the acheulean handaxe

ARTICLE

The Acheulean Handaxe: More Like a Bird’sSong Than a Beatles’ Tune?RAYMOND CORBEY, ADAM JAGICH, KRIST VAESEN, AND MARK COLLARD

The goal of this paper is to provoke debate about the nature of an iconic arti-fact—the Acheulean handaxe. Specifically, we want to initiate a conversationabout whether or not they are cultural objects. The vast majority of archeolo-gists assume that the behaviors involved in the production of handaxes wereacquired by social learning and that handaxes are therefore cultural. We willargue that this assumption is not warranted on the basis of the available evi-dence and that an alternative hypothesis should be given serious consideration.This alternative hypothesis is that the form of Acheulean handaxes was at leastpartly under genetic control.

Named after the site of Saint-Acheulin France, where they were first identi-fied, Acheulean handaxes are distinc-tive (Fig. 1). Indeed, they are sodistinctive that they are probably theone artifact that all archeologists,whatever their period of interest, arecapable of identifying. Acheulean han-daxes were produced by the bifacialreduction of a block or large flakeblank around a single, long axis. Theyhave a cutting edge in the secantplane, and range in shape from lan-ceolate through ovate to orbiculate.

Acheulean handaxes are one of thecommonest, most widely distributed,and longest-lasting archeological arti-facts. Several hundred thousandAcheulean handaxes have been recov-ered from sites in many regions of theOld World, including North, South,and East Africa; Europe; and West-ern, South, and East Asia (Fig. 2). Theoldest Acheulean handaxes date toapproximately 1.76 million years ago(Ma)1 and the youngest to between300 and 200 thousand years ago(Ka).2,3 Thus, they span several mil-

lion square kilometers, multiple eco-logical regions, and roughly ahundred thousand generations.

Acheulean handaxes are the defin-ing artifact of the Acheulean indus-try, which also includes flakes, flaketools, and cores, as well as otherlarge cutting tools such as cleavers,picks, trihedrals, and unifaces. TheAcheulean industry was preceded bythe Oldowan, which is found inAfrica and parts of Eurasia; it wassucceeded by the Middle Palaeolithicin western Eurasia and the MiddleStone Age in Africa.

Acheulean handaxes are thought tohave been produced by two extincthominin species, Homo erectus andHomo heidelbergensis. Fossils assignedto H. erectus have been recovered fromsites in East Africa, South Africa,North Africa, the Caucasus, SoutheastAsia, and East Asia.5 H. erectus issometimes subdivided into Homoergaster and H. erectus sensu stricto.5

Raymond Corbey is Professor of Epistemology and Anthropology at the Faculty of Archaeology of Leiden University and Associate Professorof Philosophy at Tilburg University, both in the Netherlands. Much of his research deals with the history and philosophy of anthropology andarcheology. Email: [email protected].

Adam Jagich is a graduate of Stony Brook University and Leiden University. Currently he is the scientific director of CommonSites, a web-based platform for the heritage sector, and is based in Lagos, Nigeria. His main research interest is Neanderthal biogeography. Email:[email protected].

Krist Vaesen is an assistant professor of philosophy at Eindhoven University of Technology and a research fellow at the Faculty of Archaeol-ogy, University of Leiden, both in the Netherlands. He leads a research program on the Darwinization of the cultural sciences. Email: [email protected].

Mark Collard is the Canada Research Chair in Human Evolutionary Studies and Professor of Archaeology at Simon Fraser University in Can-ada. He also holds a personal chair in archeology at the University of Aberdeen in the UK. Collard is an evolutionary anthropologist with awide range of interests, including the evolution of technology. He is the corresponding author for this paper. Email: [email protected].

Key words: Acheulean handaxe; cultural transmission; social learning; genetic transmission

VC 2016 Wiley Periodicals, Inc.DOI: 10.1002/evan.21467Published online in Wiley Online Library (wileyonlinelibrary.com).

Evolutionary Anthropology 25:6–19 (2016)

The former is represented by several

early African specimens and the latter

by Eurasian and later African speci-

mens. The hypodigm of H. heidelber-gensis includes specimens from East

Africa, South Africa, Europe, South

Asia, and East Asia.6 H. heidelbergensisis sometimes argued to be a

“wastebasket” taxon. Those who

choose to split this taxon into two spe-

cies usually assign Eurasian speci-

mens to H. heidelbergensis and Africanspecimens to Homo rhodesiensis.

Several issues regarding Acheuleanhandaxes are contentious. Most con-spicuously, there is disagreementabout their function. Most researchersconsider handaxes to be cutting tools,but it has also been suggested thatthey were throwing weapons.7–9 Inaddition, it has been proposed thatthey played a role in social and/or sex-

ual signaling.10,11 There is also debateabout the extent to which the form(that is, the shape and size) of Acheu-lean handaxes is deliberate. Mostresearchers assume that it is, butothers assert that resharpening greatlyaffects handaxe form.12 According tothis argument, resharpening generatessimilar forms in assemblages that aregeographically and temporally sepa-rated. Still other researchers contendthat raw material quality affected han-daxe form,13 with, for example, large,flat chunks of flint from chalk cliffsyielding handaxes of a different formthan small, elongated flint pebblesobtained from river beds.

There is even debate about thevalidity of the Acheulean handaxe asa type. The majority of researchersagree with McNabb14 that while han-daxes may be large or small, more orless refined, pointed or ovate, sym-metrical or off-set, beneath this vari-ability there are general “themes” thatare present in all handaxes. However,there are some dissenting opinions.Nicoud,15 for instance, argues thatthe Acheulean handaxe is not a uni-tary phenomenon but rather an artifi-cial grouping together of artifacts onthe basis of superficial morphologicaland technological similarities.

Thus, disagreement about Acheu-lean handaxes abounds. However, asRicherson and Boyd16 observe, thereis one thing that more or less allresearchers working on handaxesagree on, which is that the behaviorsnecessary to manufacture them werecopied from other individuals and,therefore, that handaxes are culturalobjects. Richerson and Boyd16 offeran argument against this idea. Theypoint out that both models and ethno-graphic data suggest that culturallearning in the small, relatively iso-lated groups that H. erectus and H.heidelbergensis are thought to havelived in should have resulted in rap-idly diverging traditions rather thanthe “bewildering”17:648 geographicand temporal stability exhibited bythe Acheulean handaxe. Based onthis, Richerson and Boyd16 suggestthat the conservatism of Acheuleanhandaxes may be evidence, not of cul-tural transmission, but of genetictransmission. Foley18 makes a similarpoint, arguing that since the pattern

Figure 1. Acheulean handaxes from the site of Boxgrove, England, which dates to about500 Ka. The handaxes are made of flint and are between 12 and 14.5 cm in length. Pho-tograph by W. Roebroeks; used with permission. [Color figure can be viewed in theonline issue, which is available at wileyonlinelibrary.com.]

Figure 2. Acheulean handaxes from various regions (to scale; biface 7 is 22.5 cm tall).Sites: 1) Boxgrove, England; 2) North of Bridge Acheulean, near Gesher Benot Ya’aqov,Israel; 3) Erg Tihoda€ıne, Algeria; 4) Sterkfontein, South Africa; 5) Olduvai Gorge, Tanzania;6) Bose, China, 7) Isampur, India.89,124–129 Figure by Shumon Hussain.

ARTICLE The Acheulean Handaxe: More Like a Bird’s Song Than a Beatles’ Tune? 7

http://wileyonlinelibrary.com

of handaxe variation does not matchwhat we expect to see if the behaviorsinvolved in their manufacture weresocially learned, a role for genetictransmission should be considered.

To date, the possibility that the pro-duction of Acheulean handaxes wasunder at least partial genetic controlhas not been given serious considera-tion by archeologists. In our view,this is unfortunate. Here we attemptto explain why we think the genetic-transmission hypothesis should betreated as a serious contender forexplaining the handaxe phenomenon.We review several lines of evidence infavor of the hypothesis, as well as evi-dence against the dominant culturaltransmission hypothesis. Subse-quently, we use bird behaviors thathave a substantial genetic componentto illustrate how a behavior as com-plex as handaxe production could beunder genetic control. In line withRicherson and Boyd,16 we will use theterms “social learning,” “culturallearning,” and “cultural transmission”interchangeably.

THE DOMINANCE OF THECULTURAL TRANSMISSION

HYPOTHESIS

Until the 1990s, archeologists rarelydiscussed the manner in which thebehaviors involved in the production ofAcheulean handaxes were acquired.Handaxes were simply treated as cul-tural objects and the reader was left toinfer that this meant the behaviorsinvolved in their production werelearned socially. In the 1990s, thisbegan to change and, in line with devel-opments elsewhere in the discipline ofarcheology, researchers working onAcheulean handaxes started to payattention to the transmission of thebehaviors involved in their production.Significantly, for present purposes, thetransmission mechanisms suggested tobe involved in the production of Acheu-lean handaxes in the vast majority ofthese publications, were cultural. Wehave located only a handful of publica-tions, among them the aforementionedworks by Boyd and Richerson16 andFoley,18 that suggest that somethingother than cultural transmission mighthave been involved in the acquisition

of the behaviors required to producehandaxes.

Cultural transmission has beeninvoked numerous times in the post-1990 literature to explain the stabilityof the key features of handaxes overroughly 1.5 million years. Mithen19

provides a clear example in his 1999paper, “Imitation and culturalchange: a view from the Stone Age,with specific reference to the manu-facture of handaxes.” Observing thehigh degree of similarity in formamong handaxes, Mithen argues thatimitation must have been a “necessityfor acquiring the technical skills tomanufacture them.”19:393 Similarly,

Shipton, Petraglia, and Paddayya20

state that imitation is crucial forunderstanding the conservatism ofthe Acheulean industry. They contendthat “from the outset Acheuleansocial interactions were characterizedby a propensity for true imitation andit is this, rather than any genetic pre-disposition to produce bifaces, whichhas produced the remarkable homo-geneity of the Acheulean.”20:229

Shennan and Steele21:368 agree thatimitative learning was crucial in the

production of handaxes and even sug-gest that it was “most likely aided byactive teaching.” Bar-Yosef,22 in hisconcluding remarks to a widely citededited volume on the Acheulean, sup-plies another example of the use ofcultural transmission to explain thestability of the key features of han-daxes. In a section entitled “Theconservatism of the Acheuleanknappers,” he argues that the reten-tion of predetermined solutions was“dictated by cultural concepts.”22:487

In addition to being invoked toaccount for the stability of key fea-tures of the handaxe, cultural trans-mission has been used to explain thetemporal variation in handaxe form.Over time, handaxes generallybecame smaller, thinner, and lesselongated. They also became morestandardized and more finely made.23

Furthermore, some studies havefound that handaxes became moresymmetrical as time progressed.23–25

Vaughan4 and Lycett23 used modelsfrom the field of population geneticsto account for these changes. Despiteborrowing their models from genet-ics, both authors frame their studiesin terms of cultural evolution, andtherefore implicitly assume that thebehaviors involved in the manufac-ture of handaxes were sociallylearned.

Other authors have used culturaltransmission to explain inter- andintra-regional differences in handaxes.Mithen,26:14 for example, contendsthat “[t]here are clear, well definedhandaxe types within the Acheulian,indicating cultural traditions in arte-fact form.”1 Later in the same paperhe suggests that intersite differencesmay relate to “variability in the degreeof experience and opportunities forsocial learning by individuals withinand between hominid groups owing tovariation in social interaction andgroup size.”26:17 Wynn and Tierson,27

Roe,28 and Wenban-Smith29 makesimilar arguments. Wynn and Tier-son27 hypothesize that variabilityamong handaxes from Indian, British,Israeli, and African sites reflectsregional cultural traditions, whileRoe28 and Wenban-Smith29 both positcultural transmission of handaxe tra-ditions within the British Palaeolithic.

the possibility that theproduction of Acheu-lean handaxes wasunder at least partialgenetic control has notbeen given serious con-sideration by archeolo-gists. In our view, this isunfortunate. Here weattempt to explain whywe think the genetic-transmission hypothesisshould be treated as aserious contender forexplaining the handaxephenomenon.

8 Corbey et al. ARTICLE

The strength of the belief thatAcheulean handaxes are culturalobjects is underscored by a recentpaper by Machin.30 She criticizes thestandard approach to understandinghandaxe variability, in which a singlefactor is viewed as being of primaryimportance. Instead, she contends,we need to recognize that multiplefactors would have been involved.She goes on to outline no fewer than48 factors that she believes werelikely to have influenced handaxevariability. Significantly, for presentpurposes, at no point does she ques-tion the assumption that the behav-iors required to produce handaxeswere acquired by cultural transmis-sion. The fact that even an authorwho calls on colleagues to thinkmuch more broadly about the way inwhich they approach handaxes doesnot question the assumption thathandaxes are cultural objects showsjust how deeply embedded theassumption is in archeology.

REASONS TO DOUBT THECULTURAL TRANSMISSION

HYPOTHESIS

Earlier we explained why Richersonand Boyd16 and Foley18 suggest that arole for genetic transmission in themanufacture of handaxes should beconsidered. To reiterate, they pointout that the long-term conservatism ofthe Acheulean handaxe is inconsistentwith models of cultural evolution andethnographic data, which suggest thatcultural transmission in groups of thetype H. erectus and H. heidelbergensisare thought to have lived in shouldgive rise to considerable spatiotempo-ral variability. That handaxes do notexhibit the kind of signal predicted bycultural evolutionary models and eth-nographic data, Richerson and Boyd16

and Foley18 aver, implies that han-daxes are not fully cultural objects,which in turn suggests that the behav-iors involved in their manufacturemay have been acquired at least inpart by genetic transmission.

The most obvious objection toRicherson and Boyd’s16 and Foley’s18

argument is that it ignores the possi-bility that the form of handaxes wasstable because it was strongly tied to

a specific function. However, thisrebuttal does not take into accountcopying error, which, because of thelimitations of perception, is inevitablewhen learning how to make artifactsby copying another individual.Experiments involving human sub-jects indicate that copies typicallydeviate from their target by 3%-5%,and that this error compounds overmultiple transmission events.31 Overthe course of almost 1.5 million years,3%-5% copying errors would haveresulted in a substantial amount ofspatiotemporal variation, regardlessof how closely linked handaxes wereto a particular function. This is notjust a theoretical argument. A recentstudy found that size variation in alarge sample of Acheulean handaxeswas lower than expected from copy-ing error.32 Given that copying errorwould inevitably have occurred ifindividuals had acquired the behav-iors required to produce handaxes viasocial learning, this supports the ideathat handaxes were not fully culturalartifacts and may have been under atleast partial genetic control.

As strong as we think Richersonand Boyd’s16 and Foley’s18 argumentis, we are of the opinion that it isjust one of the problems facing thecultural transmission hypothesis. Wewill outline several other reasons toquestion the assumption that thebehaviors involved in the manufac-ture of handaxes were acquiredpurely by cultural learning and totake seriously the possibility thatthey were acquired at least in partby genetic transmission. We believethat, taken together, Richerson andBoyd’s16 and Foley’s18 argument andthe evidence we discuss in this sec-tion represent a powerful caseagainst continuation of the current,uncritical approach to the nature ofthe Acheulean handaxe.

The Chimpanzee CultureDebate Demonstrates ThatMechanisms of TransmissionNeed to be Investigated andNot Simply Assumed

Another reason to question theassumption that Acheulean handaxesare cultural objects is provided by the

ongoing debate about the nature ofthe behavioral differences amongwild-living chimpanzee groups. Overthe last 40 years, numerous behaviorshave been found to vary among chim-panzee groups. Some behaviors occurat some long-term study sites but notat others. For instance, at the site ofBossou, Guinea, individuals oftendetach fronds from an oil palm anduse them to smash the tree’s crown toproduce a pulp that is edible.33 So far,this behavior has not been docu-mented at any other site.34 In addi-tion, the way in which certainbehaviors are performed variesamong sites. Nut cracking illustratesthis. At Bossou, only stone hammersand anvils are used to perform thistask,35 whereas in the Ta€ı Forest, Coted’Ivoire, two different kinds of ham-mer, wood and stone, and two differ-ent kinds of anvil, root and stone, areused to crack nuts.36 An intersitecomparison carried out a few yearsago suggested that at least 65 behav-iors vary among chimpanzeegroups.37,38

The nature of the among-groupbehavioral differences is contested.According to some researchers, many ofthe behaviors are likely to be sociallylearned and therefore meet the maincondition that is necessary for beingrecognized as cultural behaviors.34–41

Other researchers question whether thebehaviors are socially learned42–46 andhighlight the fact that many of the puta-tive cultural behaviors occur in a singlesubspecies. They point out that geneticstudies suggest that some chimpanzeesubspecies have been isolated for hun-dreds of thousands of years and con-tend that in these circumstances it isnot possible to discount a genetic originfor the behavioral differences.

Currently, it is unclear which ofthese hypotheses is correct. Both havebeen supported by recent work. Lycett,Collard, and McGrew47–49 havereported several studies in which theycompared phylogenetic trees derivedfrom chimpanzee behavioral datasetsto published genetic data pertaining tothe relationships among chimpanzeesubspecies and found that the behav-ioral trees did not match the geneticdata. Based on this, these authors47–49

conclude that the culture hypothesis ismore likely to be correct than is the


genetic one. Langergraber and col-leagues46 have reached a different con-clusion. Using the Mantel test, theyfound a significant correlation betweengenetic and behavioral data for severalchimpanzee groups. Based on thisthey argue that it is not possible to ruleout a major role for genes in generatingthe differences among the groups’behavioral repertoires. There is dis-agreement about why the results ofthese studies do not converge,50,51

which means that, at the moment, it isnot possible to state with confidencethat the behavioral differences amongchimpanzee groups are cultural orgenetic.

Thus, the nature of the behavioraldifferences among chimpanzee groupsremains unclear even after nearly 40years of discussion and completion ofseveral research projects specificallydesigned to resolve the issue. Thisclearly indicates that it is not straight-forward to infer mechanisms of trans-mission from behavioral patterns. Anobvious corollary of this is that weshould be skeptical when a mecha-nism of transmission has been con-cluded to be responsible for abehavioral pattern without anyattempt to consider alternative mecha-nisms of transmission. This is the casewith the Acheulean handaxe. Mostresearchers have simply assumed thatthe behaviors involved in the manufac-ture of handaxes were acquired by cul-tural transmission. As far as we areaware, no attempt has ever been madeto determine whether the availabledata support the cultural transmissionhypothesis better than they supportpotentially competing hypothesessuch as individual learning andgenetic transmission. Accordingly, thebasis for believing that handaxes arecultural objects is shaky.

Work on Modern Humans AlsoIndicates That Mechanisms ofTransmission Need To BeInvestigated and Not SimplyAssumed

Recent research on living humansalso suggests that we should be skep-tical when a mechanism of transmis-sion has been assumed to beresponsible for a behavioral pattern

without any attempt to consideralternative mechanisms of transmis-sion. In 2004, McElreath52 reporteda ground-breaking study in which hesought to identify the processesunderlying differences in the beliefsand behavior of farmers and pastor-alists living in the Usangu Plains ofTanzania. He focused on individualswho had changed household econo-

mies, and investigated whether theirdispositions with respect to threedomains — preference for friends orkin, respect for elders, and belief inwitchcraft — were a consequence ofindividual or social learning. McEl-reath’s52 results were mixed. Hefound that variation in the preva-lence of belief in witchcraft was bestexplained by social learning. How-ever, variation in the other twodomains was better explained byindividual learning. Variation inwhether or not individuals believedthat respect for elders is importantwas strongly influenced by individual

learning, but there was some effectof social learning. Variation in pref-erence for friends or kin was notinfluenced at all by social learning,but only by individual learning. Themain significance of these results forthe present argument is that theydemonstrate that even the beliefsand behaviors of modern humanscannot be simply assumed to be cul-tural phenomena. Many of themprobably are, but some of themlikely are not. As such, analysis isrequired before a given belief orbehavior is deemed cultural.

To put it another way, McEl-reath’s52 findings call into questionthe anthropological axiom that allthe interesting beliefs and behaviorsof modern humans are cultural. Byextension, they also challenge thedefault assumption of paleoanthro-pologists that the behaviors ofextinct hominins that are reflected inthe archeological record, includingthe behaviors that resulted in Acheu-lean handaxes, are cultural. Many ofthe behaviors of modern humanslikely are cultural, and many of thebehaviors of extinct hominins prob-ably were cultural. But it is notdefensible to simply assume that thisis the case. Rather, it needs to bedemonstrated analytically. We needto treat the processes underlying thebehaviors that anthropologists andarcheologists study as unknown anddesign research projects to testbetween the competing hypotheses,in the same way that those workingon the processes underlying behav-ioral variation among chimpanzeegroups have tested between the cul-ture hypothesis and the genetichypothesis. Given the difficulty ofassociating artifacts with homininspecies and the fragility of DNA,doing so will be difficult. However,this is not a good reason to ignorethe problem.

The Failure of AcheuleanHandaxes To TrackEnvironmental Variation IsInconsistent with Models ofCultural Evolution

As we have seen, the long-termconservatism of the Acheulean

Most researchers havesimply assumed that thebehaviors involved inthe manufacture of han-daxes were acquired bycultural transmission. Asfar as we are aware, noattempt has ever beenmade to determinewhether the availabledata support the cul-tural- transmissionhypothesis better thanthey support potentiallycompeting hypothesessuch as individuallearning and genetictransmission.


handaxe is inconsistent with a keyprediction of recent models of cul-tural evolution, namely that culturalentities should exhibit considerablespatiotemporal variability. Handaxesare inconsistent with models of cul-tural evolution in other ways.

Modeling work carried out overthe last 30 years or so sets fairlystringent conditions on the emer-gence of cultural transmission.53,54

The goal of this work was to deter-mine when cultural learning isfavored over individual learning andgenetic transmission, given that cul-tural learning is cognitively costly.For present purposes, the key findingis that cultural learning evolves onlyin moderately variable environments.In slowly changing environments,genetic transmission outperformscultural transmission, while in highlyvariable environments individuallearning beats both genetic transmis-sion and cultural transmission.Equally importantly, in order for cul-tural transmission to be favored overthe alternatives it must result inbehaviors that track environmentalconditions. If it does not give rise tosuch behaviors, genetic transmissionor individual learning will be pre-ferred by selection, depending on thespeed of environmental change.

For the behavioral routines involvedin the production of handaxes to havebeen culturally transmitted, therefore,the environments in which handaxeswere produced must have been variableand handaxes must have tracked thatvariability. The first condition is met. Itis clear that Acheulean toolmakers hadto cope with fluctuating climates.55–58

They also expanded into various newhabitats.55,56 Thus, there is reason tothink that the environments in whichhandaxes were produced were variable.In contrast, the second condition is notmet. As we explained earlier, the keyfeatures of the handaxe do not varymuch across space or through time.The implication of this is that handaxesdid not respond to the environmentalvariability that Acheulean toolmakersexperienced. Given that cultural evolu-tionary models suggest the ability totrack environmental variability is thereason why cultural transmission issometimes favored over genetic trans-mission, the fact that handaxes do not

track environmental variability presentsanother substantial challenge to the cul-tural transmission hypothesis.

Explaining the Increased Speedof Cultural Change in The LatePleistocene Is Difficult IfAcheulean Handaxes AreAssumed To Be Fully Cultural

A further problem facing the cul-tural transmission hypothesis is themuch more rapid pace of changeafter Acheulean handaxes disap-peared at 300-200 Ka. If the behav-iors involved in the production ofboth handaxes and post-Acheuleanartifacts were culturally learned, howdo we explain this very markedincrease in the speed of change?Currently, two hypotheses prevail.Unfortunately, neither is particularlyconvincing.

The first hypothesis focuses on cog-nition. According to this hypothesis,the shift in speed occurred becausethe hominin brain was “upgraded”;that is, one or more cognitive capaci-ties, such as enhanced working mem-ory, causal reasoning, or executivecontrol, were added.59–63 These herit-able biological changes introducedthe necessary cultural variation, sothat the cultural evolutionary processcould yield more complex culture.This scenario faces several problems,depending on when one sets the dateof the event. If the event is set around100 Ka, one can account for the spo-radic appearance of many markers ofmodern behavior at multiple Africansites as early as 70 to 90 ka. However,one cannot explain the disappearanceof these markers between 75-60 Kaand their later reemergence at differ-ent times in various parts of theworld. If the event is set later, sayright before the transition in Europe(45 Ka), one needs to assume that thearrival of fully modern behavior insoutheast Asia and Australia resultedfrom either convergent evolution ordispersal out of Europe. The formerseems unlikely, given the drasticnature of the change, while the latteris hard to reconcile with the availableevidence on Late Pleistocene migra-tions. Furthermore, unless conver-gent evolution is invoked, increased

cognitive ability would be denied toNeanderthals, despite compellingarguments to the contrary.64–66

The second potential explanationfor the increase in speed of artifactchange at 300–200 ka concentrateson demography. Hominin populationsize is thought to have increased inthe Late Pleistocene and it has beenargued that this would have givenrise to an increase in cultural com-plexity. The main intuition behindthe dependence of cultural changeon demography is that larger popula-tions can sustain a more complexculture because they are more likelyto contain individuals whose skilllevel is at least as good as that of theindividuals in the previous genera-tion. However, the demographicexplanation of the transition towardfully modern human behavior isproblematic in at least two respects.First, it has been shown that the for-mal model that is used most fre-quently to support the demographyhypothesis67,68 only predicts a rela-tionship between population sizeand cultural complexity in implausi-ble conditions.69,70 Second, a keyprediction of the hypothesis — thatthere should be a significant, positiverelationship between populationsize and cultural complexity inhunter-gatherers — is not supportedempirically. Several studies havefound no evidence of an impact ofpopulation size on technologicalrichness and complexity inethnographically documentedhunter-gatherers.71–74 A similar find-ing was obtained in a recent attemptto use North American archeologicaldata to evaluate the impact of popu-lation size on hunter-gatherertechnology.75 Moreover, results ofattempts to estimate the populationsizes of Late Pleistocene human pop-ulations conflict with the timing ofthe transition toward fully modernhuman behavior.76

Given that neither of the culturalhypotheses withstands scrutiny, howdo we explain the shift in speed ofartifact change 300-200 ka? If Acheu-lean handaxe production behaviorwas under genetic control, the post-Acheulean speeding up and diversifi-cation of culture perhaps markedanother event, namely greater


reliance on cultural inheritance.This, in turn, may imply either theevolution of special-purpose mecha-nisms for cultural learning ordomain-general mechanisms forlearning being “tuned,” ontogeneti-cally, to inputs from other agents.77

The first option seems to sufferfrom the same problems as thosenoted in relation to the cognitiveupgrade hypothesis. The secondoption is more promising because itleaves room for cultural differenceswithin Homo sapiens (geographicallyand diachronically) and between H.sapiens and Neanderthals, as a resultof differences in ontogeny. Accordingto this hypothesis, the tempo of cul-tural evolution depends on the extentto which attentional, perceptual,and/or motivational mechanisms areontogenetically biased toward others.The Late Pleistocene transition maybe a consequence of hominins start-ing to rely on cultural learning for awide range of behaviors, due, forinstance, to local conditions favoringincreased social tolerance and/or theemergence of new types of socialorganization in which contactbetween putative pupils and modelsis more likely.

With regard to the latter possibility,a recent study by Hill andcoworkers78 indicates that it is com-mon for contemporary hunter-gatherers to live in residential unitswith a high percentage of unrelatedindividuals. Such organizationimplies friendly visits between unitsand many more opportunities forobserving novel cultural traits than isthe case when residential units areprimarily kin-based. The emergenceof these sorts of networks may thusbe one of the ways in which humanlearning became more attuned tosocial cues. Also, if group membersare not closely related, one shouldexpect a variety of mechanisms thatactively promote cooperation (indi-rect reciprocity, sanctioning of defec-tors). Under these conditions, itseems likely that social stimuli pro-vide reliable information, which,according to Heyes,77 in turn favorshigher attentiveness to such stimuli.

Religious beliefs and practices,which have been suggested to pro-mote intragroup cohesiveness and

cooperation.79–82 may be another fac-tor facilitating the ontogenetic devel-opment of cultural learning. Incollective rituals, groups express andreaffirm shared beliefs, norms, andvalues, and this enhances communalstability and group harmony.83,84 Inaddition, belief in supernatural enti-ties that monitor their devotees andpunish them in case of transgressionshas been proposed to encourage pro-sociality towards strangers.85,86 Reli-gion would thus seem to providesuitable conditions for cultural learn-ing and therefore allow the formationof cumulative culture.

To summarize, the increase in thepace of artifact change after theAcheulean may not be the conse-quence of a cognitive upgrade orincreased population size, as archeolo-gists often suppose. Instead, it mayindicate a shift from genetic transmis-sion to increasing reliance on ontoge-netic cultural learning. This proposaloffers a plausible mechanism bywhich artifact change could acceler-ate. It also is easier to reconcile withgeographic and temporal variationthan is the cultural hypothesis.

Some Geographic Facts AboutHandaxes Are More ConsistentWith Genetic Transmission ThanWith Cultural Transmission

In the late 1940s, Movius87,88 drewattention to the geographic distribu-tion of Acheulean handaxes. Hepointed out that handaxes had beenfound in Africa, Europe, and India,but not in East and Southeast Asia.This “Movius Line” persisted formany years, but recently handaxeshave been found at various sites inthe Bose Basin and other riverbasins in China, some 1,500 kmnortheast of the Movius Line.89 Thisunexpected development presents athird major problem for the culturaltransmission hypothesis.

The unexpected occurrence of han-daxes in China has been explained interms of raw material distribution. Ithas been suggested that high-qualitylithic raw materials are rare in theeastern part of the range of H. erec-tus, but forest fires periodicallyuncovered suitable raw material in

river basins. Whenever this hap-pened, so the hypothesis goes, thehominins occupying the area pro-duced bifaces. On the face of it, thisseems like a reasonable suggestion.However, it actually ignores a crit-ically important question: How werethe behaviors involved in the produc-tion of handaxes acquired by homi-nins located 1,500 km from theMovius Line?

There are two potential answers tothis question that involve culturaltransmission, neither of which isparticularly compelling. One possi-bility is that the Chinese homininsmaintained contact with groups thatregularly produced handaxes far tothe west. The other is that Chinesehominins transmitted the knowledgeof handaxe manufacture acrossmany generations without ever pro-ducing them. Given what little weknow about population dynamics,life history traits, and patterns of dis-tribution so deep in the past bothscenarios seem unlikely, especially inview of the vulnerability of culturaltransmission to information loss.90

In contrast, the genetic transmis-sion hypothesis straightforwardlyexplains the occurrence of handaxes1,500 km from the Movius line: Thegenes responsible for the mainte-nance of such behaviors remainedunexpressed in the population untilproper raw material became avail-able, at which time the relevantgenes were expressed. The feasibilityof this scenario is illustrated by stud-ies showing that organisms are ableto respond to changing environmen-tal conditions by alterations to geneexpression.91

A fourth potential explanation forthe Chinese handaxes is worth noting.It is possible that the handaxes arethe result of convergent evolution;that is, hominins in China developedhandaxes independently from homi-nins on the west side of the MoviusLine, so that the Chinese handaxesare not, in fact, Acheulean handaxes.This is possible; over the last few dec-ades, it has become clear that conver-gent evolution is commonplace.92–94

However, the notion that the Chinesehandaxes and Acheulean handaxeswere invented independently is a less


parsimonious explanation than thegene expression-based explanation.

There is another geographic factabout handaxes that is illuminatingto consider from the perspective ofthe genetic transmission hypothesis,and that is the way in which variabil-ity in symmetry changes with dis-tance from Africa. Using a modelderived from population genetics,Lycett23 tested aspects of handaxesymmetry in relation to predictionsderived from the neutral theory ofevolution. His analyses indicatedthat the intra-assemblage variabilityof handaxe symmetry decreases withdistance from Africa and does so ina way that is inconsistent with itera-tive founder effect. On the basis ofthis, Lycett23 argues that aspects ofhandaxe symmetry were subject toselective forces. Not surprisingly,given the strength of belief in thefield that handaxes are culturalobjects, Lycett,23 throughout hispaper, assumes that the behaviorsinvolved in handaxe manufacturewere culturally learned. Lycett men-tions the possibility that geneticsinfluenced the design of handaxes,but does not take it seriously. Werespectfully suggest that this is amistake. If the variation in handaxesymmetry fits population geneticsmodels, it stands to reason that thebehaviors involved in the productionof handaxes may have been acquiredby genetic transmission.

Recent Evidence fromNeuroscience Is Consistent withthe Genetic TransmissionHypothesis

Neuroscience has shown thatpresent-day humans can gaugeintentions and feelings and can copyand predict the actions of others viamirror neurons.95,96 At first sight,this seems to support the currentapproach to Acheulean handaxes interms of domain-general social learn-ing. However, studies have begun tolink mirror neurons with domain-specific genetic preprogramming.Del Giudice, Manera, and Keysers97

argue that it is “reasonable to expectthat an ability as crucial for survivalas action recognition and learning

through observation would becomepre-programmed (‘innate’) to somedegree during phylogenetic histo-ry.”97:351 They go on to review empir-ical evidence suggesting thatHebbian learning, in which simulta-neous activation of neurons leads toincreased interconnection betweenthose neurons, and genetic preprog-ramming occur specifically in con-nection with manual action.

Another relevant insight from neu-roscience is that the neuronal cir-cuitry in the human brainresponsible for language productionnot only overlaps considerably withareas controlling manipulation butalso is under strong genetic control.Stout and colleagues98,99 argue thatboth tool-making and speech aremultilevel, hierarchically nested,sequential, and goal-directed motorsequences. They summarize theirargument as follows: “The observedpatterns of activation and of overlapwith language circuits suggest thattool-making and language share abasis in more general human capaci-ties for complex, goal-directed action.The results are consistent with co-evolutionary hypotheses linking theemergence of language, tool-making,population-level functional lateraliza-tion and association cortex expansionin human evolution.”99 We think thisline of enquiry is promising withrespect to a possible genetic compo-nent in handaxe design, but suspectthat the capacities in question maywell turn out to be much more spe-cific than Stout et al. suggest. Rele-vant points of comparison betweentool-making and speech are manualcontrol and articulatory control,action sequences and syntax, as wellas mental templates and semantics,with bird song as a promising thirdrelevant field.100

BIRD MODELS FOR HANDAXEMANUFACTURE

An obvious concern about thegenetic transmission hypothesis iswhether the production of an objectas complex as an Acheulean handaxecould really be under genetic control.We contend that the idea is notimplausible if handaxe production iscompared to some complex bird

behaviors. We begin by discussingbird song and bird tool-use as “soft”options that involve a combinationof genetic influence and social learn-ing. We then discuss a “hard” option— structure building by birds, inwhich strong genetic determinationinteracts with the availability of rawmaterial and local environmentalconditions but not with much, ifany, cultural learning.

Bird Song and Tool-Use

Bird song, a well-studied behavior,is strongly influenced by genetic pre-dispositions and implemented by dis-crete, well-defined neural circuits.101

Songbirds, hummingbirds, and par-rots have fixed, species-wide basicsongs from which individuals andregional subpopulations developtheir own varieties. Auditory-guidedvocal motor learning is important,for when members of these taxa areraised in isolation their songs remainsimple. The genetically controlledearly song of zebra finches (Taenio-pygia guttata), for instance, is plastic.The young need several months ofsocial learning to perfect their indi-vidual, adult song.102

It appears that in some bird spe-cies, tool use also combines fixedspecies-wide basic motor behaviorswith cultural learning. It has beenfound, for example, that hand-raisedCaledonian crows (Corvus monedu-loides) reliably develop leaf-toolmanufacture without ever havingobserved it in others, but neverdevelop the sophisticated behaviorsfound in wild populations whereindividuals are exposed to modelsand competitors. A study by Ken-ward and coworkers103 illustratesthis. By hand-rearing several crowsin isolation, they showed that thetool behavior of isolated individualswas so similar to that in wild onesthat it was difficult to avoid the con-clusion that the behavior is innate.They argue against “the extreme pos-sibilities that tool-use dependsentirely on social inputs (i.e., is sus-tained exclusively by cultural trans-mission and thus does not reflect adedicated evolved adaptation), andthat it has a purely individual,insight-based origin.”103:1340 If that


were the case, Kenward and col-leagues103:1340 contend, we wouldnot see “inherited action patternsthat must have evolved throughselection and that are crucial in sus-taining tool-oriented behavior inadult crows.” The same point hasbeen made with respect to the tool-use behavior of hyacinth macaws,Egyptian vultures, and woodpeckerfinches, the only other bird speciesthat is known to habitually use sticktools in the wild.104–106

Thus, if modeled on bird song andbird tool-use, the production ofAcheulean handaxes would haveinvolved both genetic transmissionand social learning. Raw materialselection, the manufacturing process,and basic design principles wouldhave been under genetic control, butfine-tuned through social learning.The latter would have been depend-ent on the presence of role modelsduring sensitive periods, but wouldnot necessarily have involved cul-tural group-specific templates orexplicit instruction. The combinationof genetic transmission and sociallearning is predicted to produce uni-formity of overall design (due to thefixed component) and slight localvariance (due to the process ofsocialization), which is the patternobserved with Acheulean handaxes.

Structure Building by Birds

It appears that one important reasonwhy archeologists tend to think thatthe behaviors involved in the produc-tion of handaxes were socially learnedis that the handaxe chaine op�eratoirerequired multiple decisions at multiplestages, from the selection of the size,shape, and quality of raw material tothe final retouch.15 Again, this argu-ment seems reasonable on its face.However, it is not as secure as it firstappears. The reason for this is that theconstruction of nests and other struc-tures by birds often involves long hier-archical sequences even though, as isgenerally agreed, the required behav-iors do not involve much, if any, cul-tural learning.

Eurasian long-tailed tits (Aegithaloscaudatus) exemplify the potentialcomplexity of nest building behavior.These birds create intricate sphericalnests from thousands of pieces oflichen, moss, and spiderweb, as wellas numerous feathers (Fig. 3). Thefinal product is accomplished througha combination of innate instructionsand individual learning.107,108 The for-mer specifies materials and methodsand a limited repertoire of repetitive,stereotyped actions. There is a chainof stimuli and responses that presup-poses construction rules and local

insight, but not necessarily complex,overall planning, conscious decision-making, or a mental image of the ulti-mate goal. The nest is an emergentproperty of the stimulus-responsechain.

The bower-birds (Ptilonorhynchi-dae) of New Guinea provide anotherexcellent set of examples. These birdscreate hut-like shelters and decoratetheir interiors and the area immedi-ately in front of them with brightlycolored objects (Fig. 4). These“bowers” are individually and ecologi-cally variable as a result of individuallearning and, possibly, some culturallearning. It is thought, however, thatthe stability of their basic form indi-cates that genes underlie their designand construction.109

As with handaxes, the bowersshow a mix of variation and uni-formity. Individual bowers displayvariation, but share a general formwithin a species. The same can besaid for handaxes: Beneath their var-iability, general themes are presentin all of them.

Structure building by birds is alsoheuristically relevant when trying toclarify exactly what was under geneticcontrol in handaxe manufacture. Itprobably was not just a simple targetform, but rather a predispositiontoward the basic behavioral routinesinvolved, such as invasive bifacialreduction while realizing cuttingedges in the secant plane, workingfrom the tip down, and keeping sym-metry. These routines would haveoperated in combination with causalunderstanding, manipulative skill,and intuitive (“folk”) physics.110

Although Gowlett111:217 thinks of han-daxes as cultural objects, our pro-posal is consistent with his approachwhen he argues against the notion ofa template as being “hard and fast”and instead talks of “instruction sets[from which] a computer could gener-ate the form of a biface mindlessly.”

Unfortunately, genetic evolution usu-ally happens too slowly for researchersto adequately study changes in geneti-cally controlled behaviors such as nestbuilding by birds. The time needed toobserve changes in evolutionary trendsextends far beyond the current lengthof study of such phenomena. Directcomparisons between the handaxe and

Figure 3. Nests created by long-tailed tits (Aegithalos caudatus), Great Britain. Long-tailed tits build their nests from thousands of pieces of lichen, moss, and spiderweb, aswell as numerous feathers. The nests are thought to be a result of innate instructions andindividual learning. Photograph A shows a nest under construction, while photograph Bshows a completed nest. (Photograph A: Alan Shearman/Wikimedia Commons; Photo-graph B: nottsexminer/Wikimedia Commons). [Color figure can be viewed in the onlineissue, which is available at wileyonlinelibrary.com.]



animal behaviors such as birds’ tooluse or structure building are thereforedifficult. Nevertheless, mounting evi-dence points to a central role for genesin maintaining aspects of animal tech-nology. We suggest that the sameshould be considered for Acheuleanhandaxes.

THE NATURE OF OTHERHANDAXE-LIKE TOOLS

It is important, under the genetictransmission hypothesis, to clarifythe nature of the other large cuttingtools in the Acheulean industry andthe handaxe-like bifaces in the indus-tries that come after the Acheulean.

Were these tools also partly geneti-cally determined or were they solelycultural phenomena?

McElreath’s52 discovery that in hisstudy-groups different mechanismsunderpinned attitudes toward friendsversus kin, respect for elders, andbelief in witchcraft clearly indicatesthat there is no reason to expect allAcheulean tools to involve the sametransmission mechanisms. Neverthe-less, we are of the opinion that theapplicability of the genetic transmis-sion hypothesis should be investi-gated in connection with the otherlarge cutting tools of the Acheulean,among them cleavers, picks, trihe-drals, and unifaces. Cleavers appear

to be a particularly good candidatefor another large cutting tool thatwas under at least partial geneticcontrol. Cleavers are bifaciallyreduced pieces with a straight cut-ting edge perpendicular to the longaxis of the piece. They were mostlymade on flake blanks struck frombifacially prepared, usually more orless tortoise-shaped cores. The clea-ver has been argued to be a subcate-gory of the Acheulean handaxe. Ifthis is the case, then parsimony sug-gests they could have been underpartial genetic control as well.

We suspect the situation is differ-ent with respect to the handaxe-liketools in the industries that cameafter the Acheulean. As we explainedearlier, in Europe the Acheulean wasreplaced by the industries of theMiddle Paleolithic between 300 and200 Ka. Several European Late Mid-dle Paleolithic industries containbifacial tools, including the FrenchMousterian of Acheulean Traditionand the Middle European Keilmes-sergruppen, also called the Mico-quian.112,113 The tools in questionresemble Acheulean handaxes, butwe think it is unlikely that they wereunder genetic control.

There are two reasons for this.First, Late Middle Paleolithic bifacesare actually only superficially similarto Acheulean handaxes. A recentmorphometric comparison of Acheu-lean and Late Middle Palaeolithichandaxes from Western Europe doneby Iovita and McPherron12:69 demon-strated that Acheulean handaxesshow an allometric pattern consist-ent with reduction of the tip, whileLate Middle Palaeolithic handaxes“show a pattern which implies amaintenance of shape throughoutthe reduction continuum.” Theauthors conclude from this that thetwo tool types are unrelated. In addi-tion, Late Middle Paleolithic bifacesseem to have been used in a differentway than Acheulean handaxes. Thelatter were primarily platforms onwhich various working edges (forexample, a point, a notch, or cuttingedge) were made, whereas MiddlePaleolithic bifaces usually served ascomplete tools with well integratedworking edges that remained inte-grated during resharpening.12,112,113

Figure 4. Bowers of the Vogelkop bowerbird (Amblyornis inornatus), western New Guinea.It is thought that the within-species stability of the basic form of these bowers indicatesthat genes underlie their design and construction. Note the organization of colorful andshiny items in the area in front of the bower in photograph A. In photograph B, note theway in which the sticks that form the bower are intricately woven together. (PhotographA: Tim Laman/naturepl.com; photograph B: Ingo Arndt/naturepl.com). [Color figure canbe viewed in the online issue, which is available at wileyonlinelibrary.com.]


http://Laman/naturepl.com

http://Arndt/naturepl.com


The other reason we think itunlikely that Late Middle Paleolithicbifaces were under genetic control isthat the industries they were part ofwere geographically restricted andrelatively short lived. Ruebens114 hasrecently reported a typo-technologicalanalysis of bifacial tools from lateMiddle Palaeolithic sites in severalEuropean countries, including Brit-ain, Belgium, the Netherlands, Franceand Germany. She found evidence ofthe existence of three differentregional entities, the Mousterian ofAcheulean Tradition, the Keilmesser-gruppen, and a new entity, which shecalls “the Mousterian with BifacialTools.” The Mousterian of AcheuleanTradition, located in southwestEurope, is dominated by handaxes.The Keilmessergruppen is found innortheast Europe and is distin-guished by backed and leaf-shapedbifacial tools. The Mousterian withBifacial Tools is located between theMousterian of Acheulean Traditionand the Keilmessergruppen and con-tains a wider variety of bifacial toolsthan either of them. That these andother Late Middle Paleolithic indus-tries were geographically restrictedand short-lived means they matchwhat we expect to see if the behaviorsinvolved in their manufacture weresocially learned.16,18 The corollary ofthis is that Late Middle Paleolithicbifaces are likely purely culturalobjects.

Based on the results of their morpho-metric analysis, Iovita and McPher-ron12 argue that Acheulean handaxesand Middle Paleolithic bifaces reflect“different and likely unrelated culturaland human evolutionary contexts.” Wesuggest a different scenario in whichgenetic transmission stopped being afactor in the behavioral routinesinvolved in the production of MiddlePaleolithic bifaces. How and why thishappened is an intriguing question but,given the connection between culturallearning and environmental variabilitydiscussed earlier, a link with anincrease in environmental variabilityseems likely.55–58

CONCLUSIONS

The remarkably conservative char-acter of the Acheulean handaxe —

limited variability in key aspects ofform over roughly 1.5 million yearsacross two continents, in almostevery type of ecology occupied byarchaic hominins — is almost unani-mously explained by cultural learn-ing, variously combined with suchfactors as individual learning, func-tional requirements, raw materialconstraints, similar ecology, reuseand resharpening, and drift. In lightof this apparent consensus, onewould expect that the status of han-daxes as cultural objects was settled

through careful study a long timeago and is backed by a substantialbody of evidence. This is not thecase, however. The assumption thatcultural transmission must havebeen the mechanism responsible forthe conservation of specific designcharacteristic of the handaxe can betraced back to the early nineteenthcentury, when they were first discov-ered in France. In the absence ofconcepts like genes or instinct, han-daxes were automatically perceivedas cultural products, similar to thoseof that period’s booming industrialtechnology. Crucially, the culturaltransmission hypothesis has goneuntested ever since. Handaxe pro-duction was, and still is, assumed bymost archeologists to be under-pinned by social learning. Therefore,

handaxes are treated as culturalobjects.

The cultural transmission hypothe-sis, however, faces several problems.It underestimates the difficulty ofinferring transmission mechanismsfrom artifacts. It conflicts with cul-tural evolutionary models and ethno-graphic data on key characteristicsof cultural objects. Moreover, thishypothesis is difficult to square withthe greatly increased rate of techno-logical change observed in post-Acheulean industries. It also fails toaccount adequately for the handaxesthat have recently been foundbeyond the Movius Line. We haveargued that these and other prob-lems can be better accommodated bywhat we call the genetic transmis-sion hypothesis, which is that han-daxe production was passed betweengenerations at least in part throughgenetic inheritance. In addition,there are two strands of positive evi-dence for the genetic transmissionhypothesis. First, handaxe variationaccords with predictions from mod-els of genetic evolution. Second, neu-roscientific findings suggest thatsimple tool use may be the result ofdomain-specific geneticpreprogramming.

In this paper, we have adopteddual-inheritance theory’s16,53,115,116

distinction between two channels ofinheritance, genetic and cultural,even though, in reality, traits mayvery well resist such neat partition-ing.16 We recognize that in doing sowe have opened ourselves up to thecharge of simply replacing one sim-plistic account with another simplis-tic implausible story. Even thoughthis may turn out to be correct, wehope that setting up the issue interms of a dichotomy will spark adebate on what has been unquestion-ably assumed by the vast majority ofarcheologists for more than acentury.

It may be helpful to clarify whatthe genetic transmission hypothesisdoes and does not entail. First, itdoes not imply strong instruction-ism, which is the idea that genescode for handaxe production in thesense of being uniquely responsiblefor specifying that behavior.117

Developmental factors may well have

there are two strands ofpositive evidence for thegenetic hypothesis. First,handaxe variationaccords with predictionsfrom models of geneticevolution; second, neu-roscientific findings sug-gest that simple tool usemay be the result ofdomain-specific geneticpreprogramming.


played an essential role in handaxemanufacture. Acheulean toolmakers,for instance, would have inheritedfrom other individuals niches con-taining stimuli related to handaxeproduction. Such niches would haveprovided cues for individual learn-ing. Their layout would also haveencouraged the task to be structuredin very specific, previously testedways.118

In accordance with much otherwork on cultural evolution, we have“black-boxed” specific mechanismsdriving the suggested genetic evolu-tion of handaxes, such as stabilizingor directional selection, genetic drift,genetic assimilation, and Baldwinianevolution. However, like Sterelny119

and others,120 we see the Baldwineffect as a plausible candidate.Learned responses to environmentalstimuli can be replaced by geneticallycontrolled behaviors that occur withlittle or no learning as a result ofselection pressure on the technicalabilities of the species. Accumulationof technology-related genes “steering”the developmental mechanism maythus gradually have resulted in a spe-cific, adaptive, extended phenotype,with basic technological behaviorsbecoming innate.

Another issue that we have leftopen is the nature of the Oldowan121

and the recently discovered Lomek-wian,122 the industries that precededthe Acheulean. It is tempting toargue that if the Acheulean handaxewas under genetic control, then themuch simpler tool types of the Oldo-wan and Lomekian were probablyalso under genetic control. However,as we explained earlier, McEl-reath’s52 finding that different mech-anisms underpin different beliefs inhis study groups means that there isno reason to expect that all tools inthe same industry involved the sametransmission mechanisms. Needlessto say, assuming that three entireindustries involved the same trans-mission mechanisms is even moreproblematic. Given this, we think theapplicability of the genetic transmis-sion hypothesis should be investi-gated in relation to pre-Acheuleanindustries, but we are uncertainabout what will be found. It is possi-ble that stone-tool technology went

from being genetically controlled topurely cultural over the course ofnearly three-and-a-half million years.But it is also possible that stone-tooltechnology went from purely culturalto genetically controlled to purelycultural again. Indeed, based onMcElreath’s52 results, we need toallow for an even more complicatedscenario in which the three indus-tries have different combinations oftransmission mechanisms.

Finally it is worth noting thatwhen nonhuman animals displaycomplex behavior, the defaultassumption is that it is under geneticcontrol. For complex behavior inhumans and other hominins, how-ever, the default position is to invokeculture and not genes. This doublestandard betrays a form of anthropo-centrism. We contend that thegenetic transmission hypothesisdeserves serious consideration forthis reason as well.123

ACKNOWLEDGMENTS

We thank Rob Foley, Wil Roebroeks,Dennis Sandgathe, John Shea, andnumerous other colleagues, includingseveral anonymous reviewers, for theirassistance with this paper. We alsothank various academic audiences fortheir feedback on presentations inwhich we have discussed the ideas out-lined here. Lastly, we thank the editorof Evolutionary Anthropology, JohnFleagle, for his encouragement in theface of considerable skepticism on thepart of certain reviewers, as well as forhis great patience. KV is supported bythe Netherlands Organization for Sci-entific Research. MC is supported bythe Canada Research Chairs Program,the Social Sciences and HumanitiesResearch of Canada, the Canada Foun-dation for Innovation, the BritishColumbia Knowledge DevelopmentFund, and Simon Fraser University.

REFERENCES

1 Beyene Y, Katoh S, WoldeGabriel G, et al.2013. The characteristics and chronology of theearliest Acheulean at Konso, Ethiopia. ProcNatl Acad Sci USA 110:1584–1591.

2 Deino AL, McBrearty S. 2002. 40Ar/39Ar dat-ing of the Kapthurin Formation, Baringo,Kenya. J Hum Evol 42:185–210.

3 Clark Y, Beyene G, WoldeGabriel G, et al.2003. Stratigraphic, chronological and behav-

ioural contexts of Pleistocene Homo sapiensfrom Middle Awash, Ethiopia. Nature 423:747–752.

4 Vaughan CD. 2001. A million years of styleand function: regional and temporal variationin Acheulean handaxes. In: Hurt TD and RakitaGFM, editors. Style and function: conceptualissues in evolutionary archaeology. Westport,CT: Bergin and Garvey. p 141–163.

5 Hublin JJ. 2014. Paleoanthropology: Homoerectus and the limits of a paleontological spe-cies. Curr Biol 24:R82–R84.

6 Stringer C. 2012. The status of Homo heidel-bergensis (Schoetensack 1908). Evol Anthropol21:101–107.

7 O’Brien E. 1981. The projectile capabilities ofan Acheulian handaxe from Olorgesailie. CurrAnthropol 22:76–9.

8 Calvin WH. 1993. The unitary hypothesis: acommon neural circuitry for novel manipula-tions, language, plan-ahead and throwing. In:Gibson KR, Ingold T, editors. Tools, languageand cognition in human evolution. Cambridge:Cambridge University Press. p 230–50.

9 Samson DR. 2006. Stones of contention: theAcheulean handaxe lethal projectile contro-versy. Lithic Technol 31:127–135.

10 Gamble C. 1997. Handaxes and Palaeolithicindividuals. In: Ashton N, Healey F, Pettitt P,editors. Stone age archaeology. Oxford: OxbowBooks. p 105–109.

11 Kohn M, Mithen S. 1999. Handaxes: prod-ucts of sexual selection? Antiquity 73:518–526.

12 Iovita R, McPherron SP. 2011. The handaxereloaded: a morphometric reassessment ofAcheulian and Middle Paleolithic handaxes.J Hum Evol 61:61–74.

13 White MJ. 1998. On the significance ofAcheulian biface variability in southern Britain.Proc Prehist Soc 64:15–44.

14 McNabb J. 2005. Hominids and the Early-Middle Pleistocene transition: evolution, cultureand climate in Africa and Europe. In: HeadMJ, Gibbard PL, editors. Early-Middle Pleisto-cene transitions: the land-ocean evidence. Lon-don: Geological Society. p 287–304.

15 Nicoud E. 2013 Le ph�enomene acheul�een eneurope occidentale: approche chronologique,technologie lithique et implications culturelles.Paris: �Editions du Comit�e des travaux histori-ques et scientifiques.

16 Richerson PJ, Boyd R. 2005. Not by genesalone: how culture transformed human evolu-tion. Chicago: University of Chicago Press.

17 McNabb J. 2005. More on Acheulean tools.Reply to Hodgson. Curr Anthropol 46:648–649.

18 Foley RA. 1987. Hominid species and stonetools assemblages: how are they related? Antiq-uity 61:380–392.

19 Mithen S. 1999. Imitation and culturalchange: a view from the Stone Age, with spe-cific reference to the manufacture of handaxes.In: Box HO, Gibson KR, editors. Mammaliansocial learning: comparative and ecological per-spectives. Cambridge: Cambridge UniversityPress. p 389–399.

20 Shipton C, Petraglia MD, Paddayya K. 2009.Inferring aspects of Acheulian sociality andcognition from biface technology. In: AdamsB, Blades S, editors. Lithic materials and Palae-olithic societies. Chichester: Wiley-Blackwell. p219–231.

21 Shennan SJ, Steele J. 1999. Cultural learn-ing in hominids: a behavioural ecologicalapproach. In: Box HO, Gibson KR editors.Mammalian social learning: comparative and


ecological perspectives. Cambridge: CambridgeUniversity Press. p 367–388.

22 Bar-Yosef O. 2006. The known and theunknown about the Acheulean. In: Goren-Inbar N, Sharon G, editors. Axe age: Acheuliantool-making from quarry to discard. London:Equinox. p 479–494.

23 Lycett SJ. 2008. Acheulean variation andselection: does handaxe symmetry fit neutralexpectations? J Archaeol Sci 35:2640–2648.

24 Lycett SJ, von Cramon-Taubadel N. 2008.Acheulean variability and hominin dispersals: amodel-bound approach. J Archaeol Sci 35:553–562.

25 Cole JN. 2011. Hominin cognitive andbehavioural complexity in the Pleistocene:assessment through identity, intentionality andvisual display. Unpublished doctoral disserta-tion, University of Southampton, UK.

26 Mithen S. 1994. Technology and society dur-ing the Middle Pleistocene: hominid group size,social learning and industrial variability.Cambr Archaeol J 4:3–32.

27 Wynn T, Tierson F. 1990. Regional compari-son of the shapes of later Acheulean handaxes.Am Anthropol 92:73–84.

28 Roe D. 2003. An overview, with somethoughts on the study of bifaces. In: SoressiM, Dibble H, editors. Multiple approaches tothe study of bifacial technologies. Philadelphia:University of Pennsylvania Museum of Archae-ology and Anthropology Press. p 273–285.

29 Wenban-Smith FF. 2006. Handaxe typologyand Lower Palaeolithic cultural development:ficrons, cleavers and two giant handaxes fromCuxton. Lithics 25:11–21.

30 Machin AJ. 2009. The role of the individualagent in Acheulean biface variability: a multi-factorial model. J Soc Archaeol 9:35–58.

31 Eerkens JW, Lipo CP. 2005. Cultural trans-mission, copying errors, and the generation ofvariation in material culture and the archaeo-logical record. J Anthropol Archaeol 24:316–334.

32 Kempe M, Lycett S, Mesoudi A. 2012. Anexperimental test of the accumulated copyingerror model of cultural mutation for Acheuleanhandaxe size. PLoS ONE 7:e48333.

33 Yamakoshi G, Sugiyama Y. 1995. Pestle-pounding behavior of wild chimpanzees at Bos-sou, Guinea: a newly observed tool usingbehavior. Primates 36:489–500.

34 McGrew WC. 1992. Chimpanzee materialculture: implications for human evolution.Cambridge: Cambridge University Press.

35 Matsuzawa T. 1994. Field experiments onuse of stone tools by chimpanzees in the wild.In: Wrangham RW, McGrew WC, de WaalFBD, et al., editors. Chimpanzee cultures. Cam-bridge, MA: Harvard University Press. p 351–370.

36 Boesch C, Boesch-Achermann H. 2000. Thechimpanzees of the Ta€ı Forest: behaviouralecology and evolution. Oxford: Oxford Univer-sity Press.

37 Whiten A, Goodall J, McGrew WC, et al.1999. Cultures in chimpanzees. Nature 399:682–685.

38 Whiten A, Goodall J, McGrew WC, et al.2001. Charting cultural variation in chimpan-zees. Behavior 138:1481–1516.

39 McGrew WC. 2004. The cultured chimpan-zee: reflections on cultural primatology. Cam-bridge: Cambridge University Press.

40 Whiten A. 2011. The scope of culture inchimpanzees, humans and ancestral apes.Philos Trans R Soc B 366:997–1007.

41 Whiten A. 2014. Animal behaviour: incipienttradition in wild chimpanzees. Nature 514:178–179.

42 Laland KN, Hoppitt W. 2003. Do animalshave culture? Evol Anthropol 12:150–159.

43 Laland KN, Janik VM. 2006. The animal cul-tures debate. Trends Ecol Evol 21:542–547.

44 Galef BG. 2009. Culture in animals? In:Laland K, Galef BG, editors. The question ofanimal culture. Cambridge, MA: Harvard Uni-versity Press. p 222–246.

45 Laland K, Kendal JR, Kendal RL. 2009. Ani-mal culture: problems and solutions. In:Laland K, Galef BG, editors. The question ofanimal culture. Cambridge, MA: Harvard Uni-versity Press. p 174–197.

46 Langergraber KE, Boesch B, Inoue E, et al.2011. Genetic and “cultural” similarity in wildchimpanzees. Proc R Soc B London 278:408–416.

47 Lycett S, Collard M, McGrew WC. 2007.Phylogenetic analyses of behavior support exis-tence of culture among wild chimpanzees. ProcNatl Acad Sci USA 104:17588–17592.

48 Lycett SJ, Collard M, McGrew WC. 2009.Cladistic analysis of behavioural variation inwild Pan troglodytes: exploring the chimpanzeeculture hypothesis. J Hum Evol 57:337–349.

49 Lycett SJ, Collard M, McGrew WC. 2010.Are behavioural differences among wild chim-panzee communities genetic or cultural? Anassessment using tool-use data and phyloge-netic methods. Am J Phys Anthropol 142:461–467.

50 Lycett SJ, Collard M, McGrew WC. 2011.Correlations between genetic and behaviouraldissimilarities in wild chimpanzees (Pan troglo-dytes) do not undermine the case for culture.Proc R Soc B 278:2091–2093.

51 Langergraber KE, Boesch C, Inoue E, et al.2011. Genetic and “cultural” similarity in wildchimpanzees. Proc R Soc B 278:408–416.

52 McElreath R. 2004. Social learning and themaintenance of cultural variation: an evolution-ary model and data from East Africa. AmAnthropol 106:308–321.

53 Boyd R, Richerson PJ. 1985. Culture andthe evolutionary process. Chicago: Chicago Uni-versity Press.

54 Boyd R, Richerson PJ. 1996. Why culture iscommon, but cultural evolution is rare. Proc BrAcad 88:77–93.

55 Potts R. 1996. Evolution and climate vari-ability. Science 273:922–923.

56 Potts R, Faith JT. n.d. Alternating high andlow climate variability: the context of naturalselection and speciation in Plio-Pleistocenehominin evolution. J Hum Evol. In press.

57 de Menocal PB. 2011. Climate and humanevolution. Science 311:540–541.

58 Maslin MA, Shultz S, Trauth MH. 2015. Asynthesis of the theories and concepts of earlyhuman evolution. Philos Trans R Soc B 370:20140064.

59 Mithen S. 1996. The prehistory of the mind:the cognitive origins of art and science. Lon-don: Thames and Hudson.

60 Klein R. 2000. Archeology and the evolutionof human behavior. Evol Anthropol 9:17–36.

61 Mellars P. 2005. The impossible coinci-dence: a single-species model for the origins ofmodern human behaviour in Europe. EvolAnthropol 14:12–27.

62 Wynn T, Coolidge F. 2007. Did a small butsignificant enhancement in working memorycapacity power the evolution of modern think-ing? In: Mellars P, Boyle K, Bar-Yosef O, et al.,

editors. Rethinking the human revolution.Cambridge: McDonald Institute Monographs.

63 Vaesen K. 2012. The cognitive bases ofhuman tool use. Behav Brain Sci 35:203–218.

64 d’Errico F, Henshilwood C, Lawson G, et al.2011. Archaeological evidence for the emer-gence of language, symbolism, and music: analternative multidisciplinary perspective.J World Prehist 17:1–70.

65 Zilhao J. 2007. The emergence of ornamentsand art: an archaeological perspective on theorigins of “behavioural modernity.” J ArchaeolRes 54:1–54.

66 Villa P, Roebroeks W. 2014. Neandertaldemise: an archaeological analysis of the mod-ern human superiority complex. PLoS ONE 9:e96424.

67 Henrich J. 2004. Demography and culturalevolution. How adaptive cultural processes canproduce maladaptive losses: the Tasmaniancase. Am Antiq 69:197–214.

68 Powell A, Shennan S, Thomas M. 2009. LatePleistocene demography and the appearance ofmodern human behaviour. Science 324:1298–1301.

69 Querbes A, Vaesen K, Houkes WN. 2014.Complexity and demographic explanations ofcumulative culture. PLoS ONE 9:e102543.

70 Vaesen K. 2012. Cumulative cultural evolu-tion and demography. PLoS ONE 7:e40989.

71 Collard M, Kemery MD, Banks SJ. 2005.Causes of toolkit variation among hunter-gath-erers: a test of four competing hypotheses. CanJ Archaeol 29:1–19.

72 Collard M, Buchanan B, O’Brien MJ, et al.2013. Risk, mobility, or population size? Driv-ers of technological diversity among recentwestern North American hunter-gatherers.Philos Trans R Soc B 368:20120412.

73 Collard M, Buchanan B, O’Brien MJ. 2013.Population size as an explanation for patternsin the Paleolithic archeological record: Morecaution is needed. Curr Anthropol 54 S8: S388–S396.

74 Anderson C, Read D. n.d. The evolution ofcultural complexity: not by the treadmill alone.Curr Anthropol. In press.

75 Buchanan B, O’Brien MJ, Collard M. n.d.Drivers of technological richness in prehistoricTexas: an archaeological test of the risk andpopulation size hypotheses. Archaeol AnthropolSci. In press.

76 Schiffels S, Derbin R. 2014. Inferringhuman population size and separation historyfrom multiple genome sequences. Nat Genet46:919–925.

77 Heyes C. 2012. What’s social about sociallearning? J Comp Psychol 126:193–202.

78 Hill KR, Walker RS, Bo�zicevic M, et al.2011. Co-residence patterns in hunter-gatherersocieties show unique human social structure.Science 331:1286–1289

79 Durkheim E. 1995. The elementary forms ofreligious life. New York: Free Press.

80 Sosis R. 2000. Religion and intra-groupcooperation: preliminary results of a compara-tive analysis of Utopian communities. Cross-Cult Res 34:70–87.

81 Irons W. 2001. Religion as a hard-to-fakesign of commitment. In: Nesse R, editor. Theevolution of commitment. New York: RussellSage Foundation. p 292–309.

82 Wilson DS. 2002. Darwin’s cathedral: evolu-tion, religion and the nature of society. Chi-cago: University of Chicago Press.

83 Sosis R, Ruffle BJ. 2003. Religious ritualand cooperation: testing for a relationship on


Israeli religious and secular kibbutzim. CurrAnthropol 44:713–722.

84 Xygalatas D. 2013. Effects of religious set-ting on cooperative behaviour: a case studyfrom Mauritius. Relig Brain Behav 3:91–102.

85 Schloss JP, Murray MJ. 2011. Evolutionaryaccounts of belief in supernatural punishment:a critical review. Relig Brain Behav 1:46–99.

86 Norenzayan A. 2013. Does religion makepeople moral? Behaviour 151:365–384.

87 Movius HL, Jr. 1948. The Lower Palaeolithiccultures of southern and eastern Asia. TransAm Philos Soc 38:329–420.

88 Lycett SJ, Bae CJ. 2010. The Movius Linecontroversy: the state of the debate. WorldArchaeol 42:521–544

89 Zhang P, Huang W, Wang W. 2010. Acheu-lean handaxes from Fengshudao, Bose sites ofSouth China. Quatern Int 223–224:440–443.

90 Mesoudi A, Whiten A, Dunbar R. 2006. Abias for social information in human culturaltransmission. Br J Psychol 97:405–423.

91 L�opez-Maury L, Marguerat S, B€ahler J.2008. Tuning gene expression to changing envi-ronments: from rapid responses to evolutionaryadaptation. Nat Rev Genet 9:583–593.

92 Sanderson MJ, Hufford L, editors. 1996.Homoplasy: the recurrence of similarity in evo-lution. New York: Academic Press.

93 Lockwood CA, Fleagle JG. 1999. The recog-nition and evaluation of homoplasy in primateand human evolution. Am J Phys Anthropol110:189–232.

94 Collard M, Wood BA. 2001. Homoplasy andthe early hominid masticatory system: infer-ences from analyses of living hominoids andpapionins. J Hum Evol 41:167–194.

95 Iacoboni M, Dapretto M. 2006. The mirrorneuron system and the consequences of its dys-function. Nat Rev Neurosci 7:942–951.

96 Keysers C, Gazzola V. 2006. Towards a uni-fying neural theory of social cognition. ProgBrain Res 156:379–401.

97 Del Giudice M, Manera V, Keysers C. 2008.Programmed to learn? The ontogeny of mirrorneurons. Dev Sci 12:350–363.

98 Stout D, Chaminade T. 2009. Making toolsand making sense: complex, intentional behav-iour in human evolution. Cambr Archaeol J 19:85–96.

99 Stout D, Toth N, Schick K, et al. 2008. Neu-ral correlates of Early Stone Age toolmaking:technology, language and cognition in humanevolution. Philos Trans R Soc B1499:1939–1949.

100 Feenders G, Liedvogel M, Rivas M, et al.2008. Molecular mapping of movement-associated areas in the avian brain: a motor

theory for vocal learning origin. PLoS ONE 3:e1768.

101 Scharff C, Adam I. 2013. Neurogenetics ofbirdsong. Curr Opin Neurobiol 23:29–36.

102 Brenowitz EA, Bleecher MD. 2005. Songlearning in birds: diversity and plasticity,opportunities and challenges. Trends Neurosci28:127–132.

103 Kenward B, Rutz C, Wiet AAS, et al. 2006.Development of tool use in New Caledoniancrows: inherited action patterns and socialinfluences. Anim Behav 72:1329–1343.

104 Borsar A, Ottoni EB. 2005. Preliminaryobservations of tool use in captive hyacinthmacaws (Anodorhynchus hyacinthinus). AnimCognit 8:48–52.

105 Tebbich S, Taborsky M, Fessl B, et al.2001. Do woodpecker finches acquire tool-useby social learning? Proc R Soc Lond B 268:2189–2193.

106 Thouless CR, Fanshawe JH, Bertram BCR.1989. Egyptian vultures Neophron percnopterusand ostrich Struthio camelus eggs: the originsof stone-throwing behaviour. Ibis 131:9–15.

107 Hansell M. 2000. Bird nests and construc-tion behaviour. Cambridge: Cambridge Univer-sity Press.

108 Hansell M. 2007. Built by animals: the nat-ural history of animal architecture. Oxford:Oxford University Press.

109 Madden JR. 2008. Do bowerbirds exhibitcultures? Anim Cognit 11:1–12.

110 Penn D, Povinelli DJ. 2007. Causal cogni-tion in human and nonhuman animals: a com-parative, critical review. Ann Rev Psychol 58:97–118.

111 Gowlett JAJ. 2006. The elements of designform in Acheulian bifaces: modes, modalities,rules and language. In: Goren-Inbar N, SharonG, editors. Axe age: Acheulian tool-making fromquarry to discard. London: Equinox. p 203–221.

112 Soressi M. 2002. Le Moust�erien de Tradi-tion Acheul�eenne du sud-ouest de la France.Bordeaux: Universit�e Bordeaux I.

113 J€oris O. 2003. Zur chronostratigraphischenStellung der sp€atmittelpal€aolithischen Keilmes-sergruppen. Versuch einer kulturgeographi-schen Abgrenzung einer mittelpal€aolithischenFormengruppe und ihr europ€aischer Kontext.Berlin R€om.-German. Kommission 84:49–153.

114 Ruebens K. 2013. Regional behaviouramong late Neanderthal groups in WesternEurope: a comparative assessment of late Mid-dle Palaeolithic bifacial tool variability. J HumEvol 65:341–62

115 Cavalli-Sforza LL, Feldman MW. 1976.Cultural transmission and evolution: a quanti-tative approach. Princeton, NJ: Princeton Uni-versity Press.

116 Durham WH. 1991. Coevolution: genes,culture, and human diversity. Stanford, CA:Stanford University Press.

117 Wheeler M, Clark A. 1999. Genic represen-tation: reconciling content and causal complex-ity. Br J Philos Sci 50:103–135.

118 Wheeler M, Clark A. 2008. Culture, embod-iment and genes: unravelling the triple helix.Philos Trans R Soc B 363:3563–3575.

119 Sterelny K. 2004. Review of evolution andlearning: the Baldwin effect reconsidered,Weber B, Depew D, editors. Evol Dev 6:295–300.

120 Weber B, Depew D, editors. 2003. Evolu-tion and learning: the Baldwin effect reconsid-ered. Cambridge, MA: MIT Press.

121 Schick K, Toth NP. 2006. An overview of theOldowan industrial complex: the sites and thenature of their evidence. In: Toth NP, Schick K,editors. The Oldowan: case studies into the ear-liest Stone Age. Bloomington: Stone Age Insti-tute Press. p 3–42.

122 Harmand S, Lewis JE, Feibel CS, et al.2015. 3.3-million-year-old stone tools fromLomekwi 3, West Turkana, Kenya. Nature 521:310–315.

123 Roebroeks W, Corbey R. 2001. Biases anddouble standards in palaeoanthropology. In:Corbey R, Roebroeks W, editors. Studyinghuman origins: disciplinary history and episte-mology. Amsterdam: Amsterdam UniversityPress. p 67–76.

124 Roberts MB, Parfitt SA, Pope, MI, et al.1997. Excavation of the Lower Palaeolithic siteat Amey’s Eartham Pit, Boxgrove, West Sussex.Proc Prehist Soc 63:303–358.

125 Sharon, G. 2011. New Acheulian localitynorth of Gesher Benot Ya’aqov: contribution forthe study of the Levantine Acheulian. In: J.-M.Le Tensorer, R. Jagher, M. Otte, editors. TheLower and Middle Palaeolithic in the MiddleEast and neighboring regions. Liege: Centre derecherches arch�eologiques de l’Universit�e deLiege. p 25–33.

126 Vaufrey R. 1955. Pr�ehistoire de l’Afrique,vol I. Paris: Masson.

127 Kuman K. 1994. The archaeology of Sterk-fontein — past and present. J Hum Evol 27:471–495.

128 Leakey M. 1971. Olduvai Gorge, vol III:excavations in Beds I and II, 1960–1963. Cam-bridge: Cambridge University Press.

129 Paddayya K, Jhaldiyal R, Petraglia MD.2006. The Acheulian quarry at Isampur, LowerDeccan, India. In: Goren-Inbar N, Sharon G,editors. Axe age: Acheulian toolmaking fromquarry to discard. London: Equinox. p 45–74.

VC 2016 Wiley Periodicals, Inc.


the acheulean handaxe

Documents