The Impossible Coincidence. A Single-Species Model for the Origins of Modern Human Behavior in Europe
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Evolutionary Anthropology 14:12–27 (2005)
ARTICLES
The Impossible Coincidence. A Single-Species
Model for the Origins of Modern Human Behavior
in Europe
PAUL MELLARS
Few topics in palaeoanthropology have generated more recent debate than the ternatives: that these patterns of be-
nature and causes of the remarkable transformation in human behavioral patterns that havior and the implied levels of as-
marked the transition from the Middle to the Upper Paleolithic in Europe.1–11 Those of sociated cognition emerged by a
us who have argued for an effective technological and cultural “revolution” at this point purely internal process of behavioral
in the Paleolithic sequence have emphasized three main dimensions1,2,9,11–14: the wide and cognitive evolution among the
range of different aspects of behavior that appear to have been affected (Fig. 1); the local European populations, extend-
relative speed and abruptness with which most of these changes can be documented ing directly through the European
in the archeological records from the different regions of Europe; and the potentially Neanderthal line; or, alternatively,
profound social and cognitive implications of many of the innovations involved. Most that at least the majority of the new
striking of all in this context is the abrupt appearance and proliferation of various forms behavioral patterns, as well as the
of perforated animal teeth, shells, beads, and other personal ornaments, and the even cognitive hardware necessary to
more dramatic eruption of remarkably varied and sophisticated forms of art, ranging support these innovations, was due
from representations of male and female sex organs, through the highly stylized animal to a major influx of new populations
and combined animal-human figures from southern Germany, to the striking wall into Europe deriving ultimately
paintings of the Chauvet Cave.8,15–18 One might add to this the similar proliferation of from either an African or Asian
more enigmatic but potentially equally significant abstract “notation” systems on bone source.29,30 It is hardly necessary to
and ivory artifacts.19 To describe the Upper Paleolithic revolution in Europe as reflect- stress the importance of this issue in
ing preeminently an explosion in explicitly symbolic behavior and expression is in no evolutionary terms. If the Neander-
sense an exaggeration, as most prehistorians would now agree. We are probably on thals did independently develop the
safe ground in assuming that symbolic behavior and expression of this level of whole range of behavior that tradi-
complexity would be inconceivable in the absence of highly structured language tionally has been regarded as the
systems and brains closely similar, if not identical to, our own.5,17,20 –28 hallmark of fully “modern” humans,
this would arguably be the most im-
portant thing we have learned about the
If we accept all of these social, ties they required emerged among Neanderthals since their original dis-
symbolic, and cognitive implications European populations becomes one covery more than 150 years ago. What
of distinctively Upper Paleolithic be- of the most critical issues in current follows is an attempt to review these
havioral patterns, then the issue of evolutionary and cognitive research. two alternatives, as briefly as possible,
exactly how these patterns of behav- Broadly, we are confronted by two in the light of the most recent archeo-
ior and the implied mental capaci- fairly stark and sharply polarized al- logical and biological research.
In a recent paper31 I have attempted
to explore the first of these scenarios
Paul Mellars is Professor of Prehistory and Human Evolution in the Department of Archaeology from an explicitly Darwinian, evolu-
at Cambridge University. Following his involvement in the notorious Mousterian debates with tionary perspective, which puts the pri-
François Bordes and Lewis Binford in the 1960s and 1970s, his research has centered on the
behavioral origins and geographical dispersal of anatomically modern human populations, mary emphasis on the complex pattern
and the extinction of the European Neandertals. He is the author of The Neanderthal Legacy, of climatic and associated environmen-
and coeditor of The Human Revolution, Modelling the Early Human Mind and other conference
volumes. He has conducted excavations on Mesolithic sites in England (Star Carr) and tal changes that occurred in Europe
Scotland (Oronsay). around the middle of the last glaciation
Department of Archaeology, Cambridge University, Downing Street, Cambridge, CB2 3DZ, (the period of oxygen-isotope stage 3,
UK. E-mail: p.a.mellars@arch.cam.ac.uk
from ca. 60,000 –25,000 BP32) and the
potential selective and adaptive effects
© 2005 Wiley-Liss, Inc. of these environmental oscillations on
DOI 10.1002/evan.20037
Published online in Wiley InterScience the demographic, social, and other cul-
(www.interscience.wiley.com). tural patterns of the local NeanderthalARTICLES The Impossible Coincidence 13
populations (Fig. 2). Reduced to its bare
essentials, this model assumes that the
occurrence of major, rapid, and re-
peated environmental fluctuations
could have precipitated repeated epi-
sodes of increased demographic and so-
cial competition between adjacent Ne-
anderthal groups for both space and
resources, which in turn would have
imposed strong selective pressures on
almost all aspects of their cultural and
behavioral adaptations, leading to a
range of associated patterns of techno-
logical, economic, and social change.33
Arguably, an increased investment in
various forms of symbolic expression
and communication could be seen as Figure 2. Possible climatic model for the “Upper Paleolithic revolution” in Europe,31 based on
one potentially direct evolutionary ad- potential technological and cultural adaptations to the rapid climatic oscillations of oxy-
aptation to cope with the increasing de- gen-isotope Stage 3.
mographic and social pressures that
emerged directly from the contempo-
raneous patterns of climatic and envi-
ronmental change.7,9,34,35 This model,
of course, carries with it the auto-
matic implication that all of the nec-
essary intellectual and neurological
capacities for these behaviors were ei-
ther already present in the indigenous
Neanderthal populations of Europe or
that these capacities emerged, pre-
sumably as a result of one or more
genetic mutations,23,36 as a further di-
rect evolutionary consequence of the
various environmental, demographic,
or other selective pressures to which
the European Neanderthals were sub-
jected.
Clearly, that kind of local evolu-
tionary model represents an impor-
tant theoretical perspective for the
possible origins of Upper Paleolithic
culture in Europe, and has been de-
bated at various times and from a
variety of different perspec-
tives.3,4,37– 40 It was, of course, an es-
sential and integral component of
the multiregional or regional-conti-
nuity model of modern human ori-
gins that largely dominated this field
throughout the 1960s and 1970s,41 at
much the same time as the strongly
“processualist” notions of the New
Archaeology and the strong reac-
tions against large-scale diffusion
and migration as an explanatory prin-
Figure 1. Early Upper Paleolithic behavioral innovations in Europe. For details, see Bar-Yosef,1,2 ciple in accounting for prehistoric
Gamble,12 Klein,66 Mellars,9,13,27,68,69 Kozlowski,137 White,15,16 Le Bon,107 Conard and Bolus.8 change. Recently, a similar viewpoint14 Mellars ARTICLES
Figure 3. Apparent dispersal routes of the earliest anatomically and behaviorally modern populations across Europe, as reflected in the
archeological data. The northern (Danubian) route is represented by the “classic” Aurignacian technologies, while the southern (Medi-
terranean) route is represented by the “Proto-Aurignacian” bladelet technologies with their inferred origins in the preceding early Upper
Paleolithic technologies in the Near East and southeastern Europe.56,60,71,107 Dates indicate the earliest radiocarbon dates for these
technologies in different areas, expressed in thousands of radiocarbon years BP. (Note that these are likely to underestimate the true
(calendrical) ages of the sites by between 2,000 and 4,000 years.104 –106)
has been argued strongly by Franceso that any fully balanced assessment of THE CORRELATION OF
d’Errico.4 the problem of the emergence of BEHAVIORAL AND
Making the maximum possible al- “modern” behavioral patterns in Eu-
BIOLOGICAL CHANGE
lowance for these arguments, how- rope or elsewhere must be based on a
ever, I continue to see a range of ma- closely integrated analysis of both
jor obstacles to attempting to view these lines of evidence. While I fully The first and most conspicuous ob-
these kinds of entirely local, indige- agree with d’Errico that the biological stacle to the “independent-evolution”
nous evolutionary processes as pro- and archeological evidence must, at model for the origins of modern be-
viding more than, at best, a partial certain levels, be treated separately havioral patterns in Europe stems
and inadequate explanation for the and, in a sense, allowed to “tell their from the extraordinary coincidence
broad sweep of radical behavioral in- own stories,” for any fully integrated between the timing of the major be-
novations that define the conventional perspective on modern human origins havioral innovations that define the
Middle-Upper Paleolithic transition in the separate dimensions of the arche- classic Middle-to-Upper Paleolithic
Europe (Fig. 1). The problems, as I see ological and biological evidence must transition in Europe and western Asia
them, stem from two separate sourc- inevitably be brought together. Stated and the generally agreed timing of the
es: first, the spate of recent informa- bluntly, we simply cannot afford the lux- dispersal of anatomically and geneti-
tion on the anatomical and genetic ury of allowing ourselves to look only at cally modern human populations
origins and geographical dispersal of one side of the scientific coin if palaeoan- across the continent (Fig. 3). The evi-
biologically “modern” populations; thropology is to move forward as a fully dence for this dispersal has been doc-
and second, a range of equally impor- integrated scientific discipline. From this umented at length in the recent liter-
tant developments in our understand- perspective, the essential considerations, ature,29,30,42 and rests on at least four
ing of the archeological evidence it- as I see them, can be summarized as fol- separate and essentially independent
self. The premise of what follows is lows. lines of evidence:ARTICLES The Impossible Coincidence 15
1. The evidence of the mitochon- Equally significant is the complete nological estimates based on the DNA
drial DNA patterns of modern Euro- skeleton of a young individual from data remain open to some debate,42
pean populations, when analyzed in the early Upper Paleolithic “Ahmar- but the occurrence of unambiguous
terms of “founder lineage” patterns, ian” levels at Ksar Akil in Lebanon, examples of fully anatomically mod-
points to an initial dispersal of fully dated by radiocarbon and associated ern, if fairly robust skeletal remains
genetically modern populations (that archeological material to well before within this time span is now beyond
is, with distinctively African-derived 35,000 BP, and most probably around dispute. This time range coincides
patterns of mtDNA) extending across 40,000 to 42,000 BP.58 – 60 There is also precisely with that of the conventional
Europe somewhere within the time a fragmentary maxilla, said to be dis- Middle-Upper Paleolithic transition in
range of ca. 40,000 to 50,000 BP,43– 45 tinctively modern in morphology, Europe and with the broad spectrum
best reflected in the distribution of the from the early Upper Paleolithic levels of technological, symbolic, social, and
U5 haplogroup. It is now clear that all at Kent’s Cavern in England, which other changes associated with this
these patterns of mtDNA are radically has been directly dated by AMS to transition.1,2,11,12 This clearly raises
different from those of the preceding 30,900 ⫾ 900 BP.55,61 Slightly less se- two critical questions:
Neanderthal populations in Europe, curely dated are the two modern cra- ● How could any major population
which have been shown from analyses nia from Mladeč in the Czech Repub- dispersal of this kind fail to bring with
of seven separate fossil samples to lic, attributed, on the basis of 14C it certain new technological or cul-
have mtDNA patterns that are totally dating of the associated calcite forma- tural elements, derived ultimately
lacking from both present-day Euro- tions, to around 34,000 to 35,000 BP62 from regions beyond Europe, presum-
pean populations45– 47 and a sample of and the two mandibles from Les Rois ably from either Asian or African
at least five early anatomically mod- in southwestern France, apparently sources?
ern humans from Europe.48,49 ● How do we account for the ex-
2. A closely similar age estimate for traordinary coincidence between the
the modern human dispersal across Eu- How do we account for timing of this population dispersal and
rope was produced by Rogers and the contemporaneous technological
Jorde50 and others from studies of
the extraordinary and cultural revolution that marks the
mtDNA “mismatch” distributions. This coincidence between Middle-Upper Paleolithic transition in
again appears to show a major popula- Europe, following a period of around
tion expansion of genetically modern
the timing of this 200,000 years of relative behavioral and
populations in Europe centred broadly population dispersal and technological stability throughout the
around 40,000 BP.51,52 preceding Middle Paleolithic peri-
the contemporaneous od?2,27,66 As I asked in an earlier paper,
3. Analyses of Y-chromosome DNA
patterns are less well calibrated in technological and “Can we really believe that after a pe-
riod of around 200,000 years of typi-
chronological terms than are those cultural revolution that cally Middle Paleolithic technology and
based on mitochondrial data and
must be handled with caution. Never-
marks the Middle-Upper behaviour, the local Neanderthal popu-
lations in western Europe indepen-
theless, studies of microsatellite and Paleolithic transition in dently, coincidentally, and almost mi-
other data once again point to an ini-
tial expansion of modern DNA pat-
Europe . . . raculously ‘invented’ these distinctive
features of Upper Paleolithic culture at
terns across Europe (as represented
almost exactly the same time as ana-
by the M89/M213 lineages) at around
tomically and behaviourally modern
40,000 to 45,000 BP, with a subse-
populations are known to have been ex-
quent expansion of the M173 lineage closely associated with the early Au- panding across Europe?”67 (p 44).
at around 30,000 BP.53,54 rignacian levels on the site, again dat-
As indicated in the title of the present
4. Evidence from fossil skeletal re- ing to around 32,000 to 34,000 BP.63
paper, this is what I am tempted to
mains over the relevant time range is From Kostienki site 14 (Markina
describe as the impossible coinci-
scarce and patchily distributed, but at Gora) in southern Russia, there is a
dence in the parallel records of hu-
least five or six discoveries point un- burial of a fully anatomically modern
man biological and cultural develop-
mistakably to the presence of fully an- skeleton dated to at least 30,000 to
ment in Europe.
atomically modern populations in 32,000 BP.64,65
both Europe and the adjacent parts of In short, we now have at least four
southwest Asia between ca. 30,000 essentially independent lines of evi- THE SCALE AND SPEED OF THE
and 45,000 BP.55,56 Most significant in dence for a major dispersal of fully UPPER PALEOLITHIC
this context are the recently discov- genetically and anatomically modern
ered remains of three separate indi- populations across Europe and west-
REVOLUTION
viduals from the Peştera cu Oase cave ern Asia somewhere within the range The second significant feature to
in Romania, which have been directly of 45,000 to 35,000 BP, that demon- emphasize in this context is the dra-
dated by radiocarbon AMS techniques strably and fairly rapidly replaced the matic scale of the so-called Upper
in two separate laboratories to preexisting Neanderthal popula- Paleolithic revolution in Europe and
34,290 ⫾ 900 BP and ⬎35,200 BP.57 tions.29,42,49 The precision of the chro- the relative speed with which it oc-16 Mellars ARTICLES
curred. As noted earlier, the Middle- THE AFRICAN EVIDENCE to all appearances identical to those
Upper Paleolithic transition is encountered in European Upper Pa-
Our knowledge of the archeological
marked by changes in effectively all leolithic sites75 (Fig. 4). Even if rare
evidence from Africa over the Upper specimens of burins have occasionally
of the archeologically visible dimen-
Pleistocene time range has expanded been claimed from the European
sions of behavior: radical innova-
tions in both the forms and tech- dramatically during the past two de- Mousterian, fully typical specimens of
niques of blade and bladelet cades.5–7,66,72–74 On the basis of this end scrapers are at best a very rare
production in stone tools; the sud- new evidence it is now possible to and debatable occurrence.81,82 The
den florescence of complex, varied, show beyond any reasonable doubt appearance of new end-scraper forms
and highly shaped bone, antler, and that many of the most distinctive ar- most probably reflects the emergence
ivory tools; the emergence of elabo- cheological hallmarks of the classic of new forms of skin-working technol-
rate notation systems on bone and Middle-Upper Paleolithic transition in ogy, while the appearance of burins
ivory artifacts; the appearance of ex- Europe can be documented at least may or may not be directly related to
tensive and organized exchange sys- 30,000 to 40,000 years earlier in cer- the appearance of shaped bone tools
tems for the distribution of both raw tain parts of Africa than anywhere at Blombos and other African sites.
materials and decorative prestige within Europe itself. In this context, 2. The appearance of a range of
items; the effective explosion of per- the evidence reported recently from carefully shaped small geometric
forated animal-tooth pendants, per- the so-called Howiesons Poort levels forms, evidently employed as insets in
forated marine shells, laboriously at Klasies River Mouth in South Af- multi-component hafted tools (Fig. 4).
shaped stone and ivory bead forms, The highly varied geometrical shapes
and other forms of personal orna- encountered in the Howiesons Poort
ments; and the emergence of highly . . . it is now possible to industries at Klasies River Mouth and
sophisticated and varied forms of show beyond any elsewhere (triangles, trapezes, cres-
both abstract and “naturalistic” cents, and obliquely blunted points)
art.1,2,11,13–16 In addition, there were reasonable doubt that not only reflect a high degree of delib-
many more inferential but appar- many of the most erately “imposed form” in stone tool
ently closely associated changes in production, with possible social and
the economic, social, and demo- distinctive archeological symbolic connotations, as discussed
graphic patterns of the human hallmarks of the classic by Wurz,83 Deacon,73 and others, but
groups2,9,12,68,69 (Fig. 1). Not only are almost certainly reflect the appear-
all these features conspicuously ab-
Middle-Upper Paleolithic ance of new forms of multi-compo-
sent from well-documented Middle transition in Europe can nent tools employed as either hunting
Paleolithic contexts in Europe, as be documented at least missiles or, possibly, as hafted insets
d’Errico4 has recently stressed, but for plant-processing tools.74,75 The de-
they show a close correlation, for the 30,000 to 40,000 years sign complexity of these tools is un-
most part, with the distribution of earlier in certain parts of questionably far greater than that of
various forms of distinctively Aurig- the occasional specimens of hafted
nacian and “Proto-Aurignacian” Africa than anywhere Mousterian points recorded from the
technologies across the continent, within Europe itself. Eurasian Middle Paleolithic.4 It
generally between ca. 40,000 and should be added that the a large part
35,000 BP.8,11,56,70,71 of the lithic industry at Klasies River
There is a sharp contrast between Mouth is manufactured on high-qual-
the relative speed and abruptness with ity raw materials deliberately im-
which all of these novel technological rica, dated on the basis of several lines ported into the site from a distance of
and cultural features appear in the ar- of evidence to around 70,000 BP,72–75 at least 20 km,74 a feature that is again
cheological records of Europe and the and the evidence from the slightly ear- hard to parallel in European contexts
apparently more gradual, piecemeal lier Still Bay levels at the nearby before the earliest Upper Paleolithic.
fashion with which similar innova- Blombos Cave (ca. 75,000 to 80,000 All in all, there is no doubt that at least
tions appear in the archeological BP)76 –79 are especially significant. the greater part of the industry from
records of Africa. In short, any at- Leaving aside the occurrence of stan- Klasies River Mouth would be unhesi-
tempt to explain the Upper Paleolithic dardized blade technology, which is tatingly classified as technologically
revolution in terms of purely local now known to occur sporadically in “Upper Paleolithic” if found in a Eu-
evolutionary processes in Europe Middle Paleolithic/Middle Stone Age ropean or southwest Asian context.
would need not only to account for contexts back to at least 200,000 to 3. The appearance of extensively
the impressive range and scale of the 250,000 BP in both Africa and Eu- shaped bone tools, exhibiting once
cultural changes in question, but to rope,2,6,27,80 this evidence can be sum- again not only a clear degree of im-
explain why these changes appear so marized as follows: posed form but a complex sequence of
much more rapidly in the archeologi- 1. The occurrence of relatively manufacturing stages to shape and
cal sequence of Europe than in that of abundant and highly typical speci- polish the tools. The best-described
Africa. mens of both end scrapers and burins, specimens are those from the Still BayARTICLES The Impossible Coincidence 17
seems to confirm their role in certain
explicitly symbolic or ceremonial ac-
tivities.78 Similar use of ochre is, in
fact, abundant in many African Mid-
dle Stone Age sites, apparently ex-
tending back, at the Twin Rivers site
in Zambia and the Kapthurin sites in
Kenya, to at least 250,000 BP.6,86,87
Whatever significance one may attach
to the sporadic occurrence of black
manganese dioxide and occasional
fragments of ochre at European
Mousterian sites,4 it is clear that the
scale of this red ochre use at African
sites vastly exceeds that recorded any-
where in Europe prior to the Upper
Paleolithic.
5. Most significant of all, the occur-
rence of a range of explicitly “artistic”
or “decorative” items, for which an
interpretation in terms of complex
symbolic communication systems
now seems beyond question. The
most significant finds are the two
large pieces of red ochre incised with
complex and repeated criss-cross de-
signs recently reported from the Still
Bay levels at Blombos.77 These are
now generally recognized as the earli-
est fully convincing examples of delib-
erate and repeated design motifs re-
corded anywhere in the world,
certainly exceeding anything at
present known from Mousterian or
earlier contexts in Europe. Recently,
the significance of these finds has
been graphically underscored by the
recovery from the same archeological
Figure 4. Stone tools from the Middle Stone Age Howiesons Poort levels at Klasies River levels of no less than 41 specimens of
Mouth, South Africa (ca. 70,000 BP), showing typical end scrapers, burins, and shaped carefully perforated seashells (Nassa-
“geometric” forms manufactured from blade segments, probably representing hafted rius kraussianus), which were appar-
inserts of composite hunting armatures (modified from Singer and Wymer75). ently introduced into the site from es-
tuarine contexts at least 20 km away
from the site and, on the basis of mi-
levels at Blombos cave, where they ap- ropean Upper Paleolithic sequence. As croscopic analyses, were intended for
parently served a range of functions, d’Errico4 has recently stressed, exten- suspension from cords or thongs79
from sharply pointed awls and leath- sively shaped bone tools of any form (Fig. 6). Even earlier occurrences of
er-piercing tools to carefully finished are as yet effectively unknown from perforated marine shells have been re-
and polished projectile points.77 Sim- well-documented Middle Paleolithic ported from the 90,000-year-old Mid-
ilar shaped bone tools have been re- sites in Europe. dle Paleolithic levels of the Qafzeh
corded more sporadically from other 4. The occurrence of large quanti- cave in Israel, where they were asso-
African Middle Stone Age sites.4,6 If ties of red ochre (including over 8,000 ciated with a veritable cemetery of es-
the dating of the highly shaped barbed pieces from the Still Bay levels at sentially anatomically modern human
bone points recovered from three sep- Blombos) including many pieces with remains, including at least one in the
arate sites at Katanda in former Zaire smoothed facets or deliberately form of a clearly ceremonial burial
can be securely attributed to around scraped surfaces, which almost cer- accompanied by red ochre and a large
90,000 BP,84,85 the levels of complex- tainly imply their use as coloring pig- pair of deer antlers.88,89 The latter
ity of bone working achieved in these ments.76,77 The presence of geometri- finds, together with those from the
early African sites will parallel any- cal designs incised on at least two nearby site of Skhul, presumably re-
thing at present known from the Eu- large pieces of ochre from Blombos flect a brief expansion of anatomically18 Mellars ARTICLES
cal records of Africa by at least 70,000
to 80,000 BP, long before their occur-
rence in Europe (Fig. 6). Exactly how
we interpret these features in cultural
and cognitive terms will no doubt re-
main the topic of lively debate. There
will no doubt be similar debate as to
how far we can trace direct continuity
of these features between the African
sites dated to around 60,000 to 70,000
BP and the earliest manifestations of
fully Upper Paleolithic culture in Eu-
rope and western Asia at around
40,000 to 45,000 BP. Perhaps the
main point to be kept in mind here is
that Africa is an extremely large and
ecologically varied continent, and
that, as Richard Klein5,66 and others
have recently stressed, well-docu-
mented archeological sites spanning
the critical period between ca. 60,000
and 45,000 BP are still virtually lack-
ing in most parts of Africa. Certainly
sites showing a similar combination
of blades, end scrapers, small “seg-
ment” forms and carefully shaped os-
trich eggshell beads are well docu-
mented from at least 40,000 BP at
sites such as Enkapune ya Muto in
East Africa,90 and may extend back to
50,000 to 60,000 BP at Mumba and
elsewhere.6,74 And of course the pre-
cise geographical source area (or ar-
eas) from which the small founder
populations of anatomically and ge-
netically modern humans colonized
Europe and western Asia from ca.
45,000 to 50,000 years onward re-
mains to be established.60,91 But in
any event, the occurrence of a wide
range of distinctively “modern” (or in-
deed “Upper Paleolithic”) behavioral
features at a much earlier date in the
continent that is known to have given
Figure 5. Perforated shells of Nassarius kraussianus from the Still Bay levels at Blombos Cave,
South Africa, dated to ca. 75,000 to 80,000 BP (reproduced with permission from Hen- rise to the evolution of anatomically
shilwood et al., Middle Stone Age shell beads from South Africa. Science 304:404. 2004 and genetically modern populations
AAAS). can hardly be dismissed.29,42,92 As I
have commented elsewhere,93 to ig-
nore these striking and well-docu-
modern populations from Africa to at least 80,000 to 90,000 BP in southern mented similarities between the Mid-
the adjacent parts of southwest Asia at Africa and the Levantine region must dle Stone Age archeological records of
an early stage in the last glacia- now be accepted as a well-documented Africa and the early Upper Paleolithic
feature of the archeological record. records of Western Eurasia would be
tion.29,30,60 At present, these finds rank
to take a strangely short-sighted view
as the earliest unambiguous examples To summarize, we now have seem- of the archeological evidence as a
of “personal ornaments” recorded in ingly unambiguous evidence that at whole.
the archeological record, and again least the majority of the most distinc-
are without parallel from European tive and widely discussed archeologi-
sites prior to ca. 40,000 BP. In short, cal features of the so-called Upper Pa-
CHRONOLOGICAL PATTERNS
the appearance of complex social or leolithic revolution in Europe can be Finally in this context we should
symbolic communication systems by firmly documented in the archeologi- note what appears to be a more gen-ARTICLES The Impossible Coincidence 19
diffusion among the later Neander-
thal communities extending effec-
tively across the whole of Europe and,
presumably, the adjacent Middle
East.3,4,40 But in the latter case, of
course, it is difficult to see how one
could exclude the possibility that
these technological diffusion pro-
cesses originated in the gradual dis-
persal of anatomically and behavior-
ally modern populations from outside
these regions—that is, ultimately
from adjacent Asian or North African
sources. As I will discuss further, this
is what I would refer to as the inevi-
table “bow-wave” effect of technolog-
ical and cultural diffusion extending
some way in advance of the actual
dispersal of anatomically modern
populations into the different regions
of Europe.
The final and perhaps most puz-
zling aspect of the local-origins model
lies in the ultimate fate of the Nean-
derthals. Proponents of the local-ori-
Figure 6. Comparison of artifacts from early Upper Paleolithic “transitional” levels in Europe gins model would, of course, dispute
and later Middle Stone Age levels in South Africa. Note that the South African forms are
the notion that there was any inherent
approximately twice as old as the typologically and technologically similar European forms
(modified from d’Errico4). cognitive or intellectual superiority of
the biologically modern populations
over those of the Neanderthals. They
would also argue that most if not all of
eral chronological cline in the dis- emergence of similar features in Africa,
the technological and symbolic inno-
persal of early Upper Paleolithic cul- this would accord much better with the
vations that traditionally have been
ture across Europe and western hypothesis of a gradual dispersal or dif-
credited to the dispersal of modern
Asia11,56,60,70 (Fig. 3). From southwest fusion of these technological elements
populations had already been devel-
Asia there is evidence of a relatively (regardless of whether carried by new oped independently among the final
sudden and sharply defined transition populations) than with a totally inde- Neanderthal communities, according
from typically Middle to typically Up- pendent evolution of the same features to the multiple-species model for the
per Paleolithic technology; (marked within the individual regions of Europe. origins of behavioral modernity.3,4,40
by a proliferation of blades, end scrap- Clearly, there is a further potential But this, of course, immediately and
ers, burins, and new “type fossil” paradox inherent in the independent inevitably begs the question of exactly
forms, together with a range of perfo- origin model here. If the argument is why and how the Neanderthals de-
rated shell ornaments) radiocarbon that characteristically Upper Paleo- clined so rapidly to extinction in the
dated at the two sites of Boker Tachtit lithic technology and culture devel- face of the modern human dispersal
in southern Israel and Ksar Akil in Leb- oped entirely independently in several across Europe. This question becomes
anon to around 45,000 to 47,000 different regions of Europe, as re- even more acute if we bear in mind
BP.1,35,60 The date of a similar transi- flected by the Chatelperronian in the fact that the Neanderthals were
tion in southeastern Europe, as at Ba- France, the Szeletian in Central Eu- the product of at least 200,000 years of
cho Kiro and Temnata in Bulgaria, rope, and the Uluzzian in Italy, then biological and behavioral adaptation
seems to center on ca. 40,000 to 43,000 this would require an extraordinary to the demanding glacial and perigla-
BP, while in western Europe there is no degree of convergent and simulta- cial environments of Europe, whereas
evidence of any substantial, analogous neous evolution in the patterns of the intrusive modern populations had
shift in technology until ca. 38,000 to technological development within evolved in biological, anatomical, and
40,000 BP, in the form of the earliest these different regions. If, on the other presumably behavioral terms to the
Aurignacian and “Proto-Aurignacian” hand, the assumption is that large- massively different tropical and sub-
technologies.1,2,8,11,56,60,71,94 If there is scale intercommunication and trans- tropical environments of sub-Saharan
indeed a significant chronological cline mission of technology between these Africa.5,29,30,95 Stated crudely, if the
in the appearance of distinctively Upper different areas accounts for these con- European Neanderthals were so cog-
Paleolithic technology from east to west vergent patterns of development, this nitively advanced and had developed
across Europe, and with a much earlier would imply large-scale technological most if not all of the elements of char-20 Mellars ARTICLES
acteristically “modern” culture and some new behavioral elements derived hearse all of the long-running debates
cognition, why did they succomb so ultimately from either African or Asian over potential contact and related “ac-
rapidly to a biologically and environ- sources? And if the Neanderthals inde- culturation” scenarios between Nean-
mentally less well adapted species pendently developed all of these fea- derthals and modern humans that
within a space of, at most, a few thou- tures, why did they so rapidly become have occupied much of the archeolog-
sand years?96 extinct in the face of a biologically and ical literature on modern human ori-
environmentally less adapted species? gins over the past decade.3,4,10,98 –102
DISCUSSION Ultimately, many of these debates will
rest heavily on the accuracy and pre-
My overall conclusion is that what- Interaction Scenarios cision of the associated dating evi-
ever weight we may attach to the ca- dence, which, in the case of radiocar-
pacity of the climatic and environ- One element that is, of course, im-
bon dates in the region of 30,000 to
mental oscillations of OIS-3 to foster plicit and inescapable in any model of
40,000 BP are notoriously problem-
adaptive changes in the behavioral modern human dispersal across Eu-
atic, due mainly to the massive prob-
patterns of later Neanderthal commu- rope is the occurrence of various
lems of contamination effects in this
nities, this still remains at best a highly forms of contact, and therefore poten-
tial interaction, between the expand- time range,103 and to the current un-
inadequate explanation to account for
ing sapiens and indigenous Neander- certainties over the precise patterns of
the broad range of radical technologi-
atmospheric 14C fluctuations and
cal, social, and cognitive changes that
their effects on associated calibration
define the classic Middle-Upper Paleo-
curves.104 –106 It is now apparent that
lithic transition in Europe and western . . . if the European many of these uncertainties will not
Asia. I certainly am not suggesting that
the technological and cultural adapta- Neanderthals were so be resolved decisively in the immedi-
ate future.
tions of Neanderthal populations were cognitively advanced The one point that is now clear is that
static throughout this period. Inevita-
bly, there would have been significant and had developed in the case of the best-documented and
most widely discussed of these accul-
adaptations in both the technology and most if not all of the turation scenarios, the French Chatelp-
economic patterns and, no doubt, the
related social organization of the hu-
elements of erronian, the totality of the available
man groups in response to the many characteristically dating evidence from radiocarbon,
episodes of climatic change throughout thermoluminescence, uranium-series,
the 200,000-year span of the Middle Pa-
“modern” culture and electron-spin-resonance, and other
leolithic sequence, as I have discussed cognition, why did they methods now leaves no significant
in detail elsewhere.27,31 The critical ob- room for doubt that at least the greater
succomb so rapidly to a part of the Chatelperronian develop-
jections to a strictly in situ model for
the emergence of fully Upper Paleo- biologically and ment in central and western France, in-
lithic culture in Europe remain simply cluding the much debated levels from
environmentally less well Arcy-sur-Cure, must be seen as broadly
those of the radical scale and complex-
ity of the behavioral changes involved, adapted species within contemporaneous with the presence of
various forms of “Aurignacian” technol-
the clear evidence for the emergence of a space of, at most, a ogies and associated anatomically mod-
most if not all of these features at a
much earlier date in Africa than in Eu- few thousand years? ern populations in the adjacent areas of
rope (in close association with the bio- both Central Europe and, almost cer-
logical emergence of our own species), tainly, along the Mediterranean coast
and what I have described as the ex- and northern Spain.8,10,56,67,107 The
traordinary “coincidence” that all of available radiocarbon dates for the
thal populations across effectively the
these behavioral innovations appear in Chatelperronian levels at Arcy-sur-Cure
whole of the continent. These contacts
the archeological records of Europe cluster around 33,000 to 35,000 BP, and
must have been repeated and must are reinforced by the results of radio-
and western Asia at almost precisely the
have occurred in all the areas occu- carbon, thermoluminescence, and elec-
same time as the well-documented ex-
pied by Neanderthals at the time of tron-spin-resonance dating from a
pansion of anatomically and genetically
modern populations across these re- the modern human dispersal. One range of other sites, including St. Cé-
gions, with a clear chronological cline critical and unavoidable issue in any saire, Les Cottés, Le Moustier, Combe
in the appearance of these elements consideration of modern human dis- Saunière, Roc de Combe, and Grotte
from east to west across the continent persal must therefore be the precise XVI, all of which put the time range of
(Fig. 3). As I discussed earlier, is it really nature of these interactions between the Chatelperronian back to, at most,
plausible that the dispersal of an en- Neanderthals and modern humans ca. 39,000 to 40,000 BP in uncalibrated
tirely new population across Europe, by and their potential reflections in the radiocarbon terms.10,67,108 By compari-
groups who are generally now seen as a archeological records of the different son, we now have multiple radiocarbon
separate biological species from the Ne- regions of Europe. dates for clear occurrences of early Au-
anderthals,29,97 would not bring with it There is hardly space here to re- rignacian technology from the sites ofARTICLES The Impossible Coincidence 21
Geissenklösterle (level III) and Keil- these items carried precisely the same sive, more technologically “advanced”
bergkirche in Germany, and Willendorf social and cultural meanings among groups. Put differently “in a contract-
(levels 2 and 3) in Austria, clearly within the final Neanderthal communities of ing, competitive, late Neanderthal
the time range of 36,000 to 39,000 BP. western Europe as they did among the world, it may have been precisely the
These points have now been docu- intrusive populations of biologically ability to copy the habits or appear-
mented clearly in recent papers by Co- and behaviorally modern people. The ance of the new, intrusive groups
nard and Bolus,8 Conard, Dippon, and alternative, of course, is that these ar- which conveyed increased social or
Goldberg,109 Haesaerts and Tey- tifacts served as various forms of per- personal prestige, or even improved
ssandier,110 and Richter and cowork- sonal or sexual display or prestige mating success, within the local or re-
ers.111 And from France itself, we now behavior within the social and demo- gional groups. If this were the case,
have a series of 14C dates for early Au- graphic context of the late Neander- then this could have had a critical im-
rignacian levels ranging between thal communities and in this way pact on the evolutionary survival
35,000 and 37,000 BP, with the possi- played a vital role in their demo- strategies of the final Neanderthal
bility of even earlier dates for the cur- graphic selection and competitive groups”67 (p 44). The exchange, of
rently undated bladelet Aurignacian strategies.10,17,116 As I have com- course, is unlikely to have been a one-
level (layer K) at the base of the long sided process; it is equally predictable
Aurignacian sequence at Le Piage.67,107 that several other aspects of behavior,
Even if we set aside the disputed inter- such as specific hunting strategies or
stratifications of Aurignacian and Chat-
. . . if the earliest the use of new raw-material supplies
elperronian levels in three separate anatomically modern would have been exchanged between
French sites (Le Piage, Roc de Combe, populations arrived in the Neanderthals and the incoming
and Chatelperron itself)112 there can be modern groups. In this sense the word
no serious doubt that the greater part of western Europe “acculturation” should perhaps be
the Chatelperronian sequence in manufacturing complex abandoned, if only because of its po-
France, including the occurrence of tential to be misrepresented in social
simple bone tools and associated bone and antler tools or socio-political terms.40 But to as-
grooved or perforated animal-tooth and wearing a variety of sume that some exchange of technol-
pendants at Arcy-sur-Cure, are contem- ogy between the indigenous and intru-
poraneous with the presence of both
personal ornaments and sive populations necessarily implies
Aurignacian technologies and appar- other items of social identical social and cognitive mean-
ently associated anatomically modern display, as ings for the technological elements in-
populations in the closely adjacent ar- volved would not simply be logically
eas of Central Europe. Similarly, there demonstrably they did, unwarranted, but positively bad an-
can be no doubt that distinctively then some exchange or thropology, as numerous ethno-
“proto-Aurignacian” bladelet industries graphic and anthropological studies
were being manufactured at the Abri replication of these of recent ethnic contact situations
Fumane and other sites along the Med- behaviors by the local have revealed.117 To argue in these
iterranean coast at broadly the same terms seems to me not so much reac-
date, between ca. 36,000 and 39,000
Neanderthal groups tionary, hidebound, and politically in-
BP.10,67,113–115 would seem an correct conservatism, as Zilhão40 has
The critical question in this context recently suggested, but the most bal-
inevitable and totally
is exactly what significance we should anced and economical way of ac-
attach to the presence of these simple predictable reaction . . . counting for the totality of the avail-
bone tools and animal-tooth pendants able archeological, biological, and
in the Chatelperronian levels at Arcy- chronological data.
sur-Cure and, in a few isolated cases, Needless to say, one would not ex-
in other French Chatelperronian sites. pect the transfer of technology be-
mented elsewhere,10 if the earliest an-
As I have discussed elsewhere,10 the tween the sapiens and Neanderthal
atomically modern populations arrived
Neanderthals were clearly expert populations (or vice versa) to involve
in western Europe manufacturing com-
craftsmen. The ability to shape simple exact replication of the various tech-
bone tools or to incise grooves or per- plex bone and antler tools and wearing nological elements in question. Each
forations in the roots of animal teeth a variety of personal ornaments and element would, no doubt, be assimi-
would have posed little challenge to other items of social display, as de- lated and integrated into the recipient
groups who could shape wooden monstrably they did, then some ex- communities in terms of their own
spears or produce elegantly controlled change or replication of these behav- preexisting technological practices
Levallois points or cordiform bifaces, iors by the local Neanderthal groups and ideological structures.102,117 This
given the opportunity to observe these would seem an inevitable and totally is clearly apparent in both the lithic
technologies (or their products at first predictable reaction, as observed in technology and the majority of bone
hand).102 Clearly, the critical issue is effectively all recent contact situa- artifacts of the late Chatelperronian
whether the production and use of tions between indigenous and intru- groups, neither of which replicates22 Mellars ARTICLES
precisely those of the earliest Aurigna- racy and precision of the associated nology, such as metal knives, guns,
cian populations. Nevertheless, Ran- dating evidence. There is, however, a and horse transportation, among the
dall White118 has shown that certain further major factor to be taken into native populations of North America
specific elements of the nonlithic arti- account in this context, which might and Canada at the time of European
facts from Arcy-sur-Cure do exhibit be most conveniently referred to as contact.117 One critical factor in the
such specific similarities to those the ripple or “bow-wave” effect of cul- rate of dispersal of any technological
from nearby Aurignacian sites (such tural and technological diffusion, po- innovations of this kind would pre-
as the frequent use of fox canines as tentially extending well in advance of sumably have been the relative func-
personal ornaments and the presence the actual dispersal of behaviorally tional efficiency of the innovations in
of distinctive bone tubes and ivory question in comparison with the pre-
and anatomically modern populations
ring-like forms) that probability that ceding Middle Paleolithic techniques.
across Europe.10,31 The premise, quite
these forms originated entirely inde- But it is not difficult to visualize how
simply, is that among the later Nean-
pendently in the two groups seems certain technological elements that
derthal populations of Europe there
virtually inconceivable. While many had strong adaptive advantages, such
of the bone artifacts at Arcy-sur-Cure must inevitably have been various as new forms of skin-working technol-
can reliably be shown to have been forms of communication or interac- ogy, reflected in the use of new end-
produced on the site,3 the possibility tion between geographically adjacent scraper forms, or simple forms of
of an actual exchange of certain items groups. Whether visualized in terms bone and antler technology, could
such as personal ornaments between of systems of local mate exchange or have dispersed in this kind of “bow-
the Chatelperronian and Aurignacian the exchange of flint or other raw ma- wave” diffusion process well in ad-
groups can in no way be ruled out, as terials,12,119,120 these linkages are vance of the dispersal of anatomically
Hublin98 and others have stressed. likely to have provided potential chan- modern populations into the more
The possibility that these exchanges central and western parts of Europe.
involved some limited degree of inter- Once again, such patterns are not
breeding between the two populations . . . among the later merely plausible, but arguably inevi-
cannot be ruled out from either the table and predictable in the social and
DNA or skeletal evidence29,47,49 and
Neanderthal populations demographic context of the late Nean-
should also be taken into account. of Europe there must derthal groups. To describe this kind
In this context, one should recall of diffusion process as “acculturation”
Francesco d’Errico’s4 recent sugges-
inevitably have been may or may not be appropriate. But it
tion that “it may have been precisely various forms of could well have served as an impor-
the new situation involving contact communication or tant factor in the earliest appearance
between anatomically modern people of certain distinctively Upper Paleo-
and Neanderthals, and the conse- interaction between lithic elements of technology over
quent problems of cultural and bio- geographically many areas of western Asia and Eu-
logical identity, that stimulated an ex- rope in the period between ca. 45,000
plosion in the production of symbolic adjacent groups. and 35,000 BP.
objects on both sides.” While this sug-
gestion seems to me astute and poten-
tially highly germane to the present “Contextual” Factors in Early
discussion, it is, of course, a specifi- Modern Culture
nels of communication for particular
cally interactive model, which implies
elements of technology or technologi- All expressions of human culture
and assumes a close contemporaneity
cal innovations extending across large are, of course, ultimately dependent
and direct interaction between the
areas of Europe, and potentially be- not only on the underlying cognitive
two groups. It is frankly difficult to
tween communities that were only and technological repertoires of the
visualize in this situation how one
distantly related in social and demo- societies involved, but also on their
would ever discriminate definitively
graphic terms. These “chains of con- interaction with purely local condi-
between the independent-evolution
nection” as John Mulvaney121 has de- tions of various environmental fac-
model versus acculturation scenarios
scribed them, have been widely tors, population densities, and so on.9
for the emergence of distinctively Up-
documented among recent hunter- Recently there has been considerable
per Paleolithic features among the fi-
gatherer groups, and are known to debate on the contribution of these
nal Neanderthal populations.
have carried both technological ideas so-called contextual factors in the
and particular elements of material varying geographical expressions of
“Bow-Wave” Diffusion Effects culture, such as prized species of ma- early “modern” culture, centering on
All of these arguments about direct rine shells and especially valued raw the archeological records of both Af-
interaction or acculturation effects of materials, over distances of several rica and Europe.5,7,8,73 In southern Af-
course rest heavily on the detailed hundred and, in some cases, thou- rica, most of the debate has focused
space-time patterning of the final Ne- sands of kilometres.122 Similar pat- on how the archeological expressions
anderthal and earliest anatomically terns can be seen in the diffusion of of new cultural and cognitive patterns
modern populations, and on the accu- specific elements of European tech- may have been influenced by chang-ARTICLES The Impossible Coincidence 23
ing population densities of local com- gument is that local climatic and eco- aspects of early modern technology
munities and the potential impact of logical conditions in this region could and culture extending from the south-
these episodes of population increase have fostered relatively high densities ern tip of Africa to the Atlantic coast
or decrease on local social and eco- of local populations, which may have of western Europe becomes, arguably,
nomic patterns.5,7,72,73 It goes without been largely sedentary over at least even more remarkable.
saying that these local contextual fac- part of the annual cycle. By analogy
tors can be invoked only in contexts with the behavior of recent hunter- The Human Revolution?
where both the innate cognitive ca- gatherer groups in similar arctic and
pacities for particular patterns of be- periglacial environments, such as the In conclusion, we might ask what
havior and the essential technological Inuit, it could be argued that this relevance all of this has to the notion
expertise for the behaviors in question could well have fostered more com- of an “Upper Paleolithic Revolution”
associated with the appearance of
are already present in the populations plex patterns of both technology and
modern humans in Europe or, indeed,
involved. various kinds of ceremonial than
a more general “human revolution”
These factors, however, may be cen- those practiced by groups in more
associated with the emergence of our
tral to understanding the varying ex- temperate or forested environ-
species as a whole, both of which have
pressions of characteristically early ments.34,125,126 One could add to their
generated lively debate in the recent
Upper Paleolithic culture in different arguments that a strong focus on spe-
literature.1,2,5–7,73 Clearly, there has
regions of Europe and Western Asia. cifically animal-centered art and tech-
been confusion in some of the recent
It has often been suggested, for exam- nology would hardly be surprising
discussions of these issues. Those of
ple, that the earliest stages of the Up- among groups who are likely to have
us who have argued for the notion of
per Paleolithic sequence in the Middle been almost entirely dependent on an-
an effective revolution in human be-
East and parts of Mediterranean Eu- imal resources for their daily and an-
havioral patterns over the period of
rope are poorly equipped in terms of nual food supplies. It is therefore un-
the conventional Middle to Upper Pa-
bone and antler technology.39 In real- likely to be a coincidence that the
leolithic transition in Europe and
ity, this is a considerable overstate- most technologically and artistically
western Asia have always tried to
ment, since a wide range of bone and complex expressions of Aurignacian
make it clear that we were visualizing
antler artifacts are well represented in culture in Europe are found in the this phenomenon essentially as a be-
the early Aurignacian levels at Hay- most northerly, periglacial parts of its fore-and-after scenario, associated di-
onim, Kebara Cave, and elsewhere in geographical range. Jochim,127 I,128 rectly with the appearance of new
Israel, dated to around 34,000 to and others have suggested similar en- populations in Europe and deriving
36,000 BP60,123 and in the similar Au- vironmentally related factors for the ultimately from regions beyond Eu-
rignacian levels (before 32,000 BP) at extraordinary concentration of both rope, in the ways discussed in the ear-
Ksar Akil in Lebanon.124 But in these Upper Paleolithic cave and mobiliary lier part of this paper.1,2,9,14,90,116 In
areas, where a large part of the food art within the densely occupied re- other words, this pattern could be
supply was almost certainly derived gions of southwestern France, the seen as a revolution in terms of its
from plant foods rather than animal Pyrenees, and Cantabrian Spain. In reflection in the archeological records
resources, and where wood is likely to my view, none of this detracts from of the classic Middle to Upper Paleo-
have been far more readily available the striking uniformity of most as- lithic transition, but emphatically not
for tool manufacture than in contem- pects of early Upper Paleolithic cul- as implying an autochtonous, in situ
porary sites in most of northern and ture over large areas of Europe and evolution of these behavioral patterns
western Europe, a reduced emphasis the adjacent Middle East, as reflected within Europe itself. My own publica-
on bone and antler for tool production above all in the widespread distribu- tions from 1989 onward have always
is no doubt largely predictable in en- tion of the highly distinctive Aurigna- tried to make this implication clear.
vironmental terms. The same could be cian and Proto-Aurignacian technolo- Whether this behavioral revolution
said of many parts of southern and gies.11,56,70,71,129 But we should originated in some closely adjacent
Mediterranean Europe, where early certainly not fall into the trap of ex- core area, such as the Nile valley or
Upper Paleolithic bone tools are sim- pecting to find an identical expression Northeast Africa, or in more distant
ilarly relatively sparse. of these early forms of Upper Paleo- parts of Africa remains the central el-
Conard and Bolus8 have recently ar- lithic culture over the whole of the ement in most of the recent de-
gued that similar environmental fac- highly environmentally diverse re- bates.1,2,5,6,9,90,93
tors are likely to have been effective in gions of Europe and southwest Asia. It How far we choose to visualize
the emergence of highly complex goes without saying that even greater what happened in Africa as reflecting
forms of bone, antler, and ivory tech- technological, economic, and social a revolution seems to me largely a
nology in the early Aurignacian sites adaptations would be expected in the question of semantics and, no doubt,
in southern Germany, and perhaps preceding dispersal of anatomically personal taste. McBrearty and
also in the florescence of both per- modern populations from Africa to Brooks6 were quite right to stress the
sonal ornaments and the impressive the vastly differing environments of contrasts between the archeological
mobiliary art objects from sites such western and northern Eurasia.93 records of Europe and those of Africa
as Vogelherd, Geissenklösterle, and When allowance is made for these fac- in this context in their influential pa-
Hohlenstein Stadel. In brief, their ar- tors, the broad similarities of many per entitled, engagingly, “The Revolu-You can also read
