As researchers searched for the fundamental nature of primate behavior [sic] in the 1960s, grey langurs were seen as a peaceful alternative to hierarchical and aggressive baboons. They rarely fought and dominance was relatively unimportant; infant-care was shared in this cooperative society. There were some reports of fighting and even infanticide by Japanese and Indian primatologists, but there was no plausible theoretical model available to "explain" such observations. The reports languished until the mid 1970s, when they were set into theoretical context and a model for sexually selected infanticide was developed; suddenly infant-killing langurs were paradigmatic exemplars of the selfishness of natural selection (e.g., Vogel, 1979).
A number of researchers did not accept this revision of langur nature, and argued forcefully that the observed violence was the result of crowding induced by humans via either deforestation or artificial provisioning (Boggess, 1984). They pointed to the locally high densities at some provisioned sites, and noted that under such circumstances conflicts could not be resolved by simple avoidance and group fission; violence begat violence and vulnerable infants died. This view was not based in any ultimate theoretical model but instead emphasized ontogenetic and social structural factors. The "behavioral pathology" explanation for male takeovers and infanticide in langurs has been abandoned by most people in light of further data from sites where langurs are not fed--but not before forcing researchers to examine carefully the ecological bases of the very real differences in social behavior and structure that had been observed, differences that tended to be ignored by proponents of the sexual selection model.
Chimpanzee behavior is deja vue all over again. Initial descriptions of non-hierarchical and unagressive apes living in open, fission-fusion groups that met occasionally for "carnivals" of display and grooming were hailed as evidence of innate human goodness; Suzuki's observation of cannibalism at Budongo in 1967 seemed an easily ignored aberation. Then in the mid-1970s "war" broke out among the chimpanzees studied by Jane Goodall and others at Gombe National Park. Males of one community seemed systematically to annihilate a neighboring group, and infanticide and cannibalism were observed repeatedly, both within and between communities. Similar events soon were observed by Toshisada Nishida and colleagues at Mahale National Park, the only other multi-decade chimpanzee study site (both are in Tanzania, about 150 km apart). These observations fit neatly into an emerging model of chimpanzee society as male-bonded and territorial, and our cousins were transformed from "the merry pranksters of the primate world" into "killers, soldiers, organized warriors" (Ghiglieri, 1987). Organized warfare, not carefree carnival, shapes chimpanzee sociality. As our closest relative, this new chimpanzee provokes some to ask, "Does war run in our genes like baldness and diabetes?" (Ghiglieri, 1987).
Now Power challenges the revision, arguing that chimpanzees are indeed the peaceful, open, jolly creatures of early reports, when not driven to warfare and cannibalism by human provisioning as at Gombe and Mahale. This tempts the reviewer in two directions: first, to analyze her argument's technical merits, and second, to discuss the form of both controversies (make it three; see Ferguson, 1992) with reference to Hobbes and Rousseau and the search for Human Nature. As a primatologist rather than philosopher/historian I focus on the first, but make no mistake--the second holds the key to a deep understanding of the book. (A prediction: sometime in the next decade bottlenose dolphin will be seen doing something really nasty, and others will blame it on stress induced by the noise of the observer's boat.) [1998 Eprint note: Infanticide has just been reported for one dolphin population by Patterson et al. 1998; we shall see if the rest of this prediction comes true.]
There are three intertwined elements to the book. The central idea is based on Woodburn's description of "immediate-return foraging systems" among gatherer-hunters; with minimal property and no formal legal, religious, political or economic institutions, "The totality of necessarily positive social relations between all members of the immediate-return societies constitutes the social structure" (p. 41, original emphasis). Such societies are necessarily egalitarian and disputes resolved by temporary fission; according to Power, this view of gatherer-hunters (which she develops based almost entirely upon the work of Woodburn and Turnbull) applies equally well to chimpanzees. Thus, both chimpanzees and humans, when living in their "natural state," are peaceful egalitarians--provided that contrary data can be explained away. The book is most interesting when she attempts to integrate her views of chimpanzees and gatherer-hunters with psychological literature and deWaal's studies of captive chimpanzees to explore the implications of immediate-return foraging for what she calls a "mutual dependence system," based on the interaction of "charismatic" and "dependent" social roles.
The bulk of the book is spent arguing that post-provisioning data gathered at Gombe and Mahale describe stressed animals and do not reflect normative patterns. To understand chimpanzees, only early Gombe and Mahale data (prior to 1965 & 1968, respectively), or data from "naturalistic" research, is reliable; she draws primarily on the work of Reynolds, Kortlandt & colleagues, Sugiyama, and Ghiglieri. The last portion of the book attempts to develop a theoretical model explaining why chimpanzee society goes wrong the way it does when faced with ecological stress, and speculates about possible "genetic factors" underlying her proposed mutual dependence system in humans, apes, humpback dolphins and perhaps mammals in general.
Power's thesis is that while provisioning itself is not so bad (just another heavily-fruiting tree in the eyes of a chimpanzee), artificially limiting access to food by enclosing bananas in boxes (Gombe) or dispensing sugarcane one piece at a time (Mahale) is intensely frustrating for animals forced to wait upon the whim of some human for a bite to eat. Frustration leads to aggression; as aggression becomes a way of life near the provisioning site, individuals lose the social skills needed to interact peacefully with relative strangers and so normal dispersal by both males and females is inhibited (some female dispersal continues, thanks to the 'passport' value of sexual swellings). Hostile communities are created and warfare between them is born.
The speed and repeatability with which these events unfold suggest a "natural" basis. Power argues that while egalitarian societies are favored when there is enough to go around, total egalitarianism in the face of severe famine might result in the death of all. Propensities toward aggression, hierarchy, cannibalism, communities and intercommunity conflict have evolved as latent safeguards; basically, when ecological disaster strikes, they kick in to break up affiliative social bonds and make life awful enough that a few individuals will disperse from their otherwise too-pleasant society into some empty habitat, thus ensuring survival of the species.
Power makes a strong and interesting case for the possible significance of frustration in shaping chimpanzee behavior at Gombe and Mahale. Her ideas deserve serious consideration by behavioral ecologists; we too often skip over the psychological processes that serve to translate genes into behavior, and to ignore completely the possibility that such processes might be affected in systematic and perhaps non-adaptive ways. Her argument, however, is seriously flawed in three general ways.
First, there is the group selection rationalization of chimpanzees' latent aggressiveness (e.g., "The ultimate goal of the entire set [of genetic attributes] is survival of the species," p. 239). Closely coupled to this is a view of a "natural," "vanished uncrowded world" in which rather than kill the neighbors, subgroups might undertake an "arduous journey" to some suitable habitat heretofore unoccupied by chimpanzees. Neither her units of selection nor the arena in which selection occurs is consistent with general evolutionary or ecological thought. This is not fatal to her criticism of the 'aggressive chimpanzee' story; she simply fails to replace it with a convincing explanation of her own (too bad; she proposes a genetic basis for TGIF parties [p. 252-3] that I found quite appealing).
Second, her scholarship is poor, partly a matter of carelessness and partly because all reports are accepted at face value. It is vital that any reader unfamiliar with the chimpanzee literature realize this--her assertions cannot be relied upon. For example: "Similarly, there was no territoriality in the sense of exclusive use of space among the Mahale apes in the early years (Azuma and Toyoshima 1962)" (p. 70). Azuma & Toyoshima in fact worked at Kabogo Point, an ecologically distinct site about 75 km from Mahale; though they do state that "groups" are non-antagonistic and non- territorial, it is clear that their "groups" are equivalent to temporary parties ("The size of a group ... varies from 2 to 3 heads to 20 to 30 heads" (Azuma & Toyoshima, 1962: 63), and furthermore they did not recognize individuals (p. 68). Minimal interpretation on Power's part would have indicated that these data do not demonstrate non-territoriality at Kabogo, let alone Mahale. Similarly, when arguing that predominantly male-inititated copulation is another consequence of provisioning (p. 78), she states that Albrecht & Dunnett (1971) "point out that only about one-quarter of the copulations in their (categorically wild) group in Guinea were initiated by males. [Unfortunately no numbers are given]." While they do make this point, they also note that their sample is "insufficient" and elsewhere indicate that "about 10" copulations were observed at Kanka Sili, the site in question (Albrecht & Dunnett, p. 39). Power either missed this or does not recognize the (in)significance of such a small sample. See also her site map's innacurracies (p. 22), blending of Budongo and Mahale (p. 97), and use of Albrecht & Dunnett's failure to see a predation during 3 months of observation in a plantation clearing as evidence that leopards are not serious threats to chimpanzees (p. 60). The nature of her hypothesis forces her to use the sparse data obtained by studies of unhabituated apes, but to use it so uncritically renders her book provocative but without substance.
The third flaw is her failure to discuss pertinent observations from Mt. Assirik or even to mention the Tai forest, both sites of longterm research on unprovisioned apes living in far more pristine habitats than several she classes as having "wild" (i.e., normal) chimpanzees. The Mt. Assirik community conducted a series of violent nightime attacks on a group of chimpanzees being rehabilitated near there, forcing the rehabilitation program to close (Goodall, 1986: 521) and the death of an adult male in the community was attributed to attacks by other chimpanzees (Baldwin, 1979: 69-71). Intercommunity relationships among the Tai chimpanzees have not been explicitly reported upon, but Boesch & Boesch (1989: 567) discuss the possible impact of "territorial fights" (29 seen in 29 months) on male grouping patterns there; carnivals and open networks seem ruled out. I could understand an attempt to explain away these reports, but to ignore them totally is either massive oversight (a book emphasizing the importance of research on unprovisioned chimpanzees overlooking the two longest qualifying studies?), or a deliberate stacking of the evidence.
However, her thesis cannot be dismissed as readily as her handling of the evidence with which she supports it. Despite its faults, it is founded on a true and troubling statement: "Despite more than 30 years of study ... there is no firm agreement as to the social organization of [chimpanzees]" (p. 1). Anyone believing this to be hyperbole should see, for example, Rodman's recent attempt to force Mahale chimpanzees into a Gombe mold (Rodman, 1991). Coming back to langurs (unimale at some sites, multimale at others), my own feeling is that this is because there is no such thing as the social organization of chimpanzees. Pace Rodman, Wrangham's model probably does not apply to all chimpanzee populations. It is difficult to see how it could work when home ranges surpass 200 km2 and densities fall below 0.01/km2, and Mahale/Gombe differences appear real, though subtle. Territorial conflict and closed communities at Gombe and Mahale do not preclude carnivals and open social networks at other sites, and vice versa. We need a better idea of precisely how ecological factors, acting through demographic variables, determine the behavioral options open to chimpanzees at different sites. Until we have such understanding it would be premature to reject the theses that (1) aggressive territoriality is not "normal" among some [all?] chimpanzee populations, and (2) "artificial" frustration may have played a significant role in structuring the intercommunity violence seen at Gombe and Mahale (though this is long odds indeed).
The Egalitarians does not live up to its forward (by Ashley Montagu; probably the most effusive ever penned), but it does set out an interesting hypothesis and remind us of some puzzling observations. Hopefully chimpanzee researchers and historians of science will find it stimulating. Others should probably ignore it, or proceed with extreme caution.
University of California, San Diego
La Jolla CA 92093
Albrecht, H. & Dunnett, S. C. (1971). Chimpanzees in Western Africa. Munich: R. Piper & Co.
Azuma, S. & Toyoshima, A. (1962). Progress report of the survey of chimpanzees in their natural habitat, Kabogo Point area, Tanganyika. Primates. 3: 61-70.
Baldwin, P. J. (1979). The natural history of the chimpanzee (Pan troglodytes verus) at Mt. Assirik, Senegal. PhD thesis, University of Stirling.
Boesch, C. & Boesch, H. (1989). Hunting behavior of wild chimpanzees in the Tai National Park. Am. J. Phys. Anthrop. 78: 547-573.
Boggess, J. (1984). Infant killing and male reproductive strategies in langurs (Presbytis entellus). pp. 283-310 IN Hausfater, G. & Hrdy, S. B. (Eds.), Infanticide: Comparative and Evolutionary Perspectives. New York: Aldine.
Ferguson, R. B. (1992). Tribal Warfare. Sci. Amer. 266: 108-113.
Ghiglieri, M. (1987). War among the chimps. Discover. 8: 66-76.
Goodall, J. (1986). The Chimpanzees of Gombe: Patterns of Behavior. Cambridge: Harvard University Press.
Rodman, P. S. (1991). Book review of The Chimpanzees of the Mahale Mountains. Int. J. Primatol. 12: 629-632.
Vogel, C. (1979). Der Hanuman-langur (Presbytis entellus), ein Parade-Exempel fur die theoretischen Konzepte der "Soziobiologie"? pp. 73- 89 IN Rathmayer, W. (Ed.), Verhandlungen der Deutschen Zoologischen Gesellschaft 1979 in Regensburg. Stuttgart: Gustav Fischer Verlag.
Patterson, I. A. P., Reid, R. J., Wilson, B., Grellier, K., Ross, H. M. & Thompson, P. M. (1998). Evidence for infanticide in bottlenose dolphins: An explanation for violent interactions with harbour porpoises? Proc. R. Soc. Lond. B. 265: 1167-1170.