YEARBOOK OF PHYSICAL ANTHROPOLOGY 42:1–30 (1999)
Evolution of Coalitionary Killing RICHARD W. WRANGHAM Department of Anthropology, Peabody Museum Harvard University, Cambridge, Massachusetts 02138
KEY WORDS
chimpanzee; lethal raiding; warfare; assessment
ABSTRACT Warf arfare are has tra tradit dition ionall ally y bee been n con consid sidere ered d uni unique que to humans. It has, therefore, often been explained as deriving from features that are unique to humans, such as the possession of weapons or the adoption of a patriarchal ideology. Mounting evidence suggests, however, that coalitional killing of adults in neighboring groups also occurs regularly in other species, including includ ing wolves and chimp chimpanzee anzees. s. This implies that selec selection tion can favor components of intergroup aggression important to human warfare, including lethal let hal rai raidin ding. g. Her Here e I pre presen sentt the pri princi ncipal pal ada adapti ptive ve hyp hypoth othesi esiss for exp explai lainin ning g the spec species ies dist distribu ribution tion of inte intergro rgroup up coal coalition itional al kill killing. ing. This is the ‘‘im ‘‘imbala balancenceof-power hypothesis,’’ which suggests that coalitional killing is the expression of a drive for dominance over neighbors. Two conditions are proposed to be both bot h nec necess essary ary and su suff ffici icient ent to acc accoun ountt for coa coalit lition ional al kil killin ling g of nei neighb ghbors ors:: (1) a state of intergroup hostility; (2) sufficient imbalances of power between partie par tiess tha thatt one party can att attack ack the oth other er wit with h impu impunit nity y. Und Under er the these se condit con dition ions, s, it is sug sugges gested ted,, sel select ection ion fav favors ors the ten tenden dency cy to hun huntt and kil killl riv rivals als when the costs are sufficiently low. The imbalance-of-power hypothesis has been criticized on a variety of empirical and theoretical grounds which are discussed. discu ssed. To To be furthe furtherr teste tested, d, studi studies es of the proximate determinants determinants of aggression are needed. However, current evidence supports the hypothesis that selection has favor favored ed a hunthunt-and-k and-kill ill propen propensity sity in chimpa chimpanzees nzees and huma hu mans ns,, an and d th that at co coal alit itio iona nall ki kill llin ing g ha hass a lo long ng hi hist stor ory y in th the e ev evol olut utio ion n of bo both th species. specie s. Yrbk Phys Anthr Anthropol opol 42:1–30, 42:1–30, 1999. 1999 Wiley-Liss, Inc.
TABLE OF CONTENTS Coalitionary Killing among Chimpanzees and Other Species .......... .................... ..................... ..................... ............... ..... 4 Species distribution of coalitionary killing .......... .................... .................... ..................... ..................... .................... ..................... ............. 5 The lethal-raiding problem .......... .................... ..................... ..................... .................... ..................... ..................... .................... ..................... ............... .... 5 Lethal raiding by chimpanzees ........... ..................... .................... ..................... ..................... .................... ..................... ..................... .................. ........ 5 Territorial defense ........... ..................... .................... ..................... ..................... .................... ..................... ..................... .................... ..................... ............... .... 6 Border patrols .......... .................... ..................... ..................... .................... ..................... ..................... .................... ..................... ..................... .................... ............ 7 Deep incursions .......... .................... .................... ..................... ..................... .................... ..................... ..................... .................... ..................... .................... ......... 7 Coalitionary attacks .......... .................... .................... ..................... ..................... .................... ..................... ..................... .................... ..................... ............. .. 8 Coalitionary kills ......... .................... ..................... .................... ..................... ..................... .................... ..................... ..................... .................... .................. ........ 8 The Imbalance-of-Power Hypothesis ........... ..................... .................... ..................... ..................... .................... ..................... ..................... ............ 11 Explaining chimpanzee violence .......... .................... ..................... ..................... .................... ..................... ..................... .................... ............... ..... 11 The significance of power imbalances ......... .................... ..................... .................... ..................... ..................... .................... .................. ........ 12 Origins of power imbalances .......... .................... .................... ..................... ..................... .................... ..................... ..................... .................... ............ 12 Group territoriality and the benefits of lethal raiding .......... ..................... ..................... .................... ..................... ............. 14 Sex differences in territoriality and aggressiveness ......... .................... ..................... .................... ..................... ................ ..... 16 Bonobos: exceptions that support the rule? .......... ..................... ..................... .................... ..................... ..................... .................. ........ 17 The imbalance-of-power hypothesis and the evolution of human warfare ...................... 18 Challenges to the Imbalance-of-Power Hypothesis .......... .................... ..................... ..................... .................... .................... .......... 20
1999 WILEY-LISS, INC.
2
YEARBOOK YEARBOO K OF PHYSICA PHYSICAL L ANTHROP ANTHROPOLOGY OLOGY
[Vol. 42, 1999
Uncertainty in the chimpanzee data .......... ..................... ..................... .................... ..................... ..................... .................... .................. ........ 20 The claim that biology is irrelevant for human warfare ........... ..................... .................... ..................... ................... ........ 21 Implications of the Imbalance-of-Power Hypothesis .......... .................... .................... ..................... ..................... .................. ........ 22 Chimpanzee and human psychology ......... .................... ..................... .................... ..................... ..................... .................... .................... .......... 22 The complexity of war .......... ..................... ..................... .................... ..................... ..................... .................... ..................... ..................... .................... ............ 23 The relation between lethal raiding and hunting ........... ..................... .................... ..................... ..................... .................. ........ 24 Morality .............................................................................................................................. 25 Conclusion ......... .................... ..................... .................... ..................... ..................... .................... ..................... ..................... .................... ..................... ..................... ............ .. 26 Acknowledgments .................... ......... ..................... .................... ..................... ..................... .................... ..................... ..................... .................... ..................... ........... 26 Literature Cited ........... ..................... .................... ..................... ..................... .................... ..................... ..................... .................... ..................... ..................... ............ .. 27
Two related but distinct hypotheses have terms of predation by carnivores and taphoproposed propos ed tha thatt war warfare fare has its origins origins in nomic processes, and the supposed bone and pre-human violence. The first is no longer horn weapons were better explained as fragsuppor sup ported ted.. Thi Thiss was the soso-cal called led ‘‘k ‘‘kill iller er ments produced by carnivores chewing bone ape’’ hypothesis, which stated that warfare (Brain, 1981; Cartmill, 1993). The killer ape spring spr ingss from an agg aggres ressiv sive e ins instin tinct ct tha thatt hypothesis fell into general disrepute in the began among austr australopi alopithecin thecines es and contin contin-- 1970s 1970s,, even though the notion that warfare ued into humans (Dart 1953; Ardrey, 1961, evolved out of complex hunting patterns has 1966; Lorenz, 1966; Tiger, 1969). Raymond not completely died (Morri (Morris, s, 1977; Ferrill, Dart based this idea on South Afric African an homi- 1985). nid fos fossil sils, s, whi which ch he int interp erpret reted ed wit with h inIn the 1980’s a second, unrelated, set of creasing pessimism after the Second World ide ideas as aro arose, se, which I col collec lectiv tively ely cal calll the War unti untill even eventuall tually y conc concludin luding g that that Austra Austra- chimp chimpanzee anzee viole violence nce hypot hypothesis hesis (CVH) (CVH).. Like lopithecus africanus not only hunted other the killer ape hypothesis, the CVH proposes mammals but also killed adult conspecifics that human warfare is built on pre-human (Dart, 1953; Dart and Craig, 1959; Cartmill, te tend nden enci cies es.. In co cont ntra rast st to th the e ki kill ller er ap ape e 1993). At the time of these ideas, intraspe- hypothesis, however, the CVH does not posit cific killing was considered to be absent in a prior history of hunting, nor an aggressive other wild mammal mammalss (inclu (including ding chimpan- instinct. These and other differences make zees Pan troglodytes) troglodytes) (Lorenz, 1966). There- the killer ape hypothesis irrelevant to the fore, killing by australopithecines was con- CVH (Table 1). The remainder of this paper sidered part of a uniquely hominid suite of is concerned with the CVH, and not with the characteristics. Lorenz (1966) lent ethologi- ki kill ller er ap ape e hy hypo poth thes esis is wi with th wh whic ich h it ha hass cal authority to Dart’s ideas by suggesting sometimes been confused (Sussman, 1997). how killing could evolve. He proposed, for The CVH proposes that selection has faexample, that the use of weapons, such as vor vored ed a te tend nden ency cy am amon ong g ad adul ultt ma male less to pebble tools, could overcome natural inhibi- assess the costs and benefits of violence, and tions tio ns aga agains instt kil killin ling g con conspe specifi cifics. cs. Lor Lorenz enz to attack rivals when the probable net ben(1966) thus followed Dart in arguing that efits are sufficiently high. It suggests that human hum an war warfar fare e had evo evolve lved d fro from m aus austra tralop lopii- this tendency occurs as a result of similar thecine aggressive instincts. conditi con ditions ons in the liv lives es of chi chimpa mpanze nzee e and The killer ape hypothesis provoked vigor- human ancestors, including a fission-fusion ous attacks (e.g. (e.g.,, Ashley Monta Montagu, gu, 1968) 1968).. sy syste stem m of gro groupi uping, ng, and int interg ergrou roup p hos hostil tility ity.. Much criticism was directed at theoretical It also raises the question of whether lethal components, such as the claim that humans raiding had a common origin in the ancestor have an innate aggressive drive that needs of chi chimpa mpanze nzees es and hum humans ans aro around und 5–6 period per iodic ic exp expres ressio sion. n. It was emp empiri irical cal evi evi-- my mya, a, or wh wheth ether er it ev evolv olved ed lat later er an and d ind indepe epenndence that felled it, however. Most impor- dently in each line. tantly, the fossils suggestive of intraspecific The CVH was stimulated by observations violence were convincingly reinterpreted in of male chimpanzees collaborating to kill or
Wrangham]
EVOLUTION OF COALITIONARY KILLING
3
for at least 5 million years. This idea of an ancient origin of warfare is supported by the Chimpanzee rarity of coalitionary lethal violence toward Killer ape violence adult conspecifics in other primates, and by hypothesis hypothesis evidence that subsequent to the split with Lethal Let hal vio violen lence ce Import Imp ortant ant in Important in gorillas Gorilla gorilla gorilla(Pil (Pilbeam beam,, 1996 1996), ), chim chim-human evoluhuman evolutionary history tionary history panzee pan zeess and hum humans ans sha share re a com common mon anc ancesesSignificance of Critical prePossible consetor around 5–6 mya. hunting cursor to intraquence of In the first part of this paper, the evidence specific viointraspecific lence violence; not a for coalitionary killing by chimpanzees and necessary prethe nat nature ure of the their ir int interg ergrou roup p agg aggres ressio sion n cursor Mechanism of Insti nc nct Strategic assessare examined. The principal adaptive explaaggression ment nation linking chimpanzee and human vioPutative reason Inadvertent Assessment that for intraspebreakdown of costs of elimilence len ce is the then n rev review iewed. ed. Th This is is the imb imbala alance nce-cific violence natural inhibinating rival of-power of-p ower hypo hypothes thesis, is, whi which ch sta states tes that tions are low coalitionary kills occur because of two facChimpanzees Nonv No nvio iole lent nt Stra St rate tegi gica call lly y vi viooconsidered to lent tors: intergroup intergroup hostil hostility ity,, and large power be asymmetries asym metries between rival partie parties. s. After Killing among Assumed to be Known to occur animals other absent considerin consi dering g separ separately ately the costs and benthan hominids efits of lethal raiding among chimpanzees, Reliance on fossil Critical Relevant, not how the imbalance-of-power hypothesis also evidence critical First ancestor of AustralopitheUnknown applie app liess to bon bonobo oboss ( Pan panisc paniscus us) and to humans supcines humans is assessed. posed to have coalitionary In the third part, objections and problems violence are consid con sidere ered. d. Arg Argume uments nts are dis discus cussed sed Evolutionary Include group Group selection that deny the relevance of biological argumechanisms selection appears favoring viounnecessary ments for unders understandin tanding g human warfare, lence including inclu ding the follow following ing claims: warfare is wholly cultural (e.g., Keeley, 1996); modern war is too complex for individual aggression brutally bruta lly wound other adults (Goodall et al., to be important (Hinde, 1993); and, nothing 1979). Most such attacks were directed touseful ful can be lea learne rned d by stu studyi dying ng spe specie ciess ward members of neighboring communities, use other er tha than n hum humans ans,, bec becaus ause e hum humans ans are in patterns reminiscent of human war raids oth already known to be violent (Leach, 1968; (Goodall, (Good all, 1986). As a resul result, t, various authors raised the possibility of functional parallels Gould, 1996). Criticisms directed specifically at the imand/or evolut evolutionary ionary continu continuities ities linkin linking g balance-ofce-of-power power hypot hypothesis hesis are also dischimpa chi mpanze nzee e vio violen lence ce and hum human an war warfar fare e balan cussed cus sed. . These The se includ inc lude e concer con cerns ns abo about ut the (Trudeau et al., 1981; Otterbein, 1985, 1997; validity of the chimpanzee data (e.g., Power, Goodall Goo dall,, 1986 1986;; Ale Alexan xander der,, 1987 1987,, 198 1989; 9; Wran ranggham, ha m, 198 1987, 7, 199 1999b;Ghig 9b;Ghiglie lieri, ri, 19 1988; 88; va van n Hoo Hooff, ff, 1991; Sussman, 1997) or about the interpretation ion of hum human an dat data a (e. (e.g., g., Kna Knauft uft,, 199 1991; 1; 1990; Hamburg, 1991; Knauft, 1991; Man- tat Spon onse sel, l, 19 1996 96), ), an and d cl clai aims ms th that at da data ta on son and Wran rangha gham, m, 199 1991; 1; Boe Boehm, hm, 199 1992; 2; van Sp bonobos bonob os undermine under mine the use of chimpanzees chimpa nzees der Den Dennen nen,, 199 1995; 5; Wran rangha gham m and Pet Peters erson, on, refe fere renc nce e sp spec ecie iess for ea earl rly y ho homi mini nid d 1996; 199 6; Boe Boesch sch and Boe Boesch sch,, 199 1999). 9). For ex- as a re ample, Otterbein (1997, p 253) noted that ancestry (Zihlman, 1997; Stanford, 1998a). They y als also o inc includ lude e acc accusa usatio tions ns of gen genetic etic similarities between chimpanzee communi- The determinism inism (e.g. (e.g.,, Regal Regal,, 1998; Sussman, ties and human bands suggest that ‘‘early determ man . . . is likely to have been organized into 1997) or societal bias (e.g., Sussman, 1997). These are import important ant consi considerati derations, ons, but locali loc alized zed gro groups ups of rel relate ated d mal males, es, gro groups ups that engaged in interg intergroup roup conflict.’’ conflict.’’ If so, they do not invalidate the comparative apOtterbein concluded, warfare has been con- proach. Therefore this article ends with a tinuous in human and pre-human ancestry bri brief ef dis discus cussio sion n of the imp implic licati ations ons of the TABLE 1. Comparison of the killer killer ape and chimpanzee violence hypotheses
4
YEARBOOK YEARBOO K OF PHYSICA PHYSICAL L ANTHROP ANTHROPOLOGY OLOGY
imbalance-of-powe imbalance-o f-powerr hypot hypothesis hesis for human psychology, warfare, and morality.
COALITIONARY KILLING AMONG CHIMPANZEES AND OTHER SPECIES Species distribution of coalitionary killing Contrary to initial assumptions (Lorenz, 1966), 196 6), res resear earch ch in rec recent ent dec decade adess has re vealed that intraspecific killing occurs in a variety of species, commonly following patterns explicable by natural selection theory. For example, among primates infanticide is widely wid ely rep report orted, ed, typ typica ically lly com committ mitted ed by non non-relatives (Hausfater and Hrdy, 1984; Palombit, 1999). Among spiders, killing of adults occurss predic occur predictably tably when resou resources rces of high value value are at stak stake e (Aust (Austad, ad, 1983). Among ants, large imbalances of power increase the probability of lethal intercolony aggression (Ho¨lldobler, ¨lldobler, 1981; Adams Adams,, 1990). Obvio Obvious us parallels can be found among humans (Daly and Wilson, 1988; Bueno de Mesquita, 1981, 1985; 19 85; Bu Buen eno o de Mes Mesqu quita ita and Lal Lalman man,, 199 1992). 2). To some extent, therefore, therefore, patter patterns ns of human killing killing app appear ear to fol follow low the ord ordina inary ry patterns patter ns of lethal aggression aggression found in other species. Not so ordinary, however, is the way that human killing occurs. Among humans most killin kil ling g occ occurs urs in war warfar fare, e, wh where ere the pre predom domiinant style of violence is coalitionary. In most animals, by contrast, even where aggression occurs occ urs at hig high h rat rates, es, let lethal hal vio violen lence ce is dya dyadic dic (one (on e ver versus sus one one)) rat rather her tha than n coa coaliti litiona onary ry (many vs. one, or many vs. many). During rut-fightin rut-fi ghting g among male prong pronghorn horn antelope ( Antilocapra americana), for example, 12% of 82 fights over mating rights to estrous females led to the death of one or both males (Byers, 1997). Likewise, in different populations of red deer Cervus elaphus, 13– 29% 29 % of ad adul ultt ma male le mo morta rtali lity ty ca came me fro from m rut-fig rut -fighti hting ng (Cl (Clutto utton-B n-Broc rock k et al. al.,, 198 1982). 2). Many similar examples occur. Deaths tend to occur in intensely escalated contests in which both opponents expose themselves to high hig h ris risk k of inj injury ury,, typ typica ically lly bec becaus ause e ‘‘a major part of a contestant’s lifetime reproductiv duc tive e suc succes cesss is at sta stake’ ke’’’ (En (Enqui quist st and Leimar, 1990). But killing is never coalitionary in these species. The exp explan lanati ation on for the wid widesp esprea read d absence of coalitionary violence is trivial. Most
[Vol. 42, 1999
species never form coalitionary alliances in any context. But coalitions are not a sufficient condition for coalitionary killing. Thus many primates form coalit coalitions ions without any evidence of adult-killing [e.g., Cercopithecus aethiops (Cheney et al., 1988)] or with fatal fightin figh ting g kno known wn onl only y fro from m dya dyadic dic int intera eracctions [e.g., Cebus capucinus] capucinus] (Miller, 1998). Indeed,, the only nonpr Indeed nonprimate imate mammal for which coalitionary violence is known to be common com monly ly res respon ponsib sible le for adu adult lt dea deaths ths is the wolf Canis wolf Canis lupus. In at least three sites, adults are known to kill other adults at high rates [Denali (Alaska), Isle Royale (Michigan), gan ), and Min Minnes nesota ota]] (Me (Mech ch et al. al.,, 199 1998). 8). Forr ex Fo exam ampl ple e in De Dena nali li,, 39 39–6 –65% 5% of ad adul ultt mortality mortal ity was due to intras intraspecific pecific killing, killing, based on 22 intraspecific killings recorded from 17–20 packs (Mech et al., 1998). This is the least disturbed study site of wolves. In northeaste north eastern rn Minne Minnesota, sota, 43% of wolv wolves es not kille ki lled d by hum human anss we were re ki kille lled d by oth other er wo wolv lves es.. Kill Ki llin ings gs te tend nded ed to oc occu curr in bu buff ffer er zo zone ness (where (wh ere ter territ ritori ories es met met), ), whi which ch wol wolves ves mos mostly tly avoided (Mech, 1994). These data suggested to Mech et al. (1998) that intraspecific killing in g is a no norma rmall co cons nseq eque uenc nce e of wo wolf lf te terr rrit itor oriiality. Occasional coalitional killing of adult conspecifi spe cifics cs in nei neighb ghbori oring ng gro groups ups has als also o bee been n recorded among other social carnivores [lions Panth Panthera era leo, spotte spotted d hyena hyenass Crocuta crocuta, cheetahs Acinonyx jubatus (Kruuk, 1972; Caro and Collins, 1986; Goodall, 1986; Packer et al., 1988; Grinnell et al., 1995)] and at least one group-territorial bird [Tasmanian native hen Gallinula mortierii (Putland and Goldizen, 1998; A. Goldizen persona so nall com commun munica icatio tion)] n)].. How Howev ever er,, the frequency of killing has not been reported for these species. Outside of mammals, only social insects are known to kill conspecifics regularly regul arly with coalit coalitional ional aggre aggressio ssion n (van der Dennen, 1995). These data suggest that animals can be divide div ided d int into o thr three ee maj major or cat catego egorie ries. s. Fir First st are species in which intraspecific killing of adults is rare (e.g., less than 1% of all adult deaths). Most species fall into this category. Second are those where killing occurs more freq fr eque uent ntly ly [o [oft ften en 10 10% % of mo more re of de deat aths hs (Enquist (Enqu ist and Leima Leimarr 1990)] 1990)],, but entirely in dyadic interactions. In these species, fatal
Wrangham]
EVOLUTION OF COALITIONARY KILLING
5
fighting is dangerous. In the third category, Abundant evidence routinely attests to the killing is also frequent, but differs by being blood-lust of the participants. polyadic (coalitionary). Furthermore, as ar Against the notion that men have a ready gued gu ed be below low,, fat fatal al fig fighti hting ng ne need ed not be dan dange gerr- app appeti etite te to att attack ack the their ir ene enemie mies, s, com combat bat-ous for the killers. ants in modern warfare are often reluctant Ants dominate this last catego category ry,, which to figh fightt (Hi (Hinde nde,, 199 1993; 3; Ehr Ehrenr enreic eich, h, 199 1997; 7; contains probably less than 10 mammalian Gro Grossm ssman, an, 199 1999). 9). This This fro frontnt-lin line e lac lack k of species speci es and perhaps no other vertebrates. vertebrates. aggr aggressiv essivenes enesss is unders understandab tandable le becau because se Chimpanzees and humans (or at least, cer- in mode modern rn warf warfare, are, unl unlike ike inte intergro rgroup up agg aggresrestain tai n pop popula ulation tionss of the these se spe specie cies) s) are the sio sion n in pri primat mates, es, sol soldie diers rs are org organi anized zed hie hierronly prim primates ates kno known wn to be freq frequent uent coal coalitio itionn- arc archic hicall ally y and are ord ordere ered d into into bat battle tle by ary killers. A possible additional candidate their superiors, regardless of their personal is the western red colobus monkey (Colobus (Colobus mot motiva ivatio tion. n. Par Partic ticipa ipation tion in rai raids ds amo among ng badius), badius ), for which at least two and possibly pre-state societies, however, is normally volfour coalitionary kills were recorded by Sta- untar untary y (Keel (Keeley ey,, 1996). Thus, reluc reluctance tance of rin (1994). In these cases, coalitions of fe- soldi soldiers ers under orders does not underm undermine ine males attacked and killed males attempting the more widespread phenomenon of male to enter their groups. eagerness for fighting. It is likely that lethal raids also occur in The lethal-raiding problem some of the carnivore species that engage in In chimpanzees, humans, and some other intraspecific killing of adults. For example, animal ani mals, s, coa coaliti litiona onary ry kil killin ling g can occ occur ur in the lethal raiding is suggested by the report of Mech ch et al al.. (1 (199 998) 8) th that at ne neig ighb hbor orss ki kill lled ed context of lethal raiding. Lethal raids are an Me unusual form of aggression because they do three members of a wolf pack, two others disapp appear eared ed uns unseen een,, and the defe defeate ated d pac pack’ k’ss not escalate from a conflict. Instead, parties dis territ ter ritory ory was wa s taken tak en over ove r by the killer kil lers. s. There The re of alli allied ed male maless colle collectiv ctively ely inva invade de a neig neighbor hbor-ing territory, territory, seek one or more vulnerable are also reports of spotted hyenas making incurs ursion ionss int into o nei neighb ghbori oring ng terr territo itorie riess to neighbors, neigh bors, appare apparently ntly asse assess ss the proba probability bility inc attack att ack neighb nei ghbors ors (Gooda (Go odall, ll, 1986). 198 6). Among Amo ng of mak making ing a suc succes cessfu sfull atta attack, ck, conduct conduct a inverteb rtebrates rates,, patte patterns rns simi similar lar to leth lethal al raid raid-‘‘surp ‘‘s urpris rise’’ e’’ attack attack tha thatt lea leaves ves one or mor more e inve ing occur in a variety varie ty of ants (Ho ¨lldobler, ¨lldobler , victims dead or dying, then return to their 1981; van der Dennen, 1995). own territory. territory. ‘‘Surprise’’ refers to the attack Among species other than humans, howoccurring without any initial conflict, withever eve r, let lethal hal rai raids ds hav have e bee been n mos mostt cle clearl arly y out escalation from a lower level, and withreported in chimpanzees. Evidence of chimout the vic victim tim int intera eracti cting ng wit with h the opp oppoopanzee pan zee rai raidin ding g has bee been n fun fundam dament ental al for nents until the attack starts. the devel developm opment ent of the CVH. CVH . I therefo the refore re Thus, lethal raids indicate an appetite for huntin hun ting g and killing killing riv rivals als that is aki akin n to review the chimpanzee data in detail. predation predat ion.. By con contra trast, st, mos mostt ani animal mal con con-flicts flic ts esc escala alate te in a ste stepwi pwise se man manner ner tha thatt allows both opponents to assess each other and to withdraw when the risks of losing appearr too high (Arch appea (Archer er and Huntingford, Huntingford, 1994). 1994 ). The ‘‘ap ‘‘appetit petite e for leth lethal al raidi raiding’ ng’’’ ther thereefore requires explanation in different terms from escalated conflicts. Among Among hum humans ans,, let lethal hal rai raids ds are wid wideespread in all forms of warfare. For example, Keeley (1996) regards small raids and ambushes as ‘‘the commonest form of combat employ emp loyed ed in pri primit mitive ive war warfar fare’’ e’’ (see als also o Turney-High, 1949; van der Dennen, 1995; Maschne Mas chnerr and Reed Reedy-Ma y-Maschn schner er,, 1998 1998). ).
Lethal raiding by chimpanzees Although lethal raiding among chimpanzees has been described more clearly than for any other mammal, few cases have been completely comple tely obser observed. ved. Furth Furthermore ermore,, all the detail det ailed ed obs observ ervati ations ons com come e fro from m a sin single gle site si te,, Go Gombe mbe Na Natio tional nal Pa Park rk in Tan anza zania nia (Goodall, 1986). Of course, lethal raids are expected to be rare, as they must be in any long-lived, slowly reproducing species. Nevertheless, erthe less, the concen concentratio tration n of obser observatio vational nal evidence at Gombe has suggested to critics of the CVH tha thatt let lethal hal raiding raiding may have been induced in Gombe by unnatural condi-
6
[Vol. 42, 1999
YEARBOOK YEARBOO K OF PHYSICA PHYSICAL L ANTHROP ANTHROPOLOGY OLOGY
TABLE 2. Ter Territorial ritorial behavior in chimpanzees chimpanzees and bonobos 1 P. t. schweinfurthii
Territorial defense Border patrols Deep incursions Coalitionary attacks Coalitionary kills Border avoidance Peaceful intercommunity association
P. t. verus
P. paniscus
Gombe
Mahale
K i bal e
Budongo
Taı¨
Wamba
Lomako
?
? ? ?
?
?
1 All
long-term studies are includ long-term included ed except for Bossou Bossou,, where the community is isolated by agricu agricultural ltural land from its neare nearest st chimpanzee neighbors. ‘‘Coalitionary kills’’ refers to adult victims only. ‘‘Coalitionary attacks’’ means non-lethal attacks by several males on a single victim. Data are from: Gombe (Goodall, 1986); Mahale (Nishida, 1979, 1986; Nishida et al. 1985); Kibale (Chapman and Wra Wrangham, ngham, 1993); Budong Budongo o (V (V.. Reyno Reynolds, lds, person personal al commun communicati ication); on); Taı ¨ (Boesc (Boesch h and Boesch, 1999); Wamba (Kano, 1992; Hashimoto et al., 1998); Lomako (White, 1996).
TABLE 3. Chimpanzees and bonobos: bonobos: lethal violence by site site 1 Site
Subspecies
Adult deaths
Infanticides
Years of study
Gombe Mahale Kibale Budongo Bossou Taı¨ Wamba Lomako Total
P. t. schweinfurthii P. t. schweinfurthii P. t. schweinfurthii P. t. schweinfurthii P. t. verus P. t. verus P. paniscus P. paniscus
6 (3) 1 (6) 2 (0) 1 (1) 0 (0) 0 (0) 0 (0) 0 (0) 10–20
6 (3) 4 (6) 1 (0) 1 (1) 0 (0) 0 (0) 0 (0) 0 (0) 12–22
38 33 11 8 22 19 24 15 170
1
Numbers include (for adult deaths) kills recorded on the basis of direct observation and/or fresh bodies as well as (in parentheses) thosefrom suspic suspicious ious disap disappeara pearances nces (seeTable 5), or (forinfanticide (forinfanticides) s) killsobserveddirectly or (in paren parentheses theses)) inferr inferred ed fromcontext (Arcadi and Wrangham 1999, updated for Mahale by Nishida (personal communication). ‘‘Years of study’’ is number of years from beginning of continuous study until 1998. Bonobo studies have been intermittent.
tions such as reduced habitat, or provisioning, in g, an and d th ther eref efor ore e th that at le leth thal al ra raid idin ing g is uncharact unch aracteris eristic tic of chimp chimpanze anzees es more gener gener-ally (Power, 1991). There are six study sites in which habituation of chimpanzees is sufficiently good to allow allo w multi multi-hou -hourr obse observati rvations ons of know known n indi viduals vid uals trave traveling ling thro througho ughout ut the home rang range e (Tables 2 and 3). These include four studies of th the e ea east ster ern n su subs bspe peci cies es ( P. P. t. sch schwei weinn furthii), furthii ), none of the two central subspecies ( P P.. t. troglo troglodytes dytes and and P P.. t. veller vellerosus osus), ), an and d tw two o of the western chimpanzee ( P. t. verus). verus). In one of the P. t. verus studies, at Bossou, the study community is ‘‘semi-isolated’’ by loss of habitat, separated by several kilometers from the home ranges of its closest neighbors bo rs (Su (Sugi giya yama, ma, 19 1989; 89; Su Sugiy giyama ama et al al., ., 1993). Consequently there is no possibility of ter territ ritori orial al beh behavi avior or or int interc ercommu ommunit nity y intera int eracti ction on fro from m Bos Bossou sou.. Thi Thiss lea leaves ves five studies of chimpanzees, varying in duration from 8 to 38 years, that permit observation of intergroup interactions (Table 3).
Six components of chimpanzee intergroup aggression are especially relevant to lethal raidin rai ding: g: ter territ ritori orial al def defens ense e (sh (showi owing ng evi evi-dence of hosti hostile le interg intergroup roup relati relationshi onships), ps), border patrols, deep incursions, coalitionary attacks, coalitionary kills, and border avoidance (Table 2). Territorial defense. Th This is ha hass be been en re re-ported in all studies in which intercommunity rela relation tionshi ships ps hav have e been des describ cribed, ed, bas based ed on some combination combination of: count counter-ca er-calling lling between parties of neigh neighborin boring g males males;; rapid trav tr avel el tow towar ard d a si site te wh wher ere e an op oppo posi sing ng pa part rty y has bee been n det detecte ected; d; avo avoida idance nce of opp opposi osing ng parties that were obviously larger; charging displays directed toward an opposing party of mal males es;; or on one e pa party rty ch chas asin ing g an anoth other er (Nishida, 1979; Goodall, 1986; Boesch and Boesch, 1999; V. Reynolds personal communication; Wrangham et al., in preparation). For example on nine occasions chimpanzees in Taı¨ hav have e bee been n see seen n in ‘‘b ‘‘back ack-an -and-fo d-forth rth attacks’’ with neighbors, in which all males
Wrangham]
EVOLUTION OF COALITIONARY KILLING
rush toward the opponents, giving loud attack calls, and opponents may flee up to 400 m. In a fu furt rthe herr ni nine ne ca case ses, s, ma male less in th the e fr fron ontt line of attack were supported by loud calls from fr om fe fema male less in a re rear ar li line ne (B (Boe oesc sch h an and d Boesch, 1999).
7
(Wrangham, 1975, Table 5.9). By contrast, in the 5 years following the extinction of the Kahama Kaha ma commu community nity (1978 (1978–1982 –1982), ), borde borderr patrols continued at a rate of 18 per individual per year (range 9–27; calculated from Goodall, 1986, Table 17.1). Tabl able e 2 sho shows ws tha thatt bor border der pat patrol rolss hav have e been bee n rep reporte orted d als also o fro from m Mah Mahale ale,, Kib Kibale ale,, and Taı¨. ¨. Border patrols have not been described in detail from Mahale, but key elements of border patrols have been reported—including scouting, and silent and cautious travel, mainly by males, in border areas (Nishida, 1979, 1990; Nishida et al., 1985). The Kibale evidence comes from border patrols seen at Kanyawara Kany awara,, invo involving lving partie partiess of males inter inter-mittently checking their territorial boundaries (Wrangham et al., in preparation). In Taı¨, ¨, border patrols occurred in 29% of 129 territ ter ritori orial al act action ions, s, nor normal mally ly inv involv olving ing at leas le astt fo four ur ma male les, s, an and d in incl clud uded ed al alll of th the e elements listed above (Boesch and Boesch, 1999).
Border patrols. The These se are an int intrin rinsic sic component compon ent of letha lethall raidin raiding, g, becau because se they put a party of individuals in a position to stal st alk k an and d to hu hunt nt a ne neig ighb hbor orin ing g vi vict ctim im.. Border Bor der pat patrol rolss are visits visits to a per periph iphera erall sector of a home range by a party of males that monitors the area. They are initiated without any immediate contact with members of the neighboring community, and are often undertaken with little or no feeding. According to Goodall (1986) and Boesch and Boesch (1999), they include some or all of the following features: (1) cautious and slow travel tra vel aro around und or acr across oss the bor border der,, inc includ luding ing long periods of gazing toward the neighboring home range; (2) nervousness shown toward unexpected sounds; and (3) inspection of signs of other chimpanzees, such as dis- Deep incursions. These were included as carded food wadges, feces, nests, or aban- part of border patrols by Goodall (1986), but doned termite-fishing tools. I dis distin tingui guish sh them them bec becaus ause e the they y inv involv olve e Bord Bo rder er pa patr trol olss we were re fir first st re repo port rted ed at deliberate travel into the neighboring terriGombe in 1971 by JD Bygott, who was the to tory ry ra rath ther er th than an mere merely ly ch chec ecki king ng of th the e first researcher researcher to condu conduct ct regul regular ar all-da all-day y border area. Deep incursions are characterobservatio obser vations ns of indiv individual idual males (Bygott, ized by (1) substantial penetration into the 1979). Most patrols at Gombe were by males neighboring territory, e.g., for one kilometer of the principal study group, the Kasekela or more; (2) silent and cautious travel durcommun com munity ity,, but the Ka Kaham hama a mal males es pa- ing periods of moving outwards from own trolled troll ed also (Byg (Bygott, ott, 1979; Wr Wrangh angham, am, 1975) 1975).. ter territ ritory ory;; and (3) noi noisy sy and vig vigoro orous us dis dis-Border Bor der patrols patrols do not occur every time a plays on return to their own territory. Deep party reaches the boundary area. In Gombe, inc incurs ursion ionss hav have e bee been n wel welll des descri cribed bed at for example, patrols occurred during 28% of Gombe and Taı¨. ¨. Boesch and Boesch (1999) 134 boundary visits made by parties from fou found nd tha thatt ‘‘m ‘‘many any pat patrol rolss wer were e pro probab bably ly the Kasekela community from 1977 to 1982 aimed at finding and attacking strangers. . . (calculated from Goodall, 1986, Table 17.1). . . . (they) were impressive by the intensity Their frequency appears to vary as a func- with which the males searched for strangtion of relations between particular commu- ers, not only entering deep (into) their terrinities nit ies.. Thu Thus, s, dur during ing and aft after er the 197 1974– 4– tory tory,, but once even heading backwards backwards to 1977 197 7 per period iod,, dur during ing whi which ch the Kas Kaseke ekela la find the neighbors.’’ Deep incursions lasted community killed the males of the neighbor- up to 6 hours, and on 5 of 129 territorial ing Kaha Kahama ma commun community ity,, Kasek Kasekela ela border interactions in Taı¨, ¨, led to attacks. All incurpatrols were disproportionately directed to- sions were led by males. wards war ds the Kah Kahama ama ter territ ritory ory (Go (Gooda odall, ll, 198 1986). 6). The recorded frequency of border patrols by Coalitionary attacks. These are interacKasekela males was highest in 1972–1973, tio tions ns in whi which ch obs observ ervers ers ass assess ess that the when 13 border patrols occurred in 58 days intent of those in the aggressive party is to victims. s. They can of ob obse serv rvati ation on,, i. i.e. e.,, a ra rate te of 82 pe perr ye year ar hurt or to kill one or more victim
8
YEARBOOK YEARBOO K OF PHYSICA PHYSICAL L ANTHROP ANTHROPOLOGY OLOGY
[Vol. 42, 1999
occur within the territory of either the ag- an adolescent male and nulliparous female gressors or the victims, or in the boundary nea nearr the ter territ ritori orial al bor border der,, but lat later er rearea. In Gombe, coalitionary attacks have tre treate ated d whe when n con confron fronted ted by fou fourr adu adult lt mal males es includ inc luded ed bot both h int intera eracti ctions ons bet betwee ween n the wel welll- from the neig neighbo hboring ring comm communi unity ty,, who chas chased ed habituat habi tuated ed Kase Kasekela kela and Kaha Kahama ma commu communini- the aggressors for 700 m (M. Muller, perties, as well as other communities. In late sonal communication). 1974, for example, a party of three Kahama Coalitio Coal itionar nary y atta attacks cks als also o occu occurr with within in commales encountered a male and female found mun muniti ities. es. For exa exampl mple, e, a bul bullyi lying ng and ins insububto the south of their territory (the ‘‘Kalande’’ ord ordina inate te you young ng adu adult lt mal male e (Ji (Jilba lba)) in Macommu co mmuni nity) ty).. Two of the Ka Kaham hama a ma males les hale’s M-group was attacked so severely by grabbe gra bbed d and att attack acked ed the Kal Kaland ande e mal male, e, but six males and two females that it took him 3 he escaped without serious injury (Goodall, months of traveling alone before he recov1986). ered ere d su suff ffici icient ently ly to rej rejoin oin the com commun munity ity In Mahale, cases included both attacks by (Nishida, 1994). M-grou M-g roup p tow toward ard K-g K-grou roup p mal males, es, and vic vice e cases es hav have e bee been n versa. For example, Nishida (1979) recorded Coalitionary kills. No cas report orted ed of dya dyadic dic vio violen lence ce lea leadin ding g to the three thr ee K-g K-grou roup p mal males es cha chasin sing g an M-g M-grou roup p rep death dea th of an adult adu lt chimpa chi mpanze nzee. e. Howeve How everr, male ma le fo forr 20 200 0 m. Two of th the e pu purs rsue uers rs ga gave ve up up,, lethall coalit coalitionary ionary attacks on adult adultss have but the third caught the victim, forced him letha been reported from all four study sites of the to th the e gr grou ound nd,, bi bitt hi hiss th thig igh, h, st stam ampe ped d on hi him, m, eastern easte rn subspecies subs pecies (Tab (Table le 4). Table 4 lists chased him as he tried to escape, but then known and inferred cases. The largest samsuddenly gave up. ple e co come mess fro from m Go Gomb mbe e an and d in incl clud udes es fiv five e In Taı¨, ¨, Boesch and Boesch (1999) reported pl observed brutal attacks followed by the disa category of attack which they called ‘‘comappear earanc ance e of the victim (four mal male, e, one mando,’’ in which a party of males (with or app female), and one case of a fresh corpse of an without females) penetrated into the neighuniden dentifi tified ed fem female ale,, con consid sidere ered d to hav have e bee been n bori bo ring ng ra rang nge e an and d at atta tack cked ed on one e or mor more e uni killed by the Kahama males. Goodall (1986) neighb nei ghbors ors,, i.e i.e., ., a com combin binati ation on of a ‘‘d ‘‘deep eep incursion’ incur sion’’’ with a coalit coalitionary ionary attack. The summarizes the observations as follows: the atta tack ckss la last sted ed at le leas astt 10 mi min n ea each ch;; th the e Taı¨ males sometimes ‘‘waited and listened at victim was always held to the ground by one silently for hours before they attacked. We twice twi ce saw the stu study dy com commun munity ity bei being ng vic victim tim or more of the assailants while others atof a co comm mman ando do at atta tack ck,, in on one e of wh whic ich h tacked; the victim was dragged in at least Mach Ma cho o (a (an n ad adul ultt ma male le)) es esca cape ped d wi with th 19 two directions, eventually gave up resisting, wounds’’ (Boesch and Boesch, 1999). A sec- and was essentially immobilized by the end ond on d fo form rm of co coal alit itio iona nary ry at atta tack ck wa wass th the e of the attack. In Mahale, Nishida et al. (1985) recorded ‘‘lateral attack’’ (seen six times), in which a party moved laterally while looking in the the deaths of all six adult males of K-group commu mmuni nity ty bet betwe ween en 196 1969 9 an and d 198 1980. 0. In Nis Nishihidirectio dire ction n of stra stranger ngers, s, then appr approach oached, ed, co chased, and on at least one occasion caught da’s words, the observers ‘‘speculate that at least st som some e adu adult lt mal males, es, par partic ticula ularly rly Soand attacked attacked one of the opponents opponents.. Thi Thiss lea appeared to be a tactic for increasing the bongo and Kamemanfu, were killed by Mgroup’s ’s chimpa chimpanzees nzees.. Severe fighting was imbala imb alance nce of pow power er by iso isolat lating ing a vic victim tim group occasionally witnessed between males of Kfrom fr om th the e re rest st of th the e pa part rty y (B (Boe oesc sch h an and d group and M-group in (their area of overBoesch, 1999). M-group’s males were somet sometimes imes In Kib Kibale ale,, no com comple plete te coa coaliti litiona onary ry at- lap). . . . M-group’ tacks have been seen, but twice parties of seen to penetrate into the core area of Kneighbors have silently charged toward iso- group’s range from 1974 onwards’’ (Nishida lated lat ed mal males es of the Kan Kanyaw yawara ara stu study dy com commumu- et al., 1985, p 288). The males who disapnity nit y, the then n vee veered red of offf on see seeing ing obs observ ervers ers,, peared were all healthy, not senile. In one suggestin sugg esting g that coalit coalitionary ionary attacks were case, M-group males were known to be very nearr to K-g K-grou roup p mal males; es; there wer were e man many y averte ave rted d by the pre presen sence ce of hum humans ans.. In a nea third case, five Kanyawara males attacked outbursts of calls; and the next day another
Wrangham]
9
EVOLUTION OF COALITIONARY KILLING
TABLE 4. All reported intraspecific intraspecific kills of adult chimpanzees 1 Aggressor ’s Result
Site
Date
Comm Co mmun unit ity y
Part Pa rty y
Victim’s Comm Co mmun unit ity y
Part Pa rty y
Victim’s ID
Death
Gombe
1974
Kasekela
7M , 1F
K ahama Ka
1M
Godi
Death
Gombe
1974
Kasekela
3M , 1F
K ahama Ka
3 M, 1F
De´
Death
Gombe
1975
Kasekela
5M
Kahama
1M
Gol ia iath
Death
Gombe
1977
Kasekela
6M
Kahama
1M
Sniff
Death
Gombe
1975
Kasekela
4M
Kahama
1F
Mada Ma dam m Be Bee e
Death Death
Mahale Bu B udongo
1995 1998
M-gp Sonso
gang gang
M-gp Sonso
1M 1M
Ntologi Zesta ( 2 injured)
Death Death Death
Kibale Kibale G ombe Go
1992 1998 1972
Rurama Kanyawara Kahama
gang g an ang ?
Kanyawara Sebitole Kalande?
1M 1M 1F
Ruwenzori Unknown 1F
Death
Gombe
1977
Kasekela
5M
Kahama
Charl ie ie
Death?
Mahale
1996
M-gp M-
?
M-gp
1M
Jilba
Death?
Mahale
1970-83
M-gp
?
K-gp
?
Some males?
Death?
Gombe
1981
Kalande
?
Kasekela
?
Humphrey killed? M att attack acked, ed, caught by 2, escaped Passion (inferred) 1 on 1 fig fight ht
1M
Attack
Gombe
1974
Kahama
3M
Kalande
M 1,
Attack
Gombe
1980
Kalande
?
Kasekela
F 1
Attack
Mahale Ma
1974
K-group
3M
M-group
M 1
1F
Ref. Goodall (1986, p. 50) Goodall (1986, p. 50) Goodall ( 19 1986, p. 50) Goodall (1986, p. 51) Good Go odal alll (1 (198 986, 6, p. 51) Ni sh shida ( 19 1996) K. Fawcett (personal communication) KCP KCP Wrangham (1975) Goodall ( 19 1986, p. 50) M. Huffman (personal communication), Hofer et al. (1998) Nishida et al. (1985) Goodall (1986, p. 51) Goodall (1986, p. 49) Goodall (1986, p. 51) Good Go odal alll (1 (198 986, 6, p. 51)
1
Parties for aggressors and victims show the number of adult males (M) and females (F). Letters in bold show the victim’s sex. KCP (citation for Kibale deaths) is records of the Kibale Chimpanzee Project.
K-group K-grou p mal male e was mis missin sing g (Ka (Kason sonta) ta) (T Nishida, personal communication). In Ki Kiba bale le,, af afte terr 3 da days ys du duri ring ng wh whic ich h ma male less from fro m the Kan Kanyaw yawara ara and Rur Rurama ama com commun muniities had been counter-calling at each other in a ho host stil ile e ma mann nner er,, a Ka Kany nyaw awar ara a ma male le (Ruwenzori) was found freshly dead in the border area. His body, huddled face down at the th e bo bott ttom om of a sl slop ope e ar arou ound nd wh whic ich h th the e vegetation had been beaten down, showed clear cle ar evi eviden dence ce of a vio violen lentt att attackby ackby chi chimpa mpannzees ze es.. In an un unre rela late ted d in inci cide dent nt,, th the e Kanyawara Kany awara males were followed by obser observvers to the fresh corpse of an individual from a nei neighb ghbori oring ng com commun munity ity (Se (Sebit bitole ole)) who had apparently been killed by chimpanzees the previou pre viouss eve evenin ning. g. The There re wer were e num numerou erouss wounds on the front of his body, his trachea had been ripped through, and both testicles
had been removed. Nine Kanyawara males had been patrolling the border on the previous evening, all of whom were present the next ne xt mo morn rnin ing g (1 (17 7 ho hour urss la late ter) r),, an and d se seve vera rall of whom beat on the victim’s body and dragged it about (M. Mulle Mullerr, perso personal nal commun communicaication). Like coal coalitio itionary nary atta attacks cks,, coal coalitio itionary nary kill killss also occur within communities. At Budongo in 1998, an adult male was killed by other males mal es of his own com commun munity ity (K. Fawcett, Fawcett, personal perso nal commun communicatio ication). n). Intrac Intracommun ommunity ity killing is also thought to have occurred in Mahale Mah ale,, whe where re adu adult lt mal male e Nto Ntolog logii was fou found nd dead dea d in the cen center ter of M-g M-grou roup’ p’ss ter territ ritory ory with wit h num numero erous us wou wounds nds on his bod body y. His death followed several coalitionary attacks on him by his for former mer sub subord ordina inates tes—af —after ter hiss de hi defe feat at as al alph phaa-ma male le of MM-gr grou oup p
10
[Vol. 42, 1999
YEARBOOK YEARBOO K OF PHYSICA PHYSICAL L ANTHROP ANTHROPOLOGY OLOGY
TABLE 5. Chimpanzee coalitionary coalitionary kills of adults adults 1
Site Gombe Gombe Gombe Kibale Kibale Mahale Mahale Budongo Total
Aggressor community
Victim’s community
Kasekela Kahama Kalande Rurama Kanyawara M-group M-group Sonso
Kahama Kalande Kahama Kanyawara Sebitole K-group M-group Sonso
Kill seen, or fresh corpse 5 1
Suspicious disappearance 2 1
1 1 1 1 10
6? 1
Reference Goodall (1986) Goodall (1986) Goodall (1986) KCP KCP Nishida et al. (1985) T. Nishida (personal communication) K. Fawcett (personal communication)
10?
1
Intra-community aggression is indicated by the victim and aggressors being in the same community. KCP, records of the Kibale Chimpanzee Project.
(Nishida, (Nishi da, 199 1996; 6; T. Nis Nishid hida, a, per person sonal al com commumunication). The chimpanzee data are summarized in Table 5. Tables 4 and 5 also list suspicious disappearances. These are cases where observers believed the most likely explanation for dis disapp appear earanc ances es was tha thatt the they y wer were e kil killed led by neighbors, because: (1) those who disappeared were healthy and not senescent; (2) other causes of death appeared improbable; and (3) the dis disapp appear earanc ances es occ occurr urred ed at a time and place where there were patently hostile hos tile rela relation tionship shipss with a neig neighbor hboring ing community. Although the reported episodes of lethal coalitionary violence are still few, the killings are noteworthy because they have been reported from four sites and, in relation to total observed adult deaths, they appear to be demog demographic raphically ally signi significant ficant.. In Gombe Gombe,, data dat a rep report orted ed by Goo Goodal dalll (19 (1986) 86) ind indica icate te that for adult males in Kasekela and Kahama, the proportion of adult male mortality from intra intraspecifi specificc coalit coalitionary ionary aggression was 30–40%. Altho Although ugh fewer kills have been bee n see seen n els elsewh ewhere ere,, it see seems ms lik likely ely tha thatt this th is va vari riat atio ion n is pa part rtly ly a fu func nctio tion n of ob obse serv rvaa-
tion ti tion time me.. Th Thus us,, Fi Figu gure re 1 sh show owss th that at in relation to observation time, the number of observed and suspected kills appears similar in the four schweinfurthii study sites sites.. The idea that as observation years accumulate, more killing will be seen, is supported by the data on infanticides, which show a similar trend (Fig. 1). In summary although the evidence needs to be substantiated by continuing observation, current evidence is that in all four populations of the eastern subspecies, adults kill each other occasionally through coalitionary violence. Figure 1 suggests a rate of approximately 0.25 adults killed per year. On the other hand, there is no evidence of lethal intraspecific aggression toward either adul ad ults ts or in infa fant ntss fr from om ei eith ther er of th the e st stud udie iess of the wes wester tern n sub subspe specie cies, s, i.e i.e., ., fro from m Taı¨ or Bossou. Because the Bossou community has no nei neighb ghbors ors and few males, males, low rates of aggression are not surprising. However, as demonstrated by Figure 1, lethal coalitionary agg aggres ressio sion n wou would ld be exp expect ected ed to hav have e been seen at Taı¨: ¨: the study is now 21 years old and should therefore have produced evidence of about five killings each of adults
Fig. 1. Chim Chimpan panzee zee intraspe intraspecific cific killings by study site. killings site. S, P. t. schwein furthii; v furthii; v,, P P.. t. veru verus. s. Se See e Tab able less 3 an and d4 for data and sources.
Wrangham]
EVOLUTION OF COALITIONARY KILLING
and infants if this population conformed to the schweinfurthii pat patter tern. n. Alt Althou hough gh the sample sam ple sizes are sma small, ll, the fac factt tha thatt Taı¨ chimpanzees show all components of lethal raidin rai ding g but no coa coalit lition ionary ary kil kills ls sug sugges gestt tha thatt the natur nature e of aggr aggressiv essive e relati relationsh onships ips differs between Taı¨ and the eas easter tern n pop popula ula-tions (below). Finally, border avoid avoidance ance is exp expect ected ed if indivi ind ividua duals ls are awa aware re tha thatt the bor border der is physically dangerous. Low frequency of use of bor border der areas has bee been n doc docume umente nted d by Boes Bo esch ch an and d Bo Boes esch ch (1 (199 999) 9),, wh who o fo foun und d th that at in Taı¨ 75 75% % of ti time me wa wass sp spen entt in th the e ce cent ntra rall 35 35% % of th the e ra rang nge. e. Wh When en Go Gomb mbe e or Ki Kiba bale le (Kanyawara (Kany awara)) chimpa chimpanzees nzees do visit border areas, they tend to do so in parties that are relative rela tively ly larg large e (Go (Gombe) mbe) and cont contain ain a hig higher her pr op ort io n of m ale s t ha n nor mal (Kanyawara (Kany awara)) (Baue (Bauerr, 1980; Chapm Chapman an and Wrangham, 1993). Finally, high prey densities in border areas have been reported for the main prey species of chimpanzees, red colobu col obuss (Colob Colobus us badius badius), ), bo both th in Go Gomb mbe e (Stanf (St anford ord,, 199 1998b) 8b) and Kib Kibale ale [Ng [Ngogo ogo (D. Watts, personal communication)]. Similarly among wolves, prey densities are higher in border areas between territories, a result of avoidance of those areas by wolves (Mech et al.,, 199 al. 1998). 8). Bor Border der avo avoida idance nce by ter territ ritory ory holders has not been reported in other nonhuman hum an spe specie cies, s, but is pre presum sumabl ably y wid wideespread in humans. In sum summar mary y, the there re are five stu study dy sit sites es (Gombe, (Gomb e, Mahal Mahale, e, Taı¨, ¨, Kibale, Budongo) in which wh ich chi chimpa mpanze nzees es hav have e nei neigh ghbor borss and where whe re inte intercom rcommuni munity ty inte interact ractions ions hav have e been at least partly described or observed (Tabl (T able e 2).At the thr three ee bes best-d t-docu ocumen mented ted sit sites es (Gombe, Mahale, Taı¨) ¨) patterns of territorial interaction appear similar: all of them show territ ter ritori orial al def defens ense e and bor border der pat patrol rolss by adult adu lt mal males, es, wit with h vio violen lentt coa coalit lition ionary ary attacks on neighbors. All these patterns are similar to the data on wolves, which is the only other nonhuman mammal with fissionfusion fus ion gro groupi uping ng and gro group up terr territo itorie riess in which whi ch int interg ergrou roup p int intera eracti ctions ons hav have e bee been n well described. These points suggest that all the major ele elemen ments ts of let lethal hal raiding raiding are routin rou tinely ely pre presen sentt in pop popula ulatio tions ns of chi chimpa mpannzees.
11
The known kills at Gombe occurred between twe en 1973 and 1977, 1977, dur during ing a per period iod of intense hostility between two communities that had recently split from a single community, and which were each dominated by two alpha-males with mutually hostile relations (Goodall, 1986). Border patrols by Kasekela males were directed mostly toward the Kahama community during this period period,, which ended end ed wit with h the extinctio extinction n of the Kahama Kahama community (Goodall, 1986). Thus, unusual demographic and social conditions applied to eli elicit cit thi thiss par particu ticular lar bou boutt of let lethal hal rai raidin ding. g. Therefore, nonlethal raiding is a routine compon com ponent ent of the chi chimpa mpanze nzee e beh behavi aviora orall repertoire. Coalitionary killing is less common. However it has been recorded in four out of five sites. This raises the question of why chimpanzees have an appetite for agonistic nis tic inte interact ractionswith ionswith memb members ers of nei neighb ghbororing communities, and why they sometimes kill opponents.
THE IMBALANCE-OF-POWER HYPOTHESIS Explaining chimpanzee violence Many reasons have been advanced to account for chimpanzee lethal raiding, including: male–male bonds, hostility toward outsiders, cooperative group living, cooperative hunting skills, power imbalances when partiess fro tie from m nei neighb ghbori oring ng com commun munitie itiess mee meet, t, large lar ge and ove overla rlappi pping ng hom home e ran ranges ges,, hig high h cognitive cogni tive abilit ability y, and innat innate e killi killing ng potent potential ial (revie (re viewed wed by van der Den Dennen nen,, 199 1995). 5). The only attempt at a cost-benefit analysis that explains expla ins the speci species es distr distributio ibution n of lethal raiding, however, is the imbalance-of-power hypothesis. This hypothesis was implied by Goodall (1986), then elaborated by Manson and Wrangham (1991), Wrangham and Peterson (1996), and Wrangham (1999b). The imbalance-of-power imbalanc e-of-power hypothes hypothesis is proposes that the function of unprovoked intercommunity aggression (i.e., deep incursions and coalitionary tiona ry attack attacks) s) is interc intercommun ommunity ity dominance. By wounding or killing members of the neighboring community, males from one commun com munity ity inc increas rease e the their ir rel relativ ative e domi domi-nance over the neighbors. According to the imbalance-o imbal ance-of-powe f-powerr hypoth hypothesis, esis, the proxi proxi-mate benefit is an increased probability of winnin win ning g int interc ercommu ommunit nity y dom dominan inance ce con-
12
YEARBOOK YEARBOO K OF PHYSICA PHYSICAL L ANTHROP ANTHROPOLOGY OLOGY
tests (no tests (nonle nletha thall bat battle tles); s); thi thiss ten tends ds to lea lead d to increased fitness of the killers through improved pro ved access access to res resour ources ces suc such h as foo food, d, females, or safety safety.. The imbal imbalance-o ance-of-powe f-powerr hypot hypothesis hesis contrastss with proposals that chimpa trast chimpanzees nzees are exceptionally capable of conducting attacks, or win par partic ticula ularly rly lar large ge rew reward ardss from int inter er-group competition.
[Vol. 42, 1999
when a pa when part rty y of th thre ree e or mo more re ma male less en enco coun un-ters a lone victim (M Wilson et al., personal communication) commun ication),, suppor supporting ting the observ observaational data from Gombe. These observations make sense because to date, there appear to be no rec records ords of any agg aggres ressor sorss rec receiv eiving ing serious wounds. In light of such data, several authors have proposed that it is the ability of a gang of three or more males to overwhelm a lone The significance of power imbalances victim, at low risk of injury to themselves, Both within and between primate groups, which at least partly explains why chimpancontests tend to be won by the larger of two zees are so ready to attack (Goodall, 1986; Manson on and Wrangh rangham, am, 1991; Wrangham Wrangham coalitions, though variables such as domi- Mans and Peters Pet erson, on, 1996; 199 6; Boesch Boe sch and Boesch Boe sch,, nance rank and geographic location are also important import ant (Chen (Cheney ey,, 1986; Chapa Chapais, is, 1995) 1995).. 1999). The logic is that a victim can be held Coalition size appears even more important down or otherwise disabled by two or more, for intera interactions ctions among chimpa chimpanzees nzees from while another aggressor can impose damage different dif ferent communities. communities. In the four longe longest st at will. This idea that the low cost of lethal studies of chimpanzees, the principal deter- aggression elicits lethal raiding is central to minant min ant of the nat nature ure of int interc ercommu ommunit nity y the imbalance-of-power hypothesis (Manson Wrangh angham, am, 1991 1991;; Wr Wrangh angham am and Pete Peterrinteractions is not the geographic location and Wr son, 1996). butt th bu the e re rela lati tive ve si size ze an and d co compo mposi siti tion on of partie par tiess whe when n the they y enc encoun ounter ter eac each h oth other er.. Origins of power imbalances This conclusion is based on direct observations tio ns at Gom Gombe, be, Kib Kibale ale,, Mah Mahale ale,, and Taı¨ All chimpanzee populations have fission(Bygott, 1979; Nishida, 1979; Goodall, 1986; fusio fusion n groupi grouping ng patter patterns, ns, with indivi individuals duals Boesch and Boesch, 1999; Wrangham et al., sometimes alone and sometimes in parties in preparation), as well as playback experi- (Fig. 2), and adult males more gregarious ments at Kibale (M Wilson et al., personal than mother motherss (W (Wrangh rangham, am, 1999a) 1999a).. Democommunication). For example, Boesch and gra graphi phic, c, soc social ial and eco ecolog logica icall var variab iables les infl influuBoesch Boe sch (19 (1999) 99) fou found nd tha thatt sma small ll par partie tiess of ence party size (Boesch, 1996). For example, males (1–3) mainly checked for the presence party size increases both with the number of of strangers by drumming and listening to females having sexual swellings, and with the response (67% of 18 occasions). Middle- the amount of fruit in the habitat (Nishida, sized siz ed par partie tiess (4– (4–6 6 mal males) es) tended to mak make e 1979; Goodall, 1986; Boesch, 1996; Wrangincurs inc ursion ionss int into o the nei neighb ghbori oring ng ter territ ritory ory ha ham, m, 19 1999a 99a). ). Pa Parti rties es ap appea pearr to be co connmore often (37% of 76 observations). Large strained by fruit availability as a result of partie par tiess (7– (7–9 9 mal males) es) ten tended ded to atta attack ck the scram scramble ble competi competition, tion, with larger parties strang str angers ers (63 (63% % of 30 obs observ ervati ations ons). ). Mor More e formed more when fruit is sufficiently abungenerally, at all sites, the probability that a dant to allow gregariousness (Chapman et party will advan advance, ce, exchange displays, displays, or al., 1995). retr re trea eatt ap appea pears rs to be we well ll pr pred edict icted ed by Neighboring communities can experience whet wh ethe herr it is la larg rger er th than an,, eq equa uall to to,, or sm smal alle lerr markedly different levels of fruit supply, a than the opposing party (Boesch and Boe- result of differences in fruit-tree density, or sch, sc h, 19 1999 99). ). Re Rela lati tive ve pa party rty si size ze is al also so a in fruiting success (Chapman et al., 1997). critic cri tical al var variab iable le amo among ng lio lions ns (Mc (McCom Comb b et al. al.,, This means that, on occasion, neighboring 1994; Grinnell et al., 1995). communities may contain parties of differThe evid evidence ence ther therefor efore e sug suggest gestss that chim chim-- en entt me mean an si size ze.. Th The e co comm mmun unit ity y in wh whic ich h panzee parties are bolder when they contain parties are able to be larger can then make relatively more males. In addition, playback lowlow-risk risk raid raidss to atta attack ck neig neighbor hbors. s. This soci sociooexperiments at Kibale support the hypoth- ecological connection has been observed in esiss tha esi thatt mal males es are more lik likely ely to att attack ack Ma Mahal hale, e, wh when en the M-g M-grou roup p com commun munity ity
Wrangham]
EVOLUTION OF COALITIONARY KILLING
13
attacked by a coalition at low risk to the aggressors. This means that populations (or seasons) seas ons) with fewer encou encounters nters betw between een soli soli-taries and large groups should have fewer violent interactions. Data on wolves at Denali (Alaska) offer a test of this prediction. The ratio of the number of parties containing three or more wolves to the number of solitary individuals was higher during winterr (5 te (5.2 .2)) th than an su summ mmer er (0 (0.1 .1). ). Th This is me mean anss th that at solitaries were much more likely to encounter a large party during winter than summer.. As exp mer expect ected, ed, win winter ter was als also o the sea season son when intraspecific kills were more likely (a sevenfold seve nfold increase in proba probability bility,, from 22 dated kills) (data calculated from Figs 5.4 Fig. 2. Party size size distribution distribution among among chimpanzees chimpanzees and 5.7, Mech et al., 1998). and bonob bonobos. os. Data sources: eastern chimp chimpanzee anzee P. t. schweinfurthii: Kibale Kibale,, Kanya Kanyawara wara commu community nity 1994– If a similar effect applies to chimpanzee verus: us: Taı¨ (Boes 1996.. We 1996 Wester stern n chimpa chimpanzee nzee P. t. ver (Boesch, ch, populations, and if the fact that Taı¨ has had 1996); Mt. Assirir (Tutin et al., 1983). Bonobo P. paniscus; Wamba (Kuroda, 1979); Lomako (White, 1988). All a low kill rate is meaningful (rather than population popu lationss show substantial substantial varia variation tion in party size stochastic), Taı¨ should have larger, less fisover time and between communities. Comparable data sioned parties than at Gombe, Mahale, or for Gombe and Mahale were not found. Kibale. Kiba le. Preli Preliminar minary y data sug suggest gest this predi predicction is qualitatively correct, because Boesch (1996) 96) fou found nd tha thatt amo among ng six chi chimpa mpanze nzee e would make seasonal forays into the terri- (19 tory of the K-group community, supplanting populations, the mean party size was highestt at Taı¨ (8 (8.3 .3,, co comp mpar ared ed to a me mean an of K-group K-gro up partie partiess and somet sometimes imes attacking es them (Nishida, 1979). 5.2 0.8 at the other five sites, including Why are chimpanzees (compared to other Gombe, Mahale, and Kibale). The percentspecies) particularly vulnerable to the pres- ag age e of lo lone ne in indi divi vidu dual alss wa wass al also so lo lowe west st in Taı¨ sures of scramble competition that lead to a (4% (4%,, com compar pared ed to 14% 3% for the three fission fiss ion-fu -fusio sion, n, rat rather her tha than n a sta stable ble-tro -troop op populations with data, Boesch, 1996, Table system of grouping? Wrangham et al. (1996) 8.2). These data thus indicate consistently argued that the important characteristic of lar larger ger par partie tiess in Taı¨ tha than n els elsewh ewhere ere,, com compat pat-chimpa chi mpanze nzees es is tha thatt eve even n whe when n fru fruits its are ible with evidence that Taı¨ is a relatively scarce sca rce,, ind indivi ividua duals ls con contin tinue ue to sea search rch for pro produc ductiv tive e hab habita itatt (Bo (Boesc esch h and Boe Boesch sch,, them. Cons Consiste istent nt with this hypothesi hypothesis, s, chim- 1999). panzees panze es (unli (unlike ke gorill gorillas) as) are restri restricted cted to Further data will test whether party size areas are as tha thatt contai contain n yea yearr rou round nd fru fruits its,, and and the frequency of high-intensity aggresspend spe nd signifi significan cantly tly gre greater ater prop proporti ortion on of sion do indeed co-vary among sites, and how their feeding time eating ripe fruits than do often chimpanzee populations tend to have sympatric frugivorous monkeys (Wrangham small parties and high rates of aggression. et al., 1998). This strategy of constant fruit Two points suggest that the high frequency search is presumably forced on chimpanzees of int intens ense e agg aggres ressio sion n see seen n at Gom Gombe be and by spec speciesies-spec specific ific dige digestiv stive e adap adaptatio tations, ns, suggested by the other eastern chimpanzee such as the rate of food passage through the populations may be unusual for the species gut and the abi abilit lity y to fer fermen mentt lon long-c g-chai hain n as a whole. First, skeletal trauma indicative carbohydrates (Milton, 1987). Whatever its of intraspecific aggression has been found at origins, it exposes them to relatively intense hi high gher er ra rate tess in a sa samp mple le of ch chim impa panz nzee ee scramble competition. crania cra nia fro from m Gom Gombe be tha than n fro from m els elsewh ewhere ere The imbalance-of-power hypothesis states (Jurmain, 1997). Second, the four P. t. schthat violence is facilitated by vulnerability, weinfurthii stu study dy sit sites es (Go (Gombe mbe,, Mah Mahale ale,, because lone individuals can be vigorously Kibale, and Budongo) are all located at the
14
YEARBOOK YEARBOO K OF PHYSICA PHYSICAL L ANTHROP ANTHROPOLOGY OLOGY
extreme extrem e eas eastt of the spe specie ciess geo geogra graphi phical cal range, ran ge, whe where re dry sea season sonss are rel relati ativel vely y lon long. g. These eastern populations may therefore be living under relatively harsh conditions of food availability compared to more western sites. In summary, chimpanzees are vulnerable to partic particularly ularly intense scramb scramble le compet competiition, apparently because of their digestive adaptations to ripe fruit. This competition forc fo rces es th them em to tr trav avel el al alon one e or in sm smal alll parties when fruits are scarce. Patchy fruit distribution can mean that one community has abundant supplies, while its neighbors have few. Demographic differences between communities (i.e., differences in the number of adult males) may also mean that parties in on one e co comm mmun unity ity ca can n be do domi mina nant nt ov over er those in the neighboring territory. Such factorss can acc tor accoun ountt for dif differ ferenc ences es bet betwee ween n populations or communities in the number of males in parties, and hence for differences in their vulnerability to attack by coalitions of neighbors.
[Vol. 42, 1999
occurs in chimpanzees. Others are (1) that the benefits of raiding rise so steeply with increased party size that it pays individuals to participate for selfish reasons or (2) that chimpanzees have evolved exceptional cooperative abilities in contexts other than lethall rai tha raidin ding g (C.van Sch Schaik aik,, per person sonal al com commumunicat ni cation ion). ). Th This is rem remain ainss an imp import ortan antt prob pr oble lem, m, as it do does es fo forr mu much ch of hu huma man n behavior (Boehm, 1999). On th the e ot othe herr ha hand nd,, th the e oc occu curre rrenc nce e of territoriali territ oriality ty among chimpa chimpanzee nzee commun communiities, and of occasional imbalances of power betwee bet ween n par partie tiess fro from m nei neighb ghbori oring ng com commun muniities, tie s, are eas easily ily exp explai lained ned.. Fir First, st, cur curren rentt theory suggests that home ranges are economically defensible if individuals can easily cro cross ss the their ir hom home e ran range ge in a day day.. Chi Chimpa mpannzees ze es ca can n do so (L (Low owen en an and d Du Dunb nbar ar,, 19 1994 94;; se see e also van Schaik, 1996). Second, a system of fission fiss ion-fu -fusio sion n gro groupi uping ng can exp explai lain n why lon lone e individuals indiv iduals occas occasionall ionally y encou encounter nter larger coalitions, and are therefore vulnerable to attack. But neither long day ranges nor fissionGroup territoriality and the benefits fusion grouping can explain lethal raiding, of lethal raiding in which individuals seek opportunities to attack ack (as opp oppose osed d to res respon pondin ding g to inv invaaUnderstanding the selective advantage of att aggres agg ressio sion n is mor more e com compli plicat cated ed for int interc ercomom- sion, escalated escalated conte contest st for resou resources, rces, etc.). munity munit y than interi interindivi ndividual dual relati relationshi onships, ps, As Goodall noted, for example, there have because any fitness benefits gained by a rise been three major invasions at Gombe and in inter intercommun community ity domina dominance nce are share shared d Mahale. ‘‘Kasekela males took over Kahama among indiv individuals iduals within the commun community ity.. ra rang nge, e, Ka Kala land nde e ma male less pu push shed ed de deep ep in into to This might be expected to favor free-riders Kasa Kasakela kela range, and M-gro M-group up moved into (individuals who would benefit from lethal K-group range. During all these invasions raiding without taking part), which would adult males (and some females) were killed lead to a suboptimal level of collective action or disappeared. Even if it is argued that the (van (va n Sch Schaik aik,, 199 1996; 6; Nun Nunn, n, 199 1999). 9). In fac fact, t, Kas Kaseke ekela la mal males es wer were e mer merely ely try trying ing to rehowever, there is no evidence of defection claim an area to which they previously had among raiding chimpanzees (Goodall, 1986; free access, the assertion does not explain Wilson et al., personal communication) [(or, the nor northw thward ard thr thrus ustt of the Kal Kaland ande e com commumuforr th fo that at ma matt tter er,, am amon ong g li lion onss in si simi mila larr in inte terr- ni nity ty or th the e ta take keov over er by th the e MM-gr grou oup p at community contexts (Grinnell et al., 1985)]. Mahale’’ (Goodall, 1986, p. 528). How lethal raiding escapes the free-rider Two kinds of hypot hypothesis hesis have been preproble pro blem m is not und unders erstoo tood. d. One pos possib sibili ility ty is sented to account for such incursions, proxithatt int tha interc ercomm ommuni unity ty con conflic flictt has bee been n so mat mate e com compet petiti ition on and dom domina inance nce dri drive. ve. intense that selection has occurred at the First, aggression may be proximately elicbetween-gr betwe en-group oup level (Boehm, 1999). How- ite ited d by res resour ource ce com compet petiti ition, on, such as for ever, this is unlikely because it would re- mat mates, es, food, or lan land d (Ma (Manso nson n and Wran ranggquire very frequent group extinctions with ham, 1991). This hypothesis is strongly supfew sur surviv vivors ors.. Ano Anothe therr is tha thatt fre free-r e-ride iders rs are por ported ted by som some e obs observ ervati ations ons,, suc such h as the policed by others in the community (Boehm, inc incurs ursion ionss by Mah Mahale ale’s ’s M-g M-grou roup p int into o the 1999), 199 9), but no evi eviden dence ce sug sugges gests ts tha thatt thi thiss K-group range. These occurred in a seasonal
Wrangham]
EVOLUTION OF COALITIONARY KILLING
15
rhythm coinci rhythm coincident dent with fruit shortages shortages in we weak aknes nesss of a dom domina inant nt du due e to ag aging ing,, wou woundndM-grou M-g roup’ p’ss ran range ge and abu abunda ndant nt fru fruits its in ing ing,, los losss of all allies ies or los losss of con confide fidence nce inK-group’s range (Nishida, 1979). Again, in creases the rate of attack by a subordinate certain certai n circum circumstanc stances es raidin raiding g might help (Bygott, 1979; de Waal, 1982; Goodall, 1986; males mal es to rec recrui ruitt you young ng fem female ales: s: thi thiss pos possib sibilil- Nishida, 1994). Selection is expected to faity is suggested by evidence at Gombe that vor the effort to rise in dominance because severe sev ere att attack ackss on the mot mother herss of nul nullip liparo arous us dom domina inant nt ind indivi ividua duals ls (or gro groups ups)) ten tend d to females fem ales in nei neighb ghborin oring g comm communi unities ties are have high fitness, and accordingly, individusometi som etimes mes fol follow lowed ed by the you young ng fema females les als opportunisticall opportunistically y take advantage of any joining the aggressor’s community (Goodall, perception of changes in power asymmetry 1986). (Chapais, 1995). Alternatively, aggression may be elicited Sometimes, Sometim es, admitte admittedly dly,, the expect expectation ation mere me rely ly by th the e op oppo portu rtuni nity ty to re redu duce ce th the e that higher dominance leads to higher fitcoaliti coa litiona onary ry pow power er of the nei neighb ghbors ors (Ma (Manso nson n ness is not met. Thus, in around half of the and Wr Wrangh angham, am, 1991 1991;; Wr Wrangh angham am and Pete Peterr- studi studies es betwee between n domina dominance nce and reprod reproducucson,, 199 son 1996). 6). Acc Accord ording ing to thi thiss ‘‘d ‘‘domi ominan nance ce tive success within primate groups groups,, there drive’’ hypothesis, no resources need be in was no relationship. However, in the other short supply at the time of the raid. Instead, hal half, f, dom domina inants nts had hig higher her fitn fitness ess tha than n unprov unp rovoke oked d agg aggres ressio sion n is fav favore ored d by the sub subordi ordinat nates es (Ha (Harco rcourt urt,, 198 1987; 7; de Rui Ruiter ter and opportunity to attack ‘‘economically,’’ ‘‘economically,’’ that is, va van n Hoo Hooff ff,, 199 1993; 3; Ell Ellis, is, 199 1995). 5). Thi Thiss mea means ns at low personal risk. If raiding leads to the that even though increased dominance does wounding or death of a neighboring male, not always lead to higher fitness, it pays on the neigh neighborin boring g commun community’ ity’ss competi competitive tive average. ability abi lity is sub substa stanti ntially ally redu reduced. ced. For exTheref The refore ore,, acc accord ording ing to the dom domina inance nce-ample, if the neighboring community has 10 drive hypothesis, a necessary and sufficient males, its fighting power is reduced by 10%. condition for intercommunity aggression is This reduction lasts for a considerable time, a perception that an opponent is sufficiently becau be cause se the sy syste stem m of mal male e ph philo ilopat patry ry mea means ns vulnerable to warrant the aggressor(s) atthat a dead male can be replaced only via tacking at low risk to themselves. births bir ths wit within hin the com commun munity ity,, whi which ch is a slo slow w The dominance drive hypothesis appears process. The aggressors’ probability of win- useful for explaining why carnivores share ning ni ng fut futur ure e int interc ercomm ommun unity ity con contes tests ts letha lethall coalit coalitionary ionary violence with chimpa chimpann(battles, not raids) will therefore be signifi- zee zees. s. As exp expect ected ed by bot both h the proxim proximate ate cantly can tly inc increa reased sed by kil killin ling g a nei neighb ghbori oring ng compet competition ition and domina dominance nce drive hypothmale. mal e. The inc increa rease se in rel relati ative ve figh fightin ting g pow power er eses, fission-fusion grouping and intergroup can be expected to enable a community to hos hostil tility ity occ occur ur in the these se spe specie ciess (T (Tabl able e 6). The enlarg enl arge e its ter territ ritory ory,, as sug sugges gested ted by pre prelim limii- proxi proximate mate competi competition tion hypoth hypothesis esis also prenary evidence of a correlation between the dic dicts, ts, how howeve everr, tha thatt the typ type e of foo food d sup suppli plies, es, numb nu mber er of ma male less an and d ter terri rito tory ry si size ze at Go Gomb mbe e mati mating ng sys system tem and/ and/or or coali coalitiona tionary ry bond (Stanf (St anford ord,, 199 1998b) 8b).. Ov Over er the lon long g ter term, m, the therere- should be similar in allowing benefits to be fore, for e, if fitn fitness ess is cor correl relate ated d wit with h ter territ ritory ory gai gained ned from rai raidin ding g or kil killin ling g nei neighb ghbors ors.. size, siz e, suc succes cessfu sfull rai raidin ding g is exp expect ected ed to in- However, the four species of carnivores in crease the raiders’ fitness. which lethal coalitionary violence has been This ‘‘be ‘‘betwee tween-co n-commun mmunity ity domin dominance ance recorded show various combinations of matdrive’’ hypothesis for explaining aggression ing sys systems tems and coal coalitio itionary nary bond bonds, s, all dif differferbetween groups is therefore similar to the ent from those found in chimpanzees (Table ‘‘within-com ‘‘wit hin-community munity dominan dominance ce drive’ drive’’’ hy- 6). For exa exampl mple, e, int interg ergrou roup p tra transf nsfer er is in pothesis, pothe sis, which contri contributes butes to explai explaining ning some species primarily by males, in others patterns of aggression between individuals primarily by females. Therefore, lethal viowithin wit hin dom domina inance nce hie hierar rarchi chies es (Po (Popp pp and lence cannot be uniformly explained as reDeVore, 1979; Chapais, 1995). Within chim- sulting from competition over females. Bepanzee communities, for example, male ag- cause the carnivore species in Table 6 vary gress gre ssion ion occ occurs urs pre predic dictab tably ly ove overr sta status tus;; in th the e ty type pe of be bene nefit fitss to be ga gain ined ed by
16
[Vol. 42, 1999
YEARBOOK YEARBOO K OF PHYSICA PHYSICAL L ANTHROP ANTHROPOLOGY OLOGY
TABLE 6. Intergroup aggression aggression in fission-fusion species with group territoriality 1 Wolf
L io n
Spotted hyena
Chimpanzee
Human
Battles Kill adults Lethal raid Food supplies
Y Y Y Dispersed
Y Y Y Variable
Y Y Y Clumped
Y Y Y? Clumped
Y Y Y? Clumped
Coalitionary bonds among
Males
Males
Pair He Help lpe ers
Fema Fe male les; s; Ma Male less
Fema Fe male less
Cheetah ? Y ? Dispersed/ clumped Male Ma less
1
‘‘Kill adults’’ is shown separately from ‘‘Lethal raid’’ because, in hyenas and lions, it is not clear if killing of neighbors occurs with lethal raids, or merely when invaders are discovered and killed by residents. ‘‘Dispersed’’ food supplies imply that individual foodpatches are not defensible, whereas ‘‘clumped’’ foods can be individually defended (e.g., carcasses). Coalitionary bonds are bonds in which adults support each other in aggression against others. Sources for mating system and coalitionary bonds are Kruuk (1972) (spotted hyenas), Grinnell et al. (1995), and McComb et al. (1994) for lions, and Mech et al. (1998) for wolves. For other sources, see text.
intergroup dominance, the dominance drive hypoth hyp othesi esiss exp explai lains ns sim simila ilarit rities ies in the their ir ten ten-denc de ncy y to us use e le leth thal al vi viol olen ence ce mo more re ea easi sily ly th than an the proximate competition hypothesis. The pro proxim ximate ate com compet petiti ition on and dom domiinance drive hypotheses are closely related, because in both cases, the ultimate benefits of dom domina inance nce are inc increa reased sed suc succes cesss in resource competition. The proximate competition hypothesis is favored if raiding is elicited by the presence of stealable resources, or if ben benefit efitss acc accrue rue imm immedi ediate ately ly aft after er a raid. On the other hand, unprovoked deep incurs inc ursion ionss and atta attacks cks on mal males es wit withou houtt any obvious reward are better explained by the dominance drive hypothesis. To differentiate tia te the these se hyp hypoth othese esess mor more e cle clearl arly y, dat data a are needed on the proximate stimuli that elicit aggression. Since current information suggests that chimpanzee raids are often initiated ate d wit withou houtt the rai raider derss per percei ceivin ving g mat mates es or food sources, the dominance drive hypothesis appears relevant to explaining the timing and direction of lethal raiding.
Sex differences in territoriality and aggressiveness Among chimpanzees, males have to date been the only observed killers and aggressors in intergroup interactions, and males are ar e al also so mo more re li like kely ly th than an fem femal ales es to be victims (Table 4). Among humans, warriors are also overwhelming male (Adams, 1983). This contrasts with spotted hyenas, where female fem aless are mor more e agg aggres ressiv sive e tha than n mal males es (Kruuk, 1972; Frank, 1986; East and Hofer, 1991; 199 1; Hen Hensch schel el and Ski Skinne nnerr, 199 1991); 1); and with wolves, where both sexes are killed at
high rates, and there is no evidence of a sex differe dif ference nce in agg aggres ressiv sivene eness ss (Me (Mech ch et al. al.,, 1998). Why, therefore, are males the principal perpetrators of aggression in chimpanzees and humans? Traditio Tr aditional nal explan explanations ations are that males are more exp expend endabl able, e, or tha thatt mal males es hav have e more mo re to ga gain in si simpl mply y be becau cause se th they ey ha have ve hig highe herr var varia ianc nce e in fit fitne ness ss th than an fem femal ales es do (r (ree viewed by van der Dennen, 1995). However, such suc h gen genera erall exp explan lanati ations ons do not acc accoun ountt for species variation in the intensity of female participation. Nor does the degree of sexual dimorp dim orphis hism m in bod body y siz size, e, bec becaus ause e amo among ng nonhuman primates, sexual dimorphism in body size is not correlated with female in volvement volvement in interg intergroup roup aggr aggressio ession n (Manson and Wrangham, 1991). Male Mal e bon bondin ding, g, whi which ch is esp especi eciall ally y pro pro-nounced among chimpanzees and humans, has often been proposed to be an important influen infl uence ce (re (revie viewed wed by van der Den Dennen nen,, 1995 19 95). ). Th This is id idea ea is su suppo pporte rted d by th the e fa fact ct th that at both in human humanss and nonhuman primates, primates, populations with more patrilocal residence (or male philopatry) have relatively greater tendency tende ncy for aggre aggressor ssorss to be male (Adams (Adams,, 1983; Manson and Wrangham, 1991). The ultimate origins of male bonding are still sti ll deb debate ated. d. In chi chimpa mpanze nzees, es, mal males es are more gregarious than mothers, possibly because, as a result of carrying and waiting for infants, mothers travel slowly (Wrangham, 1999a). 1999 a). The rela relative tive mobi mobility lity and greg gregario arioususness of males means that they can use allies to dom domina inate te acc access ess to the their ir hom home e ran ranges ges,, excluding other males and thereby forcing male mal e phi philop lopatr atry y. As a res result ult,, a sy syst stem em evo evolve lvess
Wrangham]
EVOLUTION OF COALITIONARY KILLING
in whi which ch it pay payss to era eradica dicate te mal males es from neighbor neig hboring ing commu communiti nities es (W (Wrang rangham, ham, 1999a,b). Accor Accordin ding g to thi thiss log logic, ic, the theref refore ore,, mal male e bonding has two effects. First, it contributes to the development of male philopatry and the benefits of exclud excluding ing nongroup males, thereby raising the stakes in territorial encounters. Second, it makes available allies thatt ena tha enable ble a lar larger ger party to dom domina inate te a smaller party. None of this means, however, that male participation is a necessary condition for the evolut evo lution ion of let lethal hal vio violen lence ce in terr territo itoria riall interactions. As wolves, spotted hyenas and ants show, coalitionary territoriality can be carried out by both sexes, or even primarily by femal females. es. The compa comparativ rative e evid evidence ence,, ther thereefore, for e, su sugge ggests sts tha thatt let letha hall rai raidin ding g in chi chimpa mpannzees ze es an and d hu huma mans ns ca cann nnot ot be at attr trib ibut uted ed to th the e fact that bonds among adults are primarily among males.
17
the community range borders. Among bonobos,, by con bos contra trast, st, int intera eractio ctions ns are see seen n at range ran ge bor border derss wit withou houtt the com compon ponent entss of lethal raiding. Second,, relati Second relatively vely peacef peaceful ul inter intercommucommunity nit y rel relati ations onship hipss in bon bonobo oboss may be an inci in cide dent ntal al re resu sult lt of a re redu duct ctio ion n in th the e le leve vell of within wit hin-com -commun munity ity vio violen lence ce comp compare ared d to chimpa chi mpanze nzees. es. For exa exampl mple, e, few fewer er vio violen lentt behavi beh aviora orall int intera eracti ctions ons of all kin kinds ds occ occur ur among amo ng bon bonobo obos: s: no sex sexual ual coe coerci rcion, on, no inf infananticide, no brutal fights among males or females mal es com compet peting ing for dom domina inance nce,, no mal male e beatings of females (Wrangham and Peterson, so n, 19 1996 96;; Fu Furu ruic ichi hi et al al., ., 19 1998 98). ). Th The e sa same me is true tr ue in ca capti ptivi vity ty (d (de e Waa aall an and d La Lant ntin ing, g, 1997; Stanford, 1998a). Collateral evidence comes from a survey of cranial and postcranial skeletal trauma that concluded that in chimpanzees (and gorillas), but not in bonobos, there was evidence of serious risk from interindividual aggression (Jurmain, 1997) and from the generally less robust and less Bonobos: exceptions sexually dimorphic morphology of bonobos that support the rule? than chimpa chimpanzees nzees (Zihlman and Crame Cramerr, 1978; 8; She Shea, a, 198 1984). 4). Pen Pendin ding g fur furthe therr data data,, Intercommunity relations among bonobos 197 bonobo o males consistently consistently appear to be less sometimes involve fights between large par- bonob ties ti es,, bu butt as a sp spec ecie iess th they ey ap appe pear ar to be violent than chimpanzees. substantially less hostile to each other than Accordingly, a possible hypothesis is that are chimpa chimpanzees nzees (reviewed by Wr Wrangha angham m selection may have favored a generally less and Peterson, 1996; Stanford, 1998a). First, aggr aggressiv essive e male temper temperament ament in bonob bonobos, os, bonobos have never been seen to engage in as opp oppose osed d to a los losss of mot motiva ivatio tion n spe specifi cifical cally ly lethal let hal rai raidin ding, g, nor ind indeed eed in any compo- for let lethal hal rai raidin ding. g. The rea reason sonss why male nents nen ts of suc such h beh behavi avior or (T (Tabl able e 2). Sec Second ond,, bonobos are generally less aggressive than they can include markedly peaceful interac- mal male e chi chimpa mpanze nzees es cou could ld der derive ive fro from m the tions, tio ns, in whi which ch ind indivi ividua duals ls from nei neighb ghbori oring ng dom domina inance nce of mal males es by pow powerf erful ul fem female ale– – communities rest, travel, copulate, play, and female coalitions, or the greater importance groom together (Idani, 1991; White, 1996). of mot mother herss tha than n oth other er mal males es as all allies ies for In contrast, peaceful interactions involving individual males, or other social dynamics males of neighboring communities have not occurring within communities (Kano, 1992; been seen among chimpanzees. Parish, Par ish, 1996 1996;; Wrang rangham ham and Pete Peterson rson,, 1996 1996;; In view view of the anatomic anatomical al and phylogephyloge- de Waal and Lanting, 1997). The important netic similarities between chimpanzees and point is that the reduced tendency for lethal bonobos bonob os these dif differenc ferences es are remark remarkable. able. rai raidin ding g wou would ld be vie viewed wed as an inc incide identa ntall Three kinds of expla explanation nation suggest suggest them- consequence of a more general reduction in selves. male aggression. aggression. This hypothesis hypothesis is chalFirst, the facts may be misleading; more leng lenged, ed, howe however ver,, by spec species ies that have aggr aggresesprolon pro longed ged obs observ ervati ation on may rev reveal eal let lethal hal sive interg intergroup roup intera interactions ctions despite havin having g raiding in bonobos (Stanford, 1998a). How- peace peaceful ful relati relationsh onships ips within groups (e.g. (e.g.,, ever, this eventuality appears unlikely be- female lions). Furthermore, the tendency to cause cau se at Gom Gombe, be, Mah Mahale ale,, Taı¨, ¨, and Kibale Kibale en enga gage ge in le leth thal al ra raid idin ing g se seem emss un unli like kely ly to be compo co mpone nents nts of let lethal hal rai raidin ding g we were re se seen en selec selectivel tively y neutra neutral, l, consi considering dering its poten poten-shortly after individuals were observed near tially large effe effects cts on domina dominance nce relation-
18
YEARBOOK YEARBOO K OF PHYSICA PHYSICAL L ANTHROP ANTHROPOLOGY OLOGY
[Vol. 42, 1999
shipss be ship betw twee een n gr grou oups ps an and d th the e ti time me an and d however, remains to be proven. In Lomako, male e bon bonobo oboss spe spend nd inc increa reased sed time alo alone ne effor ef fortt spe spent nt on rai raids. ds. I the theref refore ore con conclu clude de mal duringperio ingperiods ds of fru fruit it sca scarci rcity ty (Wh (White ite,, 199 1998). 8). that the low tendency for lethal raiding in dur White ite (1 (199 998) 8) su sugg gges ests ts th that at th thes ese e ma males les bonobos is not merely an incidental conse- Wh quence of the benefits of within-community choose to travel alone in order to track the increasing incre asingly ly dispe dispersed rsed female parties, but peacefulness. according accor ding to the imbalance-o imbala nce-of-powe f-powerr hypoth hypoth-The Th e th thir ird d ki kind nd of ex expl plan anat atio ion n is th that at among bonobos, important components of esis, it should be dangerous for them to do However,, if sol solita itary ry tra travel vel is a soc social ial lethal raiding has been specifically selected so. However option opt ion rather than a strate str ategy gy dictat dic tated ed by against. again st. Under what circum circumstanc stances es could this occur? According According to the imbal imbalance-o ance-off- ecological pressures, it may be possible for power hypothesis, lethal raiding is favored bonobos to restrict their solitary periods to by a combin combination ation of coalit coalitionar ionary y territo territorialrial- times and locations when they can assess ity and imb imbala alance ncess of pow power er suf suffi ficie cient nt to that they are safe. More quantitative data allow one party to kill victims of the rival will be needed to test such ideas. One of the only other species of primate in community with impunity. Since lethal raiding is absent even though coalitionary terri- which lethal raiding might be expected from toriality occurs among bonobos, the imbal- the imbalance-of-power hypothesis are spimonkeys ys Ateles spp spp., ., bec becaus ause e spi spider der ance-of-p ance -of-power ower hypo hypothes thesis is pred predicts icts that der monke monkeys have a fission-fusion grouping syscompar com pared ed to chi chimpa mpanze nzees, es, bon bonobo oboss mus mustt like e chi chimpa mpanze nzees: es: ind indivi ividua duals ls som someeexperi exp erienc ence e gre greatl atly y red reduce uced d pow power er imb imbalal- tem lik times travel alone and sometimes in larger ances between rival parties. In general, variance in bonobo party size parties. Furthermore, males are more greis less than among chimpanzees, even when garious than mothers, and they form coalitionar nary y bon bonds ds wit with h eac each h oth other er aga agains instt nei neighghaverage party size is similar (Chapman et tio boring groups (Chapman et al., 1995). On al., al ., 19 1994 94). ). Th This is is as ex expe pect cted ed fr from om th the e imbalance-of-power imbalance-o f-power hypothesis hypothesis.. However However,, the other hand, spider monkeys are wholly the more critical question is how often bono- arboreal, which may reduce their ability to aggression. Among baboons bos are forced to travel alone, because indi- use coalitionary aggression. viduals can be killed (at minimal cost to the Papio anubis, for example, coalitions of lowaggressor aggr essors) s) only when they are found alone ranking males are effective in defeating a by a rival party. The two principal bonobo single higher-ranking male on the ground, study sites both indicate that lone travel is but not in trees (Smuts, 1986). Further data effec fectt of arb arbore oreali ality ty on pow power er asy asymme mme-rarely rar ely for forced ced by eco ecolog logica icall pre pressu ssures res.. In on the ef tries tri es betwee bet ween n coalit coa lition ions s and sol solita itarie ries s is Wam amba ba,, bo bono nobo boss us usua uall lly y ra rang nge e as on one e or tw two o large lar ge mix mixed ed par partie tiess ave averag raging ing mor more e tha than n ten therefore desirable. individuals (Hashimoto et al., 1998). In LoThe imbalance-of-power hypothesis mako, where feedin feeding g compet competition ition appears and the evolution of human warfare more mor e int intens ense e tha than n in Wamb amba, a, fem female aless remain in multi multi-female -female parties even during Peace is the normal human condition, in the sea season sonss whe when n fru fruit it is lea least st ava availa ilable ble the sense that most human groups, for most (White, (Whit e, 1998) 1998).. Such obser observatio vations ns sugg suggest est of the time, are not at war (Ferguson, 1989; thatt com tha compar pared ed to chi chimpa mpanze nzees, es, the int intens ensity ity Sponsel, 1996). Nevertheless, ethnographic of feedin feeding g compet competition ition is subs substantia tantially lly re- and his histor torica icall rec record ordss cle clearl arly y sho show w tha thatt warwarduced duc ed amo among ng bon bonobo oboss (Ch (Chapm apman an et al. al.,, 199 1994; 4; far fare e is a fre freque quent nt pra practi ctice ce (Ke (Keele eley y, 199 1996; 6; Wrangham et al., 1996; White, 1998). As a Manson and Wrangham, 1991; van der Denresult, extreme imbalances of power appear nen, 1995). Incre Increasin asingly gly,, archae archaeologi ological cal data unlikely to occur between parties meeting su sugg gges estt th that at vi viol olen ence ce ha hass of ofte ten n be been en a st stat atis is-from neighboring communities. tically important source of death, and it is Whethe Whe therr the app appare arent nt dif differ ferenc ence e in the sometimes possible to infer that the violence intens int ensity ity of fee feedin ding g com compet petiti ition on bet betwee ween n was coa coaliti litiona onary ry (Ke (Keele eley y, 199 1996; 6; Lar Larsen sen,, chimpa chi mpanze nzees es and bon bonobo oboss is suf suffi ficie cient nt to 1997). In small-scale societies, the commonsatisfy satis fy the imbala imbalance-of nce-of-powe -powerr hypot hypothesis hesis,, est form of war interaction is a raid (e.g.,
Wrangham]
EVOLUTION OF COALITIONARY KILLING
Turney-High, 1949; Keeley, 1996; Maschner and Ree Reedydy-Mas Maschn chner er,, 199 1998). 8). Eve Even n if humans are routinely peaceful, therefore, war needs to be explained. Because warfare includes a variety of types of interaction (such as raids and battles), it will require multiple explanations. The myr myriad iad hyp hypoth othese esess pro propos posed ed to explain why humans practice raids and other form fo rmss of wa warfa rfare re fa fall ll in into to th thre ree e ge gene nera rall classes clas ses.. Mala Maladapti daptive ve hyp hypothe otheses ses sug suggest gest thatt war tha warfar fare e res result ultss fro from m an ori origin ginall ally y ada adapptive aggressive tendency that, as a result of subseq sub sequen uentt dev develo elopme pments nts suc such h as the inv invenention of weapo weapons, ns, became disad disadvanta vantageous geous even eve n to the win winner nerss (e. (e.g., g., Lorenz, Lorenz, 196 1966). 6). They The y suf suffer fer fro from m the the theore oretica ticall pro proble blem m that if warring tendency is indeed disadvantage ta geou ous, s, it ha hass su such ch la larg rge e ef effe fect ctss th that at it should be selected against rapidly rapidly.. However, it seems unlikely that winners fare badly. Neutra Neu trall hyp hypoth othese esess are cur curren rently tly mor more e popula pop ularr, tho though ugh the they y als also o suf suffer fer fro from m the selective-d selec tive-disadv isadvantag antage e proble problem. m. They suggestt tha ges thatt war warfar fare e sho should uld be reg regard arded ed as deriving merely from a capacity, or potential, resulting from our cognitive creativity. Neutral hypotheses consider warfare to be elicited by environmental and social stimuli that have no evolutionary significance (e.g., Bock, Boc k, 198 1980; 0; Kee Keeley ley,, 199 1996; 6; Gou Gould, ld, 199 1996). 6). The They y are often based on the (erroneous) premise that behaviors that vary among populations cann ca nnot ot be ex expl plai aine ned d in te term rmss of na natu tura rall selection selec tion witho without ut assu assuming ming genet genetic ic diff differerences between populations (see Discussion). The imbalance-of-power hypothesis exemplifies a third kind of hypothesis that views warfare as adaptive and rooted in genetic predispositions. It suggests that raiding derives from the advantages of gaining intergroup gro up dom domina inance nce and an abi abilit lity y to ass assess ess power pow er imba imbalanc lances es in an env environ ironment ment of inte interrgroup hostility hostility and power imbala imbalances nces between parties from neigh neighborin boring g commun communiities.As tie s.As in ch chimp impanz anzees ees,, it rai raise sess the qu quest estion ion of why territori territories es are def defend ended ed by mal males es rather rat her tha than n fem female ales. s. By ana analog logy y wit with h the argument argu ment for chim chimpanz panzees, ees, male rathe ratherr than female territoriality derives from the high cost co st of tra trave vell ex exper perien ienced ced by mo mothe thers rs (Wran (W rangha gham, m, 199 1999a) 9a).. Thi Thiss cos costt red reduce ucess mot mothh-
19
ers’ ability both to defend a range and to form alliances. Both Bot h int interg ergrou roup p hos hostili tility ty and a fiss fission ion-fusion fus ion gro groupi uping ng sys system tem are uni univer versal sal in contemporary human populations, whether tribal or nation-state (Rodseth et al., 1991). Whether these features were characteristic of humans in prehistory is unknown. Howeverr, wit eve with h the exc except eption ion of bra brain in siz size, e, hum human an morpholo morp hology gy has chan changed ged rela relative tively ly littl little e during the las lastt 1.9 mil millio lion n yea years rs (W (Wolp olpof off, f, 199 1998), 8), suggesting that the essential ecology of human pre prehis histor tory y may bee been n rat rather her sta stable ble pri prior or to agriculture. The essence of theories about fission-fus fissi on-fusion ion group grouping ing in chimpa chimpanzees nzees is thatt fiss tha fission ion is a res respon ponse se to hig high h cos costs ts of scrambl scr amble e comp competit etition ion (Ch (Chapma apman n et al., 1995 1995); ); scramble competition is expected to be more intense in species that depend on rare, highquality foods (Janson and Goldsmith, 1995), and humans appear adapted to high-quality foods foo ds (Mi (Milto lton, n, 198 1987; 7; Leo Leonar nard d and Rob Robert ertson son,, 1997). Following this line of argument, fission-f sio n-fusi usion on gro groupi uping ng is exp expect ected ed to hav have e been characteristic of human evolutionary history. Based on the ubiquity of xenophobia and ingroup ing roup-out -outgrou group p bias in cont contempo emporary rary populations popul ations,, interg intergroup roup hosti hostility lity is norma normally lly assume ass umed d to hav have e bee been n rou routin tine e in hum human an prehistory. The likelihood of intergroup hostility in prehistory is supported also by its prevalen prev alence ce amon among g nonh nonhuman uman prim primates ates (Ch (Cheeney ne y, 19 1986 86). ). Th The e fo form rm of ho host stil ilit ity y ca can n be inferred as being territorial, because among primat pri mates, es, territori territories es ten tend d to be fou found nd in species with long day ranges in relation to home hom e ran ranges ges,, and are predicted predicted to occ occur ur more mo re ea easi sily ly wh wher ere e gr grou oups ps ar are e sp spli litt in into to multiple parties (Mitani and Rodman, 1979; Lowen Low en and Dun Dunbar bar,, 199 1994). 4). The lon long g day ranges of contemporary forager men [e.g., 9 km (Ba (Baile iley y, 199 1991)] 1)] and the pro probab babili ility ty of fission-fu fissi on-fusion sion forag foraging ing sugg suggest est that terri territoritoriality would have been possible where home ranges were not immense. The imbal imbalance ance-of-p -of-power ower hypo hypothes thesis is is thus compatible with conventional views of human prehistor prehistory y. It can in the theory ory be cha challlenged lenge d by evide evidence nce that recen recentt prehis prehistoric toric ancestors foraged in stable parties, or had ways of reducing power imbalances between rival parties, or had little intergroup hostil-
20
YEARBOOK YEARBOO K OF PHYSICA PHYSICAL L ANTHROP ANTHROPOLOGY OLOGY
ity, though such evidence would in practice be difficult to obtain. On cladi cladistic stic grou grounds, nds, vari various ous auth authors ors have hypothesized that lethal raiding in humans and chimpanzees shared a common origin arou ar ound nd 5– 5–6 6 my mya, a, an and d ha hass be been en pr pres esen entt continuously in the subsequent evolution of each specie speciess (W (Wrangh rangham, am, 1987; Ghig Ghiglieri, lieri, 1988; Wrangham and Peterson, 1996; Otterbein, bei n, 199 1997). 7). Thi Thiss hyp hypoth othesi esiss is cur curren rently tly untestable. A key issue for human ancestry is wheth whether er austr australopit alopithecine hecine ancestors of humans foraged in temporary parties (i.e., with fission-fusion) or in stable groups. Answers to such questions are needed before we can be confident whether lethal raiding in chimpanzees chimpanzees and humans represents a synapomorphy or a homoplasy.
[Vol. 42, 1999
the forest has been disturbed by encroachment me nt or lo logg ggin ing g (K (Kib ibal ale, e, Ng Ngog ogo, o, an and d Kanyawara). However, the relative lack of observations of violence in the early years of chimpanzee studie stu diess cit cited ed by Pow Power er (19 (1991) 91) is eas easily ily und undererstoo st ood d wi with thou outt re refe fere renc nce e to th the e ef effec fects ts of disturbance. Where there was no provisioning, early observations were relatively few comp co mpar ared ed to la later ter ye year ars, s, an and d th they ey we were re mostly mos tly of poor poorly ly hab habitua ituated ted ind individ ividuals uals,, more concer con cerned ned abo about ut hum humans ans tha than n eac each h oth other er [e.g., [e. g., con contra trast st ear early ly obs observ ervati ations ons by Ghi Ghi-glieri (1984), Kibale, Ngogo community, Isabirye-Ba biry e-Basuta suta (1989 (1989), ), Kiba Kibale, le, Kany Kanyawar awara a community, Reynolds and Reynolds (1965), Budongo, Budon go, and Sugiy Sugiyama ama (1973 (1973), ), Budon Budongo] go] with those based on wellwell-habitu habituated ated indi viduals by Watts (1999) (Kibale, Ngogo comCHALLENGES TO THE munity), munit y), Wrangham Wrangham et al. (1992) (Kiba (Kibale, le, IMBALANCE-OF-POWER IMBALANCE-OF-PO WER HYPOTHESIS Kanyawa Kan yawara ra comm communit unity), y), and Newt Newton-F on-Fishe isherr Uncertainty in the chimpanzee data (1997) (Budongo). With increased observaThe Th e ev evid iden ence ce of le leth thal al ra raid idin ing g in ch chim impa pann- tion of habituated individuals, studies of the zees ze es co come mess fr from om fe few w ca case sess an and d a sm smal alll Kanyawara, Ngogo, and Budongo communinumber of populations, some of which have ties conform to the essential Gombe-Mahale experienced significant anthropogenic influ- model of dominance-motivated and strategiappears rs ence en ces. s. Th This is ha hass le led d to do doub ubts ts ab abou outt th the e cally violent males. Power (1991) appea nott to ha have ve ap appr prec ecia iate ted d th the e di diff ffic icul ulty ty of import imp ortanc ance e of let lethal hal rai raidin ding g as a spe specie ciess no traitt amon trai among g chim chimpanz panzees ees (Pow (Power er,, 1991 1991;; Suss Suss-- observing dominance behavior and violence among poorly habituated and little known man, 1997). For example, Power (1991) accepts that individuals. Provisi Prov isioni oning ng com complic plicates ates the iss issue ue bethe descriptions of chimpanzees in Gombe, cause e it confou confounds nds incre increased ased obser observabil vability ity Mahale, and Kibale as violent and status- caus with h a con concen centra trated ted foo food d res resour ource ce tha thatt is strivi str iving ng are acc accura urate, te, but reg regard ardss the beh behavav- wit iors as nonadaptive consequences of exces- liable to promote aggression. In the case of Gombe, be, Goo Goodal dall’ l’ss int introd roduct uction ion of ban banana ana sive ecological stress. A central concern for Gom Power (1991) is why there was a shift in the fe feed edin ing g in 19 1962 62 le led d to th the e ch chim impa panz nzee eess perception perce ption of chim chimpanze panzee e soci society ety from peace peace-- becoming habituated quickly, so individuals ful to violent, beginning in the 1970s based were watched at close quarters during the firstt dec decade ade.. Onc Once e the chi chimpa mpanze nzees es wer were e on obs observ ervati ations ons after the firs firstt 5 yea years rs of firs Goodall’ Good all’ss stu study dy (196 (1960–19 0–1965). 65). She argu argues es that habituated, they were observed almost entirely ely in thebanan thebanana-f a-feed eedingarea ingarea (ca (ca.. 50 50 the 1960s view of chimpanzees living in a tir approximately 1/5,000th of their terripeaceful society was the ‘‘correct’’ one, and m, approximately thatt sub tha subseq sequen uentt obs observ ervati ations ons of vio violen lence ce tory of 12 sq km or more), and not followed reflect refl ect a soc social ial env enviro ironme nment nt stress stressed ed by var varii- toward territorial boundaries. Much aggresous kinds of human-induced disturbance. In sion during intense banana-feeding years of support, she cites the fact that intense ag- 1965–1969 was clearly directed toward obgression was seen rarely in early chimpan- taining bananas (Wrangham, 1974). During 1969,bana 69,banana na fee feedin ding g wa wass red reduce uced, d, and few fewer er zee studies, especially at Gombe, Budongo, 19 aggres ressiv sive e inc incide idents nts occ occurr urred ed in the ban banana ana-and Kibale (Ngogo community). She notes agg that th at in st stud udie iess wh wher ere e vi viol olen ence ce ha hass be been en feeding area (Wrangham, 1974). Power (1991) argued that a particularly reported, either humans have provided food import ortant ant fea featur ture e of the ban banana ana-fe -feedi eding ng sys sys-for the chi chimpa mpanze nzees es (Go (Gombe mbe,, Mah Mahale ale), ), or imp
Wrangham]
EVOLUTION OF COALITIONARY KILLING
tem was that, after 1965, chimpanzees were frustrated by their lack of control. She characterized the observation period 1960–1965 as ‘‘nat ‘‘naturali uralistic’ stic’’’ (imply (implying ing undis undisturbe turbed d by humans) human s) because the banan bananas as were given freely, i.e., they were never withheld. From 1965 19 65 on onwa ward rd,, in at atte temp mpts ts to re redu duce ce th the e banana ban ana-ind -induce uced d agg aggres ressio sion n of the 196 1962– 2– 1965 era, bananas were made available in metal boxes equipped with doors that were controlled in various ways by observers. It was this system that Power argued caused chimpanzees to express their potential for aggression, because it frustrated them. In the absence of controlled experiments, no hypothesis hypothesis can be rejec rejected. ted. Two points relevant to the imbalance-of-power hypothesis can be made, however. First, even if the frustration-aggression hypothesis is correct in explaining why the Kasekela community attack att acked ed the Kah Kahama ama com commun munity ity,, it doe doess not explain why coalitional coalitional letha lethall aggre aggresssion was elicited relatively easily in these chimpa chi mpanze nzees, es, or why it occ occurs urs in mal males, es, whereas it has not been seen in any other species faced with similarly frustrating contexts.. [For example, baboons obtained batexts nanas regularly at the banana-feeding-area, thou th ough gh ob obse serv rvers ers tri tried ed to pre preve vent nt the them m (Wrangham, 1974). There has been no hint of any behavior resembling lethal raiding in these baboons, despite intense study in subsequent years]. Thus, as Power (1991) herself says, whether or not feeding frustration contributed to the social tensions at Gombe, there the re rem remain ainss a pro proble blem m to be exp explai lained ned.. Why are male chimpanzees easily prompted to adopt intense coalitionary violence as a solution to social problems? Second, the frustration frustration-aggress -aggression ion hypothesis has much against it. The idea is that ‘‘frus ‘‘f rustrat tration ion caus causes es a dis distinc tinctt beh behavi avioral oral change cha nge in the con condit dition ion of an org organi anism’ sm’’’ (Power, 1991, p 3). This implies that under natura nat urall con conditi ditions ons,, chi chimpan mpanzee zeess are not natura nat urally lly fru frustr strate ated, d, whi which ch is cle clearl arly y not true.. For exam true example, ple, inten intense se aggr aggressi essive ve compe compe-tition occurs regularly in all study sites over prized foods, such as meat (Goodall, 1986; Boes Bo esch ch an and d Bo Boes esch ch,, 19 1989 89). ). On th the e ot othe herr hand, whatever behavioral change occurred among the post-1965 Kasekela chimpanzees (compared to chimpanzees at other sites) is
21
not easily viewed as ‘‘distinct.’’ The series of atta at tack ckss th that at be bega gan n in Ja Janu nuar ary y 19 1974 74 oc oc-curred almost a decade after the start of the problematic problem atic banana banana-feeding -feeding-syst -system, em, and took place several kilometers from the feeding station. Power (1991) implies that this long lo ng de dela lay y ca can n be ac acco comm mmod odat ated ed by th the e hypothesis of a permanent behavioral/psychological cholo gical reorg reorganiza anization, tion, but clearl clearly y the behavior must also be viewed in the context of ongoing social tensions (see Goodall, 1986 for an account of the relationships between the alp alphaha-mal males es of the Kas Kaseke ekela la and Kahama ham a com commun muniti ities es tha thatt may hav have e hel helped ped precipitate the aggression). The inc incide idence nce of agg aggres ressio sion n in the ban banana ana-feeding-area was closely related to the number of bananas that chimpanzees obtained. Party Par ty siz size e inc increa reased sed in the fee feedin ding g are area a compared to the natural habitat, but within minute min utess of the chi chimpa mpanze nzees es lea leavin ving g the feedin fee ding g are area a the exp expect ected ed par party ty siz size e was restor res tored ed (W (Wran rangha gham, m, 199 1994). 4). The These se and sim simiilarr re la resu sult ltss sh show ow th that at th ther ere e we were re in inde deed ed short-t sho rt-term erm infl influen uences ces rela related ted to the ava availab ilabilility of bananas, but no long-term influences have hav e bee been n det detect ected ed (W (Wran rangha gham, m, 197 1974; 4; Goo Gooddall,, 198 all 1986). 6). The accumula accumulatio tion n of dat data a from otherr, nonp othe nonprovis rovisione ioned d sites conti continues nues to chal chal-lenge the view that chimpanzees are naturally averse to violence.
The claim that biology is irrelevant for human warfare Some critics reject evolutionary explanations of warfare out of hand, based on the misconception that the only behavioral patterns explicable explicable by biolog biology y are ‘‘ins ‘‘instincts tincts,’’ ,’’ i.e., i.e ., beh behavi aviors ors tha thatt are obl obliga igator tory y and and/or /or invari inv ariabl able. e. Acc Accord ording ing to thi thiss log logic, ic, sin since ce warfare warfa re is not ‘‘ins ‘‘instinctu tinctual,’ al,’’’ biolo biological gical adaptation tat ionss can cannot not exp explai lain n the pro propen pensit sity y for war (Keele (Keeley y, 1996; Sussman, 1997; Rega Regal, l, 1998). 199 8). Thi Thiss err error or see seems ms rem remark arkabl able, e, becaus ca use e be behav havior ioral al ec ecolo ologi gists sts hav have e lon long g stressed that psychological adaptations are expected to respond in a contingent way to appropri appr opriate ate cont contexts exts (e.g (e.g., ., Hrdy Hrdy,, 1990 1990;; Bark Ba rkow ow et al al., ., 19 1995 95;; Kr Kreb ebss an and d Da Davi vies es,, 1997 19 97). ). In th the e wo word rdss of Ot Otte terb rbei ein n (1 (199 997, 7, p 272), ‘‘Man is neither, by nature, peaceful nor warlike. Some conditions lead to war, others do not.’’
22
YEARBOOK YEARBOO K OF PHYSICA PHYSICAL L ANTHROP ANTHROPOLOGY OLOGY
[Vol. 42, 1999
The imbalance-of-power hypothesis is en- ar are e th the e pl plac aces es to se seek ek an answ swer erss ab abou outt po popu pula la-tirely compatible with the observations that tion variations. The assumption that evolutionary analymany people live wholly peaceful lives, that implie liess gen geneti eticc det determi erminis nism m see seems ms odd oddly ly some cultures have periods of peace lasting sis imp for sev severa erall gen genera eratio tions, ns, and tha thatt som some e chi chimm- old-fashioned in an era when we are beginpanzee populations have no lethal raiding ning to understand the nuances of psychofor long periods. Indeed, such variation is to neuroendocrinological adaptations of differbe expe expected cted.. The imba imbalanc lance-ofe-of-powe powerr hyp hypothoth- ent species and sexes. To Keegan (1993, p 3), esis conforms well to some theories of peace, war reaches into ‘‘the most secret places of human an hea heart, rt, pla places ces whe where re sel selff dis dissol solves ves such su ch as th the e id idea ea th that at no nonv nvio iole lenc nce e is an the hum adaptive response by societies to violence by rational purpose, where pride reigns, where emotio tion n is par paramo amount unt,, whe where re ins instin tinct ct is stronger neighbors (Dentan, 1992). Admit- emo tedly, males are expected by this hypothesis king.’’ If psychology can describe those secrett pla places ces,, it is the task of evo evolut lution ionary ary to take advantage of power over neighbors, cre anthropology to explain how they arose. especially when unfettered by social or cultural constraints. constraints. They are also expected to IMPLICATIONS OF THE probe pro be for wea weakne knesse ssess in per percei ceived ved opp oppooIMBALANCE-OF-POWER IMBALANCE-OF-PO WER HYPOTHESIS nents, and to be willing to fight in a wide Chimpanzee and human psychology varie va riety ty of cir circum cumsta stance ncess whe where re ele elevat vated ed sta sta-tus is pre predic dictab table le or per percei ceived ved opp oppone onents nts Different versions of the CVH pay varied will be wounded or destroyed at low cost. attention to such factors as cognitive ability, But the essence of the imbalance-of-power weapo weapons, ns, brain size, male-b male-bondin onding, g, territ territoriorihypothesis hypot hesis,, like other behavioral hypoth- ality, sexual dimorphism, and imbalances of esess for lar ese largege-bra braine ined d mam mammal mals, s, is tha thatt pow power er (va (van n der Dennen, Dennen, 199 1995). 5). They are expression of the behavior depends on con- unite united, d, howev however er,, in provi providing ding adaptive ratext. Whether or not an individual employs tio tional nales es for chi chimpa mpanze nzee e and hum human an vio vio- violence is expected to depend on the proxi- lence, and therefore in proposing that lethal mate mat e sti stimul muli, i, abo about ut whi which ch we sti still ll kno know w raiding has a substantial evolutionary hislittle lit tle.. Wha Whatt lea leads ds ind indivi ividua duals ls to cla classi ssify fy tory, possibly since our split from a common otherss as ‘‘opponents’’? other ‘‘opponents’’? How do socia sociall and ancestor with chimpanzees. ideolo ide ologic gical al pre pressu ssures res af affec fectt the eas ease e wit with h The implicati implication on is tha thatt the there re has been which whi ch men men,, or wom women, en, respond respond to inc incite ite-- selection for a male psyche that, in certain ments to violence? How is ‘‘dominance’’ per- circumstances, seeks opportunities to carry ceiv ce ived ed?? How ar are e ri risk skss pe perc rcei eive ved, d, fo forr in in-- out low-cost attacks on unsuspecting neighstan st ance ce,, wh when en Eg Ego o is em embe bedd dded ed wi with thin in a bors. The psych psychologi ological cal mecha mechanisms nisms that hierarchy of alliances? How do institutional would make such a complex function poswar relationships influence individual neu- si sibl ble e ha have ve no nott be been en st stud udie ied, d, bu butt a pa part rtia iall li list st roendocrinology, and vice versa? Such ques- might include: the experience of a victory tions tio ns are cri critica ticall for und unders erstan tandin ding g who thrill, an enjoyment of the chase, a tendency becomes violent, and when. for easy dehumanization dehumanization [or ‘‘dec ‘‘dechimpiz himpizaaIn the current context, the imbalance-of- tion, tion,’’’’ (Good (Goodall, all, 1986) 1986),, i.e., treati treating ng nongr nongroup oup power pow er hyp hypoth othesi esiss sug sugges gests ts tha thatt sel select ection ion members as equivalent to prey], and deindihas favor favored ed certai certain n emotio emotional nal predis predisposiposi- viduation (subordination of own goals to the tions in males that cause aggressive behav- gro group), up), read ready y coal coalitio ition n form formatio ation, n, and sop sophis his-iorr to be el io elic icit ited ed re rela lativ tivel ely y ea easi sily ly un unde derr ticate ticated d asse assessmen ssmentt of power differentials. differentials. certain circumstances. The challenge of de- Sex differences can be expected in at least fining fini ng the eli elicit citing ing cir circum cumsta stance ncess is the pro provv- some of these traits, unless developmental ince inc e of dis discip ciplin lines es tha thatt pro probe be loc local al var variaia- con constr strain aints ts int interf erfere ere.. Som Some e fea featur tures es of a tion ti ons, s, in incl clud udin ing g no nott on only ly bi biol olog ogic ical al lethal-raiding psychology are not easily preanthropology, but also social psychology, be- dic dicted ted,, suc such h as the mechanism mechanismss by whi which ch havioral havio ral ecolog ecology y, socia sociall anthr anthropolog opology y, cul- expected costs and benefits are assessed. tural ecology, or social ecology. Those disci A sharp sharp al alte tern rnat ativ ive e to th the e CV CVH H is th the e plines, plin es, toge together ther with biolo biologica gicall anth anthropol ropology ogy,, standard social science model (SSSM), that
Wrangham]
EVOLUTION OF COALITIONARY KILLING
human males have no inherent propensity to tak take e adv advant antage age of pow power er dif differ ferent ential ials. s. Instea Ins tead, d, acc accord ording ing to the SSS SSSM, M, hum humans ans merely mer ely have a cap capaci acity ty for vio violen lence, ce, and since ‘‘the range of possible cultural results is not explicable by natural selection’’ selection’’ (Bock, 1980; p. 76; cf. Gould, 1996), evolutionary history is claimed to be irrelevant. This line of thinking has several problems. It does not account accou nt for the species distribution distribution of coali coali-tional aggression. It treats biology and cultures as alternatives, rather than as mutually al ly in inter teract actin ing g infl influe uence nces. s. It doe doess no nott account for the predictability of human aggressive patterns, and it is easily subsumed under adaptive theories of violence, which can account both for the fact that individuals cho choose ose to man manipu ipulat late e oth others ers (wh (wheth ether er through ideology or other ways) and for the fact that they are so easily manipulated.
23
human gro human groups ups inc incorp orpora orate te mor more e lev levels els of sociall dynamics (Hinde, socia (Hinde, 1993). At each such level, leve l, ‘‘le ‘‘levelvel-spec specific ific prope properties rties’’’’ influ influence ence and are infl influen uenced ced by adj adjace acent nt lev levels els.. For example, amp le, gro group up pro proces cesses ses and ins instit tituti utiona onall influences modify the motivational ideals of individuals indiv iduals.. In modern nation-states, nation-states, the military-industrial-scientific complex tends to pre precip cipita itate te and mai mainta ntain in war war,, e.g e.g., ., by producing increasingly sophisticated weaponry. War as an institution can in theory be mainta mai ntaine ined d by the ine inertia rtia of sub subins institu titutio tions, ns, such su ch as th the e be beli lief ef th that at ca capi pita tali lism sm ne need edss militarism milita rism for its continued growth (Hinde, 1993 19 93). ). Dy Dyna nami mics cs li like ke th thes ese e me mean an th that at a propen pro pensit sity y for let lethal hal rai raidin ding g can cannot not be tra transnslated lat ed dir directl ectly y int into o an exp explan lanati ation on of the complexities of human warfare. Is a propensity for lethal raiding of the chimpanzee type at all relevant to human warfar war fare?Among e?Among peo people ple liv living ing in sma small ll pol politi iti-The complexity of war cally independent groups, lethal raiding appears ars st strik riking ingly ly si simil milar ar to the pat patter terns ns Even in the complex human world, some pe chimpanzees. In both cases, small of the pro proces cesses ses tha thatt reg regula ulate te agg aggres ressio sion n among chimpanzees. partie tiess of mal males es aim to mak make e und undete etected cted among large groups are analogous to those par that occur at the individual or face-to-face incursions into the ranges of neighbors, attack k uns unsusp uspect ecting ing vic victim tims, s, and ret retrea reatt wit withhlevel (Hinde, 1993). Both at the large group tac and the individual level, for example, per- out being drawn into a battle (Turney-High, sonal relations between leaders of opposing 194 1949; 9; Cha Chagno gnon, n, 199 1992; 2; Kee Keeley ley,, 199 1996). 6). Algroups can play an important role, with a though the psychological processes remain threat to the interests or values of the actor undescr undescribed, ibed, the imbalanc imbalance-of-power e-of-power hypothbeing capable of instigating aggression. Pla- esi esiss mig might ht sug sugges gestt tha thatt sel select ection ion has facatory cat ory sig signal nalss or act action ionss (e. (e.g., g., don donati ations ons)) are vor vored ed va vari riou ouss co comp mple lex x tra trait its, s, su such ch as a used to deter aggre aggression ssion.. Aggr Aggressiv essiveness eness ten tenden dency cy to cla classi ssify fy oth others ers as inin-gro group up or can be augmented by a greater asymmetry out-group, to regard members of out-groups of po powe werr, or re redu duce ced d by a pr prob obab abil ilit ity y of as po poten tenti tial al pr prey ey,, to be al aler ertt to (o (orr se sear arch ch fo for) r) punishment. Hinde (1993) found more than power asymmetries asymmetries betwe between en in-gro in-group up and 20 such analogies, analogies, of varyi varying ng sign significanc ificance e ou out-g t-gro roup up pa part rtie ies, s, an and d to be ru ruth thle less ss in and distin distinctness ctness,, linki linking ng the behav behavioral ioral in- attacking out-group parties when the perteract ter action ionss amo among ng ind indivi ividua duals ls and lar large ge ceiv ceived ed powe powerr asy asymmetr mmetry y is suf suffic ficientl iently y grea great. t. groups. This suggests that, in some ways, A list of traits such as these can in theory the log logic ic of agg aggres ressiv sive e int intera eracti ction on amo among ng describe an evolutionarily selected ‘‘propenindividuals can be applied to large groups. sity for lethal raiding.’’ Nevertheless, so many cultural and linSuch traits appear obviously relevant to guisti gui sticc novelt novelties ies com compli plicat cate e war warfar fare e tha thatt the some aspects of interg intergroup roup relations (e.g., connection to biology can appear tenuous at the planning and execution of military enbest. Among factors such as the number of gagements). Among humans, the complexity military and strategic options available, the of society means that individual propensiabilit abi lity y to dis discus cusss opt option ionss and man manipu ipulat late e ties sometimes have less direct impact on others, the adoption of cultural goals, and social outcomes than among chimpanzees. the unpredictable potential for shifting alli- Never Nevertheles theless, s, the imbala imbalance-of nce-of-powe -powerr hyances, a particularly important trait distin- pot pothes hesis is may exp explai lain n why cul cultur turall ally y der derive ived d guishing humans from chimpanzees is that inf inform ormati ation on is use used d in cer certai tain n way ways. s. For
24
YEARBOOK YEARBOO K OF PHYSICA PHYSICAL L ANTHROP ANTHROPOLOGY OLOGY
[Vol. 42, 1999
example, exampl e, as a res result ult of cul cultur tural al bel belief iefss or hu huma mans ns an and d ch chim impa panz nzee eess th than an in ot othe herr socia so ciall pre press ssure ure,, in indiv dividu iduals als can eit either her prim primates ates.. Furt Furtherm hermore, ore, the beha behavior viorss show shown n broaden or contract their concept of where by chi chimpa mpanze nzees es tow toward ard mam mammal malian ian pre prey y an in-group/out-group boundary falls, or of are partly similar to those they show toward how important important it is. Idealogu Idealogues es can per per-- conspecific victims, including quiet stalking suade their followers that sufficient power during a hunt, intense arousal during the asymme asy mmetry try exi exists sts to mak make e att attack ackss on an attac attack k phase (pilo-e (pilo-erectio rection, n, intimi intimidation dation disoutgroup worthwhile. Culture can thus ma- plays), and ambivalence toward the victim. nipulate the information an individual uses In contrast, the behaviors shown by specialto assess whether an attack is desirable. ized ize d car carniv nivore oress tow toward ard the their ir pre prey y are not The imbala imbalance-of nce-of-powe -powerr hypot hypothesis hesis can like those directed toward conspecifics. For be re reco conc ncil iled ed wi with th th the e po powe werr of cu cult ltur ure; e; example, social carnivores do not show signs therefore, if human males have a tendency of excitement when killing prey, and tend to to search for, and take advantage of, power use a killing bite (van der Dennen, 1995). asymme asy mmetrie triess suf suffi ficie cient nt to ena enable ble the them m to Such obse observat rvations ions sug sugges gested ted to Eibl Eibl-safely saf ely kil killl riv rivals, als, whi while le soc social ial pre pressu ssures res Eibesfeldt (1975) and Goodall et al. (1979) modify mod ify the con concep ceptt of ‘‘r ‘‘riva ival,’ l,’’’ ‘‘a ‘‘ally lly,’ ,’’’ and tha thatt amo among ng chi chimpa mpanze nzees es sim simila ilarr mot motiva iva-‘‘sufficient ‘‘suff icient power asymmetry.’’ asymmetry.’’ It accordingly tional factors may be involved in intraspesugges sug gests ts an exp explan lanati ation on for why hum human an cifi cificc kil killin ling g and hun hunting ting.. Eib Eibl-E l-Eibe ibesfel sfeldt dt males become dangerous when they obtain, (1975 (1975)) specifi specifically cally propos proposed ed that, ‘‘Motivaor bel believ ieve e the they y hav have, e, lar large ge pow power er adv advanan- tiona tionally lly,, hunti hunting ng behav behavior ior in chimpa chimpanzees nzees tages tag es ove overr oth others ers.. (Wh (Wheth ether er,, in nov novel el cir circum cum-- has pro probab bably ly bee been n der derive ived d fro from m int intras raspec pecific ific stances, they use such power adaptively is aggression’’ (translated and quoted by van an ope open n que questi stion. on.)) It als also o sug sugges gests ts the imp impor or-- de derr De Denn nnen en,, 19 1995 95,, p 19 192) 2).. Th The e es esse sent ntia iall lo logi gicc tance tan ce of sys systems tems tha thatt red reduce uce pow power er asy asymme mme-- is that if hunting had arisen independently, try,, such as interg try intergroup roup allian alliances ces throu through gh it should be expected to show more similaritrade, marriage or treaty. tiess to the pat tie patter terns ns dis displa played yed by soc social ial car carniniWhen Whe n lar large ge pow power er asy asymme mmetri tries es do not vores. Van Hooff (1990) agreed, suggesting occur, relationships between groups are of- that if selection favored the ability to hunt ten peac peaceful eful,, as exp expecte ected d from the imba imbalanc lancee- and kill cons conspeci pecifics fics,, the psy psychol chologic ogical al of-power of-pow er hypot hypothesis hesis (Knauft, 1991; Bueno mec mechan hanism ismss tha thatt evo evolve lved d wou would ld be eas easily ily de Me Mesq squi uita ta an and d La Lalm lman an,, 19 1992 92;; va van n de derr co-opted toward obtaining meat. Dennen, 1995). Even when there is a balNote that these ideas are opposite to the ance of power, however, lethal battles and killer ape hypothesis. The killer ape hypothwars can occur among humans, in contrast esi esiss sug sugges gested ted tha thatt int intras raspec pecific ific vio violen lence ce to the pattern among chimpanzees (Singer, evolved from hunting, whereas Eibl-Eibes1989, Boehm, 1992). Coalitionary Coalitionary aggre aggress- feldt (1975), Goodall (1986), and van Hooff sion occurring between opponents with bal- (1990) proposed that hunting evolved from anced anc ed pow power er req requir uires es oth other er kin kinds ds of exp explan lanaa- intra intraspecifi specificc viole violence. nce. As van der Denne Dennen n tion than the imba imbalanc lance-of e-of-pow -power er hyp hypothe othesis sis,, (19 (1995) 95) not notes, es, the rel relati ations onship hip bet betwee ween n int intrarasuch suc h as the cul cultur tural al exa exagge ggerat ration ion of mot motiva ivatt- speci specific fic killi killing ng and hunti hunting ng proba probably bly now ing forces or the development of self-decep- inv involv olves es mul multipl tiple e dir direct ection ions. s. For exa example mple,, in tive asses assessment sment strategies (Boehm, 1992; some human populations hunting may pro van der Dennen, 1995; Wrangham, 1999b). vide practice for warfare (Otterbein, 1997). Disentang Dise ntangling ling these relat relations ionships hips will there there-The relation between lethal raiding fore not be easy. and hunting Nevertheless the idea that violence begat Both lethal raiding and hunting are car- hunting is useful because it suggests a new ried out primarily by adult males acting in way to sol solve ve a puz puzzle zle abo about ut bon bonobo obos, s, nam namely ely,, coordinated groups: both involve otherwise that bonobos show no evidence of monkey unusual unus ual actions such as searching for large hunting. Thus, no monkey hunting or monprey,, stalk prey stalking, ing, chasi chasing, ng, seizi seizing, ng, wound wounding, ing, key eating has been recorded at the longand killing; killing; both are more ela elabor borate ated d in term bonobo sites (Wamba and Lomako), or
Wrangham]
EVOLUTION OF COALITIONARY KILLING
25
in the shorter-term studies of Yalosidi, Lake sug sugges gests ts tha thatt comp compare ared d to chi chimpan mpanzee zees, s, Tumba, or Lilungu, even though these stud- bonobos have a weaker motivation to hunt ies have provided sufficient data to record, monkeys. As a third possi possibility bility,, theref therefore, ore, Wrangfor exa exampl mple, e, ter termit mite-e e-eati ating ng in fou fourr of the ham m an and d Pe Pete ters rson on (1 (199 996) 6) no note ted d th that at th the e la lack ck sites sit es (Th (Thomp ompson son,, 199 1997). 7). Bon Bonobo oboss do eat mea meatt ha occasionally, however, which they obtain in- of monkey hunting among bonobos might be expl plai aine ned d as a co cons nseq eque uenc nce e of th thei eirr lo low w dividually divid ually by seizi seizing ng youn young g antelo antelope. pe. The ex intere erest st in int intras raspec pecific ific kil killin ling. g. For exlack of monkey hunting by bonobos is strik- int ample, e, if bono bonobos bos evo evolved lved from a chi chimpan mpanzeezeeing given that they not only prey on terres- ampl like li ke ance an cest stor or, , they th ey bega be gan n with wi th a tend te nden ency cy for trial mammals, but also complete with each lethal hal rai raidin ding g whi which ch was lost or inh inhibi ibited ted other to eat meat, and sometimes interact let acquired ed relati relatively vely stable parsocial soc ially ly wit with h mon monkey keys, s, in gro groomi oming ng and when they acquir ties.Accordingly y, the evolu evolution tion of inhibi inhibitions tions play pla y. On thr three ee occ occasi asions ons,, the they y hav have e eve even n ties.Accordingl agains instt let lethal hal rai raidin ding g may hav have e ina inadve dverrkidnapped young monkeys during play, but aga tently caused monkey hunting to be inhibnott ea no eate ten n th them em (r (rev evie iewe wed d by Wra rang ngha ham m an and d ited it ed al also so,, if th the e tw two o pa patt tter erns ns ar are e in inde deed ed Peterson, 1996). motivationally onally related (Eibl-Ei (Eibl-Eibesfeldt besfeldt,, 1975; Stanford (1998b) suggested that the rea- motivati son why male bonobos hunt rarely is that van Hooff, 1990). This proposal implies that monkey ey hunt hunting ing is moti motivati vational onally ly more simi simi-hunting has a low pay-off, because they tend monk lar to lethal raiding, and relatively distinct to lose meat to females. Against this, lowfrom m the kil killin ling g of ter terres restri trial al ung ungula ulates tes.. ranking male chimpanzees often lose meat fro Importa ortant nt simi similari larities ties betw between een monk monkey ey to hig high-r h-rank anking ing mal males, es, but sti still ll hun huntt fre fre-- Imp hunting and lethal raiding could include the quently (Goodall, 1986). Furthermore it is monkey hunting, rather than meat eating, necessity for coordination and planning and the e ab abili ility ty to as asse sess ss an ad adeq equa uate te po powe werr that appears to be lacking in bonobos, not th imbalance between predators and prey. only among males but also among females. In summary, it is admittedly speculative Stanfo Sta nford’ rd’ss pro propos posal al is the theref refore ore not sup sup-to propose that monkey hunting has been ported. lost los t in bon bonobo oboss as a result result of sel select ection ion aga agains instt Another Anoth er poss possible ible explanation explanation for the lack propen pensit sities ies rel releva evant nt to lethal lethal rai raidin ding. g. How How-of observations of monkey hunting by bono- pro ever, this idea appears to explain the facts bos is stochastic. Hunting traditions might better than alternative hypotheses. It sug vary among bonobo populations. If so, monkey hunting may be observed in the future, gests a correlation between group hunting in populations that have not yet been stud- and lethal raiding in humans, chimpanzees and d bon bonobo oboss tha thatt cha challe lleng nges es tra tradit dition ional al ied. However, there is no evidence that any an chimpanzee chimp anzee population population fails to hunt mon- thinking, and draws attention to the need keys, provided monkeys are present. Chim- for further data. panz pa nzee eess pre prey y on mo monk nkey eyss in at le leas astt 12 si site tes, s, Morality includ inc luding ing all the lon long-t g-term erm sit sites es [Go [Gombe mbe,, Mahale, Taı¨, ¨, Kibale, and Budongo (Goodall, This paper suggests that violent propensi1986)] 198 6)] as wel welll as sev seven en les lesser ser-kn -known own pop popula ula-- ties of a particular kind have been positively tions [Chambura, (B. Fahey, personal com- selected among male chimpanzees and humunication; Kahuzi-Biega, DRC (Basasose mans mans.. Bio Biologi logicall cally y, this is uns unsurpr urprisi ising. ng. Lik Likeeand Yamagiwa, 1997), Lope´ , Gabon (Tutin wis wise, e, pro propen pensit sities ies for par partic ticula ularr typ types es of and Fernandez, 1993), Mt. Assirik, Senegal alt altrui ruisti sticc and coo cooper perativ ative e beh behavi avior or hav have e (McGrew et al., 1979), Outamba-Kilimi, Si- probably also evolved through selection, and erra err a Leo Leone ne (Al (Alp, p, 199 1993), 3), Sap Sapo, o, Lib Liberi eria a (An (Ander der-- are neither more nor less important biologison et al., 1983), Tongo, DRC (A. Lanjouw, cally than violence. personal perso nal commu communicati nication). on). There is, there there-But anth anthropo ropologi logists sts’’ view viewss on vio violenc lence e tend fore, for e, a str strong ong contrast contrast bet betwee ween n the wid widee- to be int interp erpret reted ed pol politi itical cally ly.. For exa exampl mple e spread spr ead occ occurr urrenc ence e of mon monkey key hun huntin ting g in Otterbein (1997) labeled anthropologists as chimpa chi mpanze nzees es and its abs absenc ence e in bon bonobo obos. s. ‘‘Haw ‘‘Hawks’’ ks’’ or ‘‘Dov ‘‘Doves’’ es’’ according according to wheth whether er Contrary to the stochastic hypothesis, this they consider evolutionary biology relevant
26
YEARBOOK YEARBOO K OF PHYSICA PHYSICAL L ANTHROP ANTHROPOLOGY OLOGY
[Vol. 42, 1999
or irrelevant to warfare (Otterbein, 1997). understand about the evolutionary origins Although Although Doves (e.g. (e.g.,, Power Power,, 1991; Suss- and persistence of intergroup violence, the man, 1997) suggest that Hawks (e.g., Ghi- better we can predict and avert it. glieri, 1984; Goodall, 1986; Wrangham and CONCLUSION Peters Pet erson, on, 199 1996) 6) are cul cultur turall ally y bia biass ssed ed in thinking that male chimpanzees strive agDespite Despi te some important unso unsolved lved probgressively for status and use violence adap- lem lems, s, chi chimpa mpanze nzee e let lethal hal rai raidin ding g app appear earss tively in intergroup interactions, there is no gen general erally ly well expl explaine ained d by the imba imbalanc lance-ofe-ofevidence for a positive correlation between pow power er hyp hypoth othesi esis, s, whi which ch sta states tes tha thatt suc succes cesssanthropological Hawkishness (in the Otter- ful atta attacks cks on riv rivals als are fav favore ored d bec becaus ause e bein bei n sen sense) se) and pol politi itical cal bel belief iefs. s. Ind Indeed eed,, they increase the dominance status of the some notable ‘‘anthropological Hawks’’ have aggr aggressor essors. s. A combin combination ation of three points been prominent in the search for peace. For likewise suggests that selection has favored exampl exa mple, e, Ham Hambur burg g (19 (1991) 91) arg argued ued for the unpro unprovoked voked intergroup intergroup viole violence nce in human import imp ortanc ance e of bio biolog logica icall sim simila ilarit rities ies in chi chimm- mal males: es: the pre preval valenc ence e of hum human an war rai raidin ding, g, panz pa nzee ee an and d hu huma man n vi viol olen ence ce,, an and d in th the e sa same me the similarities of chimpanzee and human spirit spi rit co-c co-chai haired red a mul multiti-yea yearr ef effort fort to re- lethal raiding, and the ability of the imbalance-o e-of-p f-powe owerr hy hypot pothes hesis is to exp explai lain n the duce the frequency and intensity of interna- anc tional tion al viol violence ence (Car (Carnegi negie e Comm Commissi ission, on, 1997 1997). ). mammalian distribution of lethal violence. There is no moral high ground to be held by Until an alternative model exists, chimpanzeess and hum zee humans ans are are,, the theref refore ore,, bes bestt re virtue of being an anthropological Dove. garded ded as spe specie ciess in whi which ch a dom domina inance nce Admit Admitted tedly ly,, any the theory ory of vio violen lence ce has gar drive ve by mal male e gro groups ups has bee been n pos positi itivel vely y moral impli implicatio cations, ns, becau because se biolo biological gical analy analy-- dri ses can be misused. But no theory, however selected. If this conclusion has merit, anthropology benign or malevolent or whether based on biology, psychology, or culture, is immune to has given inadequate consideration to coalico-option by ideologues and propagandists. tionary violence as a force in human evoluWhile Whi le Ger German man mil milita itary ry phi philos losoph ophy y was tion. As anthropologists, we have a duty to backed bac ked by Dar Darwin winism ism in the Fir First st Worl orld d acknowledge the horrors of our evolutionary War ar,, Fre French nch mil milita itary ry phi philos losoph ophy y was bac backed ked past, partly for the sake of truth, and partly consi sider der how su such ch beh behav avior ior can be avo avoide ided d by Ber Bergso gson’ n’ss the theory ory of cre creati ative ve evo evolut lution ion to con future.. By combin combining ing primat primatologi ological, cal, (Tuchman, 1962). On either side, opposing in the future paleontolo paleo ntological gical and behavioral behav ioral-ecolo -ecological gical evtheo th eori ries es of ev evol olut utio ion n we were re us used ed to bo bols lste terr th the e idence, e, anthr anthropolog opologists ists can provi provide de espewaging wag ing of war war.. Mil Milita itary ry org organi anizat zation ionss can be idenc expect exp ected ed to dec deceiv eive e the themse mselve lvess and the their ir cially rich tests of evolutionary hypotheses. Thes ese e wi will ll of offe ferr a so soli lid d ba base se fr from om wh whic ich h follow fol lowers ers usi using ng any ava availa ilable ble mat materi erials als Th evolutiona evolu tionary ry anthr anthropolog opology y can work with (Wrangham, (Wrangh am, 1999b). otherr disc discipli iplines nes to unde understa rstand nd cult cultural ural I see no better course than to follow Dar- othe va varia riatio tion n and the proxim pro ximate ate stimul sti muli i thatt tha win (1871, p 405): ‘‘. . .we are not here concerned cer ned with hopes or fea fears, rs, only with the elicit violence. trut tr uth h as far as ou ourr re reas ason on pe perm rmit itss us to ACKNOWLEDGMENTS discover it.’’ Lethal violence appears strikingly frequent frequent among chimpanzees chimpanzees and huI thank Clark Larsen and the American mans, and appears explicable by relatively Association of Physical Anthropologists for simple sim ple ada adapti ptive ve rul rules. es. Cur Curren rentt evi eviden dence ce su sugg- inviting me to speak at the AAPA annual gests it has been a major selective pressure meeting in Salt Lake City, City, April 1998, which for sig signifi nifican cantt per period iodss of chi chimpa mpanze nzee e and prompted this paper. Irven DeVore, Robert human hum an evo evolut lution ion.. Unt Until il let lethal hal vio violen lence ce is Hin Hinde, de, Bil Billl McG McGrew rew,, and Kar Karen en Str Strier ier gen generershown to be a strange new phenomenon, we ously offered extended comments. For helpshould sho uld con consid sider er it su suff ffici icient ently ly anc ancien ientt to ful critique critiquess I am gra gratef teful ul als also o to Chr Chrisishave influenced the temperaments of both tophe topherr Boehm Boehm,, Chris Christophe tophe Boesch, Nancy species, particularly of males, in ways that DeV DeVore, ore, Marti Martin n Muller Muller,, Vernon Reyno Reynolds, lds, should sho uld be tak taken en ser seriou iously sly.. The mor more e we Car Carel el van Scha Schaik, ik, and Mich Michael ael Wi Wilso lson. n. Chri Chriss-
Wrangham]
EVOLUTION OF COALITIONARY KILLING
tophe Boe tophe Boesch sch,, Kat Katie ie Faw Fawcet cett, t, Mar Martin tin Mul Muller ler,, Toshis oshisada ada Nishi Nishida, da, Vernon Reyno Reynolds, lds, and Michae Mic haell Wi Wilso lson n kin kindly dly gav gave e acc access ess to unp unpubublished lish ed mater material, ial, Mich Michael ael Huf Huffman fman aided with references, and Johan van der Dennen improved the article by providing his important book.
LITERATURE LITERA TURE CITED AdamsDB. 198 AdamsDB. 1983. 3. Wh Why y the there re areso few wom women en war warrio riors. rs. Behav Sci Res 18:196–212. Adams ES. 1990. Boundary disputes in the territorial ant Azteca trigona: effects of asymmetries in colony size. Anim Behav 39:321–328. Alexa Alexande nderr RD. 198 1987. 7. The bio biolog logy y of mor moral al sys system tems. s. Hawthorne, NY: NY: Aldine de Gruyter. p. 1–301. Alexa Al exande nderr RD. 198 1989. 9. Evo Evolut lutionof ionof thehuman psy psyche che.. In: Mellars P, Stringer C, editors. The human revolution: Behavioral and biological perspectives on the origins of modern humans. Princeton, NJ: Princeton University Press. p 455–513. Alp R. 1993. Meat-eating and ant-dipping by wild chimpanzees in Sierra Leone. Primates 34:463–468. Anderson JR, Williamson EA, Carter J. 1983. Chimpanzees of Sapo Forest, Liberia: density, nests, tools, and meat-eating. Primates 24:594–601. ArcadiAC, Arc adiAC, Wr Wrangh angham am RW. 1999 1999.. Infa Infantic nticide ide in chim chimpanpanzees: review of cases and a new within-group observation tio n fro from m the Kanyawar Kanyawara a stu study dy gro group up in Kib Kibale ale National Park. Primates 40:337–351. Archer J, and Huntingford F. 1994. Game theory models and escalation of animal fights. In: Potegal M, Knutson JF JF,, edito editors. rs. The dyna dynamics mics of aggr aggressio ession. n. Hillsda Hillsdale, le, NJ: Lawrence Erlbaum. p 3–31. Ardrey R. 1961. African genesis: A personal investigation into the animal origins and nature of man. New York: Y ork: Atheneum. p 1–330. Ardrey R. 1966. The territorial imperative. New York: Atheneum. p 1–416. Ashley Montagu MF. 1968. Man and aggression. London: Oxford University Press. p 1–178. Austad Austad S. 1983. A game theo theoretic retical al inter interpret pretation ation of male combat in the bowl and doily spider ( Frontinella ( Frontinella pyramitela). pyramitela ). Anim Behav 31:59–73. Bailey RC. 1991. The behavioral ecology of Efe´ pygmy men in the Ituri Forest, Zaı¨re. ¨re. Ann Arbor Arbor,, MI: Univer Univer-sity of Michigan Press. p 1–143. Barkow Bar kow J, Cos Cosmid mides es L, Toob ooby y J. 199 1995. 5. The adapted adapted mind: Evolutionary psychology and the generation of culture. New York: Oxford University Press. p 1–668. Basasose N, Yamagiwa J. 1997. Predation on mammals by chi chimpa mpanze nzees es in the mon montan tane e for forest est of Kah Kahuzi uzi,, Zaire. Primates 38:45–55. Bauerr HR. 1980 Baue 1980.. Chimpa Chimpanzee nzee society and socia sociall dominance nanc e in evolu evolutiona tionary ry pers perspecti pective. ve. In: Omark DR, Strayer Stra yer FF FF,, Free Freedman dman D, edito editors. rs. Domin Dominance ance relations: Ethological perspectives of human conflict. New York: Y ork: Garland. p 97–119. Boehm C. 1992. Segmentary ‘‘warfare’’ ‘‘warfare’’ and the management of conflict: comparison of EastAf rican chimpanzees and patrilineal-patrilocal humans. In: Harcourt AH, de Waal FBM, editors. Coalitions and alliances in humans and other animals. Oxford: Oxford University Press. p 137–173. Boehm Boe hm C. 199 1999. 9. Thenatur Thenatural al sel select ection ion of alt altrui ruisti sticc tra traits its.. Human Nature: 10. Bock K. 1980. Human nature and history: history: A resp response onse to sociobiology. New York: Columbia University Press. p 1–241.
27
Boesch C. 1996. Social grouping in Taı¨ chimpanzees. In: McGrew WC, Marchant LF, Nishida T, editors. Great ape societies societies.. Cambridge:Cambridge UniversityPress. p 101–113. Boesch Boes ch C, Boesc Boesch h H. 1989. Hunti Hunting ng behavior of wild chimpanzees in the Taı¨ National Park. Am J Phys Anthropol 78:547–573. Boesch C, Boesch H. 1999. The chimpanzees of the Taı¨ Taı¨ Forest: Behavioral ecology and evolution. New York: Oxford University Press, in press. BrainCK. Br ainCK. 19 1981.The 81.The hun hunter terss orthe hun hunte ted?An d?An int intro roduc duc-tion toAfrican cavetaphonomy. Chica Chicago: go: Unive Universityof rsityof Chicago Press, p 1–365. Bueno de Mesquita B. 1981. The war trap. New Haven, CT: Yale University Press. p 1–223. Bueno de Mesquita B. 1985. The war trap revisited. A revise rev ised d exp expect ected ed uti utility lity mod model. el. Am Pol Polit it Sci Rev 79:156–177. Bueno de Mesquita B, Lalman D. 1992. War and reason: Domestic and international imperatives. New Haven, CT: Yale University Press. p 1–322. Byers J. 1997. American pronghorn: Social adaptations and the ghosts of predators past. Chicago: University of Chicago Press. p 1–300. Bygott Bygo tt JD. 1979. Agonistic Agonistic beha behavior vior,, domin dominance ance,, and sociall stru socia structur cture e in wild chimpanzees chimpanzees of the Gombe National Park. In: Hamburg DA, McCown ER, editors. tor s. The gre great at ape apes. s. Men Menlo lo Par Park, k, CA: Ben Benjam jamin/ in/ Cummings. p 405–428. Carnegie Carn egie Commis Commission sion on Prev Preventin enting g Deadl Deadly y Confli Conflict. ct. 1997. Preventing deadly conflict: Final report. Washington ingt on DC: Carn Carnegie egie Comm Commissio ission n on Prev Prevent enting ing Deadly Conflict. p 1–257. Caro Car o TM, Col Collins lins DA. 198 1986. 6. Mal Male e che cheeta etahs hs of the Serengeti. Nat Geogr Res 2:75–86. Cartmill M. 1993. A view to a death in the morning: Hunting and nature through history. history. Cambridge, MA: Harvard University Press. p 1–331. Chagnon NA. 1992. Yanomamo ¨: The last days of Eden. ¨: New York: Harcourt Brace Jovanovich. p 1–309. Chapais B. 1995. Alliances as a means of competition in primates: evolutionary, developmental and cognitive aspects. Yearbk Phys Anthropol 38:115–136. 38:115–136. Chapman CA, Wrangham RW. 1993. Range use of the forestt chimp fores chimpanzee anzeess of Kibale Kibale:: implic implication ationss for the understanding of chimpanzee social social organization. Am J Primatol 31:263–273. Chapman CA, White FJ, Wrangham RW. 1994. Party size in chimpanzees and bonobos: A reevaluation of theory the ory bas based ed on two sim similar ilarly ly for forest ested ed sit sites. es. In: Wrangham RW, McGrew WC, de Waal FBM, editors. Chimpanzee cultures. Cambridge, MA: Harvard Uni versity Press. p 41–58. Chapma Cha pman n CA, Wran rangha gham m RW, Cha Chapma pman n LJ. 199 1995. 5. Ecological constraints on group size: an analysis of spiderr monke spide monkey y and chimp chimpanze anzee e subgr subgroups. oups. Behav Ecol Sociobiol 32:199–209. Chapman Chap man CA, Chapm Chapman an LJ, Wr Wrangha angham m R, Isab IsabiryeiryeBasuta Basu ta G, Ben-D Ben-David avid K. 1997. Spatial and tempo temporal ral variabilit var iability y in the structure structure of a trop tropical ical forest. Afr J Ecol 35:287–302. Cheney Chen ey DL. 1986. Interactions Interactions and relat relationsh ionships ips between groups. In: Smuts BB, Cheney DL, Seyfarth RM, Wrangham RW, Struhsaker TT, editors. Primate societies. socie ties. Chicago: University University of Chica Chicago go Press Press.. p 227–239. Cheney DL, Seyfarth RM, Andelman SJ, Lee PC. 1988. Reproductive success in vervet monkeys. In: CluttonBrock Broc k TH, edito editorr. Repr Reproduct oductive ive succ success: ess: Studi Studies es of individual variation in contrasting breeding systems. Chicago: University of Chicago Press. p 384–402. Clutton-Br Clutt on-Brock ock TH, Guinn Guinness ess FE, Albon SD. 1982. Red
28
YEARBOOK YEARBOO K OF PHYSICA PHYSICAL L ANTHROP ANTHROPOLOGY OLOGY
deer: Beha deer: Behavior vior and ecolo ecology gy of two sexes. Chicago: University of Chicago Press. p 1–378. Daly M, Wils Wilson on M. 1988. Homicide. Homicide. Hawth Hawthorne, orne, NY: Aldine de Gruyter. p 1–328. Dart RA. 1953. The predatory transition from ape to man. Intern Anthropol Ling Rev 1:201–218. Dart RA, Craig D. 1959. Adventures with the missing link. New York: Harper. p 1–255. Darwin C. 1871. The descent of man and selection in relation to sex. London: John Murray. Murray. p 1–475. Dentan RK. 1992. The rise, maintenance, and destruction of peaceable polity: a preliminary essay in political ecology. In: Silverberg J, Gray JP, editors. Aggression and peacefulness in humans and other primates. Oxford: Oxford University Press. p 214–270. de WaalFBM. 198 1982. 2. Chi Chimpa mpanze nzee e pol politi itics:Powerand cs:Powerand sex among apes. New York: Harper and Row. p 1–223. de Waal FBM, and Lanting F. 1997. Bonobo: The forgotten ape. Berkeley, CA: University of California Press. p 1–210. de Ruiter JR, and van Hooff JARAM. 1993. Male dominance nan ce ran rank k and rep reprod roduct uctive ive suc succes cesss in pri primat mate e groups. Primates 34:513–524. Eastt ML, Hofer H. 199 Eas 1991. 1. Lou Loud-c d-call alling ing in a fem female ale dominated domin ated mamma mammalian lian socie society ty.. II. Behav Behavioral ioral contextss and functions of whoop text whooping ing of spot spotted ted hyenas, Crocuta crocuta. Anim Behav 42:651–669. Ehrenreich B. 1997. Blood rites: Origins and history of the passions of war. New York: Metropolitan. p 1–292. Eibl-Eibesfeldt I. 1975. Krieg und Frieden aus der Sicht der Verhal erhaltens tensforsc forschung hung.. Munic Munich: h: Piper Verlag erlag.. p 1–315. Ellis Ell is L. 199 1995. 5. Dom Domina inance nce and rep reprod roduct uctive ive suc succes cesss among nonhuman animals: A cross-species comparison. Ethol Sociobiol 16:257–333. Enquist M, Leimar O. The evolution of fatal fighting. Anim Behav 39:1–9. Ferguson Ferg uson RB. 1989 1989.. Anthr Anthropolo opology gy and war: theo theory ry,, politics, polit ics, ethics. In: Turner Turner PR, Pitt D, editors. The anthropology of war and peace: Perspectives on the nuclear age. London: Greenwood Press. p 141–159. Ferrill Ferr ill A. 1985. The origins of war from the stone age to Alexander the Great. London: Thames and Hudson. p 1–240. Frank Fra nk LG. 198 1986. 6. Soc Social ial org organi anizat zation ion of the spo spott tted ed hyaena (Crocuta ( Crocuta crocuta). crocuta ). I. Demography. Anim Behav 35:1500–1509. Furuichi T, Idani G, Ihobe H, Kuroda S, Kitamura K, Morii A, Eno Mor Enomot moto o T, Oka Okayas yasu u N, Has Hashim himotoC, otoC, Kan Kano o T. 1998. 199 8. Pop Popula ulatio tion n dyn dynami amics cs of wild bon bonobo oboss ( Pan paniscus)) at Wamba. Int J Primatol 19:1029–1043. paniscus Ghiglieri MP. MP. 1984. The chimpanzees of Kibale Forest: A field study of ecology and social structure. New York: Columbia University Press. p 1–226. Ghiglieri MP. 1988. Sociobiology of the great apes and the hominid ancestor. J Hum Evol 16:319–358. Goodall J. 1986. The chimpanzees of Gombe: Patterns of behavior. Cambridge, MA: Harvard University Press. p 1–673. Goodall J, Bandora A, Bergmann E, Busse C, Matama H, Mpongo E, Pierce A, Riss D. 1979. Intercommunity intera int eracti ctions ons in the chi chimpa mpanze nzee e pop popula ulatio tion n of the Gombe National Park. In: Hamburg DA, McCown ER, editors. The great apes. Menlo Park, CA: Benjamin/ Cummings. p 13–54. Gould SJ. 1996. The Diet of Worms and the defenestration of Prague. Natur Hist, Sept: 18–67. Grinnell J, Packer C, Pusey AE. 1995. Cooperation in male lions: kinsh kinship, ip, recip reciprocit rocity y or mutua mutualism? lism? Anim Behav 49:95–105. Grossman D. 1996. On killing: The psychological cost of learning to kill in war and society. Boston, MA: Little, Brown. p 1–367.
[Vol. 42, 1999
Hamburg DA. 1991. An evolu Hamburg evolutiona tionary ry pers perspect pective ive on human aggression. In: Bateson P, P, editor. The development and integration of behavior: Essays in honor of Robert Rob ert Hind Hinde. e. Camb Cambridg ridge: e: Cam Cambrid bridge ge Univ Univers ersity ity Press. p 419–458. Harcourt Harc ourt AH. 1987. Dominance Dominance and fertility among female primates. J Zool Lond 213:471–487. Hashim Has himotoC, otoC, Ta Tashi shiroY, roY, Kim KimuraD, uraD, Eno Enomot moto o T, Ing Ingman man-son EJ, Idani G, Furuichi T. 1998. Habitat use and ranging of wild bonobos ( Pan ( Pan paniscus) paniscus ) at Wamba. Int J Primatol 19:1045–1060. Hausfater G, and Hrdy SB, editors. 1984. Infanticide: Comparativ Compa rative e and evolu evolutiona tionary ry pers perspecti pectives. ves. Hawthorn, NY: NY: Aldine de Gruyter. p 1–596. Henschel JR, Skinner JD. 1991. Territorial behavior by a clan of spotted hyenas Crocuta crocuta. Ethology 88:223–235. Hinde Hind e RA.1993.Aggre RA.1993.Aggressio ssion n andwar: indi individu viduals,groups als,groups,, and states. In: Tetlock PE, Husbands JL, Jervis R, editors. Behavior, society, and international conflict. Oxford: Oxford University Press. p 8–70. Hofer A, Huffman MA, Zeisler G. 1998. Mahale: Begegnung mit Schimpansen. Verlag Navalon. p 1–159. Ho¨lldoblerB. ¨lldoblerB. 1981.Foragingand spat spatiote iotempor mporal al terr territoitoriess in thehone rie thehoney y ant ant Myrmecocystusmimicus Myrmecocystusmimicus Wheeler (Hymenopter (Hymen optera: a: Formic Formicidae). idae). Behav Ecol Sociob Sociobiol iol 9:301–314. Hrdy SB. 1990. Sex bias in nature and in history: a late 1980s reexamination of the ‘‘biological origins’’ argument. Yearbk PhysAnthropol 33:25–37. Idani G. 1991. Cases of inter-unit group encounters in pygmy chimpanzees at Wamba, Zaı¨re. ¨re. In: Ehara Ehara A, et al., edito editors. rs. Primat Primatology ology Today: Today: Proc Proceedin eedings gs of the XIIIth Congress of the International Primatological Society.. Amsterdam: Elsevier. p 235–238. Society Isabirye-Basuta G. 1989. The ecology and conservation status stat us of the chimpanzee chimpanzee Pan troglodytes schwein furthii in Kibale Forest, Uganda. Ph.D. dissertation, Makerere University, University, Kampala, Uganda. p 1–280. Janson CH, Goldsmith ML. 1995. Predicting group size in pri primat mates:forag es:foragingcostsand ingcostsand pre predat dationrisks ionrisks.. Beh Behav av Ecol 6:326–336. Jurmai Jur main n R. 199 1997. 7. Ske Skelet letal al evi eviden dence ce of tra traumain umain Afr Africa ican n apes, with special reference to the Gombe chimpanzees. Primates 38:1–14. Kano Kan o T. 199 1992. 2. Thelast ape ape:: Pyg Pygmy my chi chimpa mpanze nzee e beh behavi avior or and ecology. Stanford, CA: Stanford University Press. p 1–248. Keegan J. 1993. A history of warfare. New York: Alfred A. Knopf. p 1–432. Keeley Keele y LH. 1996. War before civiliz civilization ation.. New York: York: Oxford University Press. p 1–245. Knauft B. 1991. Violence and sociality in human evolution. Curr Anthropol 32:391–428. 32:391–428. Krebs JR and Davies NB. 1997. Behavioral ecology: an evolutionary approach. Oxford: Blackwell. p 1–456. Kruuk H. 1972. The spotted spotted hyena: A study of predation and social behavio behavior.Chicago: r.Chicago: ChicagoUniversity Press. p 1–335. Kuroda S. 1979. Grouping of the pygmy chimpanzees. Primates 20:161–183. Larsen CS. 1997. Bioarchaeology: Interpreting behavior from the human skeleton. skeleton. Cambr Cambridge: idge: Cambridge Cambridge University Press. p 1–461. Leach E. 1968. Don’t say ‘‘Boo’’ to a goose. In: Ashley Montagu Mont agu MF MF,, edito editor. r. Man and aggr aggressio ession. n. Londo London: n: Oxford University Press. p 65–73. Leonard Leon ard WR, Rober Robertson tson ML. 1997 1997.. Compa Comparativ rative e primate energetics and hominid evolution. Am J Phys Anthropol 102:265–281. Lorenz K. 1966. On aggression. Translated by Marjorie Latzke. London: Methuen. p 1–273.
Wrangham]
EVOLUTION OF COALITIONARY KILLING
Lowen C, Dunbar RIM. 1994. Territory size and defendability in primates. Behav Ecol Sociobiol 35:347–354. Manson JH, Wrangham RW. 1991. Intergroup aggression in chimp chimpanze anzees es and humans. Curr Anthropol Anthropol 32:369–390. Maschner Masc hner HDG, Reedy Reedy-Masc -Maschner hner KL. 1998. Raid, retreat, defend (repeat): the archaeology and ethnohistory of warfare on the North Pacific rim. J Anthropol Archaeol 17:19–51. McComb McC omb K, Pac Packer ker C, Pus Pusey ey A. 199 1994. 4. Roa Roarin ring g and numerical assessment in contests between groups of female lions, Panthera leo. Anim Behav 47:379–387. McGrew WC, Tutin CEG, Baldwin PJ. 1979. New data on meat-eating by wild chimpanzees. Curr Anthropol 20:238–239. Mech LD. 1977. Productivity, mortality, and population trends of wolves in north-eastern Minnesota. J Mammal 58:559–574. Mech Mec h LD. 199 1994. 4. Buf Bufferzonesof ferzonesof ter territ ritori ories es of gra gray y wol wolves ves as regions of intraspecific strife. J Mammal 75:199– 202. Mech LD, Adams LG, Meier TJ, Burch JW, Dale BW. 1998.. The Wo 1998 Wolves lves of Dena Denali. li. Minne Minneapolis apolis,, MN: Unive Univerrsity of Minnesota Press. p 1–225. Miller LE. 1998. Fatal attack among wedge wedge-capp -capped ed capuchins. Folia Primatol 69:89–92. Milton Mil ton K. 198 1987. 7. Pri Primat mate e die diets ts and gut mor morpho pholog logy: y: implic imp licati ationsfor onsfor hom hominidevolu inidevolutio tion. n. In: Har Harris ris M, Ros Rosss EB, editors. Food and evolution: Toward a theory of human food habits. Philadelphia: Temple University Press. p 93–116. Mitani JC, Rodman PS. 1979. Territoriality: relation of ranging patterns and home range size to defendability,, with an analy ity analysis sis of terr territori itoriality ality among primat primate e species. Behav Ecol Sociobiol 5:241–251. Morris D. 1977. Manwatching: A field guide to human behavior.. New York: Harry N. Abrams, Inc. p 1–318. behavior Newton-Fi Newt on-FisherNE. sherNE. 1997 1997.. Ta Tactic ctical al beha behaviorand viorand deci decision sion-making in wild chimpanzees. Ph.D. dissertation, Uni versity of Cambridge. p 1–325. Nishida T. 1979. The social structure of chimpanzees of the Mahale Mountains. In: Hamburg DA, McCown ER, editors. The great apes. Menlo Park, CA: Benjamin/Cummings. p 73–121. Nishida T. 1990. A quarter century of research in the Mahale Mahal e Mount Mountains:An ains:An over overview view.. In: Nishid Nishida a T, edito editorr. The chimpanzees chimpanzees of the Mahale Mount Mountains: ains: Sexual and life history strategies. Tokyo: University of Tokyo Press. p 3–36. Nishida T. 1994. Review of recent findings on Mahale chimpanzees: implications and future research directions. In: Wrangha Wrangham m RW, McGrew WC, de Wa Waal al FBM, Heltne Heltn e PG, edito editors. rs. Chimpa Chimpanzee nzee cultu cultures. res. Cambr Cambridge, idge, MA: Harvard University Press. p 373–396. Nishida T. 1996. The great chimpanzee ruler killed by a coalition coali tion of prev previous ious group mate mates: s: crue cruell polit political ical dynamicss in wild chimp namic chimpanzee anzees. s. Asahi Asahi-Shimb -Shimbun, un, Janu Janu-ary 31, 1996, evening edition. p 1. Nishida Nishid a T, Hara Haraiwa-Ha iwa-Hasegaw segawa a M, Ta Takahat kahata a Y. 1985 1985.. Group Grou p extin extinctionand ctionand femal female e tran transfer sfer in wild chimp chimpananzeess in theMahaleNationa zee theMahaleNationall Park Park,, Ta Tanza nzania.Z nia.Z Tie Tierps rpsyychol 67:284–301. Nunn CL. 1999. Collective Collective benefi benefits, ts, free-riders, free-riders, and male extragroup conflict. In: Kappeler P, editor. Male primates. Cambridge: Cambridge University Press. p 192–204. Otterbein KF. 1985. Preface to Second Edition. In: The evolution of war: A cross-cultural study. New Haven, CT: Human RelationsArea Files. p 1–165. Otterbein KF. KF. 1997. The origins of war. Crit Rev 11:251– 277. Packer C, Herbst L, Pusey AE, Bygott JD, Hanby JP, Cairns SJ, Mulder MB. 1988. Reproductive success of
29
lions. In: Clutton-Brock TH, editor. Reproductive success:: Stud cess Studies ies of indivi individual dual variation in cont contrast rasting ing breeding bree ding syst systems. ems. Chica Chicago: go: Unive University rsity of Chica Chicago go Press. p 363–383. Palombit RA. 1999. Infanticide and the evolution of pair bonds in nonhuman primates. Evol Anthropol 7:117– 129. Parish A. 1996. Female relationships in bonobos ( Pan ( Pan paniscus): paniscus ): evide evidence nce for bondi bonding, ng, cooperation, cooperation, and female dominance in a male-philopatric species. Hum Nature 7:61–96. Pilbeam D. 1996. Genetic and morphological records of theHominoid theHomin oidea ea andhomin andhominid id ori origin gins:A s:A syn synthe thesis sis.. Mol Phylogenet Evol 5:155–168. Popp J, DeV DeVore ore I. 1979. Aggressiv Aggressive e compe competitio tition n and social dominance theory. In: Hamburg DA, McCown ER, editors. The great apes. Menlo Park, CA: Benjamin/Cummings. p 317–338. Powerr M. 1991 Powe 1991.. TheEgalita TheEgalitarian rians—Hu s—Human man andchimpan andchimpan-zee: An anth anthropol ropologica ogicall view of socia sociall orga organizat nization. ion. Cambridge: Cambridge University Press. p 1–290. Prostt JH. 1985. Chimpa Pros Chimpanzee nzee behavior behavior and models of hominization. In: Kondo S, editor. Primate morphophysiology, locomotor analyses, and human bipedalism. Tokyo: University of Tokyo Press, p 289–303. Putland DA, Goldizen AW. 1998. Territorial behavior in the Ta Tasma smania nian n nat native ive hen hen:: gro group up and ind indivi ividua duall performance. Anim Behav 56:1455–1463. Regal P. 1998. Violence and sex. Q Rev Biol 73:473–476. Reynol Rey nolds ds V, Rey Reynol nolds ds F. 196 1965. 5. Chi Chimpa mpanze nzees es of the Budongo Forest. In: De Vore I, editor. Primate behavior. New York: Holt, Rinehart and Winston. p. 368– 424. Rodseth Rods eth L, Wr Wrangha angham m RW RW,, Smuts BB, Harri Harrigan gan A. 1991.. The human commu 1991 community nity as a prima primate te society. society. Curr Anthropol 32:221–254. 32:221–254. SheaBT. 1984.An allome allometric tric pers perspect pective ive on the morph morphoological logic al and evolu evolutiona tionary ry relat relationsh ionships ips betwe between en pygmy ( Pan paniscus) paniscus) and common ( Pan ( Pan troglodytes) troglodytes) chimpanzees. In: Susman RL, editor. The pygmy chimpanzee. New York: Plenum Press. p 89–130. Singer JD. 1989. The political origins of international war: a multifactorial review. In: Groebel J, Hinde RA, editors. edito rs. Aggression Aggression and war: Their biolo biological gical and sociall base socia bases. s. Cambr Cambridge:CambridgeUniversit idge:CambridgeUniversity y Pres Press. s. p 202–220. Smuts BB. 1986. Sex and friendship in babboons. Hawthorne, NY: NY: Aldine de Gruyter. p 1–303. Sponse Spo nsell LE. 199 1996. 6. The natural natural his histor tory y of pea peace: ce: a positive view of human nature and its potential. In: Gregor T, T, editor. A natural history of peace. Nashville, TN: Vanderbilt University Press. p 95–125. Stanford C. 1998a. The social behavior of chimpanzees and bono bonobos:A bos:A crit critical ical revie review w. CurrAnthro CurrAnthropol pol 39:39 39:399– 9– 420. Stanford CB. 1998b. Chimpanzee and red colobus: The ecology of predator and prey. Cambridge, MA: Har vard University Press. p 1–296. Starin Star in ED. 1994. Philopatry Philopatry and affiliation affiliation among red colobus. Behaviour 130:253–270. Sugiyama Y. 1973. Social organization of wild chimpanzees. In: Carpenter CR, editor. Behavioral regulators of behav behavior ior in prima primates. tes. Lewisburg, Lewisburg, PA PA:: Buck Bucknell nell University Press. p 68–80. Sugiyama Y. 1989. Population dynamics of chimpanzees at Bossou, Guinea. In: Heltne PG, Marqu Marquardt ardt LA, editors.Underst edit ors.Understandi anding ng chimp chimpanze anzees. es. Cambr Cambridge idge,, MA: Harvard University Press. p 134–145. Sugiyama Y, Kawamoto S, Takenaka O, Kumazaki K, Miwa N. 1993. Paternity Paternity discr discriminat imination ion and inter inter-group grou p relat relationsh ionships ips of chimp chimpanze anzees es at Bosso Bossou. u. Primates 34:545–552. Sussman RW, editor. 1997. The biological basis of hu-
30
YEARBOOK YEARBOO K OF PHYSICA PHYSICAL L ANTHROP ANTHROPOLOGY OLOGY
man behavior: A critical review. New York: Prentice Hall. p 1–382. Thompson JAM. 1997. The history, taxonomy and ecology of the bonobo ( Pan ( Pan paniscus Schwartz, 1929) with a firs firstt des descr cript iption ion of a wild populat population ion living living in a forest/savanna mosaic habitat. D. Phil. thesis, Oxford University,, p 1–315. University TigerL. Ti gerL. 196 1969. 9. Menin gro groups ups.. NewY NewYork ork:: Ran RandomHouse domHouse.. p 1–254. Trudeau Tr udeau MB, Bergm Bergmann-R ann-Riss iss E, Hambu Hamburg rg DA. 1981 1981.. Toward owardss an evolu evolutiona tionary ry pers perspect pective ive on aggr aggressiv essive e behavior: the chimpanzee evidence. In: Hamburg DA, Trudea Tr udeau u MB, edito editors. rs. Biobe Biobehavio havioral ral aspe aspects cts of aggr aggresession. New York:Alan Liss. p 27–40 27–40.. Tuchma uchman n BW BW.. 1962 1962.. The guns of August. New York: Macmillan. p 1–511. Turney-High HH. 1949. Primitive war: Its practice and concepts. Columbia, SC: University of South Carolina Press. p 1–288. Tutin CEG, Fernandez M. 1993. Composition of the diet of chimpanzees and comparisons with that of sympatric lowland gorillas in the Lope´ Reserve, Gabon. Am J Primatol 30:195–211. Tuti utin n CEG CEG,, McG McGrew rew WC, Bal Baldwi dwin n PJ. 198 1983. 3. Soc Social ial organization of savanna-dwelling chimpanzees, Pan troglodytes verus, at Mt. Assirik, Senegal. Senegal. Prima Primates tes 24:154–173. van Hoof JARAM JARAM.. 1990. Intergroup Intergroup compe competitio tition n and conflic con flictt in ani animalsand malsand man man.. In: Den DennenJMG nenJMG vander vander,, Falger VSE, editors. Sociobiology and conflict: Evolutionary perspectives on competition, cooperation, violence len ce and warfare. warfare. Lon London don:: Cha Chapma pman n and Hal Hall. l. p 23–54. van Schaik CP. 1996. Social evolution in primates: the role of ecological factors and male behavior. Proc Brit Acad 88:9–31. van van der Dennen Dennen JMG. 199 1995. 5. The origin origin of war war:: The evolution of a male-coalitional reproductive strategy. Groningen, Netherlands: Origin Press. p 1–861. Watts DP. 1999. Coalitionary Coalitionary mate mate-guar -guarding ding by male chimpanze chimp anzees es at Ngog Ngogo, o, Kibal Kibale e Natio National nal Park Park,, Ugan Uganda. da. Behav Ecol Sociobiol 44:43–55. White FJ. 1988. Party composition and dynamics in Pan paniscus. Int J Primatol 9:179–193. White Whit e FJ. 1996 1996.. Compa Comparativ rative e socio socio-ecol -ecology ogy of Pan Pan panis panis-cus. In:McGre In:McGrew w WC,March WC,MarchantLF antLF,, Nis Nishid hida a T, edi editor tors. s. Great ape societies. Cambridge: Cambridge University Press. p 29–41.
[Vol. 42, 1999
White FJ. 1998 White 1998.. Seaso Seasonalityand nalityand soci socioeco oecology: logy: the impor impor-tance of variation in fruit abundance to bonobo sociality.. Int J Primatol 19:1013–1027. ity Wolpoff M. 1998. Paleoanthropology. New York: Alfred Knopf. p 1–370. Wrangham RW. 1974. Artificial feeding of chimpanzees and baboons in their natural natural habit habitat. at. Anim Behav 22:83–93. Wrang ranghamRW hamRW.. 1975 1975.. The beh behavio avioralecologyof ralecologyof chim chimpanpanzees in Gombe National Park, Tanzania. Ph.D. dissertation, Cambridge University. p 1–280. Wrangham RW. 1987. The significance of African apes for reconstructing human social evolution. In: Kinzey WG, edito editor. r. Primate models of homini hominid d evolu evolution. tion. Albany,, NY: SUNY Press. p 51–71. Albany Wrangha Wr angham m RW. 1999a 1999a.. Why are male chimp chimpanzee anzeess more gregariou greg ariouss than mothe mothers? rs? A scra scramble mble compe competitio tition n hypothes hypo thesis. is. In: Kappe Kappeler ler P, edito editor. r. Male primat primates. es. Cambridge: Cambridge University Press. p 248–258. Wrangham RW. 1999b. Is military incompetence adaptive? Evol Hum Behav 20:3–17. Wrangham Wran gham RW, RW, Peterson D. 1996. Demonic males: Apes andthe orig origins ins of huma human n viole violence.Boston nce.Boston,, MA: Houg Houghhton Mifflin. p 1–350. Wrangha Wr angham m RW RW,, Clark AP, AP, Isabi Isabirye-B rye-Basut asuta a G. 1992 1992.. Female social relationships and social organization of the Kibale Fores Forestt chimp chimpanzee anzees. s. In: Nishida T, McGrew Gre w WC, Marler Marler P, Pic Pickfo kford rd M, de Waal FBM, editors. edito rs. Topics in prima primatolog tology: y: Human origins (V (Vololume 1). Tokyo: University of Tokyo Press. p 81–98. Wran rangha gham m RW, Cha Chapma pman n CA, Clar Clark k AP AP,, Isa Isabir biryeyeBasuta Basu ta G. 1996 1996.. Soci Social al ecol ecology ogy of Kany Kanyawar awara a chim chimpanpanzees: implications for understanding the costs of great ape groups. In: McGrew WC, Marchant LF, LF, Nishida T, editors. edito rs. Great ape socie societies. ties. Cambr Cambridge: idge: Cambr Cambridge idge University Press. p 45–57. Wrangham RW, Conklin-Brittain NL, Hunt KD. 1998. Dietary response of chimpanzees and cercopithecines to seasonal variation in fruit abundance: I. Antifeedants. Int J Primatol 19:949–970. Zihlma Zih lman n AL. 199 1996. 6. Loo Lookin king g bac back k in ang anger er.. Nat Nature ure 384:35–36. Zihlman Zihlma n AL, Crame Cramerr DL. 1978. Skeletal differences differences betwee bet ween n pyg pygmy my ( Pan panis paniscus cus)) andcommo andcommon n chi chimpa mpannzees ( Pan ( Pan troglodytes ). Folia Primatol 29:86–94.