d'errico 2003 - the invisible frontier - a multiple species model for

15
ARTICLES The Invisible Frontier. A Multiple Species Model for the Origin of Behavioral Modernity FRANCESCO D’ERRICO The alternative scenario considers behavioral modernity to be the out- come of a gradual process in Africa where anatomically modern humans originated. This process is seen as cor- responding to the technological changes observed through the African Middle Stone Age. 8–10 These changes started around 250,000 years ago, at the very end of the Acheulean, and proceeded until the transition from the Middle Stone Age to the Later Stone Age, conventionally placed at 40,000 years ago. BLOMBOS AND THE AFRICAN EVIDENCE Several recent discoveries seem to reinforce the second model. The most striking is the recent publication of two ochre fragments (Fig. 1) from the Middle Stone Age levels of Blombos, Cape Province, which bear similar en- graved geometric patterns; 11 they have been dated by two different methods to ca. 77,000 years ago. These objects were found associated with a Still Bay industry with shaped bone awls and bifacial spear points 9,12–13 (Fig. 2), a possible en- graving on bone, 14 and with more than 8,000 fragments of ochre, most of which bear clear traces of use. The presence of symbolic engravings on artifactual pigment makes it unlikely that the thousands of pigment frag- ments found at Middle Stone Age sites were strictly functional and suggests instead that they were used for sym- bolic purposes. Other engravings or notched pieces are reported from Middle Stone Age contexts at south- ern African sites such as Klasies River, Apollo 11, Hollow Rock Shelter, Bor- der Cave, and Diepkloof. 7,14 In other words, Blombos engrav- ings and bone tools may suggest that southern African human populations had already acquired behavioral mo- dernity 30,000 years before the ap- pearance of these innovations in Eu- rope. In this context, it is very tempting to consider this discovery, together with the increasing corpus of artifactual ochre, bone tools, and evi- dence of diverse food sources, not only as the final tolling of the bell for the revolution model but also as a for- mal demonstration of the gradual “Out-of-Africa” theory. Such remote evidence of symbolic culture in south- ern Africa would conclusively corre- late the origin of cultural modernity and modern language with the origin of anatomically modern humans. It is possible, however, that accept- ing this view might provide only a part of the picture. Both these scenarios share the assumption that behavioral modernity arose only in a single spe- cies. They differ in that the first model sees cultural modernity as resulting from a sudden change within this spe- cies, and in a relatively small area. Also, the first model postulates a hu- manity that was biologically modern but not culturally modern until the “revolution” brought both moderni- ties together. According to the second model, biological and behavioral mo- dernity were inextricably linked, ad- Francesco d’Errico is a researcher at the Centre National de la Recherche Scienti- fique. His main research interest is the evolution of human cognitive abilities. He has published numerous papers on the early use of bone tools, the origin of sym- bolism, the emergence of behavioral mo- dernity, the Middle-Upper Paleolithic transition, grave goods associated with Paleolithic burials, Paleolithic systems of notation, bone taphonomy, and the appli- cation of new techniques of analysis to the study of Paleolithic art objects. He leads a multidisciplinary research project funded by the Centre National de la Re- cherche Scientifique on the origin of mod- ern humans and language and partici- pates in research projects with colleagues from France, the United States, England, South Africa, Spain, and Belgium. E-mail: [email protected] Key words: Middle Stone Age, Mousterian, Ne- andertals, symbolism, modern humans Evolutionary Anthropology 12:188 –202 (2003) DOI 10.1002/evan.10113 Published online in Wiley InterScience (www.interscience.wiley.com). Two contradictory theories of human cognitive evolution have been developed to model how, when, and among what hominid groups behavioral modernity emerged. The first model, which has long been the dominant paradigm, links these behavioral innovations to a cultural “revolution” by anatomically modern humans in Europe at around 40,000 years ago, coinciding with the first arrival of our species in this region. 1–4 According to this model, the sudden and explosive character of this change is demonstrated by the appearance in the archeological record of previously unseen carvings, personal ornaments, musical instruments, depictions on cave walls, and new stone and bone technology. A variant of this model sees behavioral modernity resulting from a rapid biological change, a brain mutation producing no apparent change in skull anatomy, which occurred in Europe or, more probably, in Africa at ca. 50,000 years ago. 5,6 188 Evolutionary Anthropology

Upload: nikola-kovacevic

Post on 01-Dec-2015

80 views

Category:

Documents


2 download

DESCRIPTION

D'Errico 2003 - The Invisible Frontier - A Multiple Species Model For

TRANSCRIPT

Page 1: D'Errico 2003 - The Invisible Frontier - A Multiple Species Model For

ARTICLES

The Invisible Frontier. A Multiple Species Model forthe Origin of Behavioral ModernityFRANCESCO D’ERRICO

The alternative scenario considersbehavioral modernity to be the out-come of a gradual process in Africawhere anatomically modern humansoriginated. This process is seen as cor-

responding to the technologicalchanges observed through the AfricanMiddle Stone Age.8–10 These changesstarted around 250,000 years ago, atthe very end of the Acheulean, andproceeded until the transition fromthe Middle Stone Age to the LaterStone Age, conventionally placed at40,000 years ago.

BLOMBOS AND THEAFRICAN EVIDENCE

Several recent discoveries seem toreinforce the second model. The moststriking is the recent publication oftwo ochre fragments (Fig. 1) from theMiddle Stone Age levels of Blombos,Cape Province, which bear similar en-graved geometric patterns;11 theyhave been dated by two differentmethods to ca. 77,000 years ago.These objects were found associatedwith a Still Bay industry with shapedbone awls and bifacial spearpoints9,12–13 (Fig. 2), a possible en-graving on bone,14 and with morethan 8,000 fragments of ochre, mostof which bear clear traces of use. Thepresence of symbolic engravings onartifactual pigment makes it unlikelythat the thousands of pigment frag-

ments found at Middle Stone Age siteswere strictly functional and suggestsinstead that they were used for sym-bolic purposes. Other engravings ornotched pieces are reported fromMiddle Stone Age contexts at south-ern African sites such as Klasies River,Apollo 11, Hollow Rock Shelter, Bor-der Cave, and Diepkloof.7,14

In other words, Blombos engrav-ings and bone tools may suggest thatsouthern African human populationshad already acquired behavioral mo-dernity 30,000 years before the ap-pearance of these innovations in Eu-rope. In this context, it is verytempting to consider this discovery,together with the increasing corpus ofartifactual ochre, bone tools, and evi-dence of diverse food sources, notonly as the final tolling of the bell forthe revolution model but also as a for-mal demonstration of the gradual“Out-of-Africa” theory. Such remoteevidence of symbolic culture in south-ern Africa would conclusively corre-late the origin of cultural modernityand modern language with the originof anatomically modern humans.

It is possible, however, that accept-ing this view might provide only a partof the picture. Both these scenariosshare the assumption that behavioralmodernity arose only in a single spe-cies. They differ in that the first modelsees cultural modernity as resultingfrom a sudden change within this spe-cies, and in a relatively small area.Also, the first model postulates a hu-manity that was biologically modernbut not culturally modern until the“revolution” brought both moderni-ties together. According to the secondmodel, biological and behavioral mo-dernity were inextricably linked, ad-

Francesco d’Errico is a researcher at theCentre National de la Recherche Scienti-fique. His main research interest is theevolution of human cognitive abilities. Hehas published numerous papers on theearly use of bone tools, the origin of sym-bolism, the emergence of behavioral mo-dernity, the Middle-Upper Paleolithictransition, grave goods associated withPaleolithic burials, Paleolithic systems ofnotation, bone taphonomy, and the appli-cation of new techniques of analysis tothe study of Paleolithic art objects. Heleads a multidisciplinary research projectfunded by the Centre National de la Re-cherche Scientifique on the origin of mod-ern humans and language and partici-pates in research projects with colleaguesfrom France, the United States, England,South Africa, Spain, and Belgium. E-mail:[email protected]

Key words: Middle Stone Age, Mousterian, Ne-andertals, symbolism, modern humans

Evolutionary Anthropology 12:188–202 (2003)DOI 10.1002/evan.10113Published online in Wiley InterScience(www.interscience.wiley.com).

Two contradictory theories of human cognitive evolution have been developedto model how, when, and among what hominid groups behavioral modernityemerged. The first model, which has long been the dominant paradigm, links thesebehavioral innovations to a cultural “revolution” by anatomically modern humans inEurope at around 40,000 years ago, coinciding with the first arrival of our speciesin this region.1–4 According to this model, the sudden and explosive character ofthis change is demonstrated by the appearance in the archeological record ofpreviously unseen carvings, personal ornaments, musical instruments, depictionson cave walls, and new stone and bone technology. A variant of this model seesbehavioral modernity resulting from a rapid biological change, a brain mutationproducing no apparent change in skull anatomy, which occurred in Europe or,more probably, in Africa at ca. 50,000 years ago.5,6

188 Evolutionary Anthropology

Page 2: D'Errico 2003 - The Invisible Frontier - A Multiple Species Model For

vancing together in a long and slowdialectic. In fact, both models are de-pendent on—I would say engenderedby—the “Out-of-Africa” model for thebiological origin of our species.15–17

However, if we see archeology as anindependent discipline, we should beable to assess issues that deal withcultural change on purely archeologi-cal grounds rather than through mod-

els shaped by current hypotheses ofhuman biological evolution. This is aspecial danger if we trust archeologi-cal evidence that is ambiguous or in-consistent with suppositions as towhich human species might havebeen involved. Because archeologicalremains are numerous and ubiqui-tous, while human remains are fewand rare, it is easy in this way acciden-

tally to prejudge the relationship be-tween cultural and biological traits.

This is why, from an archeologicalperspective, the equation of biologicalmodernity with cultural modernityshould be considered as no more thana working hypothesis that needs to betested against the archeologicalrecord. Following this approach, wemay find that a fourth scenario ismore consistent with the empiricaldata.18–20 In this scenario, the traitsthat define behavioral modernity arenot peculiar to our species and aroseover a long period among differenthuman types, including Neandertals.

HOW TO IDENTIFYBEHAVIORAL MODERNITY

What features define behavioralmodernity and how do we see them inthe archeological record? McBreartyand Brooks7 (Fig. 3) argue that mod-ern behavior is characterized by fourtraits concerning both adaptation andcognition. They also provide a list ofarcheological signatures that demon-strate the acquisition of these traits(Fig. 4).

This approach would be valuable if,in a sense, the “criteria to find thecriteria” were made explicit. One canexpect, for example, that a list of traitsmight result from a cross-culturalcomparative analysis21 of various hu-man societies and that the universal-ity of a selected trait would be thecriterion for its inclusion in the list.Otherwise one should accept thatother researchers having a differentcultural affiliation could propose fea-tures that they consider to define themodern experience, features whichwe should grant an equivalent weight.

Instead of following this path, theauthors have been inspired by MiddleStone Age material culture and, to alesser degree, the European Upper Pa-leolithic, to create a list of traits thatwould demonstrate the acquisition ofbehavioral modernity. The danger ofcreating a theory that fits one’s expec-tation seems evident to me in thiscase. Also, if we model the traits sug-gesting the modern character of a cul-tural system on the archeologicalrecord of a specific region or a partic-ular human type, what heuristic valuewill those traits have when they are

Figure 1. Engraved ochres from the Middle Stone Age levels of Blombos Cave with thetracings of the incised pattern (modified from Henshilwood and coworkers11). Top: scale �1 cm; bottom, scale � 5 mm.

ARTICLES Evolutionary Anthropology 189

Page 3: D'Errico 2003 - The Invisible Frontier - A Multiple Species Model For

used to test the modern versus non-modern characters of human popula-tions living in different environmentsand developing different evolutionarytrajectories? As I will show, when Iuse these criteria, questionablethough they are, to compare the Afri-can Middle Stone Age, where all thesemodern traits are supposed to havearisen, and the contemporary mate-rial culture left by Neandertals in Eu-rope and the Near East, no dramaticdifferences appear between the tworecords.

ECOLOGY AND SUBSISTENCESTRATEGIES

There is no reason to believe thatNeandertals were less well adapted totheir environment than were MiddleStone Age populations. This is demon-strated not only by their widespread

geographical distribution over Europeand the Near East, covering, fromnorth to south, several different eco-logical and biogeographical zones,but also by the fact that between300,000 and 30,000 years ago theywere able to pass successfully throughthree main glacial and three intergla-cial periods (OIS 9-3).22,23

Binford24 has argued that Neander-tals and Middle Stone Age hominidswere obligate scavengers who lackedthe ability to hunt large mammals.Stiner and Kuhn25 have argued in-stead that Neandertals were able tohunt but that they practiced opportu-nistic scavenging more regularly thanhas been documented for modern hu-mans and that the balance betweenthese two foraging strategies wasprobably determined by ecologicaland climatic constraints. According tothis reasoning, Italian Neandertals

were primarily scavengers at sitesdated before ca. 50,000 years ago andprimarily hunters at younger sites.However, zooarcheological studiesdone in the last decade have shownthat scavenging played little or no rolein Neandertal subsistence strategies.It has also become increasingly clearthat Neandertals were expert hunterswho could hunt a wide range of largemammals, including dangerous ani-mals such as bison, rhinos, and bear,and could concentrate, if necessary,on selected species. Marean and co-workers26–28 have shown that the“head-and-foot”-dominated faunal as-semblages, interpreted as evidence ofscavenging at key sites from Europeand southern Africa, are due to the

fact that the excavators either keptonly the most easily identified bonesor if they collected all of them, ig-nored some during the study of thefaunal material. Although Stiner29 hasdebated this point, Marean’s observa-tion probably remains valid for a largenumber of sites. Chase30 had alreadyargued that Neandertals were fully ca-pable of hunting and that there wasno evidence that they scavenged. Agrowing body of evidence now indi-cates that at sites dating between125,000 to 55,000 years ago, Neander-tal subsistence strategies were basedon hunting. At the ca. 200,000-year-old site of Biache-Saint-Vaast, Nean-dertals focused on adult bovids (70%

Figure 2. Bone spear points (left) and bone awls (right) from the Middle Stone Age levels ofBlombos cave (modified from Henshilwood and coworkers).

. . . if we see archeologyas an independentdiscipline, we should beable to assess issuesthat deal with culturalchange on purelyarcheological groundsrather than throughmodels shaped bycurrent hypotheses ofhuman biologicalevolution.

190 Evolutionary Anthropology ARTICLES

Page 4: D'Errico 2003 - The Invisible Frontier - A Multiple Species Model For

of the minimum number of individu-als), but also hunted large bears, prob-ably as they hibernated.31 At the siteof Mauran, the Neandertals killedhundreds of bisons, and processed thecarcasses on site.32 A similar prefer-ence for bovids is seen at sites such asLa Borde, Champlost, Coudoulous,and Wallertheim. At the open-air siteof Salzgitter-Lebenstedt, dated to ca.58,000 to 54,000 years ago,33 Nean-dertals systematically hunted rein-deer; they also killed many animalssimultaneously in a type of huntingthat seems not to differ from that ob-served at later Upper Paleolithic sitesor among sub-Arctic Inuit groups. Arecent study also has shown that EarlyUpper Paleolithic hunting did not dif-fer from Middle Paleolithic hunting inits degree of specialization.34 Studiesthat have tried to contrast the adapta-tions of Neandertals and anatomicallymodern humans in the Near East havereached similar conclusions: thesepopulations possessed broadly com-parable organizational abilities.35 Thetwo main subsistence strategies theyseem to have adopted, a “collector”land-use strategy involving lower res-idential mobility for the former andencounter-based hunting for the lat-ter, both occur among historically ob-served hunter-gatherers.

Coastal Middle Stone Age sitesshow an intense exploitation of ma-rine resources, which has been usedto suggest the varied and moderncharacter of these societies.7 Only a

few examples of the use of marine re-sources exist at Middle Paleolithicsites in Europe. The late Mousterianlevels of Figuiera Brava, Portugal,have yielded evidence of systematiccollection of seashells, particularlyMytilus and Patella.36 Another case ofNeandertal shellfish use, and probablyseal-hunting, comes from VanguardCave, Gibraltar.37

It is a fact, however, that evidenceof the exploitation of marine re-sources, with the exception of the useof sea shells for beads, is as scant inthe Upper Paleolithic as it is in theMiddle Paleolithic of Europe. Still,nobody has used this to argue againstthe modern character of Upper Paleo-lithic societies.

What about other inland resources?Stiner38 has shown that Neandertalswere able to diversify their diet and,

Figure 3. Traits defining behavioral modernity according to McBrearty & Brooks.7

Figure 4. Archeological signatures of behavioral modernity according to McBrearty &Brooks.7

ARTICLES Evolutionary Anthropology 191

Page 5: D'Errico 2003 - The Invisible Frontier - A Multiple Species Model For

particularly after 50,000 years ago,paid more attention to sources offoods such as lagomorphs and tor-toises. Fowling has been suggested forMousterian sites in Portugal, Spain,and Italy, among them FigueiraBrava, Cova Negra, Gorham’s Cave,Archi, Mochi, Fumane, and Castelci-vita. This interpretation is based onthe birds’ large size and habitat ratherthan on the occurrence of cut marksand burning, which means we cannoteliminate the possibility that theywere rock-nesting, cliff-roosting birdsthat died naturally or were accumu-lated by owls and mammalian carni-vores. However, micro-residues offeather and other avian tissues haverecently been found on stone toolsfrom the Late Mousterian levels ofStaroselie and the Streletskayan-likestone tools from Buran Kaya level C, alevel underlying a Kiik-Koba Mico-quian.39

In sum, while zooarcheological andisotopic analyses indicate that Nean-dertals derived most of their dietaryprotein from animal sources, they donot demonstrate that Neandertalswere obligate large mammal hunt-ers.40,41 This takes us to the contradic-tory nature of the criteria used to as-sess behavioral modernity fromsubsistence strategies. Once the obli-gate24 and opportunistic25 scavengerscenarios for Neandertal economiesare abandoned, little remains to dis-tinguish the subsistence strategies ofNeandertals and anatomically mod-ern humans. The same features areused for and against the modern char-acter of an economy depending onwhich actor is under scrutiny. The ex-ploitation of a wide range of resourcesis used to suggest increased dietbreadth when sub-Saharan hominidsare concerned while, in a clear contra-diction, the hunting of many mamma-lian species is taken to demonstratethe opportunistic character of Nean-dertal economies. The focus on one ora few species of large dangerous ani-mals is at once interpreted as attestingto the “modern” organization of ana-tomically modern hunters, capable of“specialized” hunting, and as demon-strating the biologically handicappedcognition of Neandertals, who couldnot incorporate more resources intotheir diet.

LITHIC TECHNOLOGY

Laminar technologies often havebeen considered to characterize theUpper Paleolithic, behavioral moder-nity, and higher cognitive abilities,while Levallois or Mode III technol-ogy is taken to indicate a lack of plan-ning capacities.1,3,42,43 However, it hasnow become clear that blades weresystematically produced at severalMiddle Paleolithic sites in Europe andthe Near East, as well as at AfricanMiddle Stone Age sites.40

In my view, three observationsemerge from the temporal and geo-graphic dispersion of blade technol-ogy. Blade production appears as apunctuated phenomenon in both theEurasian Middle Paleolithic and theMiddle Stone Age. After these periods,the production of blades did not be-come fixed in either continent, sinceother forms of debitage continued inboth areas. The flake-based assem-blages of the ethnographic AustralianAborigines, whose approach to stone-working was characterized by little or

no blade production, flourished intovery recent times.45 New World Paleo-Indians abandoned prismatic bladetechnology as soon as they left theArctic.46 Detailed technological analy-sis of blade assemblages indicates that“blade” is quite vague as an analyticalconcept: in terms of raw-materialchoice, preparation, and apprentice-ship, the making of standardizedbladelets by pressure-flaking has littleto do with the production of Aurigna-cian blades, which are thick, elon-gated, and trapezoidal in section, andwhich differ from the short, thinblades produced by the late Neander-tals to manufacture Chatelperronianpoints. Because “blades” may resultfrom very different reduction se-quences, in which different types ofraw material are chosen and differentsequences of motions are applied, theoccurrence of blade production in thearcheological record is better ex-plained as the expression of local tra-dition than as a reflection of cognitiveevolution. Certainly blades are not asimple diagnostic marker.

Partisans of the “Out-of-Africa” and“modern-behavior” model make muchof the production of standardized for-mal tool categories (Fig. 5) at the end ofthe Middle Stone Age and of geograph-ical and temporal variations in thosecategories.7,47 This modern trend wouldbe mainly exemplified by Still Bay bifa-cial points and by Howieson’s Poortsegments. Similar standardized toolcategories and similar technologicalvariation in time and space are com-mon in the Mousterian world. This isthe case for Levallois points and forMousterian of Acheulean Tradition bi-faces. During the last forty years, manyattempts have been made to explaintechnological and typological variabil-ity in the Mousterian in terms of dif-ferential function of sites, chronolog-ical differences between assemblages,resharpening of tools, and raw-mate-rial availability, but these have allfailed.2 Also, in both Europe and theNear East at the end of the Moust-erian we observe the same trends vis-ible in the Middle Stone Age towardincreased production of standardizedtool categories. Segments very similarto those associated with the Howie-son’s Poort in southern Africa7,8,47

were produced by Chatelperronian

Because “blades” mayresult from very differentreduction sequences, inwhich different types ofraw materials arechosen and differentsequences of motionsare applied, theoccurrence of bladeproduction in thearcheological record isbetter explained as theexpression of localtradition than as areflection of cognitiveevolution. Certainlyblades are not a simplediagnostic marker.

192 Evolutionary Anthropology ARTICLES

Page 6: D'Errico 2003 - The Invisible Frontier - A Multiple Species Model For

Neandertals in France and Spain (Fig.5) and by Uluzzian Neandertals in It-aly and Greece.18,48 The Szeletian andother east European late Middle Pa-leolithic technocomplexes, which de-veloped locally from regional Moust-erian traditions,49 are characterizedby thin bifacial points that recall thosefound at South African Still Bay sites(Fig. 5). Again, we see no clear differ-ence between supposedly modernMiddle Stone Age behavior and thebehavior of the Neandertals.

HAFTING AND RELATEDCRAFT SKILLS

It has been argued that the Moust-erian was an immediate technologyinvolving a low degree of conceptual-ization.1,3,42 However, the discovery ofsix wooden spears at the Lower Paleo-lithic site of Schoningen, Germany,

dated to ca. 400,000 years ago, con-firms what was already known fromClacton and Lehringen: Middle Pleis-tocene hominids shaped wood withspecific techniques and producedspears for hunting.50,51 The finding offour wooden hafts in another part ofSchoningen also shows that compos-ite tools were used in the middle partof the Middle Pleistocene (OIS 11).

We know that hafting technologywas practiced in the Middle Paleo-lithic of Eurasia. A convergentscraper, three Levallois flakes, and acortical flake with traces of bitumenadhesive used for hafting have beenfound in Mousterian levels dated toabout 60,000 years ago at the site ofUmm El Tlel, in the El Kowm Basin,Syria.52 A blade from the Hummalian(Middle Paleolithic) levels at Hum-mal, in the same region, bears similartraces. Direct evidence of stone-tipped

spears also comes from Umm El Tlel,which produced a Levallois point em-bedded in the third cervical vertebraof a wild ass.53 Two birch-bark pitcheshave recently been found at the Mid-dle Paleolithic site of Konigsaue, Ger-many. These pieces came from twodifferent layers dated to 43,800 and48,400 years ago. One shows the im-print of a wooden haft, the other abifacial tool.54 A probable bone han-dle made of a horse metapodial hasalso been found in level C of BuranKaya III, in Crimea.55 The pre-Aurig-nacian industry from this level con-sists of clusters of chips and smallflakes derived from bifacial shapingand thinning, associated with pre-forms and bifacial points recallingthose of the Streletskaya culture.56

The maker of these objects is still un-known, but it is noteworthy that levelC underlies level B1, which contains a

Figure 5. Backed pieces and foliates associated with the late Southern African Middle Stone Age and transitional technocomplexes inEurope.

ARTICLES Evolutionary Anthropology 193

Page 7: D'Errico 2003 - The Invisible Frontier - A Multiple Species Model For

Kiik-Koba Mousterian dated to 28,600B.P. The handle has been directlydated by C14 AMS to 32,350 � 650(OxA-6869), which is well before theoldest C14 dates for the Crimean Au-rignacian.

In sum, the available evidence pro-vides no reason to believe that haftingand the use of composite tools wasdeveloped only in the Middle StoneAge. Consistent, if not more evidenceof this technology is found in Moust-erian sites from Europe and the NearEast.

BONE AND IVORY WORKING

Evidence of worked and, in somecases, decorated bone awls comesfrom Chatelperronian and Uluzziansites in France and Italy.18 But didNeandertals produce an organic speartechnology before these late stages?Villa and d’Errico,57 in an analysis ofTorralba and Ambrona ivory pointsformerly interpreted as shaped orused tools,58 found that these tuskfragments are the result of naturalphenomena. These authors also haveshown that a natural explanationmust be favored for other bone andantler points reported from Moust-erian sites such as Vaufrey, CombeGrenal, Camiac, and Pech-de-L’Aze I,and probably for those from othersites the authors were unable to study.On the other hand, a recent experi-mental study by Schmitt, Churchill,and Hylander,59 indicates that thethrusting spear was one of the princi-pal sources of strength asymmetry inboth Neandertal and Early Upper Pa-leolithic modern human male humeri.Why, then, did Neanderthals usewooden spears and stone-tippedspears but not organic spear tips? Thereason may lie in the different type ofpredatory tactics used by Middle andUpper Paleolithic hunters.57,60

Upper Paleolithic bone and stonespear-tips are different from MiddlePaleolithic stone points in their aero-dynamic properties and in theamount of kinetic energy they carry atimpact. Even when carefully shapedby retouch, Middle Paleolithic stonespear-tips have a large, thick base, im-plying a large, heavy shaft. This kindof javelin has a low velocity but highpenetration power at short distances.Similar organic points could not pen-

etrate deeply into the flesh of largeanimals because of the softer natureof the bone material. In contrast,stone and bone spear-tips used by Up-per Paleolithic hunters are similar inthat both types are thin, straight, andlight: they are made to travel at highspeed and to be cast from afar. Thisallows them to penetrate the animalbody and injure vital organs. Thus,while both Neandertals and anatomi-cally modern humans relied on thrust-ing spears before the appearance oflong-range projectile weapons, (spear-throwers) during the early part of the

Late Upper Paleolithic, Neandertalsmay have preferred more robust andheavy weapons for closer-range hunt-ing.

What impact did the production offormal bone tools by Middle StoneAge people have on their way of life?Analysis of the twenty-eight bonetools from Blombos Middle Stone Agelayers indicates that 90% of these ob-jects are awls that were made on dif-ferent types of long-bone shaft frag-ments, shaped by scraping, and then

used at the site to pierce soft materialsuch as leather.12 Three points shapedby scraping and then completely fin-ished by careful polishing probablywere projectile points made for haft-ing. These points might have been pol-ished to increase their penetrationpower and to give them a distinctiveappearance—an “added value”—to re-flect the distinct contexts of use of thetwo tool categories and the differentsocial roles of their users.

However, one may argue that bonetools are a possible, but not obligatoryoutcome of the acquisition of moderntraits. Relatively few Middle StoneAge sites have been excavated by mod-ern standards in comparison with themany Middle and Upper Paleolithicsites in Europe. But does this differ-ence fully explain why only a handfulof bone tools has been found in south-ern Africa? There are bone harpoonsfrom Katanda61–63 for which the min-imum date of 90,000 years ago re-mains controversial;6 one quite doubt-ful point from Mumbwa cave;64 a fewnotched and marked objects7,14,65

from Klasies River and Apollo 11; andone clear point from Peers cave.Singer and Wymer65 described a bonepoint from a disturbed context at Kla-sies River. Its shape is very similar tothat of Later Stone Age points, and itprobably should be attributed to amore recent period.12,57 In sum, whilewe have at Blombos clear evidence ofbone tools produced by a varied rep-ertoire of techniques and designed fordifferent purposes, we cannot use thisevidence to support either the hypoth-esis that the Middle Stone Age as awhole was characterized by system-atic production of formal bone toolsor, because little is known about theevolutionary significance of bone toolshaping the notion that the produc-tion of formal tools gives MiddleStone Age material culture a modernappearance.

THE COGNITIVE ABILITIES OFNEANDERTALS AND THEIR

RELATIONSHIP TO MODERNHUMANS IN EUROPE

In the last few years, studies havebegun to cast new light on the intel-lectual abilities of these predecessorsand the chronology of their contact

. . . one may argue thatbone tools are apossible, but notobligatory outcome ofthe acquisition ofmodern traits. Relativelyfew Middle Stone Agesites have beenexcavated by modernstandards in comparisonwith the many Middleand Upper Paleolithicsites in Europe. But doesthis difference fullyexplain why only ahandful of bone toolshas been found insouthern Africa?

194 Evolutionary Anthropology ARTICLES

Page 8: D'Errico 2003 - The Invisible Frontier - A Multiple Species Model For

with the first anatomically modernhumans to colonize Europe. Debateon this issue has generally taken forgranted that people like us arrived inthe far west of Europe from the eastabout 40,000 years ago. According tothat assumption, Cro-Magnon popu-lations, carrying a lithic and bonetechnology called Aurignacian, wouldhave triggered, through acculturation,the appearance of a new lithic tech-nology, ornaments, decorated objects,and bone tools among some late Ne-andertal groups such as the Chatelper-

ronians in France and Spain or theUluzzians in Italy.1,2,4 The manufac-ture of personal ornaments and bonetools by Neandertals has been a con-troversial issue; many researcherspreferred to attribute the presence ofsuch objects in the Chatelperronianlayers of sites like the Grotte duRenne, France, to a reworking of ar-cheological layers incorporating Au-rignacian artifacts, the Neandertalcollection of objects manufactured bymodern neighbors, or an actual trans-fer of objects to Neandertals through

trade. Recent reassessments of the ev-idence have shown that Chatelperro-nian Neandertals themselves madethe personal ornaments and bonetools from the Grotte du Renne, asdemonstrated by the presence, in thesame layers, of refittings and manu-facturing by-products.18,19 This find-ing is reinforced by a new study of thefifty Chatelperronian and nine Aurig-nacian bone awls (Fig. 6) found in theChatelperronian and Aurignacian lev-els of the Grotte du Renne.60 If thebone tools found in the Chatelperro-

Figure 6. Bone awls from the Chatelperronian layers of the Grotte du Renne at Arcy-sur-Cure, France (modified after d’Errico et al. in press).

ARTICLES Evolutionary Anthropology 195

Page 9: D'Errico 2003 - The Invisible Frontier - A Multiple Species Model For

nian layers had originated in the over-lying Aurignacian level, we would ex-pect their number to decline withdepth, but the opposite is true. Thelowest of the three Chatelperronianlayers yielded four times the numberof awls found in the Aurignacian ho-rizons of the site. The tools from thetwo cultural horizons also show a dif-ferent spatial distribution: in theChatelperronian layers the bone toolsare concentrated inside a circle ofstone located in the northwestern partof the excavated area, while the fewAurignacian awls were found in thesoutheastern part. The Chatelperro-nian tools also vary more in blanktypes, methods of blank production,and degrees of shaping. Nine Chatelp-erronian tools are marked with sets ofnotches or v-shaped motifs. Only oneAurignacian piece bears a decoration,a set of crosses. The presence of delib-erate decorations suggests that sym-bolism, rather than restricted to a fewobjects obtained through exchange,played a role in the domestic aspectsof group life. Bone tools are not iso-lated occurrences. Awls also comefrom the Chatelperronian level ofQuincay, Charente (Leveque, personalcommunication) and from several Ul-uzzian sites from Italy such as Cavalloand Castelcivita.18

The lithics tell the same story.18

While revealing the use of new knap-ping techniques and tool types, lateNeandertal technologies from differ-ent European regions show no affini-ties with the technology that modernhumans introduced into Europe; in-stead they appear as independent de-velopments from local traditions. Thechronological precedence of the Au-rignacian over these regional Nean-dertal entities, which is fundamentalto their interpretation as the outcomeof prolonged contact, is also calledinto question. Wherever archeologicallayers of both cultures are representedat the same site, the Chatelperronianalways underlies the Aurignacian,suggesting its priority. The only ex-ceptions are two instances of inter-stratification between Aurignacian,and Chatelperronian, which now havebeen formally rejected by reappraisalof these sites.67 Similarly, analysis ofall the radiometric dating shows thatthe earliest occurrences of a diagnos-

tic Aurignacian in Western Europe isno older than ca. 36,500 BP. The avail-able evidence suggests that at thattime late Neandertals were already de-veloping their own transition to theUpper Paleolithic, the Chatelperro-nian, and that this, like other late Ne-andertal cultures, emerged before anymodern humans became establishedin the same areas. This indigenous de-velopment may have included themanufacture and use of symbolic ob-jects created for visual display on thebody. The alternative hypothesis,

which I and others have proposed, isthat it was precisely the new situationinvolving contact between anatomi-cally modern people and Neandertalsand the consequent problems of cul-tural and biological identity that stim-ulated an explosion in the productionof symbolic objects on both sides.18,19

BURIALS

Thirty-five of the fifty-eight knownMiddle Paleolithic putative burials

from Europe and the Near East be-long to Neandertals.68,69 Contestingthe opinion1,70,71 that they must be in-terpreted as accidental depositions ofdead bodies would require a detaileddiscussion of each case, and that isbeyond the scope of this review. How-ever, while I can easily accept thatmost of these human remains, all ex-cavated long ago, lack the informationrequired today to assess the humanorigin of the inhumations, there is agrowing consensus on the existence ofNeandertal burials. This depends inpart on the observation72 that com-plete animal skeletons are so rare inNear Eastern and, one may add, Eu-ropean caves, as to make it unlikelythat interments were produced bycauses other than cultural processes.Also, field observations (for example,at Kebara) on the more recently foundburials help confirm the reality of thisphenomenon.73,74 In sum, manywould agree that there is enough evi-dence to believe that both anatomi-cally modern humans and Neander-tals buried their dead well before theLater Stone Age or Upper Paleolithic.

Should the Neandertal burials beseen as symbolic in nature? The onlydifference we see between the burialsof these populations is that the use ofgrave goods is apparently well estab-lished for anatomically modern peo-ple but remains controversial with re-spect to the Neandertals. Also, therecent rediscovery of the Neandertalneonate from Le Moustier, the re-mains of which were previously as-signed to La Ferrassie LF4, eliminatesfrom the literature the only known ex-ample of a Neandertal double buri-al.75 Although for these reasons thefunerary practices of anatomicallymodern humans may appear to bemore complex than those of contem-porary Neandertals, this cannot beused to suggest that the latter investedless consciously in symbolism. Manyfully symbolic ethnographic and evenmodern societies bury their dead insingle tombs with few or no durablegrave goods. The idea that these dif-ferences result from “incompletelearning” of this behavior by Neander-thals who obtained it through contactwith anatomically modern humans inthe Near East, supported by the dat-ing of ca. 100,000 for the Qafzeh buri-

. . . that the earliestoccurrences of adiagnostic Aurignacianin Western Europe is noolder than ca. 36,500 BP.The available evidencesuggests that at thattime late Neandertalswere alreadydeveloping their owntransition to the UpperPaleolithic, theChatelperronian, andthat this, like other lateNeandertal cultures,emerged before anymodern humansbecame established inthe same areas.

196 Evolutionary Anthropology ARTICLES

Page 10: D'Errico 2003 - The Invisible Frontier - A Multiple Species Model For

als, is now contradicted by the ther-moluminescence dating of Tabunlayer C to 160,000.76 This could makethe Neandertal burial therein the old-est known burial in the world or, ifone takes a minimalist view of the ev-idence, a burial as old as the oldestanatomically modern burials found inthe region.77

Again, we see no clear-cut indica-tions that anatomically modern hu-mans were culturally more “ad-vanced” than Neandertals. In fact,burials produced by both humantypes appear to resemble each othermore than they do burials dated to theUpper Paleolithic of Europe becauseof the virtual absence of ochre, bonetools, and personal ornaments. Manyfragments of ochre bearing traces ofuse and ochre-stained stone toolscome from the Qafzeh layers thatyielded the burials,78 but none ofthem were found in clear associationwith the skeletons.

COLOR

There is no traditional modern so-ciety in which the production and useof colorant is merely functional. Re-cent excavations at Kapthurin in Ke-nya79 and Twin Rivers in Zambia10,80

have yielded convincing proof thatcolorants were systematically used asfar back as the Acheulean to the Mid-dle Stone Age transition, ca. 200,000years ago. At GnJh-15 in the

Kapthurin Formation, more than sev-enty pieces of red pigment weighingmore than 5 kg altogether are associ-ated with an early Middle Stone Ageassemblage dated to 285,000 yearsago. Field work at Twin Rivers pro-duced 176 colorant fragments in lay-ers dated between 260,000 and400,000 years ago. Five different col-ors and traces of their use are attested.Layers dated to 200,000 years agoyielded 132 pieces of colorants. Thediscoveries at these two sites confirmisolated occurrences of red pigmentssignaled in the past from sites such asNooitgedacht in South Africa, Kabwein Zambia, and Charama in Zimba-bwe. Watts81 has shown that the useof red pigments increased during theMiddle Stone Age and became a con-sistent feature of MSA2b/Still Bay,Howieson’s Poort, and Late Stone Agesites. Middle Stone Age people pre-ferred strong red colorant even whenyellowish or yellowish-brown mate-rial of similar chemical compositionwas available. This choice arguesagainst a purely functional interpreta-tion of pigment use. Many MiddleStone Age colorants are shaped likecrayons, suggesting that they wereused to trace lines on soft material likeleather or to paint the body. Addi-tional evidence indicating that thespread of pigment use was not limitedto sub-Saharan regions comes fromthe Wadi Sodmein cave in Egypt,

where Van Peer and Vermeersch82

have reported a Middle Paleolithic as-semblage with Levallois debitagedated to ca. 115,000 years ago inwhich there is a Levallois flakemarked on both sides by a continuousred line perpendicular to the mainflake axis (Fig. 7).

Is pigment use exclusively a MiddleStone Age attribute? Neandertals, too,used colorants. Black pigments,mostly manganese dioxides, and to alesser extent fragments of ochre, comefrom at least seventy layers excavatedat forty Neandertal sites in Eu-rope.83,84 While most of the sites withpigments represent the CharentianMousterian and Mousterian ofAcheulean Tradition and date fromOIS 3 (60,000 to 35,000 years ago),evidence of pigment use in Europe ex-tends back to the Acheulean. The rich-est collection, comprising 451 colo-rant fragments and grinding stones,comes from the Mousterian of Acheu-lian Tradition levels of Pech-de-l’Aze I,dated to ca. 60,000 to 50,000 yearsago.84,85 Most of these pigments showtraces of use in the form of abrasionfacets like those visible on MiddleStone Age colorants; some also showevidence of use as crayons. There is noconspicuous difference between pig-ment use at Pech-de-l’Aze I and atMiddle Stone Age sites in terms of theweight, number, and proportion ofpigments used. Moreover, Pech-de-l’Aze I provides evidence of more in-tensive use of pigment than does anyMiddle Stone Age site except Blom-bos. Evidence of systematic use of pig-ment by Neandertals has recentlybeen reported from the Cioarei cave inRomania, where eight subcircularochre containers made of stalagmitefragments, associated with fifty-fiveochre fragments, were found in aMousterian layer older than 50,000BP.86

In sum, while the widespread oc-currence of pigments at Upper Pleis-tocene Middle Stone Age sites is animportant phenomenon that, follow-ing the discovery of Blombos engrav-ings, probably reflects the growingrole of symbolic activities in thesecommunities, evidence from Euro-pean Middle Paleolithic sites gives noreason to believe that Neandertalswere not using pigments in compara-

Figure 7. Levallois flake from the Middle Palaeolithic level 4 at Wadi Soldmein Cave showingan ochre band on both faces. Hatched pattern on dorsal face indicates a calcite crust. Onthe ventral face, note a circular inclusion of different color in the chert (after Van Peer,Vermeersch82).

ARTICLES Evolutionary Anthropology 197

Page 11: D'Errico 2003 - The Invisible Frontier - A Multiple Species Model For

ble activities. The chronological attri-bution of the older pigments from Af-rica (Kapthurin, Twin Rivers) andtheir association with Lupembanstone tools seem to indicate that theuse of pigments originated with Homoheidelbergensis or archaic Homo sapi-ens. If colorant use is taken as an ar-cheological indication of symbolic be-havior, then the origin of theseabilities, traditionally attributed toanatomically modern humans, has tobe considered more ancient. This isclearly consistent with the hypothesisI offer here: these abilities did not nec-essarily emerge in a single species; amodel for their origin must be primar-ily archeological and independent ofbiological scenarios for the origin ofour species.

ABSTRACT AND DEPICTIONALREPRESENTATIONS; PERSONAL

ORNAMENTS

Little evidence of abstract or depic-tional representations exists in theMiddle Stone Age.7 Apart from the ob-jects I mentioned at the beginning ofthis paper, the only evidence of depic-tions are the painted slabs with ani-mal figures from the Apollo 11 site,Namibia, found in a level overlyingassemblages with Howieson’s Poortaffinities. The chronological attribu-tion of this level, however, is unclear.The young age suggested by radiocar-bon dating, 26 to 28,000 years ago,seems to contradict the cultural attri-bution of the assemblage. No personalornaments are known at Middle StoneAge sites. The oldest traces of theirmanufacture come from the Kenyansite of Enkapune ya Muto, dated to40,000 years ago,87 and from the siteof Ntuka River 3, GvJh11, associatedwith a 29,975 year old Late Stone Agemicroblade industry (Ambrose, per-sonal communication). Perforatedand ochre-stained Glycymeris shellsapparently were found at Qafzeh inassociation with early modern humanburials dating to 90,000 to 100,000years ago; the Mousterian levels ofSkhul may have yielded similarshells.88 However, as long as no de-tailed publication of this material isavailable it is difficult to evaluate theevidence. In fact, the oldest evidenceof the production of personal orna-

ments comes from the Early UpperPaleolithic levels of Ucagizli, Turkey,and perhaps from the contemporarylayers at Ksar’ Akil, Lebanon. Recentexcavation at the former site by Kuhnand colleagues89 has yielded a largenumber of perforated marine shells ofdifferent species from levels dated toca. 39 to 41,000 years ago. The stone-tool assemblage associated with thebeads is characterized by pointedblades and small end-scrapers; itshows no Aurignacian affinities. Themakers of these shell beads are stillunknown and, considering the age ofthe layers, might well have been Ne-andertals. A morphologically modernchild was found in the Aurignacianlayers of Ksar’ Akil, but the layer fromwhich the modern remains come isdated to ca. 29,000 years ago90 andoverlies the Ahamarian layers inwhich the oldest beads occur.

It is difficult to establish whether orwhen Neandertals or earlier hominidsproduced deliberate engravings orused personal ornaments becausemany objects that have been de-scribed as such are actually the resultof natural phenomena.91 This is thecase with the Pech-de-l’ Aze II rib andseveral purported engraved bonesfrom Cueva Morin, Stranska Skala,and other sites. It is also true of per-forated bones from Pech-de-l’Aze,Bois Roche, Kulna, Bockstein-schmide, and Repolusthohle, as wellas putative musical instruments likethe “flute” from the Slovenian site ofDivje Babe. The putative engravingsare blood-vessel impressions, the pu-tative pendants are actually bone frag-ments regurgitated by hyenas, and theperforations on the so-called flute arepunctures produced by cave bears.92

Although no reports have been pub-lished regarding some of them, a lim-ited number of bone and stone objectsfrom Acheulean and Mousterian sitesin Europe and the Near East do seemto bear deliberate engravings in theform of sequences of more or less par-allel incisions. These include the well-known mammoth shaft fragmentfrom Bilzingsleben, which has an en-graved fan-like motif; the Tata “en-graved” nummulite and polishedmammoth dental plate; the parallellines on the Temnata slab; parallel in-cisions on bone or antler from Ermit-

age, Ferrassie, Vergisson IV, and Vau-frey; shaft fragments with dozens ofparallel lines from the late Mousterianlevels of the French sites of Unikote,La Chapelle-aux-Saints, and Marillac;and shaft fragments with criss-crosspatterns from Peyrere.

We also have seen an increase in thenumber of sites dated to ca. 35,000 to40,000 years ago and located in peri-arctic regions that have yieldedMiddle Paleolithic or “transitional”stone-tool industries associated withsequentially notched bone and ivoryworking.93 At the moment, nothingdemonstrates that these assemblageswere not produced by Neandertals. Inspite of the consistent, albeit discon-tinuous, presence of anatomicallymodern humans, near Eastern Mous-terian sites are as spare as those fromEurope in potentially symbolic ob-jects. The only two examples are, sofar, a flint cortex engraved with a setof concentric lines, which was foundat Quneitra94 in a level dated to ca.60,000 years ago, and another cortexwith a set of parallel incisions, whichwas found at Qafzeh in the same levelsas the burials.95 The only evidencefrom the Near East that might offersymbolic expressions from periodscomparable with the most ancient Af-rican evidence of systematic use ofpigments is the so-called BerekhatRam figurine.7,79–81 This piece comesfrom an Acheulean layer sandwichedbetween two well-dated volcanic de-posits, indicating that the humanpresence at the site is older than230,000 years and probably lies be-tween 250,000 and 280,000 yearsago.96–97 A recent microscopic analy-sis has shown that the object was pur-posely modified by humans, but this,of course, does not demonstrate itssymbolic nature.98

It is noteworthy that most of themore convincing “representations”found in Europe come from relativelylate Neandertal sites, indicating an in-crease in the production of possiblesymbolic objects more or less at thesame time as the increase in the pro-duction of engravings and pigmentuse in the southern African MiddleStone Age. We see no difference in thefrequency and nature of such objectsbetween Europe and the Near East,where anatomically modern humans

198 Evolutionary Anthropology ARTICLES

Page 12: D'Errico 2003 - The Invisible Frontier - A Multiple Species Model For

were present since at least 100,000years ago. As with the Middle StoneAge, very little evidence exists for theuse of personal ornaments by Nean-dertals at sites before the end of theMousterian. At the end of the Nean-dertal period, however, Neandertalsdid produce different types of per-sonal ornaments and decorated bonetools with sets of notches. This ap-pears clearly at the Grotte du Renne.18

Personal ornaments also come fromthe Chatelperronian level of Quincayin the Charente.99 Shell beads havealso been reported from several Uluz-zian sites in Italy such as Cavallo andCastelcivita.18 Several small rodsmade of hare and wolf long bones thatbear no trace of use as tools and thatmight well have been used as pen-dants come from the level C of BuranKaya III in the Crimea.55 (See the sec-tion on hafting and related craft skillsfor the context and dating of theseobjects). This review of the evidencemakes it clear that, with the possibleexception of the shell beads fromQafzeh, there is no convincing proofof the use of personal ornaments be-fore 40,000 years ago and no dramaticdifference in the amount and natureof depictional and abstract represen-tations between the archeologicalrecords produced by the two humantypes.

SINGLE-SPECIES OR MULTIPLE-SPECIES ORIGIN OF

BEHAVIORAL MODERNITY

The application of the criteria usedso far to identify behavioral moder-nity in the material culture of Nean-dertals and contemporary anatomi-cally modern humans does not seemto support the single-species or single-population model for the origin ofthese modern traits. Neandertal sub-sistence strategies and technologicaland symbolic traditions do not signif-icantly differ from those of contempo-rary human populations in Africa andin the Near East. Submitted to closescrutiny, comparable evolutionarytrends common to the two geograph-ical areas may be detected in a num-ber of domains (Fig. 8). It may be ar-gued that it is not the mere presenceof “advanced” behaviors that matters,but rather the frequency, consistency,

and association of those behaviors.Considering, for example, that rela-tively few southern African MiddleStone Age sites have been excavatedas compared with the number ofMousterian sites in southwesternFrance, the association at Blombos ofbone tools, engravings, and a largequantity of pigments can be inter-preted as highly significant. One cancounter, however, that none of theother eleven South African MiddleStone Age sites with fauna have pro-vided bone tools like those found atBlombos. And if we accept that therelevant comparison is between thewhole Mousterian in Europe and thewhole Middle Stone Age in Africa,there are at least sixty Middle StoneAge sites throughout Africa that couldbe expected to show what Blombosdoes, and none do. Although there aremany more excavated sites in Europe,there are enough in Africa to showthat “modern” markers are no morecommon in the Middle Stone Age thanthey are in the Mousterian of Europe.

Also, it is dangerous to equate thefrequency of a type of archeologicalmaterial with its ancient social signif-icance. The amount of pigment recov-ered in an excavation depends on ta-phonomic factors, including thetechnique used to prepare the pig-ment, the media on which the pig-ment was applied, and the frequencyof the activities in which pigmentswere used. The presence of used pig-ment indicates that other colorantsmay also have been used but did notsurvive archeologically. Thus, can thepresence of pigment at “only” fortyMousterian sites be used to suggestthat symbolic activities were less im-portant in Europe than in Africa? Thesame applies to most of the othermodern traits. We can note only thatsome of them, such as burials, “for-mal” tools, hafting, blade technology,and even pigments, are clearly presentin the two records, and well before theappearance of transitional techno-complexes such as the Chatelperro-nian. Other traits such as deliberatemarkings are rare in both records.Still others, such as personal orna-ments and geometric stone tools,seem to appear only at a relatively latestage in both records, while depic-tional images are virtually absent

from both. Some behavioral innova-tions, such as a bone-tool technology,seem to appear in the Middle StoneAge of Africa before they do in theMousterian of Europe, but their usedoes not seem to have been wide-spread. Also, the chronology of thecultural entities associated with thesenew behaviors on both continents isstill uncertain. Some “transitionaltechnocomplexes” from Europe, tra-ditionally considered to postdate thebeginning of the Aurignacian, are nowconsidered to be older and so arecloser in age to some Middle StoneAge assemblages.19 And even if a moreprecise chronology does demonstratethat Middle Stone Age people had pri-ority in some of these innovations,should this be taken as supporting asingle-species model for the origin ofbehavioral modernity? It is possiblethat the more precocious appearanceof some of these traits in Africa wasfundamentally a question of that con-tinent’s greater size and larger humanpopulation,100 which created greateropportunities for innovations to de-velop and survive.101

On the other hand, the traits used toidentify behavioral modernity are nomore than a list of the major archeo-logical features that characterize theUpper Paleolithic in Europe. Theproblem is that this behavior is highlyderived within Homo sapiens. It doesnot consistently characterize the be-havior of the earliest Homo sapienspopulations nor does it appear inmany parts of the world (much of Af-rica, most of Asia, all of Australia) un-til long after its appearance in Europeca. 40,000 years ago. Rather than ac-cept that this complex of behaviorsreflects adaptive strategies that wereunique to the problem of colonizingEurope, many archeologists cling tothe notion that the course of humanbehavioral evolution can be modeledin terms of a simple progression fromarchaic to modern behavior.

We might reconstruct the conceptof modernity encompassing behav-ioral universals among ethnohistorichumans and excluding behavioralconvergences revealed by primateethology. However, this would resultin the exclusion from “modernity” ofmany recent Homo sapiens archeolog-ical populations and, as I have shown

ARTICLES Evolutionary Anthropology 199

Page 13: D'Errico 2003 - The Invisible Frontier - A Multiple Species Model For

here, the inclusion of various Nean-dertal or earlier hominid populations.After that, could we continue to usemodernity as a metaphor for the hu-man condition? Behavioral modernityhas been a useful concept to highlightthe inconsistencies of the revolutionmodel, but we now need to go further.

“Modern” traits may have appearedin different regions and among differ-ent groups of humans, much as hap-

pened later in history with the inven-tions of agriculture, writing, and statesociety. Two hypotheses, which arenot mutually exclusive, may explainboth convergences and differences be-tween the two populations on which Ihave focused here. The first is that thetwo populations reacted in compara-ble ways to comparable ecologicalpressures. The other is that, as theirsimilar lithic technology in the Near

East suggests, cultural barriers, andperhaps biological ones, betweenthese populations were permeable.The limited amount of fossil DNAavailable seems to indicate17,101 thatdifferences between Neanderthals andrecent humans were of the order oftwo or three times those found withinrecent humans. But even in this case,the data can be used to support theplacement of Neandertals and recent

Figure 8. Occurrence of “modern” traits in the African, Near Eastern, and European archeological records. Interrupted lines indicatediscontinuous presence.

200 Evolutionary Anthropology ARTICLES

Page 14: D'Errico 2003 - The Invisible Frontier - A Multiple Species Model For

humans either in the same species orin different ones, given the recent or-igin of common ancestry. And Europewas, at all times, a cul de sac.Handaxes arrived there one millionyears after their invention in Africa,and agriculture, in some areas of Eu-rope, 7,000 years after its invention inthe Near East. This demonstrates thatwe do not need to assume differentcognitive abilities to explain gaps inthe appearance of some behaviors inthe two populations.

ACKNOWLEDGMENTS

I am grateful to Richard Klein forinviting me contribute this paper andfor his stimulating critical reading ofthe manuscript. I also thank Christo-pher Chippindale, Paola Villa, JohnFleagle, and two anonymous refereesfor their constructive comments onearlier drafts. This work was pro-duced in the framework of the CentreNational de la Recherche Scienti-fique/European Science Foundationprogram “Origin of Man, Languageand Languages.”

REFERENCES

1 Stringer CB, Gamble C. 1993. In search of theNeandertals. London: Thames & Hudson.2 Mellars P. 1996. The Neandertal legacy: an ar-chaeological perspective from Western Europe.Princeton: Princeton University Press.3 Mithen SJ. 1996. The prehistory of the mind: asearch for the origins of art, religion, and science.London: Thames & Hudson.4 Bar-Yosef O. 1998. On the nature of transi-tions: the Middle to Upper Paleolithic and Neo-lithic Revolution. Cambridge Archaeol J 8:141–163.5 Klein RG. 1999. The human career, 2nd ed.Chicago: Chicago University Press.6 Klein RG. 2000. Archaeology and the evolutionof human behaviour. Evol Anthropol 9:17–36.7 McBrearty S, Brooks AS. 2000. The revolutionthat wasn’t: a new interpretation of the origin ofmodern human behavior. J Hum Evol 39:453–563.8 Deacon HJ. 1989. Late Pleistocene palaeoecol-ogy and archaeology in the southern Cape, SouthAfrica. In: Mellars P, Stringer CB, editors. Thehuman revolution: behavioural and biologicalperspectives on the origins of modern humans.Princeton: Princeton University Press. p547–564.9 Henshilwood CS, Sealy JC. 1997. Bone arte-facts from the Middle Stone Age at BlombosCave, southern Cape, South Africa. Curr An-thropol 38:890–895.10 Barham LS. 1998. Possible early pigment usein south-central Africa. Curr Anthropol 39:703–710.11 Henshilwood CS, d’Errico F, Yates R, JacobsZ, Tribolo C, Duller GAT, Mercier N, Sealy JC,Valladas H, Watts I, Wintle AG. 2002. Emergence

of modern human behavior: Middle Stone Ageengravings from South Africa. Science 295:1278–1280.

12 Henshilwood CS, d’Errico F, Marean C, MiloR, Yates R. 2002. An early bone tool industryfrom the Middle Stone Age at Blombos Cave,South Africa: implications for the origins of mod-ern human behaviour, symbolism and language.J Hum Evol 41:631–678.

13 d’Errico F, Henshilwood CS, Nilssen P. 2001.An engraved bone fragment from 70,000-year-oldMiddle Stone Age levels at Blombos Cave, SouthAfrica: implications for the origin of symbolismand language. Antiquity 75:309–318.

14 Wurz S. 2000. The Middle Stone Age at Kla-sies River, South Africa. Ph.D dissertation, Uni-versity of Stellenbosch.

15 Cann R, Stoneking M, Wilson A. 1987. Mito-chondrial DNA and human evolution. Nature 325:31–36.

16 Ingman M, Kaessmann H, Paabo S, Gyllen-sten U. 2000. Mitochondrial genome variationand the origin of modern humans. Nature 408:708–713.

17 Stringer CB. 2002. Modern human origins:progress and prospects. Philos Trans R Soc Lon-don B 357:563–579.

18 d’Errico F, Zilhao J, Baffier D, Julien M, Pele-grin J. 1998. Neandertal acculturation in westernEurope? a critical review of the evidence and itsinterpretation. Curr Anthropol 39:S1–S44.

19 Zilhao J, d’Errico F. 1999. The chronologyand taphonomy of the earliest Aurignacian andits implications for the understanding of Nean-dertal extinction. J World Prehist 13:1–68.

20 d’Errico F. n.d. Not just us: a multiple-speciesmodel for the origin of behavioral modernity,symbolism and art. In: Chippindale C, editor.Paleoart. Washington: Smithsonian InstitutionPress. In press.

21 Ember CR, Ember M. 2001. Cross-culturalresearch methods. Lanham: AltaMira Press.

22 Tzedakis PC, Andrieu-Ponel V, de BeaulieuJL, Crowhurst S, Follieri M, Hooghiemstra H,Magri D, Reille M, Sadori L, Shackleton NJ, Wi-jmstra TA. 1997. Comparison of terrestrial andmarine records of changing climate of the last500,000 years. Earth Planetary Sci Lett 150:171–176.

23 de Beaulieu JL, Andrieu-Ponel V, Reille M,Gruger E, Tzedakis PC, Svobodova H. 2001. Anattempt at correlation between the Velay pollensequence and the Middle Pleistocene stratigra-phy from central Europe. Quaternary Sci Rev20:1593–1602.

24 Binford LR. 1989. Isolating the transition tocultural adaptations: an organizational ap-proach. In: Trinkaus E, editor. The emergence ofmodern humans: biocultural adaptations in thelater Pleistocene. Cambridge: Cambridge Univer-sity Press. p 18–41.

25 Stiner MC, Kuhn SL. 1992. Subsistence, tech-nology, and adaptive variation in Middle Paleo-lithic Italy. Am Anthropol 94:306–339.

26 Marean CW. 1998. A critique of the evidencefor scavenging by Neandertals and early modernhumans: new data from Kobeh Cave (ZagrosMountains, Iran) and Die Kelders Cave 1 Layer10 (South Africa). J Hum Evol 35:111–136.

27 Marean CW, Assefa Z. 1999. Zooarchaeologi-cal evidence for the faunal exploitation behaviorof Neandertals and Early Modern Humans. EvolAnthropol 8:22–37.

28 Marean CW, Kim SY. 1998. Mousterian fau-nal remains from Kobeh cave (Zagros Moun-tains, Iran): implications for Neanderthals andearly modern humans. Curr Anthropol 38:79–113.

29 Stiner M. 2002. On in situ attrition and verte-brate body part profiles. J Archaeol Sci 29:979–991.30 Chase PG. 1986. The hunters of Combe Gre-nal: approaches to Middle Paleolithic subsistencein Europe. Oxford: BAR International Series 286.31 Auguste P. 1995. Chasse et charognage auPaleolithique moyen: l’apport du gisement de Bi-ach-Saint-Vaast (Pas-de-Calais). Bull Soc PrehistFr 92:155–167.32 Farizy C, David F, Jaubert J. 1994. Hommes etbisons du Paleolithique moyen a Mauran. Paris:CNRS.33 Gaudzinski S, Roebroeks W. 2000. Adult only:reindeer hunting at the Middle Paleolithic siteSalgitther-Lebenstedt, northern Germany. JHum Evol 38:497–521.34 Grayson DK, Delpech F. Specialized early Up-per Paleolithic hunters in southwestern France?J Archaeol Sci 20:1439–1449.35 Shea J. 1998. Neandertal and early modernhuman behavioral variability: a regional-scaleapproach to lithic evidence for hunting in theLevantine Mousterian. Curr Anthropol 39:S45–S78.36 Antunes M. 1992. O Homen da gruta daFigueira Brava (ca. 30000 BP). Memoria AcadCienc Lisboa 31:487–536.37 Fernandez-Jalvo J, Andrews P. 2000. The ta-phonomy of Pleistocene caves, with particularreference to Gibraltar. In: Stringer CB, BartonRNE, Finlayson JC, editors. Neanderthals on theedge: 150th Anniversary Conference of theForbes’ Quarry Discovery, Gibraltar. Oxford: Ox-bow Books. p 171–182.38 Stiner MC, Munro ND, Suravell T, TchernovE, Bar-Yosef O. 1999. Paleolithic populationgrowth pulses evidenced by small animal exploi-tation. Science 283:190–194.39 Hardy BL, Kay M, Marks AE, Monigal K.2001. Stone tool function at the paleolithic sitesof Starosele and Buran Kaya III, Crimea: behav-ioral implications. Proc Natl Acad Sci USA 98:10972–10977.40 Bocherens H, Billiou D, Mariotti A, Patou-Mathis M, Otte M, Bonjean D, Toussaint M.1999. Palaeoenvironmental and palaeodietaryimplications of isotopic biogeochemistry of LastInterglacial Neanderthal and mammal bones inScladina Cave (Belgium). J Archaeol Sci 26:599–607.41 Richards MP, Pettitt PB, Trinkaus E, SmithFH, Paunovi M, Karavani I. 2001. Neanderthaldiet at Vindija and Neanderthal predation: theevidence from stable isotopes. Proc Natl Acad SciUSA 98:6528–6532.42 Noble W, Davidson I. 1996. Human evolution,language and mind: a psychological and archae-ological inquiry. Cambridge: Cambridge Univer-sity Press.43 Foley R, Lahr MM. 1997. Mode 3 technologiesand the evolution of modern humans. CambridgeArchaeol J 7:3–36.44 Bar-Yosef O, Kuhn SL. 1999. The big dealabout blades: laminar technologies and humanevolution. Am Anthropol 101:322–338.45 Bonnichsen R, Turmire KL. 1999. Ice Agepeoples of North America. Eugene: Oregon StateUniversity Press.46 Mulvaney DJ. 1975. The prehistory of Austra-lia, 2nd ed. London: Penguin.47 Wurz S. 1999. The Howiesons Poort backedartefacts from Klasies River: an argument forsymbolic behaviour. S Afr Archaeol Bull 54:38–50.48 Koumouzelis M, Ginter B, Kozlowski JK,Pawlikowski M, Bar-Yosef O, Albert RM, Lityn-ska-Zajac M, Stworzewicz E, Wojtal P, Lipecki G,

ARTICLES Evolutionary Anthropology 201

Page 15: D'Errico 2003 - The Invisible Frontier - A Multiple Species Model For

Tomek T, Bochenski ZM, Pazdur A. 2001. Theearly Upper Paleolithic in Greece: the excava-tions in Klisoura Cave. J Archaeol Sci 28:515–539.49 Allsworth-Jones P. 1986. The Szeletian andthe transition from Middle to Upper Paleolithicin central Europe. Oxford: Oxford UniversityPress.50 Thieme H. 1997. Lower Paleolithic huntingspears from Germany. Nature 385:769–771.51 Thieme H. 2000. Lower Paleolithic huntingweapons from Schoningen, Germany—the oldestspears in the World. Acta Anthropol Sinica 19:S136–S143.52 Boeda E, Connan J, Muhesen S. 1998. Bitu-men as hafting material on Middle Paleolithicartifacts from the El Kowm Basin, Syria. In: Aka-zawa T, Aoki K, Bar-Yosef O, editors. Neander-tals and modern humans in western Asia. NewYork: Plenum Press. p 181–204.53 Boeda E, Geneste JM, Griggo C, Mercier N,Muhesen S, Reyss JL, Taha A, Valladas H. 1999.A Levallois point embedded in the vertebra of awild ass (Equus africanus): hafting, projectilesand Mousterian hunting weapons. Antiquity 73:394–402.54 Grunberg JM. 2002. Middle Paleolithic birch-bark pitch. Antiquity 76:15–16.55 d’Errico F, Laroulandie V. 2000. Bone tech-nology at the Middle-Upper Paleolithic transi-tion: the case of the worked bones from Buran-Kaya III level C (Crimea, Ukraine). In: OrschiedtJ, Weniger GC, editors. Neanderthals and mod-ern humans—discussing the transition: centraland eastern Europe from 50,000–30,000 BP.Mettmann: Neanderthal Museum. p 227–242.56 Mark T, Monigal K. 2000. The Middle to Up-per Paleolithic interface at Buran-Kaya-III, East-ern Crimean. In: Orschiedt J, Weniger GC, edi-tors. Neanderthals and modern humans—discussing the transition: central and easternEurope from 50,000–30,000 BP. Mettmann: Ne-anderthal Museum. p 212–226.57 Villa P, d’Errico F. 2001. Bone and ivorypoints in the Lower and Middle Paleolithic ofEurope. J Hum Evol 41:69–112.58 Howell FC, Freeman LG. 1983. Ivory pointsfrom the earlier Acheulean of the Spanish Me-seta. In: Homenaje al Prof. Martin AlmagroBasch. Madrid: Ministerio de Cultura. p 41–61.59 Schmitt D, Churchill SE, Hylander WL. 2003.Experimental evidence concerning spear use inNeanderthals and early modern humans. J Ar-chaeol Sci 30:103–114.60 Shea J. 1997. Middle Paleolithic spear pointtechnology. In: Knecht H, editor. Projectile tech-nology. New York: Plenum Press. p 79–106.61 Brooks AS, Helgren DM, Cramer JS, FranklinA, Hornyak W, Keating JM, Klein RG, Rink WJ,Schwarcz H, Leith Smith JN, Stewart K, ToddNE, Verniers J, Yellen JE. 1995. Dating and con-text of three Middle Stone Age sites with bonepoints in the Upper Semliki Valley, Zaire. Sci-ence 268:548–553.62 Yellen JE, Brooks AS, Cornelissen E, Mehl-man MJ, Stewart K. 1995. A Middle Stone Ageworked bone industry from Katanda, UpperSemliki Valley, Zaire. Science 268:553–556.63 Yellen JE. 1998. Barbed bone points: traditionand continuity in Saharan and sub-Saharan Af-rica. Afr Archaeol Rev 15:173–198.64 Pinto A, Andrews P, Barham LS. 2000. Bonetools. In: Barham LS, editor. The Middle StoneAge of Zambia, south central Africa. Bristol:Western Academic & Specialist Press. p 122–128.

65 Singer R, Wymer J. 1982. The Middle StoneAge at Klasies River Mouth in South Africa. Chi-cago: Chicago University Press.

66 d’Errico F, Julien M, Liolios D, Baffier D, Van-haeren M. n.d. Les poincons en os des coucheschatelperroniennes et aurignaciennes de laGrotte du Renne (Arcy-sur-Cure, Yonne): com-paraisons technologiques, fonctionnelles et de-cor. In: Approches fonctionnelles en prehistoire.Actes du XXVe Congres Prehistorique de France,Nanterre (Hauts-de-Seine), 24–26 Novembre2000.

67 Bordes JG. 2001. Chatelperronian/Aurigna-cian interstratifications at Roc-de-Combe and LePiage (Lot, France): lithic taphonomy, strati-graphic re-evaluation and archaeological impli-cations. In: Pre-actes. XIV Congres de l’UISPP,Liege, 2–8 Septembre 2001. p 135–136.

68 May F. 1986. Les sepultures prehistoriques.Paris: CNRS.

69 Riel-Salvatore J, Clark GA. 2001. Grave mark-ers. Middle and early Upper Paleolithic burialsand the use of chronotypology in contemporaryPaleolithic research. Curr Anthropol 42:449–460.

70 Gargett RH. 1989. Grave shortcomings: theevidence for Neandertal burial. Curr Anthropol30:157–190.

71 Gargett RH. 1999. Middle Paleolithic burial isnot a dead issue: the view from Qafzeh, Saint-Cesaire, Kebara, Amud, and Dederiyeh. J HumEvol 37:27–40.

72 Belfer-Cohen A, Hovers E. 1992. In the eye ofthe beholder: Mousterian and Natufian burials inthe Levant. Curr Anthropol 34:463–471.

73 Bar-Yosef O, Vandermeersch B, Arensburg B,Belfer-Cohen A, Goldberg P, Laville H, MeignenL, Rak Y, Speth JD, Tchernov E, Tillier AM,Weiner S. 1992. The excavations in Kebara Cave,Mt Carmel. Curr Anthropol 33:497–550.

74 Hovers E, Kimbel HW, Rak Y. 2000. TheAmud 7 skeleton—still a burial: response to Gar-gett. J Hum Evol 39:253–260.

75 Maureille B. 2002. A lost Neanderthal neonatefound. Nature 419:33.

76 Mercier N, Valladas H, Froget L, Joron JL,Ronen A. 2000. Datation par la thermolumines-cence de la base du gisement paleolithique deTabun (Mont Carmel, Israel). C R Acad Sci 330:731–738.

77 Grun R, Stringer CB. 2000. Tabun revisited:revised ESR chronology and new ESR and U-series analyses of dental material from Tabun C1.J Hum Evol 39:601–612.

78 Hovers E, Ilani S, Bar-Yosef O, Vandermeer-sch B. n.d. Different strokes for different folks:the use of ochre by early modern humans inQafzeh Cave, Israel. Curr Anthropol. In press.

79 McBrearty S. 2001. The Middle Pleistocene ofEast Africa. In: Barham L, Robson-Brown K, ed-itors. Human roots: Africa and Asia in the MiddlePleistocene. Bristol: Western Academic and Spe-cialist Press. p 81–92.

80 Barham LS. 2002. Systematic pigment use inthe Middle Pleistocene of south-central Africa.Curr Anthropol 31:181–190.

81 Watts I. 1999. The origin of symbolic culture.In: Dunbar R, Knight C, Power C, editors. Theevolution of culture. Edinburgh: Edinburgh Uni-versity Press. p 113–146.

82 Van Peer P, Vermeersch PM. 2000. The Nu-bian Complex and the dispersal of modern hu-mans in North Africa. In: Krzyzaniak L, KroeperK, Kobusiewicz M, editors. Recent research intothe Stone Age of Northeastern Africa. Poland:

Poznan Archaeological Museum. p 47–60. (Stud-ies in African Archaeology 7)

83 Demars PY. 1992. Les colorants dans le Mous-terien du Perigord. L’apport des fouilles de F.Bordes. Bull Soc Prehist l’Ariege 67:185–194.

84 d’Errico F, Soressi M. 2002. Systematic use ofmanganese pigment by the Pech-de-l’Aze Nean-dertals: implications for the origin of behavioralmodernity. J Hum Evol 42:A13.

85 Soressi M, Armand D, d’Errico F, Pubert E,Jones H, Pubert E, Rink J, Texier JP, Vivent D.2002. Pech-de-l’Aze I (Carsac): nouveaux travauxsur le Mousterien de tradition acheuleenne. BullSoc Prehist Fr 99:1–7.

86 Carciumaru M, Moncel MH, Anghelinu M,Carciumaru R. 2002. The Cioarei-Borosteni Cave(Carpathian Mountains, Romania): Middle Pa-leolithic finds and technological analysis of thelithic assemblages. Antiquity 76:681–690.

87 Ambrose SH. 1998. Chronology of the LaterStone Age and food production in East Africa. JArchaeol Sci 25:377–392.

88 Bar Yosef O. 1992. The role of western Asia inmodern human origins. Philos Trans R Soc Lon-don 337:193–200.

89 Kuhn SL, Stiner MC, Reese DS, Gulec E.2001. Ornaments of the earliest Upper Paleo-lithic: new insights from the Levant. Proc NatlAcad Sci 98:7641–7646.

90 Tillier AM, Tixier J. 1991. Une molaired’enfant aurignacien a Ksar’ Aquil (Liban). Pale-orient 17:89–93.

91 d’Errico F, Villa P. 1997. Holes and grooves:the contribution of microscopy and taphonomyto the problem of art origins. J Hum Evol 33:1–31.

92 d’Errico F, Villa P, Pinto A, Ruiz Idarraga R.1998. A Middle Paleolithic origin of music? usingcave bear bone accumulations to assess the DivjeBabe I bone “flute.” Antiquity 72:65–79.

93 Pavlov P, Svendsen JI, Indreli S. 2001. Humanpresence in the European Arctic nearly 40,000years ago. Nature 413:64–67.

94 Marshack A. 1995. A Middle Paleolithic sym-bolic composition from the Golan Heights: theearliest known depictive image. Curr Anthropol37:356–365.

95 Hovers E, Vandermeersch B, Bar-Yosef O.1997. A middle Paleolithic engraved artefactfrom Qafzeh cave, Israel. Rock Art Res 14:79–87.

96 Goren-Inbar N. 1986. A figurine from theAcheulian site of Berekhat Ram. Mitekufat Hae-ven 19:7–12.

97 Marshack A. 1997. The Berekhat Ram figu-rine: a late Acheulian carving from the MiddleEast. Antiquity 71:327–337.

98 d’Errico F, Nowell A. 2000. A new look at theBerekhat Ram figurine: implications for the ori-gins of symbolism. Cambridge Archaeol J 10:123–167.

99 Granger J-M, Leveque F. 1997. Parurecastelperronienne et aurignacienne: etude detrois series inedites de dents percees et comparai-sons. C R Acad Sci Paris 325:537–543.

100 Relethford J, Jorde L. 1999. Genetic evi-dence for larger African population size duringrecent human evolution. Am J Phys Anthropol108:251–260.

101 Shennan S. 2001. Demography and culturalinnovation: a model and its implications for theemergence of modern human culture. Cam-bridge Archaeol J 11:5–16.

© 2003 Wiley-Liss, Inc.

202 Evolutionary Anthropology ARTICLES