federal do Vale do Sao˜ Francisco (UNIVASF), Campus Serra de Capivara, Rua Joao˜ Ferreira dos Santos, s/n, Bairro Campestre, Sao˜ Raimundo Nonato–PI, 64770-000, Brazil 8 Archaeo: Pesquisas Arqueologicas,´ Av. Caranda´ı 99, Parque Georgia, Cuiaba´ – MT, 78085-485, Brazil 9 Departamento de Arqueologia, Universidade Federal de Pernambuco, Av. Prof. Moraes Rego 1235, Cidade Universitaria,´ Recife – PE, 50670-901, Brazil 10 Instituto Goiano de Pr´e-Historia´ e Antropologia, Pontificia Universidade Catolica´ de Goias,´ Av, Universitaria´ 1.440, Setor Universitario,´ Goianiaˆ – GO, 74605-010, Brazil
Introduction Late Pleistocene human occupation in South America has been the focus of incessant debate for more than 20 years. Human presence in South America before 11 000 cal BC is now, as in North America, widely accepted by most of the scientific community (Miotti 2003; Bonnichsen et al. 2005; Waters & Stafford 2007, 2013; Goebel et al. 2008; Dillehay 2009; Collins et al. 2013). In addition to the celebrated Chilean site of Monte Verde (Dillehay 1997), several sites now show the occupation of South America during the Late Pleistocene, including Taima-Taima in Venezuela (Ochsenius & Gruhn 1979; Ranere & Lopez´ 2007), Arroyo Seco 2 in Argentina (Steele & Politis 2009) and Huaca Prieta in Peru (Dillehay et al. 2012). Acceptance of sites dating to earlier than 14 000 cal BC, however, seems to be much more difficult (Roosevelt et al. 2002; Dillehay 2008; Goebel et al. 2008). All discoveries made have been systematically and legitimately discussed, which is justified, but also systematically and surprisingly rejected, without exception. This has occurred, in particular, for sites in north-eastern Brazil, in the Serra da Capivara region (Meltzer et al. 1994; Meltzer 2009; Farina˜ et al. 2014). Is there some sort of curse that affects the common sense both of archaeologists making the discoveries, and their colleagues, at the announcement of an age older than 14 000 cal BC? In hypothetico-deductive reasoning, verifiable hypotheses are constructed on the basis of a series of assumptions, the confirmation of which would support the postulates or premises on which the hypotheses are based (Clark 1968). In other words, if one questions the assumptions, the construction of the hypothesis must also be questioned. Why then, despite abundant and varied data collected during the last two decades, are the new results systematically rejected? Is the importance of the hypothesis greater than the validity of the assumptions? Has the hypothesis become a paradigm with its own structure of thinking? Another issue concerns the value given to the initial assumptions when it comes to stone tools. Indeed, it would appear that when tool types familiar from later periods, like spear or arrow points, are found with Pleistocene ages, a multitude of objections are immediately raised concerning both the ages obtained and the stratigraphy. In contrast, when artefacts are outside our modern memory references, it is their human origin that is questioned and not the stratigraphic context or dating. From a strictly scientific viewpoint, this attitude is quite unsettling, raising the question that if scientific criteria are valid for the Holocene, why would they not be equally valid for late Pleistocene periods? Despite such difficulties, research in the Sao˜ Raimundo Nonato region (Piau´ı, Brazil) continued after the discovery of the Upper Pleistocene and Holocene sequence at Boqueirao˜ C Antiquity Publications Ltd.
928 ers e reat eercvrdaogwt odcacasta aebe ae to dated been have that charcoals wood with along recovered were artefacts few a metres, 000clB seTbeS) eoddt ncaca,fo ae tlatoemetre one least at layer a from charcoal, on date of second age A an provided S4). surface, Table the (see below BC cal 000 20 a icvrdi 98drn apino etecvtoso h meit periphery immediate the on excavations test of campaign a ofthesiteofBoqueir during The Furada. 1998 Pedra in da Vale discovered artefacts S at was the conducted with we in sequences study located the Holocene is here site and present We Pleistocene 1). (Figure new dates providing and Furada, Pedra da Vale hsts xaain(Bo excavation test this aPdaFrd nte18s(udn&Dlbis18;Gio 99 Parenti 1989; Guidon 1986; Delibrias & (Guidon 1980s the in Furada Pedra da sites. adjacent and Furada Pedra da Vale of Location 1. Figure aly ti oae ttebs fatlso adtn okalrsligfo rso of erosion from 10–35 resulting of rockfall slope sandstone a Baix of with the talus talus, of rocky a bank The of left base the the on at a site located open-air is an It is Valley. Furada Pedra 2011 da Vale in Furada Pedra da Vale at Excavations for material supplementary online (Lahaye (see Peia sites Tira new three details): of further discovery the with 2001) Parenti rmmnmna ocnieal mle.Teebok r ie ihsdmnsformed sediments with mixed are blocks These smaller. considerably to monumental from cuesta hc eei oeta 0 wyadsrie sarsda ig Fgr S1). (Figure ridge residual a as survives and away 30m than more is here which od er uaa(eie20,20) tadpho oeta two than more of depth a At 2002). 2000, (Felice Furada Pedra da ao ˜ oRiud oaorgo nteeg fteSrad aiaa It Capivara. da Serra the of edge the on region Nonato Raimundo ao ˜ eda ¨ tal et 2013). . tal et c 300clB seTbeS) n21,w reopened we 2011, In S4). Table (see BC cal 000 13 . rcBo Eric ◦ 03,S 2013), . scmoe fsnsoebok agn nsize in ranging blocks sandstone of composed is , 929 ¨ eda tal. et ´ tod eo(Guidon Meio do ıtio C niut ulctosLtd. Publications Antiquity od er Furada Pedra da ao ˜ tal et 94 and 1994) . tal. et 1996; c .
Debate A new late Pleistocene archaeological sequence in South America by quartz and quartzite cobbles, as well as sand resulting from both regressive erosion of the cuesta and the marine conglomerate overlying the Devonian sandstone (Pellerin 1983). Erosion of this conglomerate continues on the valley floor, but no longer contributes to the talus within the area of the site. Only during rainy years do two caldeiroes˜ (waterfalls) drain the plateau surface waters onto the valley floor. The larger of these creates every year an intermittent creek that changes slope from around 10◦ in this first 10m to become much flatter well before reaching the valley floor. The site is located on the left bank and at a distance of 15m from this ephemeral creek (Figure S1). The excavated surface was located over the earlier 30m2 test excavation, including its south and east stratigraphic profiles, and continued into the unexcavated lower layers (Figure S2). All of the lithic artefacts, knapped or not, were recovered. Three-dimensional coordinates were recorded for knapped artefacts, and the location of non-anthropic objects logged by quarter-metre in 50mm thick spits. Systematic sieving to 2mm enabled recovery of abundant small fragments of knapping debris and retouch flakes. Each charcoal fragment was individually recorded to better discern the sequence of burning. Within the 2.6m stratigraphy, eight sedimentary layers have been identified, although the base of the archaeological sequence has not yet been reached. The first layer— C1—is about 0.2m thick and corresponds to the modern surface. Below this, from bottom to top there is a succession of coarse and fine deposits that can be grouped into two broad units. The upper unit, 0.8–1m thick, contains a single layer—C2—dominated by siliciclastic silty sands, without visible lamination or other internal synsedimentary structures (Figure 2). The lower unit, currently excavated to a depth of 1.5m, includes six sedimentary layers, from C8 to C3, showing the alternation between episodes of fine, essentially sandy sedimentation (C8, C6 and C4) and episodes of coarser deposition (C7, C5 and C3) composed of a mixture of gravel with some cobbles and boulders in a sand-granule matrix. The three sandy layers are of different colours, evidence of different depositional and post-depositional processes. Layer C8 (base not yet reached) is powdery, resulting from the disintegration of large sandstone blocks on the talus against which the site is set. The base of layer C6 (>0.3m thick) contains, depending on the area excavated, several thousand wood charcoal fragments (mm and cm in size) concentrated in certain zones and making the sediment grey in colour. Sandstone blocks of different sizes (50–200mm) are present at the base of layer C6, lying directly on gravelly layer C7 or separated from it by a sandy ash layer a few centimetres thick (Figure S3). No quartz or quartzite cobbles, except for artefacts, are associated with these sandstone blocks. The origin of these blocks, some of which weigh 3–4kg, is problematic. Indeed, their presence cannot be explained exclusively by colluvial processes because other lithic materials would have been associated with them. Some blocks show evidence of heating that caused their breakage, which indicates a temporal correlation with the charcoal. A disc-shaped block some 150mm thick, currently being analysed, shows traces of use including smoothing/polishing and areas of pecking on the upper surface. Layer C4 varies in thickness (0.3–0.4m) depending on location and is defined uniquely by redder sand. Layer C7, 0.3m thick, differs from the preceding sandy layers by the more varied size of the cobbles it contains (some more than 0.3m across), and by their more diversified raw material. Quartzite cobbles, nearly absent in this layer, are found in the upper gravel layers. C Antiquity Publications Ltd.
930 ato h lfs ti ec o osbet bev h culadntrlpoessof deposition processes the natural that and deduce actual we the however, observe lower research, to the the possible of in not stage paintings this hence rock At is recent reworking. It more alluvial cliffs. the the of of display part and conservation archaeological weathering. with overlying by materials the clays local and produce few cliff would are that the There silt composition of explanation. mineralogical and sandstone viable the a Devonian sand a note the provide but where may as palaeoclimate, to soils, conglomerate, such Pleistocene environment marine the sediments, the the about in tropical of information and of source local sediments the kind detailed Pleistocene no this is the in There in dominate. interest both of more clay, rarely is of are cobbles It shortage interstratification The an long. thick. of of 50–100mm 40–50mm lenses composed sort sandy than is any and/or thick, indicate granular 0.8m to with C3, macro-traces gravels of Layer no surfaces. are 100mm supports their There this most of manuports. and at process, modification are mass depositional is they in same that difference which the Such hypothesis layer, by weight. explained the this in be 5–8kg in cannot blocks 1:100) yet sandstone than varied long, rectangular (more are few 50–60mm C5, a layer than find in we more only thick, rarely found but exclusively different almost sand size yellow cobbles, in bright Quartz and layers. granules other quartz all of from is cobbles these surrounding matrix (C8 The C3–C8 unit lower and C2; and reached). C1 been unit yet upper not the has showing base section its Furada: Pedra C7; da below Vale directly of is zone Southern 2. Figure h rgnlcluildpst r ocae ytearneet aefrthe for made arrangements the by concealed are deposits colluvial original The rcBo Eric 931 ¨ eda tal. et C niut ulctosLtd. Publications Antiquity
Debate A new late Pleistocene archaeological sequence in South America of the sediments was not a continuous process. It coincided rather with sporadic humid periods, during which enough water was running from the upper sections of the cliffs for local alluvial deposition that reworked the colluvium near the base of the cliffs. There are no structures such as lenses with cross-stratification or imbrication of the clasts that would indicate medium to high energy sedimentation environments. Thus, the energy of ephemeral streams was not sufficient to move the coarse material, producing lag deposits where the only remobilised sediment was the sand. This was followed by a long dry period without deposition, during which humans occupied the site. The large boulders are not genetically related to the alluvium, and represent either remains of earlier colluvium or human transport. The alternation between periods of no deposition and those with very low energy colluvium reworking by ephemeral streams would allow the preservation of the archaeological remains, including perhaps hearths. It is even possible that intense human activity broke down the original sedimentary structure of these deposits. Several archaeological layers have been identified. Artefacts were discovered in both the upper and lower units. The upper unit, excavated over an area of 5m2, contains two sparse archaeological horizons (C2a and C2b), separated by a 0.2m-thick sterile horizon (Figure 3). The two horizons are associated with extremely abundant charcoal scatters that in themselves do not suggest any human involvement. The archaeological material consists only of lithics and includes worked quartz and quartzite cobbles and flakes. The lower unit is much richer. Sedimentary unit C3, excavated over an area of 2m2, includes a minimum of four archaeological horizons (C3a to C3b) separated by sandy lenses. The archaeological material is generally found at the interfaces, but given the loose sandy context, the clear demarcation between each horizon and the thickness of the sandy lenses, the deposits are more or less easy to follow. A few charcoal fragments are also present in the archaeological horizons. Horizon C5a is found in gravel layer C5. It contains rectangular sandstone blocks weighing several kilograms (Figure 2) that cannot be explained by natural deposition processes. Unit C7 includes three distinct horizons (7a, 7b and 7c) of artefacts, separated by sterile horizons. The first horizon, 7a, excavated over an area of 6m2, directly overlies gravel layer C7. It is contained in sandy, charcoal-rich sediment forming the base of C6 in places and reaching a thickness of 100mm. This horizon has partially overlapping concentrations of charcoal fragments (several thousand per m2). A large number of sandstone blocks and chunks are scattered across the ground. Some show evidence of exposure to fire. The presence should be noted of two slabs (0.3×0.3m) with highly altered upper surfaces, unlike the surfaces of other smaller blocks. The two remaining archaeological horizons (7b and 7c), excavated over an area of 1m2, are completely different from each other and separated by sterile deposits. Horizon 7c includes an ashy concave elliptical zone (0.4×0.6×0.1m) at its centre with abundant charcoal and artefacts around its periphery. Unlike horizon 7a, no sandstone blocks or chunks are present. The ashy and charcoal-rich concentrations of horizons 7a and 7b differ in intensity of burning and in their structure. Those of horizon 7a do not appear to have been structured while horizon 7b has a central depression with abundant charcoal. Are all of the burning zones in layers 6base and 7 due to human activity, as our initial observations suggested? That cannot yet be definitively determined but we are awaiting the results of taphonomic and anthracological analyses of the charcoal, as well as sedimentary data from these zones. C Antiquity Publications Ltd.
932 nqatie h a aeilcmsfo h imnln ftemrn aesabove layers marine the of dismantling rarely the more cobbles, from quartz comes on material made raw mainly are The Furada quartzite. Pedra on da Vale from artefacts The assemblage lithic The (circles). locations sampling and (diamonds) results dating (triangles), horizons Archaeological 3. Figure rcBo Eric 933 ¨ eda tal. et C niut ulctosLtd. Publications Antiquity
Debate A new late Pleistocene archaeological sequence in South America the Devonian sandstone that has accumulated at the base of the cliff in the mass of fallen rocks, dispersed by intermittent run-off fed by the caldeiroes˜ . But this proximity of the raw material does not imply that the procurement territory was so restricted. Technological analyses have demonstrated that the selection was based on the quality of the quartz and quartzite blocks and on very specific morphological criteria that are much less common than one would believe, as we have verified by experimentation. Poor-quality flint is available in the Piau´ı terraces a few kilometres away but was not exploited, although it was present in low quantities in the Holocene industries where it is used for the production of specific tools such as lesmas. It may be recalled that the artefacts from Boqueirao˜ da Pedra Furada were considered the products of natural breakage when they fell from the cliff, despite scientifically rigorous technical and taphonomic arguments by Parenti (2001) confirming their human origin. To avoid once again becoming embroiled in endless discussions, we decided to extend the original study by undertaking four associated analyses: taphonomic, technological, experimental and functional. The source of the cobbles at Vale da Pedra Furada is clearly the same as for Boqueirao˜ da Pedra Furada, although the site is further from the cliffs. Taphonomic analysis was carried out first on material from Boqueirao˜ da Pedra Furada. (Details of the method can be found in the online supplementary material.) We were then able to compare these results with the material from Vale da Pedra Furada. It should be noted that throughout the excavation of Vale da Pedra Furada and regardless of sedimentary context, all objects greater than 20mm (fractured or not) were kept, sampled, weighed and technically analysed, forming an assemblage of several thousand objects. Through this comparison a series of 294 objects could be identified by the presence of technical traits that showed them to be different from those that taphonomic analysis had demonstrated to be of natural origin. This considerable work allows us to assert with confidence the undeniably human origin of the artefacts recovered from Vale da Pedra Furada. We also analysed the technology. Three processes have been identified in each of the archaeological assemblages: one involving shaping and two involving knapping. Analysis of the artefacts from the different Pleistocene layers shows the diversified production of tools on flakes and cobbles. Two methods dominate: bipolar reduction sensu stricto, and a reduction scheme to shape cobbles that had been carefully selected on the basis of certain criteria. The assemblages differ from one layer to another, evidencing differences in site function and/or different cultural facies (Figures 4 & 5, and supplementary Figures S7–S10). Functional analysis was also conducted to complete the study (Semenov 1964). Using low- and high-power magnification, we analysed a sample (n = 18) of quartz artefacts from Vale da Pedra Furada, from archaeological horizons C3 and C7. This experimental method allows us to analyse tools made on different raw materials and from all chronological periods (Keeley 1980; Keeley & Toth 1983; Clemente-Conte 1997). Some artefacts from both C3 and C7 present evidence of butchery activities. Some have marks linked to sawing and scraping of medium-hard materials such as wood. Some of the C3 artefacts also present marks showing they had been used to perforate hard animal materials and others show traces of activities on soft to medium materials such as hide (Figure 6). These results confirm that the artefacts were humanly shaped and were used by people in their everyday activities. C Antiquity Publications Ltd.
934 ehd ol eapid h ml iceac ewe h S and OSL the between both discrepancy wherever We small results, observed. The OSL are applied. and inversions radiocarbon be relation stratigraphic the could in No between methods ages correlation S16). OSL good (Figure the the of sediments remark coherence the also good of very the depth note the first We to 3. Figure and S6 Table material. in supplementary online the in included are occupations the the of chronology at These IRAMAT-CRP2A C8. at to radiocarbon (OSL) C2a In luminescence Universit from stimulated by France. at units, optically Gif-sur-Yvette, by accessible dated dated the in then of thus were l’Environnement all materials in were de samples sediment et C6base/C7 collected Climat we and parallel, du C4 Sciences Furada. des C3, Pedra Laboratoire da C2b concentrations. Vale different of the C2a, from analysis purpose Horizons chronological this for detailed collected a were samples was charcoal Different project the of goal Another Chronology (n edge transverse bevelled double with half-cobble worked C3b: Layer 4. Figure gsotie hudb oe:OLand OSL noted: be should obtained ages Fgr 1) h eane ftevle r oeet n h hoooyo h ieis site the of chronology and sediments the OSL coarser and of coherent, combination and are the fine values by between the defined limit of well the remainder sample C3 at The OSL the S13). situated of (Figure was rate dose sample annual the This of underestimation (BR2011-11). the from result may C3 layer h hoooia aaotie ybt ehd rdoabnadOL r reported are OSL) and (radiocarbon methods both by obtained data chronological The odax3 rne eal ftemtrasadmtosue oetbihthe establish to used methods and materials the of Details France. 3, Bordeaux e ´ rcBo Eric 14 935 oehrstaelyrC rud60 BC, 6000 around C2 layer situate together C ¨ eda 14 tal. et gs ial,teaslt auso the of values absolute the Finally, ages. C ◦ 192158). C niut ulctosLtd. Publications Antiquity 14 Cresultsfrom
Debate A new late Pleistocene archaeological sequence in South America
Figure 5. Asymmetric convergent pieces: on split cobbles: 1 (n◦ 192157) and 3 (n◦ 191220), layer 3c; on cobble, 2 (n◦ 192377), layer 7a. layer C3 between 12 500 and 17 500 BC, layer C4 between 15 000 and 17 000 BC, and finally layers C6 and C7 at around 22 000 BC.
Conclusion The site of Vale da Pedra Furada is a succession of open-air human occupations near monumental sandstone blocks at the base of a cliff and next to an intermittent stream. The combination of 14C and OSL dates situate the upper and lower units during Oxygen Isotope Stages 1 and 2. The upper unit is an early Middle Holocene deposit laid down between 9000 and 7000 BC. The artefacts within it cannot be attributed to a specific techno-cultural facies. The lower unit is clearly late in age, OIS 2, during the Last Pleniglacial. Two sedimentary deposits (C3 and C7) from this unit contain nearly all of the archaeological assemblages. Their 14C and OSL ages correlate them with two humid phases at 15 000 and 24 000 BP documented by marine core GeoB 3104-1 (Behling et al. 2000). These observations are corroborated by results obtained from nearby sites: Sitio do Meio (new excavations) (Boeda¨ et al. in press), Tira Peia (Lahaye et al. 2013) and Boqueirao˜ da Pedra Furada (new C Antiquity Publications Ltd.
936 hroladlrenmeso erh r vial opoieeiec fhmnactivities. human of evidence provide to available Only are zones. hearths sediments limestone of the the in numbers of sites large acidity at and present high charcoal are The these identified. although remains, be faunal to destroyed yet has resource materials other lithic and of materials animal patterns in tools continuity of set supports Early common Holocene micro-region. the a this the in during of management to disappear transcendence Pleistocene This Some Holocene. the sites. da the Vale Pleistocene from from into other assemblages persist the the selection other by of Holocene, blank features shared in generally, are More tools, Furada size. of Pedra in representation broadly and statistical rostres. are shaped) and the or Furada pieces in bevelled (knapped Pedra appear double da or differences Vale single contrast, as the In such from tools The characteristic layers made. certain be Pleistocene with can similar, different observations the several However, from large. toolkits is sufficiently not presence are Human assemblages times. the other at occupations of intensity. variable possibility of although the but periods, continuous, preclude two these not during does concentrated be this to seem occupations The excavations). n of analysis Use-wear 6. Figure ihrsett ucin h nlssso vdneo rcsigwo,hd,hard hide, wood, processing of evidence show analyses the function, to respect With because artefacts these of attribution cultural the on comment to possible yet not is It ◦ 907fo ae 7,idctn utn ciiy butchery. activity: cutting indicating C7b, layer from 198057 rcBo Eric 937 ¨ eda tal. et C niut ulctosLtd. Publications Antiquity
Debate A new late Pleistocene archaeological sequence in South America
These hearths have not been found in all of the archaeological assemblages and those that have vary in appearance. These results give the strong impression that each assemblage does not necessarily indicate the same functionality or the same duration of occupation. This is clearly evident for layer C6base, which is particularly rich and appears to be a palimpsest. The other assemblages occur in small clusters, as we observed at Boqueirao da Pedra Furada, located about 100m downstream, when we reopened excavations in the remaining part of the site. These similarities extend to the typo-technological and functional composition of the assemblages that we have recently recovered. Yet, having reached only the layers dated to 14 580 BC, it is still too early to conclude complete similarity between the sequences. Comparison of artefacts from Vale da Pedra Furada with the artefacts from the previous excavations shows a high degree of technological similarity, but as the excavation techniques were not the same, we can compare only artefacts with artefacts and not assemblages as a whole. In addition, we have only recently reached the uppermost Pleistocene layers. For broader comparison, it is impossible to include Santa Elina in Brazil (Vilhena Vialou 2005) because of the very low number of lithic artefacts there. The sites of Monte Verde in Chile (Dillehay 1997), Arroyo Seco and Los Toldos in Argentina (Cardich et al. 1973; Miotti & Salemme 2004; Steele & Politis 2009), Huaca Pietra (Dillehay et al. 2012) in Peru and Taima-Taima in Venezuela (Ochsenius & Gruhn 1979; Ranere & Lopez´ 2007) are too geographically distant to allow useful comparison; further, the raw materials are different. The problem of raw materials in the Piau´ı region has often been used as a reason to reject human modification of the material. Why were quartz cobbles used while much better materials could be found elsewhere? We respond to this with several arguments:
1) Regardless of period or place, groups exploit only raw materials from the immediate proximity; so why not quartz at Piau´ı? In East Asia, Europe and Africa quartz was commonly employed by Palaeolithic societies, yet does not present a problem for researchers. 2) The fallacy that considers cobble tools synonymous with archaic, and thus the Lower Pleistocene, prevents us from realising that the cobble is above all as much the basis for a tool as are shaped or retouched blades and flakes. 3) The term quartz covers a wide diversity of material and tools are made on the best-quality quartz. 4) The taphonomic analysis has demonstrated that it is impossible to confuse natural breakage and human production.
We should consider the world of cobbles, and quartz in particular, as the expression of one technological option among others. This technical orientation existed at different places and at different times. It is thus quite difficult to reconstruct possible population dispersals and their cultural links from their lithic artefacts. We can, on the other hand, identify the geographic limits of different technological options. Such is the case for the Capivara region, where we have an increasing number of exceptional sites located within a 10km radius and beyond, in different sedimentary contexts. The 14C and OSL dates, while not directly dating the humanly modified materials, allow us to situate these industries in time and thus to compare different periods. As already C Antiquity Publications Ltd.
938 B B igl nvriy(argn,San.Tak r u oJ-.Dmui n .Mra o M facilities. i AMS Rovira for at Moreau Service C. Technical and and Dumoulin Science J.-P. the to in due Unit are Microscopy Thanks the Spain). in (Tarragona, University MEB Virgili the at analyses with assistance his ’qian n h U Isiu nvriar eFac)frfnigadtcnclspot hsproject this Investissements support; the technical under Oll Agency A. and Mr National thank funding Research also French We for the ANR-10-LABX-52). (reference France) by Program managed d’Avenir de help Universitaire State French (Institut from benefited IUF the and Piau d’Aquitaine in mission Franco-Brazilian the of financing annual References (Fundac FUMDHAM and Affairs Foreign of Minister French the thank We not that should Acknowledgements why so direction, this in populations? have point ancient significant would more all without to differentiation data apply areas Cultural Ethnological stronger. densities. vast the patchy demographic across all low of dispersed been very periods, rather the dealing time to is are owing long We proposed contact them. over image between groups would links The we small visible few it. occurred, with with had for space that across evidence If disseminated significant occupations continents. found two these have of already invasion systematic and massive archaeological the at traces discovery. visible no and leave interpretation the which of not of population gives scale some that and should forms, us meaning many show events take with history can these and loaded movements Ethnology of too movements. migration is population visibility ‘migration’ of mass term of the waves impression the Yet these Indeed, record. of others. archaeological the striking obscure the most data. the in of successive Only visible of lack result irregular. the remain a were as to occupied owing that were America explain new events South to and migration these North difficult technical that general, still accept is to of In need that now populations. continuum We history of successive the complex group of a and a indicate rootedness frame, results typify long-term time and the another single and one a traditions They follow within disappear. that others assemblages innovations while archaeological technological industries, their Holocene the and of in both Some 000 found reveal homogeneous. 25 are functionally between and and typo-technologically known persist are currently characteristics age sites of years the 000 from 15 industries Pleistocene the observed, O EHLING EDA ¨ uhmyb h aefrtevr rtmgain,wihovosyddntivlethe involve not did obviously which migrations, first very the for case the be may Such litcn nutiso Piau of industries Pleistocene etrfrteSuyo h is mrcn,Texas Americans, First University. the A&M of Study the for Center (ed.) Waters odyssey Paleoamerican M.R. & Ketron C.V. Graf, K.E. S0277-3791(99)00046-3 Reviews 3104-1. GeoB core from marine inferences Brazil, northeastern in climate dynamics and vegetational quaternary Late 2000. V .D A. F ONTUGNE IANA . .L A. E., , ,H.,H.W.A A ,I.C C 9 8–4 http://dx.doi.org/10.1016/ 981–94. 19: OSTA ,S.H LEMENTE OURDEAU &M.P OELTZ RZ ,J.P -C 2–5 olg Station: College 425–45. : ,C.L AGLI ONTE ,N.G ATZOLD ¨ 03 h late The 2013. . AHAYE utrayScience Quaternary ,M.P ´ rzl e aa in data, new Brazil: , ı UIDON ,G.D.F &G.W INO .M P A.-M. , ,M. ELICE EFER rcBo Eric ESSIS . ,S. , 939 ¨ eda ´ ;Breu-otin nvriy h osi R Conseil the University, Bordeaux-Montaigne ı; C B C B tal. et ONNICHSEN O LARK ARDICH EDA ¨ olg tto:Cne o h td fteFirst University. the A&M of Texas Study Americans, the for Center Station: College el uv eLsTlo SnaCruz, (Santa Toldos Los Argentina). de 3 Cueva la de Methuen. press. L euni arqueol Secuencia W donn rne @rch France: OURDEAU ATERS ..1968. D.L. , ,E.,G.D.F ´ e.Msinfac br franco Mission ees. .. ..C L.A. A.R., , e nutispl industries Les 2005. . . ..L B.T. R., , ,C.L ¸ Relaciones oMsud oe mrcn)for Americano) Homem do Museu ao ˜ eo- ,rsace tIHSTraoafor IPHES-Tarragona at researcher e, ´ ´ ELICE nltclarchaeology Analytical aeaeia rgn:byn Clovis beyond origins: Paleoamerican editions. ´ AHAYE ogicaycronolog ´ ARDICH ,M.F C EPPER ´ eistoc :85–123. 7: niut ulctosLtd. Publications Antiquity ,M.P ´ slen 2010/11. esilienne ,D.J.S ` nsd Piau du enes ONTUGNE &A.H AGLI TANFORD ´ &S.V aradiocarb ıa AJDUK ,S.H London: . ´ .Nouvelles ı. IANA 1973. . OELTZ &M.R. egional ´ .In onica ´ ,A. .
Debate A new late Pleistocene archaeological sequence in South America
CLEMENTE-CONTE, I. 1997. Los instrumentos l´ıticos de KEELEY, L.H. 1980. Experimental determination of stone Tunel´ VII: una aproximacion´ etnoarqueologica´ tool uses. A microwear analysis.Chicago(IL): (Treballs d’Etnoarqueologia 2). Madrid: CSIC. University of Chicago Press. COLLINS, M.B., D.J. STANFORD &D.L.LOWERY. 2013. KEELEY,L.H.&N.TOTH. 1983. Microwear polishes on North America before Clovis: variance in early stone tools from Koobi Fora, Kenya. Nature temporal/spatial cultural patterns, 27,000–13,000 293: 464–65. http://dx.doi.org/10.1038/293464a0 cal y B.P., in K.E. Graf, C.V. Ketron & M.R. LAHAYE,C.,M.HERNANDEZ,E.BOEDA¨ ,G.D.FELICE, Waters (ed.) Paleoamerican odyssey: 521–39. College N. GUIDON,S.HOELTZ,A.LOURDEAU,M.PAGLI, Station: Center for the Study of the First A.-M. PESSIS,M.RASSE &S.VIANA. 2013. Human Americans, Texas A&M University. occupation in South America by 20,000 BC: the DILLEHAY, T.D. 1997. Monte Verde: a late Pleistocene Toca da Tira Peia site, Piau´ı, Brazil. Journal of settlement in Chile. Vol. 2: the archaeological context Archaeological Science 40: 2840–47. and interpretation. Washington, D.C.: Smithsonian http://dx.doi.org/10.1016/j.jas.2013.02.019 Institution Press. MELTZER, D.J. 2009. First peoples in a new world: – 2008. Profiles in Pleistocene history, in H. Silverman colonizing Ice Age America. Berkeley: University of & W.H. Isbell (ed.) The handbook of South California Press. American archaeology: 29–43. New York: Springer. MELTZER, D.J., J.M. ADOVASIO &T.D.DILLEHAY. – 2009. Probing deeper into first American studies. 1994. On a Pleistocene human occupation at Pedra Proceedings of the National Academy of Sciences of the Furada, Brazil. Antiquity 68: 695–714. USA 106: 971–78. http://dx.doi.org/ MIOTTI, L.L. 2003. Patagonia: a paradox for building 10.1073/pnas.0808424106 images of the first Americans during the DILLEHAY, T.D., D. BONAVIA,S.L.GOODBRED,JR,M. Pleistocene/Holocene transition. Quaternary PINO,V.VASQUEZ´ &T.ROSALES THAM. 2012. A International 109–110: 147–73. http://dx.doi.org/ late Pleistocene human presence at Huaca Prieta, 10.1016/S1040-6182(02)00210-0 Peru, and early Pacific Coastal adaptations. MIOTTI,L.L.&M.C.SALEMME. 2004. Poblamiento, Quaternary Research 77: 418–23. movilidad y territories entre las sociedades http://dx.doi.org/10.1016/j.yqres.2012.02.003 cazadoras-recolectoras de Patagonia. Complutum FARINA˜ , R.A., P.S. TAMBUSSO,L.VARELA,A. 15: 177–206. CZERWONOGORA,M.DI GIACOMO,M.MUSSO,R. OCHSENIUS,C.&R.GRUHN. 1979. Taima-Taima: a BRACCO &A.GASCUE. 2014. Arroyo del Vizca´ıno, late Pleistocene Paleo-indian kill in northernmost Uruguay: a fossil-rich 30-ka-old megafaunal locality South America: final reports of 1976 excavations. with cut-marked bones. Proceedings of the Royal Berlin: CIPICS & South America Quaternary Society B: Biological Sciences 281(1774): 20132211. Documentation Program. http://dx.doi.org/10.1098/rspb.2013.2211 PARENTI, F. 2001. Le gisement quaternaire de Pedra FELICE, G.D. 2000. S´ıtio Toca do Boqueirao˜ da Pedra Furada (Piau´ı, Br´esil), stratigraphie, chronologie, Furada, Piau´ı – Brasil: estudo comparativo das ´evolution culturelle. Paris: Recherche sur les estratigrafias extra s´ıtio. Unpublished MA Civilisations. dissertation, Universidade Federal do Pernambuco. PARENTI,F.,M.FONTUGNE &C.GUERIN´ . 1996. Pedra – 2002. A controversia´ sobre o s´ıtio arqueologico´ Toca Furada in Brazil and its ‘presumed’ evidence: do Boqueirao˜ da Pedra Furada, Piau´ı–Brasil. limitations and potential of the available data. FUMDHAMentos 2: 143–78. Antiquity 70: 416–21. GOEBEL,T.,M.R.WATERS &D.H.O’ROURKE. 2008. PELLERIN, J. 1983. Missao˜ geomorfologica´ em Sao˜ The late Pleistocene dispersal of modern humans in Raimundo Nonato, sudeste do Piau´ı, Brasil. the Americas. Science 319: 1497–502. Cadernos de Pesquisa (S´erie Antropologia-II) 3: http://dx.doi.org/10.1126/science.1153569 203–23. GUIDON, N. 1989. On stratigraphy and chronology at RANERE, A.J. & C.E. LOPEZ´ . 2007. Cultural diversity in Pedra Furada. Current Anthropology. 30: 641–42. late Pleistocene/early Holocene populations in GUIDON,N.&G.DELIBRIAS. 1986. Carbon-14 dates northwest South America and lower Central point to man in the Americas 32,000 years ago. America. International Journal of South American Nature 321: 769–71. http://dx.doi.org/ Archaeology 1: 25–31. 10.1038/321769a0 ROOSEVELT, A.C., J. DOUGLAS &L.BROWN. 2002. The GUIDON,N.,F.PARENTI,M.F.DA LUZ,C.GUERIN´ & migrations and adaptations of the first Americans: M. FAURE. 1994. Le plus anciens peuplement de Clovis and pre-Clovis viewed from South America, l’Amerique: le Paleolithique´ de Nordeste bresilien.´ in N.G. Jablonski (ed.) The first Americans: the Bulletin de la Soci´et´ePr´ehistorique Franc¸aise 91: Pleistocene colonization of the New World: 159–235. 246–50. San Francisco: California Academy of Sciences.