Capuchin Monkey (Cebus apella) Grips for the Useof Stone Tools

GREGORY CHARLES WESTERGAARD* AND STEPHEN J. SUOMILaboratory of Comparative Ethology, National Institute of Child Healthand Human Development, Poolesville, Maryland 20837

KEY WORDS capuchin; hominid; precision grip; power grip; tooluse

ABSTRACT This research examined capuchin monkey (Cebus apella)grips for the use of throwing, nut-cracking, and cutting tools. We providedsubjects with stones and apparatus that accommodated the use of stones astools. Our subjects exhibited five grips, two of which the animals used whenforce was the primary consideration (power grips) and three of which theanimals used when accuracy of sensory judgment and instrumentation wasrequired (precision grips). We believe that the range of contexts in whichcapuchins use stone tools, combined with the ability of capuchins to employboth power and precision grips as part of their tool repertoire, indicate thatCebus apella can be used to identify grips that facilitated hominid lithictechnology.Am J PhysAnthropol 103:131–135, 1997. r 1997 Wiley-Liss, Inc.

Several authors have argued that a com-plete understanding of the relation betweenmanual grips and tool use in extant nonhu-man primates would help us to identifypotential grips and tool behaviors in fossilhominids (Boesch and Boesch, 1993; Marzkeand Wullstein, 1996). Boesch and Boesch(1993) report that chimpanzees exhibit sixdifferent grips when they crack open nutswith tools. The range of chimpanzee nut-cracking grips led these authors to postulatethat Pan troglodytes is morphologically ableto perform most tool tasks that have beenattributed toHomo habilis.Capuchins are arboreal New World pri-

mates whose manipulative skills approachthose of chimpanzees (Visalberghi, 1990;Westergaard, 1994). Capuchins use tools inisolated instances in nature, and habituallyin captivity when provided with appropriatematerials and incentives (Visalberghi, 1990;Westergaard, 1994). Examples of capuchintool behavior in captivity include aimedthrowing of stones, the use of stones asnut-cracking tools, and the use of stoneflakes as cutting tools (Anderson, 1990;West-

ergaard and Suomi, 1994, 1995, 1996). Capu-chins lack true thumb opposition as thethumb is incapable of rotation about its ownlong axis (Napier, 1960). However, capu-chins, unlike other New World primatessuch as squirrel monkeys, use both precisionand power grips to manipulate objects (Cos-tello and Fragaszy, 1988). To the best of ourknowledge capuchin grips for the use of toolshave not been described.This research examined capuchin grips

for the use of throwing (Experiment 1),nut-cracking (Experiment 2), and cuttingtools (Experiment 3). We provided subjectswith stones and apparatus that accommo-dated the use of stones as tools and recordedgrips using a modified version of the classifi-cation scheme outlined byMarzke andWull-stein (1996; see Table 1). We further distin-guished between grips used when force was

Contract grant sponsor: National Institute of Child Health andHuman Development; Grant name: Intramural Research Train-ingAward.*Correspondence to: Gregory Charles Westergaard, National

Institutes of Health Animal Center, P.O. Box 529, Poolesville,MD 20837. e-mail: WESTERGG@lce.nichd.nih.govReceived 26April 1996; accepted 5 March 1997.

AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 103:131–135 (1997)

r 1997 WILEY-LISS, INC.

the primary consideration (power grips) andthose used when accuracy of sensory judg-ment and instrumentation was required(precision grips; Napier, 1960). Our goal inthis research was to establish the range ofgrips required for a simple nonhuman pri-mate stone-tool technology. We chose to ex-amine capuchin grips for the use of stonetools, as opposed to tools made of vegetationor other materials, because stone tools arethe most ancient homonid artifacts pre-served in the archaeological record (Schickand Toth, 1993).

METHODSSubjects

The subjects were four tufted capuchins(Cebus apella). These animals were chosenas subjects because of their proficient use oftools across a number of different contexts(Westergaard and Suomi, 1994, 1995, 1996).Included were one adult female (Carlina)and three adult males (Morris, Jeb, andVirgil) that ranged in age from 7 to 16 years.Subjects were housed socially in indoor cages(2 3 1 3 2 m). Food and water were avail-able ad libitum.

Throwing

At the onset of each test session we filled aplastic container (15 3 12 cm high) withpeanut butter and placed it on the room floor

40 cm from the front of a group’s cage. Wethen placed a video camera (Canon ES2000Hi-8mm video camcorder) outside the cagedirectly behind the apparatus. At the onsetof each trial we placed a stone (between 20and 40 g in weight) on the cage floor. Anobserver noted from videotape grips used bythe focal subject to throw a stone towardsthe container, and whether the stone landedin the container (‘‘hit’’) or on the room floor(‘‘miss’’). We calculated the accuracy rate foreach grip using the formula [(number ofhits/number of throws) 3 100]. We con-ducted 40 trials per subject over a 2-weekperiod.

Nut-cracking

The apparatus was a plastic container(10 3 10 cm high) with a round opening (2.5cm in diameter) on top. The apparatus helda walnut in a fixed position. At the onset ofeach test session we attached the apparatusto the cage floor, placed a video cameraoutside the cage directly behind the appara-tus, and distributed four stones (each be-tween 40 and 60 g in weight) on the cagefloor. At the onset of each trial we mounted awalnut in the apparatus so that it protrudedthrough the opening. An observer notedfrom videotape grips used by the focal sub-ject to crack open a walnut with a stone. Ifan animal successively struck a walnut withtwo or more stones the observer only re-corded grips used to hold the stone thatcracked the walnut’s shell. We conducted 40trials per subject over a 2-week period.

Cutting

The apparatus was a plastic container(10 3 10 cm high) with a square opening(3 3 3 cm) on top. The container held peanutbutter in a small well (2 3 1 cm high). At theonset of each test session we attached theapparatus to the cage floor, placed a videocamera outside the cage directly behind theapparatus, and distributed four stones (eachbetween 20 and 60 g in weight) on the cagefloor. Each stone had at least one sharp edge.At the onset of each trial we placed anacetate barrier over the apparatus so that itcovered the well. The animals could onlyobtain food by cutting through the barrierwith a stone. An observer noted from video-

TABLE 1. Grips and operational definitions

Operational definition

Power gripsJaw-chuck Grip in which an object is held

tightly against the palm byflexed fingers. The thumb eitheropposes the fingers or supportsthe object from the side.

Bimanualjaw-chuck

Jaw-chuck grip in which an objectis held by both hands.

Palm push Grip in which an object is pushedby an open palm.

Precision gripsCup Grip in which an object is cupped

in an upturned palm, andpropped by flexed fingers and anopposed thumb.

Pad-to-side Grip in which the thumb props orholds an object against the sideof the index finger.

Bimanualtip-to-tip

Grip in which object is heldbetween tips of fingers on eachhand.

132 G.C. WESTERGAARD AND S.J. SUOMI

tape grips used by the focal subject to cutthrough acetate with a stone tool. If ananimal successively placed two or morestones in contact with the barrier the ob-server only recorded grips that the animalused to hold the stone that penetrated thebarrier. We conducted 40 trials per subjectover a 2-week period.

RESULTSThrowing

The capuchins used three different gripsto throw stones (see Table 2). Four subjectsused the jaw-chuck grip in 74% of the overalltrials (accuracy rate 5 50%).Acapuchin typi-cally held a stone against a downturnedpalm by flexed fingers, and propelled thestone by downward extension of the arm. Asecond capuchin throwing grip was the cupgrip in which an animal held a stone in anupturned palm and propelled the stone byextension of the arm away from the body.One subject used the cup grip in 24% of theoverall trials (accuracy rate 5 82%). A thirdcapuchin throwing grip was the pad-to-sidegrip in which an animal held a stone be-tween the thumb and side of the indexfinger, and propelled the stone by downwardextension of the arm. Two subjects used thepad-to-side grip in 2% of the overall trials(accuracy rate 5 67%). Subjects used onlyone type of throwing grip within each trial.

Nut-cracking

The capuchins used two different grips tocrack open nuts with stone tools (see Table3). Four subjects used the jaw-chuck grip in94% of the overall trials. A capuchin typi-cally held a stone in a downturned palm byflexion of the fingers against an opposedthumb. Involvement of the fourth and fifthdigits was dependent upon the size and

shape of the stone. Greatest involvement ofthese digits occurred when an animal usedlarge cylindrical stones. A second capuchinnut-cracking hand posture was the bi-manual jaw-chuck grip. Three subjects usedthis grip in 24% of the overall trials. Capu-chins held stones either with hands side-by-side or with one hand overlapping the other.It appeared that subjects used the bimanualjaw-chuck grip to increase the force of thestone on the walnut. The mean number ofgrips used per trial was 1.2.

Cutting

The capuchins used four different grips tocut through acetate barriers with stone tools(see Table 4). Four subjects used the jaw-chuck grip in 100% of the overall trials.Capuchins used the jaw-chuck grip to placestones against acetate, to push down onstones, and to remove stones from the appa-ratus. A second capuchin hand posture forcutting was the palm push. Four subjectsused this grip in 94% of the overall trials.Capuchins used the palm push to forcestones through acetate barriers. A thirdcapuchin hand posture for cutting was thepad-to-side grip. Four subjects used this gripin 24% of the overall trials. Capuchins usedthe pad-to-side grip to place stones againstacetate, to reposition stones during theiruse, and to remove stones from the appara-tus. A fourth capuchin hand posture forcutting was the bimanual tip-to-tip grip.Three subjects used this grip in 7% of theoverall trials. Capuchins used the bimanualtip-to-tip grip to remove stones from theapparatus. The mean number of grips usedper trial was 2.2. Analysis of variance indi-cated a significant main effect of task on themean number of grips per trial (F(3,2) 5

TABLE 2. Capuchin grips for aimed throwing of stones

Subject

Percentage of trials inwhich subject used this grip

Meannumberof gripsper trialJaw-chuck Cup Pad-to-side

Carlina 5 95 0 1.0Jeb 98 0 2 1.0Morris 100 0 0 1.0Virgil 95 0 5 1.0Mean per subject 74 24 2 1.0

TABLE 3. Capuchin grips for nut-crackingwith stone tools

Subject

Percentage of trials inwhich subject used this grip Mean

numberof gripsper trialJaw-chuck

Bimanualjaw-chuck

Carlina 95 12 1.1Jeb 100 0 1.0Morris 100 10 1.1Virgil 82 75 1.6Mean per subject 94 24 1.2

133CAPUCHIN GRIPS FOR STONE TOOL USE

22.86, P , .05). Fisher’s protected least sig-nificant difference (PLSD) post hoc pairedcomparison test (P , .05) indicated that sub-jects used a greater mean number of gripsper trial for cutting than for nut-cracking orthrowing (see Fig. 1).

DISCUSSION

Our results indicate that capuchins em-ploy a variety of grips when they use stonesfor throwing, nut-cracking, and cutting. Oursubjects exhibited five grips, two of whichthe animals used when force was the pri-mary consideration (power grips) and threeof which the animals used when accuracy ofsensory judgment and instrumentation wasrequired (precision grips; Napier, 1960). Webelieve that the range of contexts in whichcapuchins use stone tools, combined withthe ability of capuchins to employ both powerand precision grips as part of their toolrepertoire, indicate that Cebus apella can be

used to identify grips that facilitated homi-nid lithic technology (Parker and Gibson,1977; Westergaard and Suomi, 1994).Marzke and Wullstein (1996) note three

elements of gripping that distinguish hu-mans from chimpanzees. These elementsare (1) the relative force of precision grips(pinch vs. hold), (2) the relative ability totranslate and rotate objects by the thumband fingers (precision handling), and (3) therelative ability to orient a cylindrical objectso that it functions effectively as an exten-sion of the forearm (power squeeze). Capu-chin precision grips were weak relative tothose of humans, and did not involve transla-tion or rotation of objects at the joints of thethumb and fingers. Capuchin power gripsdid not suggest active use of the palm thatwould facilitate increased forearm leveragewhen using an object as an extension of thearm. Preliminary studies of grips by orangu-tans (Marzke andWullstein, personal obser-vation), baboons (Guthrie, 1991; Jude, 1993,cited in Marzke and Wullstein, 1996), andcapuchins (Costello and Fragaszy, 1988) in-dicate a lack of these gripping elements inPongo, Papio, and Cebus, providing furtherevidence that precision grip force, precisiongrip maneuverability, and power squeeze ofcylinders differentiate the gripping behaviorof humans from that of great apes andmonkeys.Given the apparent discontinuity between

humans and nonhuman primates concern-ing gripping, it is noteworthy that capu-chins, particularly because of the pseudo-opposable status of their thumbs, employ arange of power and precision grips whenthey use tools. Our results indicate thatballistic movements such as those involvedin throwing and nut-cracking require amorelimited range of grips than do the moreprecisemovements involved in cutting. There

TABLE 4. Capuchin grips for cutting with stone tools

Subject

Percentage of trials in which subject used this grip

Mean number ofgrips per trialJaw-chuck Palm push Pad-to-side

Bimanualtip-to-tip

Carlina 100 98 10 10 2.2Jeb 100 88 5 0 1.9Morris 100 98 68 12 2.8Virgil 100 90 12 5 2.1Mean per subject 100 94 24 7 2.2

Fig. 1. Mean number of grips per trial exhibited bycapuchins for the use of stones as throwing, nut-cracking, and cutting tools. *Mean number of grips forcutting was significantly greater than the mean numberof grips for throwing or nut-cracking (Fisher’s PLSDtest, P , .05).

134 G.C. WESTERGAARD AND S.J. SUOMI

is little direct evidence indicating the rangeof contexts in which early hominids usedtools, although it is believed that stoneswere used for throwing, nut-cracking, andcutting (Schick and Toth, 1993). Susman(1994) has noted that the earliest Australo-pithecines, who had ape-like thumbs, didnot leave behind evidence of a lithic technol-ogy but that species with a human-likethumb (for example, Paranthropus robus-tus) did. The rationale underlying this argu-ment is that long stout thumbs facilitatedrefined precision grasps and production ofartifacts that are recognizable in the archaeo-logical record. Capuchins and bonobos pro-duce simple stone tools by striking stonesagainst hard surfaces using actions similarto those involved in nut-cracking with stones(Toth et al., 1993; Westergaard and Suomi,1994). Precise human-like finger move-ments are simply not required for this as-pect of tool behavior. Similarities in tool-using propensities between capuchins, apes,and hominids have been attributed to conver-gent processes (Westergaard and Suomi,1994), and the extent to which the hand andwrist morphology of Cebus differentiates thetool-behavior of capuchins from that of hu-mans remains to be determined. We believethat the data presented in this report areconsistent with Napier’s (1962) assertionthat the ability to form a precision grip is notan essential requisite for use and productionof simple stone tools, and Susman’s (1995)assertion that the earliest hominid technol-ogy emerged prior to the development oftrue thumb opposition and correspondingremnants in the archaeological record.

ACKNOWLEDGMENTS

Subjects were housed in Laboratory ofComparative Ethology facilities at the Na-tional Institutes of Health Animal Center inPoolesville, Maryland. We thank ClaraMenuhin-Hauser and Lily Westergaard for

commenting on an earlier draft of thismanu-script.

LITERATURE CITED

Anderson JR (1990) Use of objects as hammers bycapuchin monkeys (Cebus apella). Folia Primatol.54:138–145.

Boesch C, and Boesch H (1993) Different hand posturesfor pounding nuts with natural hammers by wildchimpanzees. In Preuschoft H, and Chivers DJ (eds.):Hands of Primates. New York: Springer-Verlag, pp.31–43.

Costello MB, and Fragaszy DM (1988) Prehension inCebus and Saimiri: I. Grip type and hand preference.Am. J. Primatol. 15:235–245.

Guthrie EA (1991) Variability of the primate trapezio-metacarpal articulation: Description and functionalsignificance. MA Thesis, Arizona State University,Tempe.

Jude J (1993) Manipulative behavior of hamadryasbaboons. Senior Thesis, Arizona State University,Tempe.

Marzke MW, and Wullstein KL (1996) Chimpanzee andhuman grips: A new classification with a focus onevolutionary morphology. Int. J. Primatol. 17:117–139.

Napier JR (1960) Studies of the hands of living pri-mates. Pro. Zool. Soc. Lond. 134:647–657.

Napier JR (1962) Fossil hand bones fromOlduvai Gorge.Nature 196:409–411.

Parker ST and Gibson KR (1977) Object manipulation,tool use and sensorimotor intelligence as feedingadaptations in Cebus monkeys and great apes. J.Hum. Evol. 6:623–641.

Schick KD, and Toth N (1993) Making Silent StonesSpeak: Human Evolution and the Dawn of Technol-ogy. NewYork: Simon & Schuster.

Susman RL (1994) Fossil evidence for early hominid tooluse. Science 265:1570–1573.

Susman RL (1995) Thumbs, tools, and early humans.Science 268:589.

Toth N, Schick KD, Savage-Rumbaugh S, Sevcik RA,and Rumbaugh DM (1993) Pan the tool-maker: Inves-tigations into the stone tool-making and tool-usingcapabilities of a bonobo (Pan paniscus). J. Arch. Sci.20:81–91.

Visalberghi E (1990) Tool use in Cebus. Folia Primatol.54:146–154.

Westergaard GC (1994) The subsistence technology ofcapuchins. Int. J. Primatol. 15:899–906.

Westergaard GC, and Suomi SJ (1994) A simple stone-tool technology in monkeys. J. Hum. Evol. 27:399–404.

Westergaard GC, and Suomi SJ (1995) The stone tools ofcapuchins (Cebus apella). Int. J. Primatol. 16:1017–1024.

Westergaard GC, and Suomi SJ (1996) Hand preferencefor stone artifact production and tool-use by monkeys:Possible implications for the evolution of right-handedness in hominids. J. Hum. Evol. 30:291–298.

135CAPUCHIN GRIPS FOR STONE TOOL USE