See discussions, stats, and author profiles for this publication
at:http://www.researchgate.net/publication/226864910The Emergence
of Ornamentsand Art: An ArchaeologicalPerspective on the Origins
ofBehavioral ModernityARTICLEinJOURNAL OF ARCHAEOLOGICAL RESEARCH
FEBRUARY 2007DOI:
10.1007/s10814-006-9008-1CITATIONS104DOWNLOADS80VIEWS2501
AUTHOR:Joo ZilhoUniversity of Barcelona112 PUBLICATIONS
1,992CITATIONS SEE PROFILEAvailable from: Joo ZilhoRetrieved on: 24
July 2015J Archaeol Res (2007) 15:154DOI
10.1007/s10814-006-9008-1ORI GI NALPAPERThe Emergence of Ornaments
and Art:An Archaeological Perspective on theOrigins of Behavioral
ModernityJo ao Zilh aoPublished online: 30 January 2007C Springer
Science+Business Media, LLC 2007Abstract Theearliest knownpersonal
ornamentscomefromtheMiddleStoneAgeofsouthernAfrica, c.
75,000yearsago, andareassociatedwithanatomicallymodernhu-mans. In
Europe, such items are not recorded until after 45,000 radiocarbon
years ago, inNeandertal-associated contexts that signicantly
predate the earliest evidence, archaeolog-ical or paleontological,
for the immigration of modern humans; thus, they represent
eitherindependentinventionoracquisitionoftheconceptbylong-distancediffusion,
implyingin both casescomparable levels of cognitivecapability and
performance. The emergenceof gurative art postdates c. 32,000
radiocarbon years ago, several millennia after the timeof
Neandertal/modern human contact. These temporal patterns suggest
that the emergenceof behavioral modernity was triggered by
demographic and social processes and is not aspecies-specic
phenomenon; a corollary of these conclusions is that the
corresponding ge-netic and cognitive basis must have been present
in the genus Homo before the evolutionarysplit between the
Neandertal and modern human lineages.Keywords Art . Modern humans .
Neandertals . OrnamentsIntroductionOver the last quarter century,
it has become clear that the ancestry of present-day
humanpopulations can be traced back to African people of the late
Middle Pleistocene. In this con-text, the long-lasting geographical
segregation between Neandertals and African modernsand the ultimate
replacement of the former by the latter have led many scholars to
acceptthe notion that the two taxa should be given species status.
This view has been challengedin recent years, especially by the
nding of early European modern human fossils bearingarchaic traits,
which suggests extensive admixture with Neandertals at the time of
contact(Trinkaus,2005).ThissuggestionisconsistentwithrecentgeneticstudiesofthenuclearJ.
Zilh ao (
)Department of Archaeology and Anthropology, University of
Bristol,43 Woodland Road, Bristol BS8 1UU, United Kingdome-mail:
[email protected] J Archaeol Res (2007)
15:154genome of living populations, which indicate that we carry
genetic material inherited fromEurasian (in particular, east Asian)
populations that had differentiated hundreds of thousandsof years
before the mid-Late Pleistocene out-of-Africa dispersal of early
modern humans(Templeton, 2002, 2005).Given that hybridization
between closely related species is well known among mammalsin
general and primates in particular, the evidence for admixture does
not necessarily imply,however, that signicant biological
differences, perhaps at the species level, did not
existbetweenNeandertalsandmodernhumans. Moreover, under
theparadigmaticviewthatspecies must differ in behavior as much as
in morphology (Henshilwood and Marean, 2003),that
evidencealsodoesnot sufcetoexcludethepossibilitythat signicant
behavioraldifferences, with attendant cognitive implications,
separated anatomically modern peoplefrom coeval archaic humans. In
fact, the notion that such a separation existed
underliesspeculations that certain features of complex human
culture that are undocumented in
thearchaeologicalrecordoftheMiddlePleistocenesuchasartorritualburialmusthaveemerged
as a by-product of the biological processes involved in the
speciation of the Africansapiens (Klein,1998, 2003; Mellars,2005;
StringerandGamble, 1993). The assumptionis that the absence of
those features reects the lack of the required cognitive
capabilitiesand that it is only after the acquisition of the latter
by the rst modern humans that thecorresponding behavioral
correlates could be externalized in archaeologically visible
ways.At the empirical level, this approach initially tended to date
such an acquisition to thetime of the transition from the Middle to
the Upper Paleolithic, the latter being dened as apackage of
cultural traits appearing rather suddenly and at about the time
when, in Europe,Neandertals were replaced by moderns. Among the
listed traits, aspects of subsistence, settle-ment, and lithic
technology used to feature prominently in different versions of the
denitionof the Upper Paleolithic (for instance, Mellars, 1973;
White, 1982). Recently, however, awide consensus seems to have been
achieved that logistically organized hunting, as well asthe
reliance on blade technology or the long-distance procurement of
raw materials, are tobe found at different times and places during
the Middle Paleolithic, and in
unquestionableassociationwitharchaichumans(Bar-Yosef,2004;Bar-YosefandKuhn,1999;Burke,2004;
Marean and Kim, 1998; R evillion and Tuffreau, 1994). On the other
hand, the rstevidenceforcarefullyshapedbonetools, shellornaments,
andabstractmarkingsisnowknown to come from the Middle Stone Age
(MSA) of southern Africa, not from the UpperPaleolithic of Europe
(Henshilwood et al., 2001, 2002, 2004).In this context, Henshilwood
and Marean (2003), following up on Wadley (2001), arguedfor a
modern human behavior different from that of the Neandertals, to
which they proposedthe designation of fully symbolic sapiens
behavior; in the archaeological record, it wouldmanifest itself
when artifacts or features carry a clear symbolic message that is
exosomaticforexample, personal ornaments, depictions, orevenatool
clearlymadetoidentifyitsmaker. In this review, I use Henshilwood
and Mareans denition to assess the distributionin space and time of
the earliest evidence for behavioral modernity (ornaments and
art),and the extent to which the human groups involved in the
production of such early evidencewere biologically modern or
archaic. In particular, I discuss the different
explanationsthathavebeenproposedforthefactthatbothornamentsandartareknownamongsuchunquestionably
anatomically nonmodern populations as the late Neandertals of
Europe.In the following, calendar dates derived from the oceanic or
ice-cap records or obtainedby thermoluminescence (TL),
electromagnetic spin resonance (ESR), and uraniumthorium(UTh)
methods are given in years or thousands of years (ka) BP, and
radiocarbon dates areexpressed in years or thousands of years
(ka)14C BP. The recognition that oscillations inthe production of
atmospheric14C at this time were not as dramatic as once thought
makesSpringerJ Archaeol Res (2007) 15:154
3preliminarycalibrationpossible, andit isnowwell establishedthat,
inthistimerange,radiocarbon underestimates true calendar ages by
three to ve millennia (Fairbanks et al.,2005; Hughen et al., 2004;
Shackleton et al., 2004; Weninger and J oris, 2005). Because
therelative ordering of the events is not affected, and to keep the
discussion of chronologicalissues within reasonable limits, only
uncalibrated ages are used here for the 3045 ka14CBP
interval.Temporal and geographical patternsAfricaAs shown by
different authors (Barham, 2002a, b; Henshilwood et al., 2001;
McBrearty andBrooks, 2000; Villa et al., 2005), many of the
innovations traditionally associated with theEuropean Upper
Paleolithic are now known to appear signicantly earlier in Africa.
This isthe case in particular with bone tools (such as the harpoons
from Katanda, Congo, and theawls from Blombos, South Africa), but
it also applies to such features of lithic technology asthe
manufacture of geometrics (the lunates of the South African
Howiesons Poort industry)and the production of bladelets from
prismatic cores (documented in level RSP of the Sibudurockshelter,
South Africa). Enough reliable dating evidence is now available to
place thesedevelopments before c. 50 ka BP and, in some cases, even
before c. 70 ka BP. However,these innovations did not form a
package of co-occurring traits and did not become a stablefeature
of human culture once they appeared. Instead, for many thousands of
years thereafter,they were abandoned as piecemeal and suddenly as
they were rst introduced, and the sameapplies to ornaments and
abstract
markings.Wherethelatterareconcerned,thekeyevidencecomesfromtheseasidecavesiteofBlombos,
southern Cape (dErrico et al., 2003a, 2005; Henshilwood et al.,
2002, 2004). Thissite features a sequence where the uppermost
MSAlevel (M1) belongs to the Still Bay culture,characterized by
foliate points, and is separated from the surcial Late Stone Age
(LSA)deposits by a thick sterile sand dune. This stratigraphic
conguration precludes contaminationfrom overlying, later
occupations as an explanation for the presence of personal
ornamentsand decorated pieces of ochre in level M1, dated to 74.9
3.8 ka BP by optically stimulatedluminescence (OSL), and to 74 5 ka
BP by TL (Tribolo et al., 2005). The number ofutilized pieces of
ochre is in excess of 8000, and two of them, in the shape of
crayons,
bearunequivocalabstractdesigns(engravedcross-hatchedmotifs)ononeofthefacets.LevelM1
also yielded personal ornaments, all perforated shells of the
marine mollusk
Nassariuskraussianus(Fig.1).Forty-onesuchitemshavebeendescribedsofar;allwerefoundinclusters
of 217 beads showing similar size, color, wear, and perforation
type, suggestingthat each cluster may correspond to a single
beadwork item.IntheSouthAfricanculture-stratigraphicscheme,
theStill Bayis replacedbytheHowiesonsPoort industry, whichTriboloet
al. (2005)
TL-datedto563kaBPatKlasiesRiverMouth(southernCape)andto5565kaBPatDiepkloof(westernCape).These
results are consistent with the AAR (amino acid racemization) and
ESR ages in thec. 6070 ka BP interval obtained for the
corresponding levels of the Border Cave
sequence,northernKwazulu-Natal, byMilleretal. (1999), Gr
unandBeaumont(2001), andGr unet al. (2003). The latter also discuss
(and reject) the possibility that the securely provenancedhuman
remains found in this cavethe near complete infant skeleton BC3,
and the largelycomplete lower jaw BC5could represent intrusions of
later Pleistocene or even Holoceneage. Indeed, direct ESRdating of
an enamel fragment fromBC5 yielded a result of 74 5 kaSpringer4 J
Archaeol Res (2007) 15:154Fig. 1 African personal ornaments: (a)
modern Nassarius kraussianus shell; (b) N. kraussianus shell
beadfrom MSA level M1 of Blombos (after Henshilwood et al., 2004,
modied); (c) ostrich eggshell bead from theMSA site of Loiyangalani
(after Hathaway, 2004, modied). Marine shells used as ornaments in
the IUP andthe Early Ahmarian of the Near East: (df) perforated
Nassarius gibbosula beads from layer H ofUca gizli(after Kuhn et
al., 2001, modied)BP, which is consistent with similar results for
faunal samples from the same levels. Thisevidence in turn
strengthens the hypothesis that the BC3 burialwhose grave pit is
reportedto have been entirely cut into the underlying MSAdeposits
and to have had its lip lying belowan ash horizon at the very base
of the Howiesons Poort levelsalso was in situ. Given
itsstratigraphic position and accompanying dating evidence, it is
thus quite possible that thisburial was broadly contemporary with
the Still Bay occupation of Blombos. A perforatedConus bairstowi
sea shell was reportedly associated with the BC3 skeleton and may
havebeen a bead worn by the dead infant, in which case Border Cave
would add a further ritualdimension to the use of personal
ornaments at this time.Forthenext30,000years, however,
nosimilarndsareknownineitherHowiesonsPoort or post-Howiesons Poort,
later MSA contexts. Secure evidence for ornaments turnsup again
only in eastern Africa, where the rockshelter of Enkapune ya Muto,
Kenya, yieldedostricheggshell beadsinanearlyLSAcontext,
withfragmentsfrombeadmanufacturedirectlydatedtoc.
3740ka14CBP(Ambrose, 1998).
McBreartyandBrookss(2000)reviewoftheAfricanevidencementionssimilarndsinBoomplaas,inassociationwithstatistically
identical (in the range of 42 ka14C BP) dates on charcoal, but in
an MSA notLSA context, as is also the case at the recently reported
but as yet undated Tanzanian site ofLoiyangalani (Hathaway, 2004).
An ostrich eggshell fragment (but no beads) was found in theburial
pit containing skeleton 1a from Nazlet Khater, in Upper Egypt,
dated on
associatedcharcoaltoc.38ka14CBP(Vermeersch,2002).Thesesitesarealllocatedfarfromthecoast,
which could explain the absence of marine shell beads in the
inventories. However, atleast where Boomplaas is concerned, the
distance in question (c. 80 km) is identical to thatwhich separates
Border Cave from the sea. The scant evidence available indicates
that onlyperforated marine shells were in use c. 75 ka BP, and only
ostrich eggshell beads were in usec. 40 ka BP; thus, changes
through time in mobility patterns, exchange systems, or
culturalpreferences also may have been involved.SpringerJ Archaeol
Res (2007) 15:154
5InasecureHowiesonsPoortcontextfromDiepkloof,Parkington et
al.(2005)foundabstract markings on small fragments of ostrich
eggshells thought to have been used as waterasks. They noted that,
although the fainter marks could result from use wear, the
deeperones were clearly intentional and, in a few cases, formed
compositions akin to the abstractdesigns made on the Blombos ochre
crayons. The patterns, however, are said to be moresuggestive of
intentional marking to denote ownership than of artistic
decoration.In Africa, the earliest gurative art is represented by
the much later painted slabs fromApollo 11 Cave in Namibia
(Vogelsang, 1998; Wendt, 1974, 1976; Fig. 2). As argued byWendt,
these hand-sized slabs are not exfoliated fragments of wall
paintings but mobiliaryart. Their diverse geological nature and
their shape are not consistent with the local bedrock,although
similar slabs can be found in nearby slopes. Moreover, in at least
two instances, therepresentations occupy the center of the slab,
implying a pre-existing frame, and, in somecases, traces of color
also could be observed on the face opposite that containing the
gures.On three of the slabs, such gures can be identied and portray
what seemto be a rhinoceros,a zebra, and a large animal, probably a
feline with human-like hind legs.The site features an approximately
2-m-thick MSA-to-LSA sequence, and the slabs
wererecoveredtowardtheupperpart ofthedeposits,
some50cmbelowthesurface, at theinterfacebetweenthelatest
MSAandtheearliest LSAlevel. Conventional
radiocarbonresultsforassociatedcharcoalsamplesdatetheseslabstoc.2628ka14CBP,withthePta-1040
result (26,300 400 BP)obtained on a single large piece of
carbonized woodrepresenting in all likelihood the best
approximation of their chronology. In any case, thestratigraphic
consistency of the series leaves no doubt that the slabs date to
between c. 18and c. 34 ka 14C BP [only sample Pta-1032 is
anomalous, probably due to the incorporationof younger material
brought down by rodents nesting in adjacent sediments (Wendt,
1974,p. 36)].McBrearty and Brooks (2000) remark that the dates are
anomalously young for an MSAcontext andarguethat theApollo11art
issignicantlyolder basedonthe59kaBPostrich eggshell AAR age
obtained by Miller et al. (1999) for the sites MSA deposits,
inagreement with a direct AMS radiocarbon date of>41 ka14CBP for
a single ostrich eggshellfragment. As Miller et al. caution,
however, this apparent discrepancy does not invalidate
theradiocarbon chronology, because the ostrich eggshell samples
they analyzed were collectedin deposits from the mouth of the cave,
where the MSA sequence may be abbreviated bycomparison to that
observed in the area further inside from where the slabs came.
Moreover,as Miller et al.s dating work also shows, individually
dated eggshell fragments moved upand down the sequence as a result
of intensive human occupation combined with very slowsedimentation
rates (2 cm/millennium); thus, they cannot be relied on as a tool
to date, byassociation, the different archaeological levels.
Finally, the radiocarbon results obtained forthe immediate context
of the painted slabs are not unexpectedly young; in the region,
theMSA lasts until c. 20 ka14C BP (Deacon and Deacon, 1999), and
the anomaly diagnosedby McBrearty and Brooks (2000) most likely
resides in their expectations, not in any realproblems with the
dating of the site.The only securely provenanced human remains
fromthis time range in southern Africa arethose recovered from the
SAS member of Klasies River Mouth, dated to c. 100 ka BP.
Theirtaxonomic afnities are controversial. As Trinkaus (2005) sums
up, the problem is that thedearth of comparable material precludes
adequate assessment of whether the Klasies RiverMouth remains are
modern or simply a southern African equivalent of late archaic
humans,antedating the dispersal into the region of the anatomically
modern populations that
haddifferentiatedineasternAfricaduringthelaterMiddlePleistocene.
ThemorecompleteBorder Cave material, however, compares well with
the present-day San (Rightmire, 1984).Springer6 J Archaeol Res
(2007) 15:154Fig. 2 Top: slab from the late MSA levels of Apollo 11
cave (Namibia), representing a predatory big cat.Bottom:
stratigraphic sequence in the 1972 extension to the main trench
excavated in 1969; the provenience ofthe radiocarbon samples
collected in this extension is indicated (dotted contours; note
that sample Pta-1040corresponds to a single, large piece of wood
charcoal), as is the exact location of the three painted slabs
foundin situ during its excavation (black lled contours); layer 3
=layer D of the main trench (Early LSA), layer4 =layer E of the
main trench (Latest MSA) (after Wendt, 1974, modied)If
BC3andBC5areindeedinsitunds,
thenpeoplewhowerefullymodernintheiranatomy had evolved in (or
dispersed into) southern Africa by c. 75 ka BP, and the
personalornaments and abstract designs from Blombos level M1 are
indeed representative of theirbehavior.SpringerJ Archaeol Res
(2007) 15:154 7AsiaThe Near East before c. 50 ka BPAlthough
conceivable, the notion that two shell beads from the cave of Skuhl
in northernIsrael (Vanhaeren et al., 2006) are of oxygen isotope
stage (OIS) 5 age is controversial and,as discussed in the
following section, at present is not the most parsimonious reading
of theevidence. McBrearty and Brooks (2000) also mention the
presence of perforated shells inassociation with Homo sapiens in
nearby Qafzeh Cave, c. 100 ka BP. However, as recentlyshown by
Taborin (2003), the perforations in these items (Glycymerys shells)
are natural,and they were used as recipients for ochre, not as
ornaments. Processing of ochre at the siteis particularly important
in level XVII, which contained ve intentional burials. On the
basisof this context, Hovers et al. (2003) argue that the Qafzeh
ochre reects color symbolism,but use wear analyses of broadly
contemporary South African MSAmaterial showthat ochrecould have
served more practical functions (e.g., in the tanning of hides and
the productionof hafting pastes), and that even when the abandoned
pieces have a crayon shape, symbolism(for instance, related to body
painting) is not necessarily involved (Wadley, 2005; Wadleyet al.,
2004).Theoccupationof theNear East byearlymodernhumansat that
timeispart of anortheastern extension of African environments and
ceases with the return of cold conditionsduring OIS-4, after c. 75
ka BP. Human remains dated to OIS-4 and to the earlier part ofOIS-3
(after c. 59 ka BP) come from the sites of Amud and Kebara in
Israel and Dederyiehin Syria, and all are of Neandertals; the
youngest in chronology is the nearly complete adultskeleton buried
in level B1 of Amud. The level is dated by TL and coupled ESR/U-Th
toc. 53 ka BP(Kaufman, 2002; Rink et al., 2001; Valladas et
al.,1999), which provides aterminus post quem for the burial itself
and, hence, for the replacement of Neandertals bymodern humans in
the region.Fromthe point of viewof lithic technology and
adaptation, the cultural remains associatedwith OIS-5 moderns and
OIS-4 Neandertals in the Near East are virtually
indistinguishable(Shea, 2003). Where symbolic artifacts are
concerned, if unambiguous evidence for personalornaments is
lacking, the regional evidence for art and abstract design before
the UpperPaleolithic is equivocal at best: a gurine from Berekhat
Ram, an Acheulian open air siteintheGolanHeights,
andtwoengravedcorticalfacesofintartifactsfromQafzehandQuneitra,
another Golan Heights open air site of late Middle Paleolithic
age.TheBerekhat Ramgurineisa3.5-cm-longpieceofbasalt
whoseshapeevokesthefemale body and is vaguely reminiscent of the
well-known Venus gurines of the Gravettian(Soffer et al., 2000). A
recent study by dErrico and Nowell (2000) conrms some level
ofdeliberate human modication (abrasion and grooving) but does not
reject the hypothesisthat it served mere utilitarian purposes. In
any case, the object dates to >200 ka BP and,therefore, if
symbolic, it relates to archaic not modern people (the same applies
to thenatural pebble from the Middle Acheulian site of Tan Tan,
Morocco, described as a gurineby Bednarik, 2003a). The Qafzeh piece
is part of the c. 100 ka BP context of the sites earlyHomo sapiens
burials and consists of a broken Levallois core, 6.2 cm long, that
bears a setof incised lines on its cortical face. The analysis of
these lines by dErrico et al. (2003a)concludedthat theycouldnot
beasimpleby-product oftasksperformedonthatsurfacewith cutting tools
(such as butchering), but it produced no evidence that they
represent adeliberate composition or part of some abstract design.
The same applies to the Quneitraobject, a tabular piece of int
cortex of broadly the same size and aspect and bearing
fourconcentric semicircles surrounded by vertical lines (Marshack,
1996); the site is ESR-datedSpringer8 J Archaeol Res (2007)
15:154to 4055 ka BP, which means that, depictive or not, this image
could relate to either thelatest Neandertal or the earliest modern
OIS-3 populations of the region.The Near East after c. 50 ka BPFrom
level 1, at the bottom, to level 4, at the top, the Israeli open
air site of Boker Tachtit inthe Negev desert (Marks, 1983; Marks
and Ferring, 1988) provides a detailed picture of theregional
technological transition fromthe Middle to the Upper Paleolithic.
Level 4 is identicalto levels XXIXXV of the long sequence at the
Lebanese rockshelter of Ksar Akil, the otherkey site for the
transition in the Levant (Bergman and Stringer, 1989; Marks and
Ferring,1988). Typologically, these assemblages are characterized
by the Emireh pointelongated,triangular, morphologically Levallois
items that, in the southern Levant, often bear ventral,thinning
retouch of the base. In the northern Levant, the so-called
chamfered pieces also areindex fossils of this assemblage type. For
lack of a better term, these occurrences at
presentareconsideredpartofasingleNearEasterntransitionaltechnocomplexdesignatedasInitial
Upper Paleolithic (IUP) (Bar-Yosef, 2000; Kuhn, 2002, 2003).Two
conventional charcoal dates for basal level 1 of Boker Tachtit
place it at approximately47 ka14CBP, in spite of their large
standard deviations, but no precise chronology is availablefor
uppermost level 4. The latter must in any case date to>35
ka14CBPgiven the radiocarbonresult of 35,055 410014C BP (SMU-579),
obtained on a charcoal sample from whichhumates could not be
extracted, which means the result is probably a minimum age
only(Marks, 1983) (Table 1). No dates are available for levels
XXIXXV of Ksar Akil, but thecontemporaneity with the Negev site
suggested by the lithics is consistent with a conventionalresult of
c. 44 ka14C BP obtained for the immediately underlying Middle
Paleolithic levelXXVI (Bergman and Stringer, 1989).The southern
Turkish cave site ofUca gizli (Kuhn, 2002, 2003; Kuhn et al., 2001)
providesa better x on the chronology of the Near Eastern IUP.
Levels G and H, with a lithic
industryidenticaltothatinKsarAkillevelXXI,
yieldedaconsistentseriesofacceleratedmassspectrometry(AMS)radiocarbonresultsoncharcoal,placingtheirdepositioninthe3641
ka 14C BP interval. AtUca gizli, as elsewhere in other stratied
Near Eastern occurrences,the IUP is followed by the Early Ahmarian,
a fully Upper Paleolithic technocomplex. AtKsar Akil,
suchEarlyAhmarianassemblagesarefoundinlevelsXVIXX,
whicharestratigraphically and technologically very close to the
preceding IUP (Kuhn, 2003). Theseindications of continuity are
further strengthened by the resemblance between the industryfrom
uppermost level 4 of Boker Tachtit and that contained in the nearby
single-level site ofBoker A, which is clearly of Early Ahmarian
afnities (Jones et al., 1983; Monigal, 2003).Two conventional
charcoal results of>33.5 ka14CBPare available for Boker A, which
agreewith the single nite date of 37,920 2810 14C BP (SMU-578),
also on charcoal. Given thelarge standard deviation of the latter,
the three results are consistent with the stratigraphicevidence
that places the Early Ahmarian after the IUP and, hence, with a
radiocarbon ageapproximately in the 3635 ka14CBPrange or younger.
Technologically, the Early Ahmarianfeatures a single platform,
soft-hammer production of blades and bladelets extracted
fromprismatic cores in the framework of a continuous reduction
system, and typologically it ischaracterized by the so-called
El-Wad points, which are made on long, slender bladelets orsmall
blades and laterally bear direct inverse or alternate retouch
extending along at leastone of the blanks edges.At
KebaraCaveinnorthernIsrael,
theEarlyAhmarianlevels(UnitsIIIIV)yieldedsomewhat
olderdates(asearlyasc. 43ka14CBP)(Bar-Yosef, 2000;
Bar-Yosefetal.,1996). However, the results are widely scattered,
and only one (of 35,600 160014C BP,SpringerJ Archaeol Res (2007)
15:154
9Table1RadiocarbondatesfortheMiddletoUpperPaleolithictransitionintheNearEastSiteProvenienceMaterialMethodLabno.ResultBPCultureUcagizliFcharcoalAMSAA3526034000690IUPFcharcoalAMSAA3762435020740IUPGcharcoalAMSAA37626391001500IUP(=KsarAkilXXI)HcharcoalAMSAA37623330401400IUP(=KsarAkilXXI)HcharcoalAMSAA3526135670730IUP(=KsarAkilXXI)HcharcoalAMSAA27995389001100IUP(=KsarAkilXXI)HcharcoalAMSAA27994394001200IUP(=KsarAkilXXI)HcharcoalAMSAA37625414001100IUP(=KsarAkilXXI)KsarAkilIV[layer9a]charcoalAMSOxA-180330250850AurignacianVI[layer10lower]charcoalAMSOxA-1804312001300AurignacianVI[layer11bm]charcoalAMSOxA-1805324001100AurignacianVIcharcoalconventionalMC-1192320001500AurignacianVIIVIII(67m)shellconventionalGrN-219528840380AurignacianXXVIdarkclay(charcoal?)conventionalGrN-2579437501500MousterianKebaraIcharcoalconventionalPta-426832200630AurignacianIsubsurfacecharcoalconventionalPta-424722900250AurignacianIbasecharcoalAMSOxA-397434510740AurignacianIItopcharcoalAMSOxA-397533920690AurignacianII,inburrowcharcoalconventionalPta-426331400480AurignacianII,inburrowcharcoalconventionalPta-426928700450AurignacianIIf(Q16d,4.70m,hearth)charcoalAMSOxA-1230360001600AurignacianIIfabovehearthcharcoalAMSGif-TAN-9015132670800AurignacianIIfhearthcharcoalAMSGif-TAN-90028343001100AurignacianIIfhearthcharcoalconventionalGx-17276428004800AurignacianIIIBcharcoalAMSOxA-3976435002200EarlyAhmarianIIIBcharcoalconventionalPta-4267361001100EarlyAhmarianIIIBfcharcoalAMSOxA-3977>43800EarlyAhmarianIIIBfcharcoalAMSGif-TAN-90037>42500EarlyAhmarianSpringer10
J Archaeol Res (2007)
15:154Table1ContinuedSiteProvenienceMaterialMethodLabno.ResultBPCultureIIIBfcharcoalAMSGif-TAN-90168>41700EarlyAhmarianIIIBf(Q16d,5.38m,hearth)charcoalAMSOxA-1567356001600EarlyAhmarianIVBcharcoalconventionalPta-5002425001800EarlyAhmarianIVBcharcoalconventionalPta-4987421002100EarlyAhmarianIVB(adjacenttoburrow)charcoalAMSOxA-397828890400EarlyAhmarianIVVinQ16b/Q15dcharcoalconventionalPta-5141437001800Mousterian/EarlyAhmarianinterfaceVcharcoalAMSOxA-3979>44000MousterianVcharcoalAMSOxA-3980>44800MousterianVwcharcoalAMSGif-TAN-90030>46900MousterianV(Q16a/b,6.17m)charcoalAMSOxA-1568380002100MousterianBokerTachtit4charcoal(nohumatesremoved)conventionalSMU-579>350554100EarlyAhmarian1charcoalconventionalGY-3642>34950IUP(Emiran)1charcoalconventionalSMU-184>45570IUP(Emiran)1charcoalconventionalSMU-259469302400IUP(Emiran)1charcoalconventionalSMU-580472809050IUP(Emiran)BokerA1charcoalconventionalSMU-260>33420EarlyAhmarian1charcoalconventionalSMU-187>33600EarlyAhmarian1charcoalconventionalSMU-578379202810EarlyAhmarianUmmelTlelII2bcharcoalAMSGifA-9321232000580AurignacianXII(=II4?)charcoalconventionalGif-9004030790760Aurignacian+AhmarianIII2acharcoalAMSGifA-9321634530750IUPSpringerJ
Archaeol Res (2007) 15:154 11OxA-1567; Hedges et al., 1990) was
obtained on a hearth sample (in Unit IIIBf); signicantly,it agrees
well with the chronometric and stratigraphic evidence fromUca
gizli, Boker A, andKsar Akil. The excavators report problems with
the integrity of the charcoal lenses fromwheretheKebarasamplescame,
andthedatedmaterial maywell includeasignicantcomponent derived from
underlying Middle Paleolithic Unit V. This hypothesis is
consistentwith the fact that in the western part of the south prole
the Upper Paleolithic levels ll a1.5-m-wide erosional channel cut
into the Mousterian deposits; such a major unconformitymay also
explain why no IUP contexts were recognized at
Kebara.TheavailablechronostratigraphicevidencethereforeplacestheIUPoftheNearEastapproximately
in the 3644-ka14C BP interval and the Early Ahmarian in the
subsequenttwomillennia, c. 3635ka14CBP.
Itwasatsometimeduringthesetenmillenniathatpersonal ornamentation,
abundantlydocumentedinthecorrespondinglevelsof
thekeysitesofKsarAkilandUca gizli(Fig. 1),
rstappearedintheregion;theearliestactualevidence is that from level
H ofUca gizli, for which available dates average c. 39 ka 14C
BP.According to Kuhn et al. (2001), all such items, in both sites
and in both the IUP and theEarly Ahmarian, are perforated marine
shells, mostly from only three speciesNassarius(
=Arcularia)gibbosula,Columbellarustica, andGlycymerissp.,
althoughthelatter, asdiscussed for Qafzeh, are more likely to
represent containers rather than actual ornaments.Excluding
themfromthe counts, 194 beads were recovered in IUP levels
XXIXXIVof KsarAkil, 75% N. gibbosula and 11% C. rustica; the
corresponding gures for Early Ahmarianlevels XIVXVIII are 364, 53%,
and 36%, respectively. The published count for the IUPlevels ofUca
gizli is 108, but the total is now several hundred, 90% of which
belong to asingle species, N. gibbosula (Kuhn, personal
communication,
2005).TheonlyevidenceconcerningtheauthorshipoftheIUPandtheEarlyAhmarianisEgbert,
a juvenile modern human skeleton uncovered in 1938 at Ksar Akil, at
a depth of11.46 m below datum; this elevation indicates that the
bones pertain to the Early Ahmarianstrata between level XVI and the
base of level XVIII (Bergman and Stringer, 1989). Theskeleton is
now lost (only a cast of the skull is preserved in the Natural
History Museum ofLondon), so direct dating is impossible, and the
hypothesis that this was an intrusive burialfrom overlying
occupations cannot be tested. As pointed out by Mellars (2004),
however,the thickness of the deposits (the bones appear to have
come from more than 1 m below thesurface of the uppermost
unquestionable Early Ahmarian deposits, level XVI) argues
againstthat possibility. No counterparts of the Ksar Akil human
remains exist for the IUP, but itis not unreasonable to assume, on
the basis of the apparent continuity in lithic technologybetween
the latest IUP and the Early Ahmarian, that the people who
manufactured the latteralso made the former. However, it cannot be
excluded that Neandertals also were involved;the lesson from the
Near Eastern record of OIS-5 and OIS-4 is that no necessary
correlationexists between archaeological culture and physical
types, and this caveat must hold as wellwhen interpreting the
evidence from early OIS-3.Vanhaeren et al. (2006) argue that the
two perforated N. gibbosula found at Skhul are fromlayer B (which
contained the remains of ten anatomically modern humans) and,
therefore,thattheirageshouldbeintherangeof100135kaBP.Theyfurtherarguethatanothersuch
bead from Oued Djebbana (Algeria), the type site of the North
African Aterian, is ofsimilar age. If their arguments are correct,
personal ornamentation emerged at least 25,000years earlier than
suggested by the evidence from Blombos. As they acknowledge,
however,the chronology of the Aterian and Skhuls layer B is
controversial. The Aterian is currentlyestimated to fall in the
3590 ka BP range (Deb enath, 2000; Wrinn and Rink, 2003),
whereasthe U-Th chronology (Gr un et al., 2005) and the morphology
of the skeletons (Stringer, 1998)indicate that two periods are
represented in layer B of Skhul, which yielded several
datesSpringer12 J Archaeol Res (2007) 15:154in the 3050 ka BP
interval. Because an overlap with the chronology of the IUP is
clear inboth cases, and because the IUP features large amounts of
the bead type in question, it isquite possible, and at least cannot
be excluded at present, that the perforated N.
gibbosulafromSkhulandOuedDjebanna,insteadofbeingoftheproposedOIS-5age,areinfactcontemporary
with those fromUca gizli and Ksar Akil.Russia and central
AsiaIUP-like assemblages are known in the Alta and other parts of
central Asia in associationwith dates as early as c. 43 ka14C BP.
Given the arguments in favor of an association ofthe Near Eastern
IUP with modern humans, it is conceivable that such occurrences
representafurtherrangeextensionofthelatterintomorenorthernlatitudes,buttheissueremainscontroversial
(KrivoshapkinandBrantingham, 2004; Rybin, 2004).
Becausethedirectlydated human material (mandible and postcrania)
from Tianyuandong (near Beijing, China)documents people with a
modern anatomy in the Far East c. 35 ka 14C BP (Trinkaus,
2005),inbroadcontemporaneitywithKsarAkilsEgbert,itmakessensetoassumethattheintervening
regions of central Asia and the Alta also were settled by modern
humans at thattime. Conversely, if Neandertals still inhabited the
Near East c. 50 ka BP, as suggested bythe Amud data, any spread of
modern humans into central Asia via a Near Eastern route canhave
occurred only at a later date. In sum, the replacement process must
have taken place incentral Asia somewhere between c. 50 and c. 35
ka BP but, as in the Near East, constrainingit with greater
precision is impossible at present.Inanycase,
onecancertainlyexpectmodernhumangroupsdispersingoutofAfricatohavecarriedwiththemthesocialorganizationandcorrespondingsociofactsthattheirancestors
had developed. A rather convincing indication that an inux of
ultimate Africanorigin is involved in the East Asian process is
provided by the presence of ostrich eggshellbeads in the Mongolian
site of D or olj 1 (Jaubert et al., 2004), dated to c. 32 ka14C
BP.Aclearconnectionwithcultural developmentsintheNearEast
alsoisapparent afewmillenniaearlierinsiteswest oftheUrals.
Forinstance,
aperforatedColumbellashell,modernrepresentativesofwhichareconnedtotheMediterraneanbasin,wasrecoveredinculturallayerIVb(welldatedbyAMSoncharcoalsamplestoc.36.5ka14CBP)ofKostenki
14 (Markina Gora), now situated more than 700 km from the shores of
the BlackSea (Sinitsyn, 2003, 2004). Although the technological and
typological features of the
lithicassemblagerecoveredthereinareofafullUpperPaleolithicnature,itsculturalafnitiesremain
unclear, and an isolated tooth is reportedly of modern human
afnities. Sinitsyn alsodescribes an apparently shaped piece of
mammoth ivory recovered in this level as the headof a female
gurine; he acknowledges, however, that the surface is covered with
tracesof natural damage and that the object is an obviously
unnished product broken duringmanufacture. Thus, as with the
Berekhat Ram gurine, the art may well be in the eye ofthe
beholder.At an even earlier date, bone tools and ornaments are
reported by Derevianko and Rybin(2003) from IUP-like contexts in
Denisova cave (layer 11) and Kara-Bom (Horizon 5), buttheactual
anatomical
afnitiesofthemanufacturersoftheseassemblagesareunknown,and the
ornaments (animal tooth pendants and bone beads) are not of the
kind seen in
theNearEastatthattime(whenonlymarineshellbeadswereinuse). Moreover,
theexactstratigraphicprovenienceofthendsisnot devoidofambiguity.
Amajordiscontinuityseparates OIS-3 layer 11 of Denisova from the
immediately overlying OIS-2 level 9, andthe contact between the two
is signicantly disturbed. Because the range of ornaments fromlevel
11 is identical to that found in both level 9 and the pockets
containing level 9 lithics thatSpringerJ Archaeol Res (2007) 15:154
13penetrated deeply into level 11 (Derevianko and Shunkov, 2003,
Fig. 7), their associationwith the IUP is questionable. At
Kara-Bom, the material (namely, one perforated bovid toothand a
pear-shaped bone bead) was found in a small depression that
contained signicantamounts of goethite pigment; this feature was
located 1 m away from a hearth excavated in1987 by Okladnikov in
his Stratum 3, now correlated with the lower part of
lithologicallevel 6, which contains Occupation Horizons 6 and 5.
These occupations are both AMSdated on charcoal to c. 43 ka14C BP,
but no more than approximately 30 cm above
andinthesamelithologicalunit isthesignicantlyyounger
OccupationHorizon4(c.34 ka14C BP). The excavation plan (Derevianko
and Rybin, 2003, Fig. 8) makes it clear that
thedepressionwiththepigmentandtheornamentswasbeyondtheboundariesofthelithicscatterassociatedwiththehearth,andthenatureofthendsisstronglysuggestiveofacache.
Stratigraphically, this cache was excavated into the hearth level
and, therefore, thetwo are not necessarily coeval; all that can be
securely said is that the dates for Horizons 6and 5 provide a
terminus post quem, and those for Horizon 4 a terminus ante
quem.EuropeSymbolism in the Lower and Middle Paleolithic?As inthe
Near Eastern, Russian, andcentral Asianregions reviewedabove, the
evi-dence for symbolic artifacts before the Upper Paleolithic in
European regions west of theRussian/Ukrainianplainsalsoisambiguous.
WheretheLowerPaleolithicisconcerned,claims have been made that a
small ensemble of animal bone remains fromthe open air site
ofBilzingsleben (Germany), dated to>300 ka BP, are marked with
motifs that carry a symbolicmeaning (Bednarik, 2003b; Mania and
Mania, 1988; Meller, 2003). The markingsgroupsof ne strokes whose
broadly parallel disposition indicates that they are unlikely to
derivefrom ordinary utilitarian activities such as butchering or
cuttingare clearly anthropic; thebest piece, a percussion tool
manufactured from a spall of elephant tibia, bears two groups
ofmarks, one with 7 strokes and another with 14, forming a
suggestive rhythmical arrangement.However, unlike the ochre pieces
from Blombos, it is not evident that these markings weremade to
obtain a predesigned graphic composition with a specic even if
elusive meaning.Where the Middle Paleolithic is concerned, two
important objects come from the Hungar-ian open air site of Tata,
dated to>70 ka BP (Moncel, 2003). One is a silicied
nummulitecrossed at right angles by engraved lines on both sides,
forming + motifs fully inscribedin the objects circular outline
(Bednarik, 2003b). The other is an ivory plaque carefully
sep-arated from a mammoth molar, shaped, beveled, and rubbed with
red ochre. The edge-wearpolish indicates long-term use, and the
overall shape evokes the sacred churinga (stonesor wooden boards
associated with the wanderings of mythological ancestors) of
AustralianAborigines (Marshack, 1976, 1989). It is not obvious,
however, that the engraving on thenummulite is decorative, and a
utilitarian explanation for the churinga (bone tool usedin the
framework of ochre-processing tasks?) cannot be excluded either.
Representationalstatus has been claimed for a int nodule featuring
a natural tubular perforation into which abone splinter is wedged
(Marquet and Lorblanchet, 2003); this Neandertal face,
however,ismostlikelyanunmodied
pierre-gure,andnaturalprocesscannotberuledoutasanexplanation for
the wedged bone (Pettitt, 2003).Clear evidence for complex abstract
thinking involving graphic modication of objectsin connection with
ritual activities comes from the Mousterian graveyard of La
Ferrassie inFrance (Deeur, 1993; Peyrony, 1934) (Fig. 3). Seven
individuals (one fetus, two infants, twochildren, and two adults)
were buried in the Ferrassie Mousterian levels of this
rockshelter;Springer14 J Archaeol Res (2007) 15:154Fig. 3
LaFerrassie:Above:Planandproleoftheburial ofindividual 6,
a35-year-oldchild; belowleft, detail of the lower face of the stone
slab that covered the burial pit, decorated with cupules
virtuallyidentical to those found in blocks scattered in the
habitation levels of the Evolved Aurignacian at the top ofthe
stratigraphic sequence (see Fig. 10). Below right: Engraved bone
found with the adult skeleton in burial 1(after Peyrony, 1934,
modied)available dating evidence from southwestern France as a
whole suggests that occurrencesofthisassemblagetypealldatetothec.
6570kaBPinterval(Mellars, 1996). TheLaFerrassie 1 individual, an
adult male, was buried in a shallow pit together with a
cylindricalbone fragment decorated with four sets of parallel
incisions; the La Ferrassie 6 individual, a35-year-old child, had
three int tools (a point and two very large sidescrapers)
carefullySpringerJ Archaeol Res (2007) 15:154 15placed on top of
his dead body, which had been interred in a deep pit covered by a
limestoneslab whose inferior face was decorated with cupules.La
Ferrassie thus sufces to establish a level of symbolic expression
among EuropeanNeandertals at least identical to that seen in the
African lineage at the same time; however,thefact that
nocounterpartsoftheBlombosbeadshaveeverbeenfoundintheMiddlePaleolithic
of Europe is a major difference between the two continents, and one
that is
allthemoresignicantbecauseofEuropescomparativelymuchlongerandmoreintensiveresearchhistory.
Moreover, theirabsencefromthehundredsofMiddlePaleolithiccaveand
rockshelter sites with favorable preservation settings excavated in
Europe over the last150 years precludes taphonomic explanations;
thus, it is legitimate to conclude that, in thiscase, the absence
of evidence should indeed be considered as evidence of
absence.Upper Paleolithic culture-stratigraphic frameworkThe
earliest Upper Paleolithic of Europe corresponds to a diverse array
of cultural entitiesfeaturing lithic technologies that, in one way
or the other, t at least some aspects of thetechnological denition
of the period and are often collectively designated as
transitional.In the Franco-Cantabrian region there is the
well-known Ch atelperronian, where blade pro-duction is oriented
toward the production of blanks for curve-backed Ch atelperron
points andknives. In Italy and Greece, there is the Uluzzian, a
ake-based industry that also featuressome production of
non-Levallois blade blanks but is mostly dened by the manufacture
ofstandardized backed microlithsthick arched pieces, truncations,
lunates, and some trapeze,all trimmed with sur enclume retouch. In
Bulgaria there is the Bachokirian, where the Up-per Paleolithic
cachet is mostly due to the preponderance of endscrapers that are
made
onLevalloisbladeblanks.InMoraviaandsouthernPolandthereistheBohunician,charac-terizedbytheproductionofmorphologicallyLevalloispointsobtainedbynon-Levalloismethods.
Finally, in different parts of central and northern Europe, from
southern EnglandtoPoland, thereistheSzeletian(anditsGermancousin,
theAltm uhlian), characterizedbytheproductionofblattspitzen,
whicharecarefullyaked, thin, fullybifacial foliatepoints,
plano-convex or, more typically, biconvex in cross section; these
foliate point com-plexes come after the Bohunician and then evolve
to such unifacial blade point industriesas the so-called Lincombian
of England and the Jerzmanovician of eastern Germany andPoland.In
the wake of the extensive taphonomic critique of the evidence by
dErrico et al. (1998),Zilh aoanddErrico(1999, 2003a, b),
Rigaud(2001), Bordes(2002, 2003), Teyssandier(2003), and others,
suggestions of a long-term contemporaneity between these earliest
tran-sitionalUpperPaleolithicentitiesofEuropeandtheAurignacian,basedonradiocarbondates
and on patterns of putative interstratication (Bernaldo de Quir os,
1982; Bordes andLabrot, 1967; Champagne and Espitali e, 1981;
Gravina et al., 2005), have now been largelyabandoned(Zilh aoet
al., 2006). Inparticular, themost vocal proponent of that
notionhashimselfrecentlyconceded(Mellars, 2006)all
themajorpointsmadebyZilh aoanddErrico (1999, 2003a, b) on the
issues of interpretation raised by the application of
radio-carbontothistimerange.
Oncethenumeroussourcesoferrorareadequatelyltered, aclear picture
emerges (Zilh ao, 2006a, b, c). (1) The transitional
technocomplexes eitherunderlieorpredatetheearliest
occurrencesoftheAurignaciananywhereinEurope. (2)The development of
these technocomplexes took place in the interval between c. 45 and
c.35 ka14C BP, whereas the earliest Aurignacian dates to no more
than c. 36.5 ka14C BP(Table2). (3)Theslight
chronometricoverlapisaninevitableconsequenceofthepoorprecisionofdatingtechniquesandofthefactthattheCh
atelperronianisalmostentirelySpringer16 J Archaeol Res (2007)
15:154Table2RadiocarbondatesfortheChatelperronian(AMSonly)andtheProtoaurignacian(AMSonbone,AMSandconventionaloncharcoal)aSiteLevelMaterialMethodLabnumberResultBPCultureKlissoura1V,hearth42organicresidueofhearthconventionalGd-10714>31100UluzzianV,hearth53organicresidueofhearthconventionalGd-10715>30800UluzzianV,hearth42burntboneAMSGifA-9916840010740UluzzianAbriDubalen(Brassempouy)EBC2boneAMSGifA-10104536130690ChatelperronianChatelperronB5boneAMSOxA-1362239150600ChatelperronianB5boneAMSOxA-1432039240380ChatelperronianB5boneAMSOxA-1362140650600ChatelperronianGrotteduRenneIXboneAMSOxA-3465451002800ChatelperronianXboneAMSOxA-346433820720ChatelperronianXaY11bone(mammoth)AMSOxA-8450/Ly-89325820280ChatelperronianXb1Y10bone(horse)AMSOxA-8451/Ly-894383001300ChatelperronianXb1Y10bone(reindeer)AMSOxA-9122/Ly-105533400600ChatelperronianXb2Y11bone(mammoth)AMSOxA-8452/Ly-89534450750ChatelperronianXcY11bone(mammoth)AMSOxA-8453/Ly-89633400600ChatelperronianCaunedeBelvis[7]boneAMSAA-7390354251140ChatelperronianCombeSauni`ereXboneAMSOxA-6503(tripeptide)381001000ChatelperronianGrotteXVIBboneAMSGifA-95581350001200ChatelperronianBboneAMSAA-2997381001670ChatelperronianBboneAMSAA-2674>39800ChatelperronianLaQuina,aval4boneAMSOxA-10261/Ly-136735950450ChatelperronianRoc-de-Combesq.K9,level8boneAMSGif-10126439540970Chatelperroniansq.K9,level8boneAMSGif-101266400001300Chatelperroniansq.K9,level8boneAMSGif-101265451002100ChatelperronianRoche-au-Loup5[b]charcoalconventionalGif-2414>40000ChatelperronianSpringerJ
Archaeol Res (2007) 15:154
17Table2ContinuedSiteLevelMaterialMethodLabnumberResultBPCultureKrems-HundsteigbrownlayerwithhearthscharcoalconventionalKN-654355002000ProtoaurignacianGrottadiFumaneA2,nearhearthS14charcoalAMSUtC-204836500600ProtoaurignacianA2base,hearthS14charcoalAMSUtC-268836800/+1200/1400ProtoaurignacianA2base,hearthS14charcoalAMSOxA-805234120460ProtoaurignacianA2base,hearthS14charcoalAMSUtC-268935400/+1100/1300ProtoaurignacianA2base,hearthS14charcoalAMSUtC-269034200900ProtoaurignacianA2base,hearthS14charcoalAMSOxA-805333640440ProtoaurignacianRiparoMochieasttrench1959;G,Cut5657charcoalAMSOxA-359034680760Protoaurignacianeasttrench1959;G,Cut59charcoalAMSOxA-359135700850Protoaurignacianeasttrench1959;G,Cut60charcoalAMSOxA-359234870800ProtoaurignacianEsquicho-GrapaouSLC1bcharcoalconventionalMC-2161345402000ProtoaurignacianIsturitzU27,level4dburntboneAMSGifA-9823236510610ProtoaurignacianV126,level4dburntboneAMSGifA-9823334630560ProtoaurignacianMorin8charcoalAMSGifA-9626336590770ProtoaurignacianaResultsforlArbredaarenotincludedbecauseofuncertaintyregardingthestratigraphic/artifactualassociationsofthedatedsamplesand,whereFumaneisconcerned,onlysamplesfromlevelsA1A3andcollectedinsidethedriplineareincluded(foradiscussion,seeZilhaoanddErrico,1999,2003a).Springer18
J Archaeol Res (2007) 15:154dated on samples of bone that were not
pretreated with the recently developed
ultraltra-tiontechnique(BronkRamsey et
al.,2004)andthusareunderestimated(thekeysiteofthe Grotte du Renne
is a particular case in point). (4) At the continental scale, it
remainspossible that the Jerzmanovician/Lincombian may have emerged
or survived in the northernEuropean plains at the time of the
earliest Aurignacian settlement of European regions to thesouth,
and it is certain (contra J oris et al., 2003) that the Middle
Paleolithic continued untilmuch later in Iberian regions south of
the Cantabro-Pyrenean mountain range (Zilh ao, 1993,2000,
2006a).Recent technological studies in France (Bon, 2002; Bordes,
2002; Chiotti, 1999; Lucas,2000)havealsoconrmedtraditional
typology-basedviewsofAurignaciansystematics.Moreover, the evidence
now clearly shows that the so-called Protoaurignacian,
originallydened by G. Laplace and Italian authors (Palma di
Cesnola, 1993) and generally
consideredtobeacultural/geographicMediterraneanfaciesoftheclassicalAurignacian(Bon,2002),
correspondsinsteadtoachronological phase.Infact,
resultsfromtherecentre-excavationofthekeycavesiteofIsturitz(NormandandTurq,2005),ingoodaccordwiththerevisedstratigraphyofLePiage(Bordes,2002),suggestthatinFrance,aswellas
in Italy and Spain, this Protoaurignacian stratigraphically and
chronometrically precedestheclassical EarlyAurignacianor
AurignacianI. Theformer ischaracterizedbyFont-Yvespointsandlong,
slenderDufourbladeletsofDemarsandLaurents(1989)Dufoursubtype, which
are extracted fromunidirectional prismatic cores in the framework
of a single,continuous reduction sequence for both blades and
bladelets. The latter is characterized
bysplit-basedbonepointsandbytheuseof
carinatedscrapersasspecializedcoresfortheextractionof straight or
curvedbladelet blanksthat remainlargelyunretouched. Inthesubsequent
EvolvedAurignacianor AurignacianII, thepreferredtypesof
bladeletcores are thick burins (carinated or busked) and
thick-nosed scrapers, which generatecharacteristic small, twisted
blanks retouched into a particular Roc-de-Combe subtype ofDufour
bladelets; other types of points made of ivory, bone, or deer
antler emerged in thislater facies, all with massive bases, mostly
featuring at or oval cross sections and an overalllozengic
morphologythe Mlade c (Lautsch) points.Late Neandertals, early
moderns, and their cultural associationsThe chronostratigraphic
framework for the Middle to Upper Paleolithic transition in
Europeand its correlation with Near Eastern developments proposed
in Fig. 4 provide the backgroundfor the discussion on the cultural
associations of the human remains from the period. Theearly-mid
Upper Pleistocene evidence from the Near East cautions against the
establishmentof biunivocal correspondences between hominid taxa and
archaeological cultures, but suchreservations do not apply in the
same way to the geographical cul-de-sac of the Old Worldrepresented
by the European continent, where the Neandertal lineage is
unanimously agreedto have differentiated and evolved. As a result,
it is legitimate to assume that until a time whenthe presence of
modern humans is rst and unambiguously documented in the
continentalfossilrecordbydiagnosticskeletalremains,Neandertalsareconsideredtobethehumanactors
responsible for the features of the archaeological record.At
present, the earliest such modern human material is the complete
mandible recoveredin the cave site of Oase (Romania), directly
dated to c. 35 ka 14C BP (Trinkaus et al., 2003a,b, 2006). The cave
sites of Muierii and Cioclovina, also in Romania, have produced
slightlylater modern human material (in the c. 2930 ka14C BP
range), and the age of the largeensemble from the Moravian site of
Mlade c has now been conclusively established by theSpringerJ
Archaeol Res (2007) 15:154 19Fig. 4 Chronostratigraphic correlation
scheme between key Early Upper Paleolithic stratied sequences
ofEurope and the Near EastSpringer20 J Archaeol Res (2007)
15:154direct dating of human teeth from four different individuals
to c. 31 ka14C BP (Trinkaus,2005; Wild et al., 2005). Direct dating
of other modern human fossils from central Europetraditionally
considered to be of early Upper Paleolithic age has shown that, in
fact, they allare of Magdalenian, Mesolithic, or even later
prehistoric times (Conard et al., 2004a; Smithet al., 1999;
Svoboda, 2003; Svoboda et al., 2002; Terberger and Street, 2003a,
b).In western Europe, the only diagnostic modern human remains
likely to predate c. 30 ka14C BP are at present the juvenile
mandibles from La Quina Aval and Les Rois (Trinkaus,2005). The Les
Rois sequence belongs entirely to the Evolved Aurignacian, whereas
the LaQuina material comes from level 3 of the old excavations, at
the interface between the sitesEarly and Evolved Aurignacian
levels; a sample from the former, collected in the context ofmodern
testing work, yielded an AMS bone date of 32,650 85014C BP
(OxA-6147/Ly-256) (Dujardin, 2001) that provides a good terminus
post quem for the mandible. The dentalmaterial from the Aurignacian
I levels of Brassempouy, dated to c. 32 ka14C BP, also maybe of
modern human afnities, but the issue remains controversial (Bailey
and Hublin, 2005;Henry-Gambier et al., 2004). In any case, the
conclusion is that, given the stratigraphic anddating context, none
of these French fossils is older than c. 33 ka14C BP.Conversely,
nowhere in Europe north of the Ebro River basin have Neandertal
remainsbeen found for which an age postdating 36 ka14C BP can be
suggested on rm
grounds.Twoputativeexceptionsforwhichdirectradiocarbondatesofc.
2829ka14CBPhavebeenreported: thematerial fromlevel G1of
theCroatiancavesiteof Vindija(Smithet al., 1999) and the infant
skeleton from the cave of Mezmaiskaya in the northern
Caucasus(Ovchinnikov et al., 2000). Where the latter is concerned,
the excavators convincingly arguedthat the skeleton was found below
intact Mousterian deposits reliably dated to>36 ka14CBP and that
the direct date for the infant was therefore simply a minimum age,
the burialbeing signicantly earlier (Golovanova et al., 1999).
Where Vindija is concerned, severallines of reasoning also
indicated that the results were minimumages (Zilh ao, 2006b), and
thisinference has now been vindicated by redating of the original
samples (Higham et al., 2006).This evidence is consistent with the
hypothesis that after c. 35 ka 14C BP the archaeologicalrecord of
Europe (except parts of the Iberian Peninsula) is entirely related
to the activity ofanatomically modern people; by the same token, it
also implies that one can legitimatelyassume that the
technocomplexes of the earliest Upper Paleolithic (i.e., those
predating theProtoaurignacian) were manufactured by Neandertals
(Fig. 5).Sound evidence for this scenario is available for the Ch
atelperronian, given the unques-tionably Neandertal afnities of (1)
the individual buried in level EJOPsup of the St.-C
esairerockshelter, TL-dated to 36.5 2.7 ka BP (average of six
measurements on burnt ints), and(2) the fragmentary dental and
cranial material from the Ch atelperronian levels of the
GrotteduRenneatArcy-sur-Cure(BaileyandHublin, 2006;Hublinetal.,
1996;L ev equeandVandermeersch, 1980). This conclusion has recently
been strengthened by the direct datingtoc. 3841ka14CBP(i.e.,
inthetimerangeoftheCh atelperronian)ofthediagnosticNeandertal
remains from the El Sidr on cave in Asturias, at the western end of
the Franco-CantabrianregiontowhichtheCh
atelperronianisconned(Fortea et al.,2003;Laluezaet
al.,2005).HumanremainsassociatedwiththeUluzzianarelimitedtotwodeciduousteethfoundinlevelEoftheCavallocave,whicharesimilartoNeandertalteethinsize,cusp
morphology, and taurodontism; this latter feature, in particular,
is often present in Ne-andertal deciduous molars but has never been
observed in early modern human juveniles,suggesting that the most
parsimonious interpretation of this scarce material is that it
belongsto Neandertals as well (Churchill and Smith, 2000).In
central Europe, several important Neandertal fossils, most notably
the two individualsfromthe type site itself, are directly dated to
c. 3940 ka14CBP, but no direct evidence existsSpringerJ Archaeol
Res (2007) 15:154 21Fig. 5 Key sites documenting the archaeological
associations of late Neandertals and early European moderns(in
Iberian regions south of the Ebro basin, Neandertals survived until
well after the time of contact elsewherein Europe). Above: Latest
reliably dated Ch atelperronian, late Micoquian, and Uluzzian sites
(circles); siteswith Neandertal remains reliably directly dated
to