Tool assemblages

This sequence suggests an iconic backed geometric particularly crescent-shaped microliths increasing emphasis on backed technology through time from a are not the dominant form of retouched tool in many HP assem- local origin, and nothing suggests an abrupt arrival of a radically blages. Where backed tools are common, these are often not of different technology.

At this locality the lithic assemblage consists of three et al. While in the case of Diepkloof, backed tools components Leakey et al. In the early to middle Howiesons Poort at material excavated in There are only six complete microliths been concealed by rising sea levels. The Indian sites such as Patne and Jwalapuram 9 JWP 9 illustrate assemblage appears to be a local variant of a Late Palaeolithic in- incremental changes in lithic technology from the Middle Palae- dustry Leakey et al.

James' Mumba rockshelter is one of the classic East African archaeo- technological analysis of lithics from the site of Patne one of the logical sites. In seen as emerging locally and gradually out of the preceding MP. However, attribute analysis of the entire sequence Eren et al. Similarly, blade use appears to be a et al.

Aside Misra and Bellwood, This again suggests that geometric microliths emerged Bellwood, In fact, proponents of the occurred Ambrose, , This dated. This is described as a unique assemblage, but with extremely complex of sites includes the youngest MP sites known from South low artefact densities. To reiterate, H. As none of the in MP and LP assemblages. The assemblage is indeed domi- veys in some areas including those with very narrow coastal nated by microblade production, but strikingly only two retouched shelves such as southern Arabia, the third component of their artefacts have been found.

Mishra et al. De distributions are not provided for any samples, nor are strategies or the character of retouch James and Petraglia, The reader is therefore unable to Instead an initial phase of blade and microblade production is fol- evaluate the distributions for high skewness, which may indicate lowed by a proliferation of microlithic technology around 35 ka, as that partial bleaching has occurred, and would therefore necessi- would be expected with an in situ origin of the Late Palaeolithic.

It Please cite this article in press as: Groucutt, H. Location of key dated Late Middle Palaeolithic sites. These sites Finally, it can be noted that the early Sri Lankan microlithic as- are found across southern Africa, along the Rift Valley and in the semblages, some of the earliest in South Asia, are generally non- Middle East through to India.

While some of the Middle Eastern geometric in form. If, as seems highly likely, many of these late MP Deraniyagala, ; Perera et al.

The Middle Palaeolithic in various spatial and temporal contexts during the Late Pleisto- cene. This is the dispersal of H. Aside from making the general Cave, Iran Tsanova, The Middle Palaeolithic Middle Stone Age of Africa we argue that the different models are not necessarily contradic- tory as our default expectation should be that there were repeated 4.

The Middle Palaeolithic of South Africa population dispersals when environmental conditions allowed it. Given that more than km separates many of these Please cite this article in press as: Groucutt, H. Tryon, Shea a, pp. A number of recurrent centripetal and preferential reduction.

Porraz et al. For instance, a. The differs from areas such as East Africa. For example, to Porraz et al. Wurz like- ; contra Usik et al. Examples of recurrent centripetal Levallois cores have Discriminant Function Analysis to suggest similarities between been described in the literature as discoidal cores, asymmetric South African and South Asian Middle Palaeolithic core forms discoidal cores, disc cores, radial cores, and other terms. Given such Petraglia et al.

Given the lack of published attribute data on points, than in East Africa e. This retouched tools sensu e. Scerri, a; Scerri et al. The key point is not the exact frequency of these observed. The much discussed Howiesons Poort and Still Bay tool forms, but the dominance of these basic types and the rarity of technocomplexes are associated with late MIS 5 through to MIS 3 others. But there is also ; Shea, a; Sahle et al. Previous increasing evidence for variability within the Howiesons Poort.

Wendorf and Schild, , pp. Will et al. As further sites are discovered and additional et al. These late Middle Palaeolithic assemblages are chronometric estimates become available the picture of continuity technologically varied and further research is needed to under- becomes even more pronounced.

For example, Shea a em- stand the spatial and temporal characteristics of this variability. The earlier and later ends of this 4. The Middle Palaeolithic of East Africa spectrum are represented by Gademotta and Porc Epic respectively, with recent research suggesting that the former may actually date The oldest known fossils of H.

Shea a suggests that the small differences important role in the origin of our species e. Basell, ; Shea, between the Omo Kibish sites he analysed and other, similar, as- a. Wurz , pp. In what follows we therefore erential methods SOM 10 Fig. Levallois methods of producing population movements SOM 10 discusses the presence and im- points were also used, but much less frequently. As Levallois plications of different forms of technological packages elsewhere. East African Middle Palaeolithic lithics.

The Middle Palaeolithic of Northeast Africa area can meaningfully be subsumed in such a technocomplex has been debated e. Kleindienst, Hawkins clear similarities with the East African Middle Palaeolithic. Along- et al. Wendorf et al. Marks, of increased humidity see e. Drake et al. These sites are a,b; Van Peer et al.

Blome et al. Van Peer, ; Rose et al. Key examples come , but the extent to which much of the archaeology of the from the MIS 5 Bir Tarfawi and Bir Sahara palaeolakes of Please cite this article in press as: Groucutt, H. Map of key sites with similarities to East African Middle Palaeolithic sites.

A model of last interglacial precipitation is used to represent humid phases. As described and illustrated in the core reduction strategies and the retouched tool forms typical considerable detail by Wendorf et al. It is important to The Khormusan industry was discovered at several excavated note that compared to the MP of East Africa, there is already a sites in northern Sudan Marks, b.

As well as large lithic as- considerable reduction in the frequency of points, and this helps semblages Fig. Unfortunately the age of the Khormusan assem- Levant. Both the debitage and retouched tools are like- blages.

Sites such as Fig. Examples of lithics from several phases of occupation at Bir Tarfawi, Egypt. Northwest Africa Libya. Ongoing work is clarifying the nature of the lithic assem- blages at the site, but published illustrations appear to show lithics As our focus here is on dispersal out of Africa, space prevents a with the same basic features we highlight as characterising the East thorough discussion of the northwest African record.

Most research in African MP e. McBurney, , pp. Reynolds this area has addressed the Aterian technocomplex e. Bouzouggar suggests similarities of the Haua Fteah lithics to assemblages in the and Barton, ; Scerri, a; Scerri et al.

Traditional Levant and northeast Africa, rather than North African tech- attention has focussed on the typology of retouched artefacts e. Relatively little information is available on the core compared to the other sampled sites. Available evidence suggests reduction strategies at most Aterian sites. Likewise, little information the absence of the type fossils of the Aterian and Nubian Complex is available on the technology of pre-Aterian MP sites such as Jebel from Haua Fteah.

Douka et al. We acknowledge that it is quite possible that the Maghreb dating programme of the site, showing that the MP at the site is played an important role in the background to dispersal out of Africa, present by MIS 5 and extends through to MIS 3.

Examples of lithics from northeast African Middle Palaeolithic sites. AeH: , Sudan. IeP: Station One, Sudan. I,J: Levallois cores. QeW: 34A, Sudan. One of the few sites where good information is available is that Levallois and appears similar to that of East Africa, with both of Ifri N'Ammar, which has been described in a detailed monograph preferential and recurrent centripetal methods being well repre- Nami and Moser, The earlier Middle Palaeolithic at the site sented.

Core reduction is heavily tanged points at Ifri N'Ammar, and that the latter can be seen as a Please cite this article in press as: Groucutt, H. Tanging can be seen as a particular form of hafting rebian sites such as Contrabandiers e.

Dibble et al. Examples of lithics from S, Sinai Peninsula. Lithics from Qafzeh Cave. Sites along possible routes between East and northwest Africa, MP, and is found exclusively with fossils of H. We see the Levantine Late An example of a site 5.

The Middle Palaeolithic of the Levant Zaidner et al. The site appears to span MIS 6 to late MIS 5, Most of our knowledge on prehistory in Southwest Asia although as the artefact bearing sediments at the site have been currently comes from the Levant, which has a diverse MP record deposited through processes such as slope wash there is a risk of e.

Shea, , ; Hovers, ; Groucutt, , suggesting partial bleaching of grains and age overestimation. Along with its manufacturers is weak. Change through time in key lithic forms at Qafzeh Cave.

The chronology of these sites matches the MIS 6 Levallois cores are technologically and morphologically similar to wet phase that saw northwards expansions of populations in Africa cores from East Africa Fig.

Recurrent and preferential cores at e. Qafzeh are similar in basic size and shape features, indicating that Several Levantine assemblages dating to MIS 5 resemble the the differences between them do not relate to reduction intensity East African MP, including some associated with H. Hovers, Shea, , ; Hovers, Qafzeh. As elsewhere, its similarities with other sites are obscured by East African MP than the later ones are.

An important transition variable nomenclatures. Layer XV is also of interest as this is very unusual for Qafzeh described as Levallois cores.

Forms of Levallois reduction present in having a large number of Levallois points, the frequency of which include recurrent centripetal and preferential with centripetal rapidly increases and then rapidly decreases. This phase of point preparation Fig. Shea, that tively be seen as retouched points. Clearly more research is needed resulted in demographic changes affecting lithic technology, as well on the site and the assemblage, but it is certainly consistent with as numerous other factors indicating changes in how the site was the hypothesis proposed in this paper Fig.

The buried lithics, used Hovers, , pp. This suggests that chronology seems parsimonious. Qafzeh dates to mid rica, while core technology stays more constant.

With multiple layers, Qafzeh is a rare dence relates to the geographical distribution of sites Fig. It is example allowing the study of the nuances of change through time commonly implied that H. Stewart and African MP Figs. Stringer, While not an unreasonable hypothesis, this pattern Please cite this article in press as: Groucutt, H. Examples of Middle Palaeolithic lithics from the Arabian Peninsula. AeD: Jubbah, northern Saudi Arabia. EeG: Mundafan, southwest Saudi Arabia.

Research in Saudi Arabia being conducted by the common in the northern and central, but apparently rare in the Palaeodeserts Project is revealing a number of assemblages, dating southern, Levant. This may indicate that populations from published illustrations to include high frequencies of recur- leaving Africa primarily moved initially eastwards through Arabia, rent centripetal Levallois reduction and other features common in before moving northwards to the Levant along the TigriseEuphrates East African MP assemblages Fig.

Given the persistence of arid areas even during peak et al. Similar assemblages are known from other surface interglacials Jennings et al. In the case of Jebel Faya, Drake et al. The Middle Palaeolithic of Arabia allows for a more robust consideration of the site within the model proposed here.

Bretzke et al. Illustrations show both recurrent centrip- phase of recent developments focussed on the southern and etal and preferential with centripetal preparation core forms.

Such eastern extremities of the Peninsula. These revealed assemblages reduction methods, dating to MIS 5, are clearly congruent with the interpreted as either autochthonous developments Armitage et al. The rich, but cation. The younger post-MIS 5 assemblages at Jebel Faya show a undated, archaeological record of the Huqf area of Oman presents diversity of typo-technological features, congruent with population another distinctive variant of the Arabian Palaeolithic Jagher, The fast changing picture of the Arabian Middle Palaeolithic In Saudi Arabia, knowledge on vast areas is limited to rapid Groucutt and Petraglia, , therefore suggests repeated reconnaissance surveys conducted during the 's.

Arabia, as suggested by many of the illustrations of these surveys Within this pattern we emphasise the evidence for East African MP- published in the journal Atlal. In particular, Zarins et al. Many increasing evidence suggests the amelioration of Arabia occurred of these appear to be recurrent centripetal Levallois cores Fig. The most important 5. Otte et al. The increasing knowledge of the Iranian MP is being matched by its Many of the other sites at Jubbah relate mostly to raw material increasing diversity.

Smith, Until recently, little was known and only an early stage of reduction remains to be fully understood. Not much progress was semblages from the Nile Valley using a suite of multivariate sta- made on clarifying either the variability within Iran, or how this tistics.

Their results demonstrate that the North African and related to surrounding regions. Arabian sites do not neatly separate into their respective The sites of the more southern Zagros area, such as Kunji Cave, geographical areas.

This result indicates that lithic variability demonstrate technological traits similar to the East African MP within the Saharo-Arabian belt does not match modern political Fig. Core technology is dominated by recurrent centripetal structures, and is congruent with a northeast African background of Levallois cores, while the presence of large ovoid preferential the makers of at least some of the Jubbah assemblages. The dominant retouched compo- African MP Petraglia et al.

Others, such as Shanidar in closely neigh- blages characterised by both recurrent centripetal and preferential bouring northern Iraq Solecki and Solecki, and Warwasi Cave, centripetal Levallois reduction Fig. Rosenberg et al. In this sense Palaeolithic sites in the western Nefud desert also feature charac- technology at Warwasi is primarily characterised by a combination teristics analogous to the East African MP.

The centripetal to MIS 5 by Rosenberg et al. The lithic assemblage at this Levallois cores that dominate the Kunji assemblage are all but absent site, while again small, is dominated by the key forms also found in from Warwasi. Understanding the character of this emerging tech- the East African MP. Armitage et al.

Lithics from Kunji Cave, Iran. Misra, ; James and Petraglia, Due to the number of sites in the Thar Desert Blinkhorn et al. The recently excavated site of combine Levallois and discoidal debitage schemes with a wide Katoati Blinkhorn et al. Continued presence of Middle Acheulean populations. The variability of point production Palaeolithic assemblages from MIS 5 and into MIS 3 is also observed methods is particularly notable, including both debitage various at Jwalapuram Clarkson et al.

As Blinkhorn et al. Discoidal cores at JWP 22 and this arguably parallels aspects of technology associated with H. Retouched types at JWP sapiens to the west. Although the number of Dispersal SOM 10 are found in Indian Middle Palaeolithic sites, well dated sites currently prohibits an analysis of spatial diversity suggesting that populations dispersed eastwards in a basically within the Middle Palaeolithic of South Asia, evidence of diachronic longitudinal manner.

The current lack of evidence for Neanderthals trends, such as the emergence of blade production, are evident e. Important sites include Jwa- African-like features to more localised forms of technology, which lapuram 22 JWP 22 Haslam et al. James and Petraglia, ; Jones and Pal, ; Clarkson et al. Along with OSL This occurred after MIS 5, and the basic pattern of techno- estimates clustering around 85 ka from the archaeological layer logical continuity before and after the Toba eruption has been itself, the occupation is well dated to MIS 5a.

The technology at this described in several publications Petraglia et al. Available evidence suggests that there is fying patterns such as adaptation and dispersal. In ferences does not necessarily disprove demographic or cultural distinguishing such spatialetemporal entities we must rely on the connections.

Beyond super- quencies. Marks, a; nology of when these technologies emerged. South Asia present strong evidence for the in situ emergence of the Shea, , ; Hovers, and the Arabian Peninsula Late Palaeolithic. In the case of South Asia, for instance, blade Armitage et al. The earliest One must ask, dle Palaeolithic technologies through time.

South Asia. Problematically, the lithic methodology of Mellars et al. For example, need of further research see also e. Wallace and Shea, ; their analysis is based on literature review and the comparison of Meignen et al. Such an approach ignores not East Africa, in contrast to other regions which saw repeated ex- only recent research in areas such as South Africa and South Asia tinctions e. Shea, b. Elston and Kuhn, ; Hiscock et al. In Arabia and sise that the similarities highlighted in this paper corresponds with South Asia hominins were living inland at the time cited by Mellars the dispersal of H.

Like- This hypothesis, which clearly demonstrates that alternatives are wise, once people reached Sahul they rapidly expanded to a di- possible to the model of Mellars et al. Clarkson, the types of attribute analyses, which are proving useful elsewhere , pp. Similarly, H.

Tostevin, ; Scerri, a; Scerri et al. Basell, , perhaps indicating improved dispersal abilities e. Even if evidence is found to support such dispersals, they McBrearty and Brooks, After MIS 5, populations in many are likely to merely represent one phase of repeated dispersals out areas appear to have contracted until MIS 3 in the face of climatic of Africa, which we contend began by MIS 5.

Likewise, if evidence is deterioration. Shea, ; dispersal. Scerri et al. The environ- mental context of the origins of modern human diversity: a synthesis of microlithic technologies. We therefore cannot agree to the presence regional variability in African climate ,, years ago. In: Monographie du the early dispersals of H. This is fully congruent with sub- CRA, vol. CNRS, Paris. Levallois: volumetric construction, methods, a technique.

In: sequent dispersals taking place, perhaps including the frequently Dibble, H. Har HariL 5. NahaJ Zin: 6. Erq-el-Ahmar; 7. Mazaraq an Na 'aj; 8. El Quseir ; 9. Qafza; Phillips of the University of Illinois. Although the Bordean method thic sites Bar-Yosef and Belfer ; Gisis and has been criticised, the graphic representation of Gilead ; Gilead Fig.

The type list used here is the per Paleolithic occurrences have been discovered, one proposed in the London symposium of as most of them still unpublished. However, since a presented by Hours The list lacks vast territory separated the two above mentioned essential information such as definitions and des- areas, it was thought that the area of Kadesh Bar- criptions, but it has been adopted here because of nea, geographically located between them should several reasons, the first of which was to avoid the be studied.

Since the surroundings of Quseime al- confusion that could result from the use of several ready yielded sites attributed to the Upper Palaeoli- personal lists ibid.

The reason for this choice was that the tools were survey is carried out by O. Bar-Yosef of the He- numbered in groups, and thus the modification of a brew University and by l. Gilead of Ben-Gurion tool in one group would not effect the numbering University in the Negev. The list has been slightly modified for it seemed This ample evidence is supplemented by more that it would better fit the typological characteris- occurrences previously collected Nahal Lavan in tics of the Negev and Sinai assemblages.

I and new sites being discovered at nale behind it will be published elsewhere. Howe- present as part of a large scale salvage project ver, the main modification affected groups 0 and conducted in the Negev by the Israeli Department H which were reorganized into a group that inclu- of Antiquities. To the first were also added the El-Wad also provide evidence of palaeoenvironmental mo- points, designated here as H II and Aurignacian bla- difications Horowitz ; Tchernov ; Gold- des H 12 , The first five items of group H were berg , ; Shmida To sum up, the the various backed blades and the following five quantity and quality of Upper Palaeolithic evidence were the retouched blade varieties.

The main re- from the Negev and Sinai equal and even surpasse sults of these typological analyses are presented as that from northern Israel. A synthesis of this time cumUlative graphs. This is only one aspect of the span which would not take into account the Negev assemblages which will be supplemented by addi- and Sinai as well as the Northern Israel, would fail tional technological studies.

Results Most of the sites represented in the cumulative Typology of Assemblages graphs Fig. Nahal Zin , A. Generally two basic graph patterns can be discer- Many of the site reports use similar terms for the ned : diagonal and concave. The first Fig. The other groups of some importance are the the three major areas studied. The tion of the "endscraper-burin" assemblages is Aurignacian endscrapers are not common and their known from the site of Ein Aqev as well as from maximum frequency is ca.

The second, concave pattern of the cumulative One of the most interesting facts concerning the graphs is most common in Kadesh Barnea and archaeological contexts of the assemblages domina- Gebel Maghara Fig. This pattern means Nahal Zin are known only from in situ sites Boker that the most common tool types are the various A, BE. The same situation occurred in Kadesh retouched and backed blades and points group H Barnea ,9 , where the two richer sites were or the retouched and backed bladelets group J.

Among the blades, metres of silt-sand accumulation. The different graph patterns are not territorially The published assemblages dominated by bladelet restricted. More unpublished material demonstrates the wide distribution of simi- C.

Baruch, personal Several C" dates are now available and some communication and Central Sinai Kosloff The transitional of them are defined as "Lagaman" Bar-Yosef and assemblages are not part of this discussion which is Belfer , and in Kadesh Barnea. The blade concerned with the Upper Palaeolithic stricto sensu assemblages are not represented in the graphs from and the earliest date available therefore is that from Nahal Zin since new sites known from this locality Baker A which is older than 31, BC 0 7,?

Recently a The dates for the Lagaman cluster are between new site has been briefly described - Boker, area 29 , , BC Haas The assemblage A Marks b : b - dominated by blade from level 11 at Boker BE which is considered as a tools as well as by blade waste. Similar properties variant of Boker tradition Marks b is dated are assigned to assemblages from Baker BE ibid.

On the basis of the available Assemblages rich with bladelets were found in descriptions the sites from Kadesh Barnea are more Gebel Maghara Gisis and Gilead ; Gilead similar to assemblages from Baker A than to those Fig.

In Nahal from Boker BE. On typological grounds Kadesh Zin, the site of Ein Aqev East Ferring pro- Barnea should therefore be closer to 30,, duced a variant of a rich blade let assemblage than to 25 , BC.

A fourth Upper Palaeolithic assemblage At Nahal Zin three assemblages demonstrate that with numerous bladelets is reported also from the "endscraper-burin" group is later that the blade Wadi Sudr, Western Sinai U. Baruch, personal tool assemblages. Level I at Boker BE stratigraphi- communication. Lagama XV N I -. Area C of the same site it is also strati- provinces differing in such aspects as climate, graphically later than both Boker BE I and n. The landscape, vegetation and possible sources of sub- site of Ein Aqev has three dates clustering between sistance.

Although climatic fluctuations during the ca. Middle and Upper tially defined, but the following must be mentio- Palaeolithic faunal remains from the Negev are ned. In Gebel Maghara the sites of Lagama IIID indicative of an open landscape Tchernov ; ans X are partially in situ occurrences and are during the periods in which deposition occured, in therefore contemporaneous with the basal sand!

The esti- ver exeeded mm Goldberg , A re- mated date for the termination of this phase is cent study of the Lisan lake sediments Begin et al. Moreo- part. This could imply that in the South, palaeoeco- ver, high frequencies of bladelets characterize the logical circumstances probably required specific Lagaman assemblages and in Lagama V this is the prehistoric cultural adaptations, differing from most common group Fig. In Nahal Zin the Bladelet assemblages are repre- These two different provinces may have deman- sented by Ein Aqev East which is undated.

This ded different adaptive strategies, neither separated in -situ occurrence is broadly contemporaneous nor isolated one from the other. Passing from fer- with Ein Aqev on geological grounds Ferring tile to arid zones in Israel is a matter of a few hours , This argues against blade assemblages in the Negev and Sinai are treatment of the Negev and Sinai as independent considerably earlier than the "endscraper-burin" areas which exhibit separate cultural manifestations assemblages.

The latter were introduced at a time not found in the North. Department of Anthropology. Degree Name Master of Arts M. Physical Description 1 online resource 2, ix, pages. Abstract Fifteen bone tool assemblages from shell midden sites were compared. Rights In Copyright.

Comments If you are the rightful copyright holder of this dissertation or thesis and wish to have it removed from the Open Access Collection, please submit a request to pdxscholar pdx. Included in Archaeological Anthropology Commons. Find Enter search terms:. Digital Commons.