Evolution and Human Behavior 37 (2016) 202–209

Contents lists available at ScienceDirect

Evolution and Human Behavior

j ourna l homepage: www.ehbon l ine .org

Original Article

Hadza handedness: Lateralized behaviors in a contemporary

hunter–gatherer population

Tammany Cavanagh a,⁎, J. Colette Berbesque a, Brian Wood b, Frank Marlowe c

a Centre for Research in Evolutionary and Environmental Anthropology, Roehampton University, London, UKb Department of Anthropology, Yale University, New Haven, CT, USAc Department of Archaeology and Anthropology, Cambridge University, Cambridge, UK

a b s t r a c ta r t i c l e i n f o

⁎ Corresponding author.E-mail address: tammany.cavanagh@oxon.org (T. Cav

http://dx.doi.org/10.1016/j.evolhumbehav.2015.11.0021090-5138/© 2016 Elsevier Inc. All rights reserved.

DownlFor personal use

Article history:Initial receipt 22 June 2015Final revision received 19 November 2015

Keywords:AsymmetryLateralityHandednessHadzaHunter–gatherersTool-use

Handedness has been studied extensively in Western populations and in non-human primates. Seldom,however, has this focus been extended to non-industrialized populations. This study examines handednessfrom an evolutionary perspective in an ecologically meaningful setting. The sampled population, the Hadza ofTanzania, are a contemporary hunter–gatherer people whose handedness behaviors represent an important,and yet neglected, avenue of investigation. Ethological sampling methods were utilized based on recordedvideo footage of an ethnographic nature. This footage was coded and analyzed to document dominant handuse in a range of evolutionarily relevant activities, such as hunting and foraging tasks. Strength of lateralizationwas compared across task categories. The results demonstrate that the Hadza are extremely lateralized in tool-use tasks, with 96% of all tool-use tasks being executedwith the right hand. Importantly, strength of lateralizationwas not influenced by degree of tool complexity. They were less strongly lateralized for more commonplace,empty-handed tasks (such as eating or communicative gestures). Theywere the least lateralized in objectmanip-ulation tasks. These results support theoretical scenarios which place tool use at the centre in the evolution ofhuman right-handedness.

anagh).

oaded from ClinicalKey.com at Yale New Haven Healthonly. No other uses without permission. Copyright ©201

© 2016 Elsevier Inc. All rights reserved.

1. Introduction

Handedness, particularly right-handedness, is a phenomenon thatcan be seen across the entire human species, and to the best of ourknowledge no other organism displays such consistent, species-levelhand dominance (Braccini, Lambeth, Schapiro, & Fitch, 2010; Cashmore,Uomini, & Chapelain, 2008; Cochet & Byrne, 2013; Corballis, 2003;Forrester, Quaresmini, Leavens, Mareschal, & Thomas, 2013; Frayeret al., 2012; Marchant & McGrew, 2013; McGrew & Marchant, 1997;Steele & Uomini, 2005; Uomini, 2009). Handedness in non-human pri-mates has been documented at group-level (e.g., a troop of baboons)but has not been conclusively demonstrated at population-level(e.g., all chimpanzees) (Marchant & McGrew, 2013). Human patternsof handedness, then, are conspicuous in relation to other species.

Complex bimanual tasks have been studied extensively in non-human primates (Meguerditchian, Vauclair, & Hopkins, 2013), withstrong individual- and group-level biases being reported (Canteloup,Vauclair, & Meunier, 2013; Hopkins et al., 2011; Llorente et al., 2011;Mangalam, Desai, & Singh, 2014; Mosquera, Geribàs, Bargalló, Llorente,& Riba, 2012). Hopkins and colleagues' (2011) examination of handed-ness behaviors of 777 (captive) great apes claims to show population-level biases exist for one specific extraction task in chimpanzees and

gorillas (which exhibit right-handedness), and orang-utans (whichshow left-handedness). Meguerditchian, Vauclair, and Hopkins's(2010) analysis of intra-species communicative gestures in chimpan-zees showed a group-level right-hand bias. This has been taken as evi-dence of a precursor to left-lateralized language dominance in humans(Corballis, 2014; Corballis, Badzakova-Trajkov, & Häberling, 2012;Meguerditchian et al., 2010). This approach is further supported by neu-roimaging studies, which report structural asymmetries in the brains ofgreat apes in areas homologous to Broca's area in humans (Cantalupo &Hopkins, 2001). But see Cochet and Byrne (2013), as well as Marchantand McGrew (2013), for studies that do not show lateralized behaviorsin these contexts. At present, and in opposition to the consistency ofdata from human studies, the overall picture of handedness in non-human primates is one which lacks a general consensus.

There is an abundance of literature concerning handedness in con-temporary humans, however, much of the research is concerned pri-marily with industrialized populations (Marchant, McGrew, & Eibl-Eibesfeldt, 1995). Few attempts have been made to sample non-industrialized populations. Exceptions are Marchant and colleagues'(1995) study of three contemporary non-industrialized groups, ofwhich the G/wi San of Botswana were the only hunter–gatherers, andFaurie, Schiefenhövel, Bomin, Billiard, and Raymond's (2005) study ofthe Baka of northern Gabon, for which only one handed behavior (ma-chete use)was recorded—making this studymore ameasure of handedtask specialization (see Methods). Clearly, there is still a distinct lack of

System April 21, 2016.6. Elsevier Inc. All rights reserved.

203T. Cavanagh et al. / Evolution and Human Behavior 37 (2016) 202–209

research in this area. These studies do, however, validate the assump-tion that similar handedness ratios exist across cultures and geographi-cal locations, although there is some variation in the ratio of left- andright-handers in any given population (Faurie et al., 2005; Raymond &Pontier, 2004).

No human societies exist in which the majority of individuals are si-nistral (or left-handed) (Marchant & McGrew, 2013), neither are theresocieties in which left-handedness is completely absent (Annett,2002). While the frequency of sinistrality may vary cross-culturally, itnever approaches the 50% or above expected from a neutral phenotype(Faurie et al., 2005; Raymond & Pontier, 2004). Faurie and colleagues'(2005) study of four “traditional” societies estimated variation incross-cultural frequencies of left-handers from approximately 3.3% toan unusually high 26.9% — although some of these data may perhapsbe more accurately described as representing task preference thanhandedness as, in some cases, only one task was sampled (Marchant &McGrew, 2013— see Methods). In industrialized populations, however,research has shown a more stable ratio of right- and left-handers ataround 9:1 (Annett, 2002), documented for as long as the historical re-cord allows (Corballis, 2014).

Here, we analyze data on handedness collected on a contemporarypopulation of hunter–gatherers, the Hadza of Tanzania. We follow themethodological approach suggested by Marchant and McGrew (2013),and investigate how handedness is expressed by recording “multiplemeasures of performance on evolutionarily meaningful tasks”(McGrew & Marchant, 1997: 215). Handedness information from theHadza is highly relevant because “…the clearest data are likely tocome from societies still practicing traditional lifestyles, not yet subjectto the imposition of foreign constraints, such as classroom education ornewly arrived industrial technology” (Marchant et al., 1995: 241).Schooling has been found to have an effect on hand preference (DeAgostini, Khamis, Ahui, & Dellatolas, 1997; Geuze et al., 2012), as has in-creased sedentism associated with an industrialized country(Tutkuviene & Schiefenhövel, 2013). These factors are largely absentor limited in the Hadza population. In order for a study of human hand-edness to be relevant, the datamust be ecologically valid. “The key issueis ecological validity… [e]ach step removed from the real world …makes more difficult a functional explanation of laterality as a productof natural and cultural selection” (Marchant et al., 1995: 240). The aimof this study is to make a contribution to the cross-cultural data in thisarea. We also aim to determine if there is a significant difference in lat-eralization between categories of ecologically salient tasks which mayhave deep roots in our evolutionary history.

Despite much research, the timing and evolutionary circumstancesunder which manual laterality arose in Homo sapiens remain unclear(Cashmore et al., 2008). It appears that the earliest, reliably-established record of right-hand dominance at group level comes fromHomo neanderthalensis skeletal remains (Frayer et al., 2012; Lozano,Bermúdez de Castro, Carbonell, & Arsuaga, 2008; Uomini, 2009), al-though there are some tentative inferences of hand preference at the in-dividual level from earlier fossils (Uomini, 2009). According to someauthors, however, examination of the archeological record shows evi-dence of right-hand bias going back 500 kya (Mosquera et al., 2012)or even 2 mya (Cashmore et al., 2008). Handedness has been inferredin the archeological record in a variety of ways, however, many ofthese have proved problematic and are contested (Bax & Ungar, 1999;Cashmore et al., 2008; Uomini, 2009).

It is believed by many that both language and right-handpopulation-level dominance are so explicitly, and exclusively, expressedin humans (Steele & Uomini, 2005), that these two aspects must be re-lated within an evolutionary framework (e.g., Annett, 2002; Falk, 1987;McManus, 2002). These theories are supported by the common associ-ation of both right handedness and the production of speech and lan-guage with the left hemisphere of the brain, particularly Broca'sarea — although this is by no means a universal human characteristic(Knecht et al., 2000). If these traits are, in fact, related, studies of

Downloaded from ClinicalKey.com at Yale NewFor personal use only. No other uses without permission. C

human patterns of handedness may also have implications for theoriesregarding the evolution of language (Falk, 1987; Mosquera et al., 2012).Many other authors support a neuroanatomical link between languagecapacity and hand lateralization (e.g., Frayer et al., 2012; Knecht et al.,2000), although the association may be more complex than was previ-ously recognized (Corballis et al., 2012).

The “tool use theory”, and variations thereof (Ambrose, 2001;Kimura, 1979), suggest that lateralization may have “… developed inconjunction with the asymmetric activity of the two limbs during tooluse, the left hemisphere … becoming the hemisphere specialised forprecise sequential limb positioning” (Kimura, 1979: 203). In this scenar-io, cognitive development increased as tool making techniques movedfrom reductive to composite (Ambrose, 2001). Another proposal for se-lection favoring increased cognitive abilities is Calvin's (1983) theory oflateralization evolving through dependence on throwing actions and re-lated targeted movements. In a similar vein, Fagot and Vauclair (1991)proposed “task complexity” model argues that lateralization is depen-dent on the cognitive and manipulatory demands of the task, withstronger lateralization being exhibited for novel or complex tasks thanfor simple ones.

If task complexity is associatedwith strong lateralization in primates(Fagot & Vauclair, 1991), and if this is homologous to humans, thearcheological record should show a progression from simple tools asso-ciated with hand preference, to complex tools associated with consis-tent, group-level lateralization (Uomini, 2009). A study of three non-industrialized societies identified differences in strength of lateraliza-tion according to the nature of the task or activity, with tool-use evokingthe greatest bias (Marchant et al., 1995),which lends support to this hy-pothesis. This effect of enhanced lateralization is linked to bimanual,complementary tasks, in which both hands operate to perform coordi-nated but asymmetrical functions (Braccini et al., 2010; Guiard, 1987).This configuration of handednessmay represent a significant advantagein hand function specialization, in the use of tools, and in the executionof complex tasks, such as extractive foraging.

2. Methods

2.1. Previous handedness methods

Evaluation of handedness is complicated by the variety of methodsused thus far that do not always allow direct comparison across sam-ples, including between humans and other primates (Cochet & Byrne,2013). Considering non-human primates, while there is a body ofwork demonstrating manual and/or task specialization in certain indi-viduals and groups, most studies that claim handedness of a populationhave been criticized as being fraught with methodological issues(e.g., only one task assessed), and contextual problems, such as wildversus captive populations (Marchant & McGrew, 2013; McGrew &Marchant, 1997). Lateralized hand function is, furthermore, most ofteninduced experimentally, which has been criticized on multiple grounds(Marchant & McGrew, 2013).

To complicate matters further, much of the data that constitute ourknowledge of human handedness are derived from industrialized pop-ulations (Faurie et al., 2005), and data from non-Western groups areoften markedly context- or task-specific (Marchant et al., 1995). Ques-tionnaires are often unsuitable (Steele &Uomini, 2005), unreliable as in-dicators (Cochet & Byrne, 2013) and/or biased towardWestern culturalframeworks (Marchant et al., 1995). A further issue confronting com-parison of studies of handedness is the generalization of manual behav-iors that donot represent explicit handedness of a population.Marchantand McGrew (2013) specify four categorical types of manual lateralityaccording to subject and task: (1) hand preference, defined as onehand preferred by one subject for one type of task; (2) task specializa-tion, defined as one hand preferred by most subjects for one type oftask; (3) manual specialization, defined as one hand preferred by onesubject for most tasks; and (4) handedness: one hand preferred by

Haven Health System April 21, 2016.opyright ©2016. Elsevier Inc. All rights reserved.

204 T. Cavanagh et al. / Evolution and Human Behavior 37 (2016) 202–209

most subjects for most tasks. According to these authors, many studiesconfuse these categories, leading them to claim false positive results(Marchant &McGrew, 2013). They propose that the only way to clearlyidentify handedness bias in a population is to focus on normative hand-use for most tasks performed by most individuals.

While it may not be possible to infer explicit handedness of theHadza from the limited sample size of this study, data on manual andtask specialization are also important to understanding lateralization(Dunham & Davenport, 2012). We here aim to follow the methodolog-ical approach suggested by Marchant and McGrew (2013), and investi-gate how handedness is expressed by recording “multiple measures ofperformance on evolutionarilymeaningful tasks” (McGrew&Marchant,1997: 215).

2.2. Subjects and setting

The sampled population of this study, the Hadza of Tanzania, are ahunter–gatherer group who live in savannah/woodland areas aroundLake Eyasi in the north of the country. Hadza ecology probably hasmuch in common with pre-agricultural human societies (Marlowe,2010) and therefore represents away to examinehandedness in an eco-logically salient context. The Hadza inhabit an area of East Africa thathas an extensive history of hominin occupation (Marlowe, 2002). Fur-thermore, the environmental resources exploited by the Hadza todayare probably similar to those exploited by our ancestors 5 mya. In com-parison to other contemporary foragers, Hadza ecology is probablyamong the most evolutionarily meaningful. With few exceptions, thetools they use today may most closely resemble those used byhunter–gatherer humans for much of our evolution in Africa. Whiledata concerning tools such as bow and arrow may only be analogousto a relatively recent lifestyle, data on the use of digging sticks, rocksand perhaps knives may be relevant to a much more distant homininpast (Marlowe, 2010). The Hadza lack role specialization except in re-gard to a strong sexual division of labor. Hadza men make their ownbows and arrowswith which to hunt endemic wildlife. Thewomen dis-play the same characteristic behaviors of self-sufficiency in manipulat-ing their environment — producing their own digging sticks withwhich to extract tubers from the ground (Marlowe, 2010). A previousstudy has indicated that the vast majority of Hadza are right-handedwhen asked to throw a rock, with only 2% reported as being left-handed (Marlowe, 2010). While learning may be unconsciously right-hand guided, the Hadza do not appear to have a cultural prejudiceagainst sinistrality and have not been observed to actively discourageleft-handed behaviors in children (personal communication, Berbesque,2014). About 40% of Hadza under 30 years old have attended school,and those that do seldom remain in schooling for longer than a yearor two (Marlowe, 2010). Some Hadza are literate, but many of these in-dividuals choose to live andwork outside the study area (Blurton-Jones,2013).

2.3. Materials and procedure

This is an ethological studymaking use of footage of theHadza goingabout everyday activities. In this way, a measurement of laterality offunction for culture-specific behaviors can be generated (Faurie et al.,2005). We analyzed video footage from multiple digital cameras overa four-year time period spanning 2005–2009. The filmingwas intendedfor other purposes andwas analyzed for handedness post-hoc. A total of42 individuals (males:N=22, females:N=20)were sampled from thevideo footage. Previous handedness studies offer contradictory resultsas to whether sex differences exist in handedness patterns in bothhumans and other primates (Cochet & Byrne, 2013). Due to the natureof this study (between-task comparative rather than between-subjectcomparative) and the small sample, adults of both sexeswere combinedin the analysis.

Downloaded from ClinicalKey.com at Yale For personal use only. No other uses without permissio

2.4. Procedure for footage coding

A total of 972.66 min (16.21 h) were analyzed. Adults of both sexeswere sampled, with qualifying factors being whether an individual wasrecorded displaying unambiguous handedness behavior/s. Adult is de-fined as being older than 20 years. Focal subjects were sampled whenidentified as executing at least one complex bimanual action definedin the ethogram (see Appendix, available on the journal's Web site atwww.ehbonline.org). All subsequent relevant actions by that individualwere then recorded —consequently this is not a comprehensive recordof all handedness behaviors exhibited in the footage. Units of behaviorobserved and recorded were those showing asymmetry of hand use,and activities were excluded in which both hands were “’yoked’ in si-multaneous, identical roles” (McGrew & Marchant, 1997: 204). Inorder to quantify these handedness behaviors, the activities observedwere put into categories according to task, separated into:

• object manipulation• tool-use• non tool-use, non-object manipulation categories.

Object manipulation and tool use categories were then further sub-

categorized into unimanual and (complementary) bimanual categories(Schaafsma et al., 2012). In bimanual activities, the dominant handwasrecorded as representing the data point. The dominant hand was de-fined as performing the more precise, manipulatory motions while thesubordinate hand provided static stability (McGrew&Marchant, 1997).

To ensure independent data points, bouts of handedness activitywere only recorded when separated by an event of a different nature(Marchant et al., 1995; McGrew & Marchant, 1997), and events depen-dent on previous events – for example offering food with left handafter eating foodwith left hand –were excluded. Frequencies of indepen-dent bouts of handedness actionwere sampled— recording only the firstin a sequence of repeated, identical events (Chapelain, Hogervorst,Mbonzo, & Hopkins, 2011; Marchant et al., 1995; McGrew & Marchant,1997). In certain circumstances, events were grouped together to avoidexaggerated frequencies of bouts, for instance, if an individual was re-peatedly turning and rearranging tubers around a fire, rather thanbeing recorded as “pick up” and “place”, it would be categorized as “ob-ject manipulation” — for categories see Appendix (available on thejournal's Web site at www.ehbonline.org). Handedness events wereonly considered as legitimate data points when both handswere equallyavailable to be used. Events were excluded when either hand wasdeemed to be unavailable for use — if, for example, an infant wasblocking its mothers left or right torso, a reached-for object was locatedto the side of an individual, or one hand was already in use.

Individuals with hand or arm injuries were excluded from sampling(Chapelain et al., 2011). As with most studies of laterality, inter- andintra-observer errors are not considered to play a significant role dueto the binary nature of the data: left vs. right hand (McGrew &Marchant, 1997). When demonstration of hand used was ambiguous,the data point was discarded.

2.5. Calculation of individuals' handedness indices

An individual's handedness index (HI) was calculated using the for-mula: HI = (R − L)/(R + L); subtracting the number of left-handedbouts from the number of right-handed bouts, divided by the total num-ber of bouts observed (Chapelain et al., 2011; Potier, Meguerditchian, &Fagard, 2013). A left-hand bias is reflected in the negative spectrum andright-hand bias in the positive (Chapelain et al., 2011; Potier et al., 2013).

3. Results

In total, 2388 data points (observations of handed activity)were col-lected for 42 subjects (males: N=22, females:N=20). Of these, 74.8%were performed with the right hand.

New Haven Health System April 21, 2016.n. Copyright ©2016. Elsevier Inc. All rights reserved.

205T. Cavanagh et al. / Evolution and Human Behavior 37 (2016) 202–209

3.1. Unimanual hand movements

Unimanual actions (defined as empty-handed, directed handmove-ments without tools or object manipulation involved), were observedacross a variety of behaviors (see Figs. 1 and 2). Four categories (eat,hold, pick up, and communicative gesture) were the most frequentlydisplayed unimanual, empty-handed movements (see Fig. 3). In total,1119 instances of unimanual activity were recorded. Of these, 72.6%were right-hand exclusive. Carrying was done seemingly at random,with the left hand used in only 2 more instances than the right. Holdingan object (without further acting on it) was performedmore often withthe left hand (60.2%). Throwing (including targeted throwing andtossing/lobbing) was executed with the right hand in 74.3% of the in-stances. 69.6% of all instances (N=125) of pointing or directed gestur-ing were performed right-handed. In 345 instances of picking up ofobjects, 88.4% were initiated with the right hand. In 357 occurrencesof eating, 68.9% of the time this was donewith the right-hand. As an in-teresting anecdote, in 14 observed occurrences of smoking, 13 of thesewere performed with the left hand.

3.2. Object manipulation hand movements

Object manipulation activities were divided into unimanual and bi-manual categories. These included any instancewhen the handwas uti-lized as a tool might be to act upon an object, such as extricating seedsfrom fruit (usually bimanually performed) or squeezing liquid honeyfrom the comb (usually unimanual) (Fig. 4). Unimanual object manipu-lation was recorded in 264 instances. In 70.5% of these, the right handperformed the action, while the left handwas uninvolved. In 536 objectmanipulations where two hands were used, in 63.1% of instances theright hand was the dominant hand, performing fine motor functionsof manipulation while the left hand stabilized the object. Combiningunimanual and bimanual categories, the right handperformed thedom-inant function in 65.5% of 800 instances of object manipulation.

3.3. Tool use hand movements

Handed tool-use activities were also categorized according tounimanual and bimanual activities (Figs. 5 & 6). These actions involvedthe use or manipulation of any object, naturefact or artifact (Oswalt,1973), in performing action upon another object or (in the case ofhunting) subject. These included subsistence activities involved in for-aging and hunting, or otherwise manipulating the environment (for

0

5

10

15

20

25

30

Less Common Unimanu

Right

Fig. 1. Less common unimanual, directed hand mov

Downloaded from ClinicalKey.com at Yale NewFor personal use only. No other uses without permission. C

categories of actions see Table 3 in Appendix, available on the journal'sWeb site at www.ehbonline.org).

In all instances recorded inwhich onehandheld a rock used to hit anobject (N=40) the right handwas used. In the 12 instances of bow andarrow use observed, all used the right hand to manipulate the arrow.Axe/machete usewas also entirely right-handed: in nine instances of bi-manual axe use (when onehandheld an objectwhile the other acted onit with the axe); in eight instances of two-handed bimanual axe use(where both hands held the axe at the same time); in 63 instances ofunimanual axe use. Related to this were the two unimanual and threebimanual right-handed instances of striking a tree with a hoe. Someexamples of hand use were interesting but extremely rare: as ananecdotal aside, in 6 recorded events of needle-and-thread sewing,both the left and the right hand executed the dominant manipulatoryfunction equally. So, too, in the only observed instance of rope manipu-lation: the right hand performed the dominant motion of the action 8times, while the left hand took over this dominant role 7 times duringthis episode.

One of the most striking observations concerned knife use: no in-stances were recorded in which the subjects used their left hand togrip the knife. In two instances of unimanual knife use, both wereright-handed. In 237 occurrences of bimanual knife use (holding an ob-ject in one hand and manipulating it with the knife held in the otherhand), all were right-handed.

Overall, in seven categories of unimanual tool use (N= 114), 96.5%were performed right-handed. Similarly, 12 categories of bimanual tooluse (N = 352), reveal a 96% right-hand bias. In a combined total, 466data points were recorded for tool use tasks, and 96.1% of these wereexecuted with the right hand. This considerable bias in tool-use tasksis noteworthy.

3.3.1. Unimanual vs. bimanual tasksIn combining unimanual object manipulation tasks and unimanual

tool-use tasks (N = 378), 78.3% of these were performed with theright hand. This is very similar to combined bimanual tasks (object ma-nipulation and tool-use), in which the right hand was the dominanthand in 76.1% of instances (N = 888).

3.3.2. Handedness of individualsIn total, overall HI was calculated for 16 individuals (males: N=10,

females: N= 6), all of which appear to be right-handed. Individuals' HIwas used when the number of data points recorded for themwas abovethe conservative threshold of 50 measures, recommended by re-searchers (Marchant & McGrew, 1991), with which to make reliable

al Empty-handed Actions

Left

ements without tools or object manipulation.

Haven Health System April 21, 2016.opyright ©2016. Elsevier Inc. All rights reserved.

Fig. 2. Example of communicative gesturing. Fig. 4. Example of unimanual object manipulation — squeezing liquid honey from thecomb.

206 T. Cavanagh et al. / Evolution and Human Behavior 37 (2016) 202–209

inferences of laterality of an individual. Number of data points per sub-ject varied between 56 and 240. Of these, all used their right handmoreoften than their left— that is, their HIwas calculated to be positive. If thethreshold is relaxed to 30 measures, an additional 5 individuals (males:N = 1, females: N = 4) can be classified as being right-handed.

4. Discussion

The results of this study indicate that Hadzamen andwomendisplaydominant right-handedness in their spontaneous foraging activities(Fig. 7). This lateralization was most apparent in tool-use tasks, with96% of all occurrences being right-handed. Hadza handedness appearsto havemuch in commonwith other human societies, and the data sup-port the conclusion of right-hand dominance across the human species.Random frequencies were closest to being realized in object manipula-tion tasks. The conspicuous right-hand bias in tool-use tasks is in keep-ing with results from earlier studies of non-industrialized groups(Faurie et al., 2005; Marchant et al., 1995). This consistency across geo-graphically and genetically separate (Marlowe, 2010) groups implies anancient origin, and distinguishes tool-use as themost lateralized humanbehavior in non-industrialized activities.

Of those individuals whomet the conservative conditions for whichhandedness index could be calculated, all were classified as being right-handed. Taken together with the 2% left-handers for throwing docu-mented by Marlowe (2010), these results are indicative of a wider

0

50

100

150

200

250

300

350

Eat (n = 357) Hold (n = 83)

Frequent Unimanual

Right

Fig. 3. Frequent unimanual, directed hand movem

Downloaded from ClinicalKey.com at Yale For personal use only. No other uses without permissio

pattern of strong right-handdominance. Alongwhatmight be describedas the spectrum of human handedness, theHadza appear to bemarked-ly right-handed. This is, however, a conclusion which needs more sup-port in the way of data, because, as has been noted, handedness of apopulation cannot be inferred through the measurement of one task(for example, throwing) or limited numbers of individuals (Marchant& McGrew, 2013). Whether or not population-level handedness of theHadza can here be inferred from such a small sample size is debatable.The consistency of right-hand use observed in this instance and inMarlowe's (2010) study, however, may point in that direction.

Our study aimed to investigate handedness trends in a hunter–gatherer group over multiple categories of tasks. While the sampled in-dividuals appeared to display an overall consistency of right-hand pref-erence, there was some variation in results between the types of tasksmeasured. This has implications for both the origins of right-handedness in humans, and for the further study of handedness trendsin non-human primates. Object manipulation as a category was closerto randomassortment than the tool-use category, whichwas almost en-tirely right-hand dominant. If lateralized hand use can be more easilydetected in tool-use tasks then this particular type of task should be pri-oritized in the continuing search for definitive primate handedness con-tinuity. Many non-human primate studies have attempted to measurelateralization through complex manipulatory tasks that do not involvetools. Furthermore, researchers have, in the past, assigned multiple

Pick up (n = 344) Communicative gesture(n = 125)

Empty-handed Actions

Left

ents without tools or object manipulation.

New Haven Health System April 21, 2016.n. Copyright ©2016. Elsevier Inc. All rights reserved.

Fig. 5. Example of unimanual tool use — use of stick to remove bees.

207T. Cavanagh et al. / Evolution and Human Behavior 37 (2016) 202–209

definitions to complexity of task (Uomini, 2009). This study demon-strates that different results are obtained by analyzing tool and non-tool use categories separately, and that complex manipulatory tasksare not as lateralized as tool-use tasks.

A recent meta-analysis indicates that right-handers of both sexeshave a small but statistically significant advantage in "overall spatialability" over left-handers (Somers, Shields, Boks, Kahn, & Sommer,2015). If spatial ability allows humans to manipulate our circumambi-ent environment, tool-use is the ability to use that environment to fur-ther manipulate it to our advantage. Tool-use handed behaviors aretasks in which essentially a third variable has been introduced, andthe data from this study are clear in that the type of tool used doesnot affect the preferred hand. In the Hadza it is not necessarily the com-plexity of the tool that elicited the strongest laterality, as the exclusiveuse of the right hand to hammer rocks against objects indicates. Remov-ing individual bees from honey with fingers in a precision grip is argu-ably as complicated a task. This, however, was almost as likely to bedone with the left hand as the right. It appears as if the left hand itselfmay not be as disadvantaged as we tend to think (Guiard, 1987), ratherit is its ability to handle the additional introduced variable that may belacking in comparison to the right. It may be that lateralization ofhand function “depends on the extent of hand-role differentiation forthe task” (Potier et al., 2013: 578). This does not, however, appear tobe the case with the Hadza: in almost all instances of bimanual tool-use, irrespective of whether or not the two hands were performing

Fig. 6. Examples of bimanual tool

Downloaded from ClinicalKey.com at Yale NewFor personal use only. No other uses without permission. C

mechanically parallel but functionally different roles (i.e. two-handedgrip), the right hand was dominant. The Hadza appear to be reasonablymatched in strength in their right and left limbs, possibly because bothsexes engage in foragingbehaviors that cultivate ambidextrous strength(Marlowe, 2010). This effect of working behaviors has also been shownin Western populations; manual laborers have very little difference ingrip strength between hands, whereas office workers do (Steele &Uomini, 2005). Grip strength does not appear to correlate with domi-nant hand (Clerke & Clerke, 2001).

In 125 communicative gestures, the right hand was used in 69.6% ofinstances. This action was far less lateralized than eating or picking up.While right-handdominance for communicative gestureswould appearto be in keeping with non-human primate studies that show a signifi-cant right bias in this action category, it does not correlate with claimsthat this category shows more right-hand lateralization than tool-usetasks, as reported by Meguerditchian et al. (2013). It should be addedhere that many occurrences of gestural communication observed in-volved both hands and were therefore excluded from analyses. Thisstudy finds limited support for theoretical scenarios which posit gestur-al communication as a causal link to right-handedness in humans.While throwing was not a commonly recorded behavior in this study,it was primarily executed with the right hand and the results thereforeshow some support for Calvin's (1983) hypothesis. As throwing any-thing, however, involves the use of it, the distinction between throwingas a form of tool-use and throwing as a separate function is unclear.

In 466 instances of tool-use the right hand was used. One of the moststriking examples of this bias is that in 239 occurrences of knife-use allwere right-handed. Overall, the results of this study offer little support forthe “task complexity theory”; rather the data appear to further substantiatethe “tool use theory”. It is clear that, in future studies, the effort should bemade to define complexity as it relates to tool-use or lack thereof to differ-entiate between complexmanipulatory tasks and complex tool-using tasks.

Cross-cultural normative right-handedness implicates a biologicalrather than cultural base for handedness, although its expression is un-doubtedly influenced by culture. Handedness may be genetically(Annett, 2002; McManus, 2002; Medland et al., 2009) or epigenetically(Crow, 2010; Klar, 2004) determined. Factors, apart from genetics, thatmay affect handedness expression in humans include levels of testoster-one present in ontogenic development and twinnedness (Ooki, 2014).These effects are also seen in other animals: for example, in chicks; “ge-netic expression may determine the presence (or absence) oflateralisation at the individual level, whereas … light and steroid hor-mone levels are influencing factors that cause variation among individ-uals in the presence or strength of lateralisation” (Vallortigara & Rogers,2005: 585–586). Lateralized motor behaviors in humans appear tomanifest very early on in development (van Dongen et al., 2014):

use — use of rock and knife.

Haven Health System April 21, 2016.opyright ©2016. Elsevier Inc. All rights reserved.

Unimanual, empty-handed(N = 1119)

Object manipulation (N =800)

Tool use (N = 466)

Right 72.6% 65.5% 96.1%

Left 27.4% 34.5% 3.9%

0.0%

100.0%

Comparison of Categories

Right Left

Fig. 7. Comparison of categories according to type.

208 T. Cavanagh et al. / Evolution and Human Behavior 37 (2016) 202–209

directional thumb-sucking in 15-week-old fetuses has been shown tocorrelate with later handedness (Hepper, Wells, & Lynch, 2005), andlimbs show asymmetrical development in weeks 10–20 of gestation(van Dongen et al., 2014). Handedness for unimanual actions, as wellas symmetrical bimanual actions, occurs earlier in infants than bimanu-al coordinated behaviors (Potier et al., 2013). This would indicate thatcoordinated bimanual actions require more developed capabilities —either cognitive or motor. The data here show, however, that therewas very little difference in unimanual and bimanual handedness be-haviors. This surprising result is at odds with non-human primate stud-ies which have highlighted increased lateralized behavior in bimanualcompared to unimanual actions (e.g., Meguerditchian et al., 2013).

As right-handedness appears, in fact, to be a species-typical trait thatcan be discerned in many divergent cultures and geographic locations,we can assume that it was present in early Homo populations beforethe radiation out of Africa. Hadza ecology and technologies show paral-lels with earlier human societies before the advent of agriculture or sys-tematic schooling. Their cultural autonomy and forager lifestyle may,therefore, represent a reliable proxy for “unmoulded” (Marchant et al.,1995) handed behaviors. This ethological studymeasured spontaneous,culture-specific behaviors across a wide range of categories. The datafrom the Hadza clearly indicate that tools are used almost exclusivelywith the right hand. Right-hand dominance was absolute in simpletool-use actions like rock-hammering, suggesting that this dominancecould have emerged with early hominin use of simple technologies.Humans are pervasive tool-users, as were our hominin ancestors, andis may explain the extreme lateralization that these tasks elicit(Uomini, 2009). The use of a digging stick allows Hadza women to ex-tract deeply buried tubers that are unavailable to baboons. In earlyhumans, the ability to access this staple food source may have provideda distinct advantage and strong selective pressures for continued tool-use (Marlowe & Berbesque, 2009). The data from this study would ap-pear to contradict previous theories that claim Western cultural pres-sures enforce right-hand norms (Previc, 1991). Whether or not thespeculative right-hand dominance present in the Hadza may be influ-enced by cultural processes would have to be the subject of anotherstudy. It may in fact be the case that the homogeneity (in terms oflack of role specialization and technology) of theHadza population is re-sponsible in part for the homogeneity of their handedness.

5. Conclusions

The results of this study are consistent with the view that humanright-handedness is a trait that evolvedwith tool-use, and puts its origin

Downloaded from ClinicalKey.com at Yale For personal use only. No other uses without permissio

in the very distant past. That lateralization is elicited by tool-use tasks inhumans is consistent with studies of other great apes (Hopkins et al.,2011; Llorente et al., 2011). Lateralized tool-use in great apes may in-deed represent a precursor to right-handedness in H. sapiens— the dis-crepancies in hand-use between humans and other primates may be “adifference of degree” (Marchant et al., 1995: 256). More studies areneeded inwild populations, but there appears to be grounds for a tenta-tive inference of continuity among great ape species (includinghumans) for tool-use tasks. That both complex and non-complex tool-use is the most lateralized of behaviors in a contemporary hunter–gatherer group has significant evolutionary implications for the selec-tive processes that favor human right-hand dominance at species level.

Supplementary Materials

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.evolhumbehav.2015.11.002.

Acknowledgments

Wewould like to thank two anonymous reviewers for their commentswhich helped to improve the manuscript. We also thank Daudi Petersonfor assistance, and Costech for permission to conduct research. This re-search was supported by the Leakey Foundation, Wenner Gren, and NSF(Nos. 1062879 and 0544751, 0242455, and 0650574). Finally, we wouldlike to thank the Hadza for their enduring tolerance and hospitality.

References

Ambrose, S. H. (2001). Paleolithic technology and human evolution. Science, 291(5509),1748–1753.

Annett, M. (2002). Handedness and brain asymmetry: the Right Shift Theory. PsychologyPress.

Bax, J. S., & Ungar, P. S. (1999). Incisor labial surface wear striations in modern humansand their implications for handedness in Middle and Late Pleistocene hominids.International Journal of Osteoarchaeology, 9(3), 189–198.

Blurton-Jones, N. G. (2013). Hadza demography and sociobiology. http://www.sscnet.ucla.edu/anthro/faculty/blurton-jones/hadza-ch9.pdf

Braccini, S., Lambeth, S., Schapiro, S., & Fitch, W. T. (2010). Bipedal tool use strengthenschimpanzee hand preferences. Journal of Human Evolution, 58(3), 234–241.

Calvin, W. H. (1983). A stone's throw and its launch window: timing precision and its im-plications for language and hominid brains. Journal of Theoretical Biology, 104(1),121–135.

Cantalupo, C., & Hopkins, W. D. (2001). Asymmetric Broca's area in great apes. Nature,414(6863), 505-505.

Canteloup, C., Vauclair, J., & Meunier, H. (2013). Hand preferences on unimanual and bi-manual tasks in Tonkean macaques (Macaca tonkeana). American Journal of PhysicalAnthropology, 152(3), 315–321.

New Haven Health System April 21, 2016.n. Copyright ©2016. Elsevier Inc. All rights reserved.

209T. Cavanagh et al. / Evolution and Human Behavior 37 (2016) 202–209

Cashmore, L., Uomini, N., & Chapelain, A. (2008). The evolution of handedness in humansand great apes: a review and current issues. Journal of Anthropological Sciences,86(2008), 7–35.

Chapelain, A. S., Hogervorst, E., Mbonzo, P., & Hopkins, W. D. (2011). Hand preferences forbimanual coordination in 77 bonobos (Pan paniscus): replication and extension.International Journal of Primatology, 32(2), 491–510.

Clerke, A., & Clerke, J. (2001). A literature review of the effect of handedness on isometricgrip strength differences of the left and right hands. The American Journal of Occupa-tional Therapy: Official Publication of the American Occupational Therapy Association,55(2), 206–211.

Cochet, H., & Byrne, R. W. (2013). Evolutionary origins of human handedness: evaluatingcontrasting hypotheses. Animal Cognition, 16(4), 531–542.

Corballis, M. C. (2003). From mouth to hand: gesture, speech, and the evolution of right-handedness. Behavioral and Brain Sciences, 26(2), 199–208.

Corballis, M. C. (2014). Left brain, right brain: facts and fantasies. PLoS Biology, 12(1),e1001767. http://dx.doi.org/10.1371/journal.pbio.1001767.

Corballis, M. C., Badzakova-Trajkov, G., & Häberling, I. S. (2012). Right hand, left brain: ge-netic and evolutionary bases of cerebral asymmetries for language and manual ac-tion. Wiley Interdisciplinary Reviews: Cognitive Science, 3(1), 1–17.

Crow, T. J. (2010). A theory of the origin of cerebral asymmetry: epigenetic variationsuperimposed on a fixed right-shift. Laterality: Asymmetries of Body, Brain andCognition, 15(3), 289–303.

De Agostini, M., Khamis, A. H., Ahui, A.M., & Dellatolas, G. (1997). Environmental influencesin hand preference: an African point of view'. Brain and Cognition, 35(2), 151–167.

Dunham, J., & Davenport, T. (2012). Handedness: theories, genetics and psychology.Hauppauge, New York: Nova Science Publishers.

Fagot, J., & Vauclair, J. (1991). Manual laterality in nonhuman primates: a distinction be-tween handedness and manual specialization. Psychological Bulletin, 109(1), 76–89.

Falk, D. (1987). Brain lateralization in primates and its evolution in hominids. AmericanJournal of Physical Anthropology, 30(S8), 107–125.

Faurie, C., Schiefenhövel, W., Bomin, S., Billiard, S., & Raymond, M. (2005). Variation in thefrequency of left‐handedness in traditional societies. Current Anthropology, 46(1),142–147.

Forrester, G. S., Quaresmini, C., Leavens, D. A., Mareschal, D., & Thomas, M. S. (2013).Human handedness: an inherited evolutionary trait. Behavioural Brain Research,237(2013), 200–206.

Frayer, D. W., Lozano, M., Bermúdez, D. C., Carbonell, E., Arsuaga, J. L., Radovčić, J., ...Bondioli, L. (2012). More than 500,000 years of right-handedness in Europe.Laterality, 17(1), 51–69.

Geuze, R. H., Schaafsma, S. M., Lust, J. M., Bouma, A., Schiefenhövel, W., & Groothuis, T. G.(2012). Plasticity of lateralization: schooling predicts hand preference but not handskill asymmetry in a non-industrial society'. Neuropsychologia, 50(5), 612–620.

Guiard, Y. (1987). Asymmetric division of labor in human skilled bimanual action. Journalof Motor Behavior, 19(4), 486–517.

Hepper, P. G., Wells, D. L., & Lynch, C. (2005). Prenatal thumb sucking is related to post-natal handedness. Neuropsychologia, 43(3), 313–315.

Hopkins, W. D., Phillips, K. A., Bania, A., Calcutt, S. E., Gardner, M., Russell, J., ... Schapiro, S. J.(2011). Hand preferences for coordinated bimanual actions in 777 great apes: implicationsfor the evolution of handedness in hominins. Journal of Human Evolution, 60(5), 605–611.

Kimura, D. (1979). Neuromotor mechanisms in the evolution of human communication.Neurobiology of social communication in primates: an evolutionary perspective.New York: Academic Press, 197–220.

Klar, A. J. S. (2004). An epigenetic hypothesis for human brain laterality, handedness, and psy-chosis development'. Cold Spring Harbor Symposia on Quantitative Biology(69), 499–506.

Knecht, S., Dräger, B., Deppe, M., Bobe, L., Lohmann, H., Flöel, A., ... Henningsen, H. (2000).Handedness and hemispheric language dominance in healthy humans. Brain,123(12), 2512–2518.

Llorente, M., Riba, D., Palou, L., Carrasco, L., Mosquera, M., Colell, M., & Feliu, O. (2011).Population-level right-handedness for a coordinated bimanual task in naturalistichoused chimpanzees: replication and extension in 114 animals from Zambia andSpain. American Journal of Primatology, 73(3), 281–290.

Lozano, M., Bermúdez de Castro, J. M., Carbonell, E., & Arsuaga, J. L. (2008). Non-masticatory uses of anterior teeth of Sima De Los Huesos individuals (Sierra DeAtapuerca, Spain). Journal of Human Evolution, 55(4), 713–728.

Downloaded from ClinicalKey.com at Yale NewFor personal use only. No other uses without permission. C

Mangalam, M., Desai, N., & Singh, M. (2014). Division of labor in hand usage in free-ranging Bonnet macaques, Macaca radiata. American Journal of Primatology, 76(6),576–585.

Marchant, L. F., & McGrew, W. C. (1991). Laterality of function in apes: a meta-analysis ofmethods. Journal of Human Evolution, 21(6), 425–438.

Marchant, L. F., &McGrew,W. C. (2013). Handedness ismore than laterality: lessons fromchimpanzees. Annals of the New York Academy of Sciences, 1288(1), 1–8.

Marchant, L. F., McGrew, W. C., & Eibl-Eibesfeldt, I. (1995). Is human handedness univer-sal? Ethological analyses from three traditional cultures. Ethology, 101(3), 239–258.

Marlowe, F. (2002). Why the Hadza are still hunter–gatherers. In S. Kent (Ed.), Ethnicity,Hunter gatherers, and the ‘Other' (pp. 247–281). Smithsonian Institution Press.

Marlowe, F. (2010). The Hadza: hunter–gatherers of Tanzania. Berkeley: University ofCalifornia Press.

Marlowe, F. W., & Berbesque, J. C. (2009). Tubers as fallback foods and their impact onHadza hunter–gatherers. American Journal of Physical Anthropology, 140(4), 751–758.

McGrew, W. C., & Marchant, L. F. (1997). On the other hand: current issues in and meta‐analysis of the behavioral laterality of hand function in nonhuman primates.American Journal of Physical Anthropology, 104(S25), 201–232.

McManus, C. (2002). Right hand, left hand: the origins of asymmetry in brains, bodies, atomsand cultures. London: Weidenfeld & Nicolson.

Medland, S. E., Duffy, D. L., Wright, M. J., Geffen, G. M., Hay, D. A., Levy, F., ... Whiteh, V.(2009). Genetic influences on handedness: data from 25,732 Australian and Dutchtwin families. Neuropsychologia, 47(2), 330–337.

Meguerditchian, A., Vauclair, J., & Hopkins, W. D. (2010). Captive chimpanzees use theirright hand to communicate with each other: implications for the origin of the cere-bral substrate for language. Cortex, 46(1), 40–48.

Meguerditchian, A., Vauclair, J., & Hopkins, W. D. (2013). On the origins of human hand-edness and language: a comparative review of hand preferences for bimanual coordi-nated actions and gestural communication in nonhuman primates. DevelopmentalPsychobiology, 55(6), 637–650.

Mosquera, M., Geribàs, N., Bargalló, A., Llorente, M., & Riba, D. (2012). Complex tasks forcehand laterality and technological behaviour in naturalistically housed chimpanzees:inferences in hominin evolution. Scientific World Journal, 2012(514809), 1–12.

Ooki, S. (2014). An overview of human handedness in twins. Frontiers in Psychology,5(10), 1–5.

Oswalt, W. H. (1973). Habitat and technology: the evolution of hunting. New York: Holt,Rinehart and Winston.

Potier, C., Meguerditchian, A., & Fagard, J. (2013). Handedness for bimanual coordinatedactions in infants as a function of grip morphology. Laterality, 18(5), 576–593.

Previc, F. H. (1991). A general theory concerning the prenatal origins of cerebral laterali-zation in humans. Psychological Review, 98(3), 299–334.

Raymond, M., & Pontier, D. (2004). Is there geographical variation in human handedness?Laterality, 9(1), 35–51.

Schaafsma, S. M., Geuze, R. H., Riedstra, B., Schiefenhövel,W., Bouma, A., & Groothuis, T. G.G. (2012). Handedness in a nonindustrial society challenges the fighting hypothesisas an evolutionary explanation for left-handedness. Evolution and Human Behavior,33(2), 94–99.

Somers, M., Shields, L. S., Boks, M. P., Kahn, R. S., & Sommer, I. E. (2015). Cognitive benefitsof right-handedness: a meta-analysis. Neuroscience & Biobehavioral Reviews, 51,48–63.

Steele, J., & Uomini, N. (2005). Humans, tools and handedness. In V. Roux, & B. Bril (Eds.),Stone knapping: the necessary conditions for a uniquely hominin behaviour(pp. 217–239). Cambridge: McDonald Institute for Archaeological Research.

Tutkuviene, J., & Schiefenhövel, W. (2013). Laterality of handgrip strength: age‐ and phys-ical training-related changes in Lithuanian schoolchildren and conscripts. Annals ofthe New York Academy of Sciences, 1288(1), 124–134.

Uomini, N. T. (2009). The prehistory of handedness: archaeological data and comparativeethology. Journal of Human Evolution, 57(4), 411–419.

Vallortigara, G., & Rogers, L. J. (2005). Survival with an asymmetrical brain: advantagesand disadvantages of cerebral lateralization. Behavioral and Brain Sciences, 28(4),575–588.

van Dongen, S., Galis, F., Ten Broek, C., Heikinheimo, K., Wijnaendts, L. C. D., Delen, S., &Bots, J. (2014). When right differs from left: human limb directional asymmetryemerges during very early development. Laterality, 19(6), 1–11.

Haven Health System April 21, 2016.opyright ©2016. Elsevier Inc. All rights reserved.