Violence and Sociality in Human Evolution

CHICAGO JOURNALS

Wenner-Gren Foundation for Anthropological Research

Violence and Sociality in Human Evolution [and Comments and Replies]

Author(s): Bruce M. Knauft, Thomas S. Abler, Laura Betzig, Christopher Boehm, Robert Knox

Dentan, Thomas M. Kiefer, Keith F. Otterbein, Tohn Paddock, Lars Rodseth

Source: Current Anthropology, Vol. 32, No. 4 (Aug. - Oct., 1991), pp. 391-428

Published by: The University of Chicago Press on behalf of Wenner-Gren Foundation for

Anthropological Research

Stable URL: http://www.jstor.org/stable/2743815

Accessed: 03/10/2008 21:15

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher ?publisherCode=ucpress.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the
scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that
promotes the discovery and use of these resources. For more information about JSTOR, please contact support@jstor.org.

The University of Chicago Press and Wenner-Gren Foundation for Anthropological Research are collaborating
with JSTOR to digitize, preserve and extend access to Current Anthropology.

STOR

®
http://www.jstor.org

Current Anthropology Volume 32, Number 4, August-October 1991

© 1991 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved ooii-3204/9i/3204-ooo2$2.50

Violence and
Sociality in Human
Evolution'

by Bruce M. Knauft

Patterns of violence and sociality found among simple human
societies are compared and contrasted with those documented
among the four great-ape species and among more complex
prestate human societies. An evolutionary scenario is developed
by assessing comparative trends with respect to food/resource
sharing, dominance structure, male competition for sexual access
to females, violence and intergroup competition, and overall pat-
terns of violence and sociality. Simple human societies contrast
with both great-ape and middle-range human societies in exhib-
iting a relative absence of competitive male dominance hierar-
chies and of systematic violence between closed social groups,
while being more egalitarian among adult males politically, sex-
ually, and in terms of resource sharing. A U-shaped evolutionary
trajectory of selected features of human violence is proposed,
with the trough of the curve persisting throughout most of
Homo sapiens evolution. Simple human societies constitute a
major anomaly for models which propose evolutionary similarity
between great-ape and prestate human patterns of violence and
suggest a view of human evolution that takes seriously the
group-selection potential of symbolic transmission. It is sug-
gested that more consideration be given to cultural capacities in
interaction with ecological constraints and biogenetic selection
in studying the evolution of human sociality and violence.

BRUCE M. KNAUFT is Associatc Profcssor of Anthropology at Em-
ory University (Atlanta, Ga. 30322, U.S.A.) and during 1991-92 a
Fellow at the Center for Advanced Study in the Behavioral Sci-
ences (202 Junipero Serra Blvd., Stanford, Calif. 94305, U.S.A.).
Born in 1954, he was educated at Yale University (B.A., 1976)
and the University of Michigan (Ph.D., 1983). Among his re-
search interests are the evolution of violence, practice theory,
Melanesian ethnography, sorcery, spirit mediumship, sexuality,
and the comparative study of simple societies. He has published
Good Company and Violence: Sorcery and Social Action in a
Lowland New Guinea Society (Berkeley: University of California
Press, 1985), "Ritual Form and Permutation in New Guinea"
{American Ethnologist 12:321-40), "Text and Social Practice"
{Ethos 14:252-81), "Homosexuality in Melanesia" {Journal of
Psychoanalytic Anthropology 10:155-91), "Melanesian Warfare"
{Oceania 60:250-311), "Divergence between Cultural Success
and Reproductive Fitness in Preindus trial Cities" {Cultural An-
thropology 2:94-114), and "Reconsidering Violence in Simple
Human Societies" (current anthropology 28:457-500). The
present paper was submitted in final form i iii 91.

I . Financial support for the research upon which this study is based
from The Harry Frank Guggenheim Foundation is gratefully ac-
knowledged. Supportive comments and critical suggestions were
offered by Robert Carneiro, Jeremy Dahl, Arthur Demarest, Frans
de Waal, Johann Galtung, Thomas Gregor, Robert Paul, Marc Ross,
Bradd Shore, E. O. Smith, and Donald Tuzin; shortcomings remain
my own. Previous versions of this paper were presented at the 1987
annual meeting of the American Political Science Association and
at a 1990 H. F. Guggenheim conference in Charleston, S.C.

In simple human societies, lethal violence may be high
in aggregate statistical terms, but the pervasive ethos is
one of active cooperative affiliation among diverse
groups of relatives and nonrelatives (Knauft 1987a). The
cultural norms of sociality in these societies seem to be
both strong and prone to lethal contravention within the
local group. The violence that does occur has relatively
little to do with territorial rights, property, ritual status,
or male leadership concerns and is based more on con-
sensually approved status leveling among men than on
status elevation. Rather than being valued or associated
with kin-group or ethnic oppositions, violence emerges
sporadically among local cooperative groups, especially
as a social-control mechanism or as an expression —
commonly displaced — of male sexual frustration. Such
incidents are relatively uncontrolled and likely to result
in homicide.

This pattern of violence and sociality contrasts in very
broad terms with that found in more complex prestate
societies. In these ''middle-range'' societies, including
complex hunter-gatherers and what used to be called
''tribes'' and "chiefdoms" (Price and Brown 1985, Fein-
man and Neitzel 1984, Upham 1990a, Fried 1975,- cf.
Service 1971, Sahlins 1968, Carneiro 1981), sedentism,
property ownership, and male status differentiation are
more developed, and conflict tends to arise from overt
and chronic political status competition, both within
and between groups, and from competition over access
to resources. In contrast to that in simpler human
groups, violence in middle-range societies tends to be
valued as a dimension of masculinity, frequently takes
the form of collective reciprocating conflict (i.e., war-
fare), and is often linked with fraternal interest groups,
social boundedness, and ethnocentrism. In the evolu-
tion of Homo sapiens sapiens, it is likely, as Gell-
ner (1989:521) suggests, that coercion and violence as
systematic means of organizational constraint devel-
oped especially with the increasing socioeconomic com-
plexity and potential for political hierarchy afforded by
substantial stored food surplus and food production (see
Testart 1982).

Given these trends, the causes of severe violence com-
monly adduced for middle-range societies — political sta-
tus competition, resource or population pressure, terri-
torial disputes, and the existence of fraternal interest
groups — are less effective in explaining patterns of vio-
lence in simpler human societies (Knauft 1987a; con-
trast Otterbein 1968, Otterbein and Otterbein 1965,
Thoden van Velzen and van Wetering 1960; cf. Ross
1985, 1986; Price 1984). At the same time, sociobiologi-
cal explanations of violence emphasizing the impor-
tance of male sexual competition for reproductive suc-
cess have difficulty explaining the extreme political and
sexual egalitarianism among men in simple societies
(see Leacock and Lee 1982; contrast Chagnon 1988; cf.
Knauft 1987a). This raises an important larger question:
what is the overall trajectory of violence and sociality
in human evolution?

To begin answering this question, a broad comparison
of great-ape and human societies is here conducted with

391

39^ I CURRENT ANTHROPOLOGY Voluuie 32, Nuuiber 4, August- October 1991

regard to food/resource sharing, dominance structure,
male competition for sexual access to females, violence
and intergroup competition, and overall patterns of vio-
lence and sociality. Theoretical assessments and ele-
ments of an evolutionary scenario are developed in the
course of data presentation and analysis.

Qualitative contrasts between simple and more com-
plex prestate human societies raise a host of defini-
tional and methodological questions (Knauft 1990a;
i987a.-478-79, 481-82,- Woodburn 1980, 1982; Testart
1985; Feinman and Neitzel 1984). For present purposes,
''simple human societies'' are societies that lack recog-
nizable leadership roles and status differentials among
adult men. Egalitarianism tends to be pervasive in such
societies. This pattern is characteristic of many no-
madic foraging societies and a very few forager-
horticulturalists. In contrast, ''complex'' hunter-
gatherer societies, particularly those with sedentary
residence and/or rich resources, may exhibit elaborate
economic and political status-differentiation systems,
including rank distinctions and chiefs (e.g.. Price and
Brown 1985; Carneiro 1981:49). In evolutionary terms,
complex hunter-gatherer societies were most common
after 12,500 b.p., usually transitional between simple
hunter-gathering and agricultural systems, and of rela-
tively short prehistoric duration, usually no more than
two to three millennia (Henry 1985:366).^ A very few
migratory foragers, particularly Australian Aboriginal
societies, recognized political inequality among men on
the basis of ritual grade and age distinctions associated,
as Woodburn (1980) has suggested, with distinctive pat-
terns of delayed-return reciprocity.^

As Netting (1990) has recently noted, the development
of nonegalitarian political organization in cultural evo-
lution is not a determinate function of ecological and
demographic factors. Correspondingly, the core dimen-

2. Complex hunter-gatherer adaptations were present in selected
areas of Europe during the Upper Paleolithic (e.g., Mellars 1985)
and common during the Mesolithic (e.g., Zvelebil 1986; Bailey
1983: chaps. 8 and 9; Price and Brown 1985). Evidence of such com-
plexity in the Middle East dates to 14,000-10,000 b.p. in the Levant
Natufian (Henry 1985), and there is an exceptional case from the
central Russian Plain as early as 26,000 b.p. (Soffer 1985, McBurney
1976). Some complex hunter-gatherer formations survived and/or
developed during the contemporary historical period, among them
Northwest Coast fishing adaptations and Great Plains domes-
ticated-animal adaptations (the latter not involving sedentism but
potentiating competitive property rights and material status differ-
entiation).

3. Pronounced adult male status differentiation, polygyny, a greater
incidence of domestic and intergroup fighting, and a more devel-
oped sense of ethnic territorial identity (e.g., based on totemic/
ritual identification) were as a broad cluster incipient among Aus-
tralian Aboriginal cultures but rare among nonintensive hunter-
gatherers elsewhere. In the Australian context, raiding and warfare
were most developed among the complex hunter-gatherers who
inhabited resource-rich coastal areas (e.g., in Arnhem Land [e.g.,
Warner 1930, 1969 (1937); Hiatt 1965; Bemdt 1965) and among the
Tiwi [Hart and Pilling i960)). Reciprocal raiding and warfare were
much less developed in arid central Australia (e.g., among the
Walbiri [Meggitt 1962) and the Aranda [Strehlow 1947, 1965)) and
extremely arid western Australia (e.g., among the Pintupi [Myers
1986; cf. Hamilton 1982)).

sions of ''simple societies'' are sociopolitical and not
completely reducible to factors of subsistence or popula-
tion density (Layton 1986, Ingold 1987, Woodburn 1980).
It is, however, worthwhile to note ecological, economic,
and demographic characteristics that are seldom found
in association with simple societies as here defined: pop-
ulation densities greater than 2-3 per km^, year-round
residence at a single site, pronounced food storage (Tes-
tart 1982), substantial delayed-return reciprocity sys-
tems (Woodburn 1980, 1982, 1988), substantial material
wealth, and intensive reliance upon domesticated ani-
mals or fishing."^

These exclusions focus consideration on the ethno-
graphically documented societies that are more likely to
be representative of social dynamics among humans in
the period before complex and semisedentary hunter-
gatherer societies began to become prominent. These are
smaller and more decentralized human groups than have
been considered in most previous comparative assess-
ments of prestate or hunter-gatherer violence and social-
ity (e.g.. Ember 1978, Ferguson 1984a, Otterbein 1968,
Otterbein and Otterbein 1965, Haas 1990; see also Man-
son and Wrangham 1991; cf. Wrangham 1987).

The use of known ethnographic patterns to infer evo-
lutionary trajectories is of course fraught with problems
(see critiques by Schrire 1984; Headland and Reid 1989;
Leacock and Lee 1982: pts. 2 and 3,- Wobst 1978; Lewin
1988; Foley 1988:220; Myers 1988), and the number of
well-documented simple societies is relatively small. If
carefully undertaken, however, such an enterprise can
be a useful point of departure for further, more empiri-
cally sophisticated comparative research. Indeed, careful
articulation of ethnographic with prehistoric research is
indispensable to the development of a penetrating view
of human sociocultural development (e.g., O'Connell,
Hawkes, and Blurton Jones 1988a, b; see Foley and Lee
1989). Ethnographically studied simple societies are not
here equated with prehistoric ones or great apes with
our hominoid ancestors. Rather, it is suggested that the
striking differences among these more contemporary
congeries of social systems may have analogies in earlier
phases of human evolution.

It would be a mistake to write off our best data about
simple societies as a function of Western state encroach-
ment and projection (see Solway and Lee 1990, Wood-
burn 1988; contrast Wilmsen 1989). If anything, the his-
torical changes that have impacted hunter-gatherers in
this century have served to reduce rather than artifi-

4. In the Standard Cross-Cultural Sample, societies with significant
evidence of class stratification or wealth distinctions as tabulated
by Murdock (1981:101-2, col. 67) cannot be considered "simple."
Also excluded are intense riverine/coastal adaptations, which oc-
curred relatively late in human evolutionary history and frequently
supported extensive sedentary populations, food storage, and/or so-
ciopolitical hierarchies. High hunter-gatherer dependence on fish-
ing is correlated with such class stratification (x^ = 8.226, d.f. =
I, p < .01). As might have been expected from Ross's (1985:554;
1986) finding that socioeconomic complexity and external warfare
CO vary, there is also a positive (though not statistically significant)
correlation between warfare frequency and fishing dependence
among hunter-gatherers (cf. Ember 1978).

KNAUFT Violence and Sociality in Human Evolution \ 393

cially increase the distinctiveness of these societies.
Leadership and residential centralization, individualistic
property ownership, and status competition tend to in-
crease as foragers are impacted by the trade networks
and political status differentiation of horticultural and
state societies (e.g., Cashdan 1980, 1983, 1986; Hitch-
cock 1987; Kent 1989; Knauft 1990a; cf. the classic anal-
ysis by Murphy and Steward 1955). At the same time,
decentralized leadership, diffuse and flexible interband
alliance, generalized reciprocity, and adult male status
equality tend on the whole to be more common in more
autonomous and more decentralized foragers (e.g.,
Mbuti net-hunters versus Mbuti archers, !Kung San ver-
sus Basarwa [Cashdan 1980, 1986]). In terms of develop-
mental trajectories, then, the distinctive features of sim-
ple societies are unlikely to have been a function of
contemporary developments (see Woodburn 1988).

Food/Resource Sharing

One key common feature in many scenarios of human
evolution is the regular sharing of food and other re-
sources among adults and by adults with children (e.g.,
Tooby and DeVore 1987; Isaac 1978; Lovejoy 1981,
1982; Foley 1989; Foley and Lee 1989:905; Tanner 1987;
cf. Hill 1982, Foley 1982). Such social investment of re-
sources is consistent with highly altricial and late-
maturing offspring and a distinctive sexual division of
labor.^ As described by Foley (1984:99-103,- 1987; 1989,-
Foley and Lee 1989) and Kurland and Beckerman (1985),
the conditions faced by hominids in savanna environ-
ments included patchy resource distribution and the
need for complex and flexible foraging strategies, a broad
dietary niche, high mobility, and protection against
predators. Divergent interpretations proliferate concern-
ing subsistence patterns and social organization in hu-
man evolution, for example, concerning the importance
of hunting and carnivory versus scavenging (e.g., Isaac
1984; Potts 1987, 1988; Shipman 1983, 1986; Tanner
1987; Binford 198 1, 1983, 1989). Hyperscepticism aside,
it remains quite likely, at least for Homo, that coopera-
tive strategies including significant reliance on hunt-
ing were pronounced over a significant portion of hu-
man evolutionary history (e.g.. Hill 1982; O'Connell,
Hawkes, and Blurton Jones 1988a, b; Foley 1982; Ni-
tecki and Nitecki 1987). Isaac's (1978) scenario of band
dispersal and reaggregation for protection and the shar-
ing of food and information, while unlikely to be appli-
cable to all hominids, is likely to remain important for

5 . Frequent separation of males from their mates — entailed by dis-
persed food resources and the sexual division of labor — offers the
opportunity for sexual cheating and presents a threat to paternity
certainty. Under such conditions, a man risks "wasting" his paren-
tal investment on someone else's offspring even if he limits his
provisioning to his mate's children. How intensive male provi-
sioning could have evolved under conditions of sex-biased diurnal
dispersal is thus a key question for many models of hominid be-
havior.

an evolutionary understanding of H. sapiens and possi-
bly Homo in general.

In all ethnographically known simple societies, coop-
erative sharing of provisions is extended to mates, off-
spring, and many others within the band. Communal
sharing of food, especially of prized food items such as
meat, is highly developed and indeed crucial, serving,
among other things, as a collective means of ecological
risk reduction (Cashdan 1980, 1985; cf. 1990; Smith
1988). This sharing takes place well outside the range of
immediate kin, viz., among the diverse array of kin and
nonkin who constitute the typical residence group of
25 -f- persons.^ Archeological evidence suggests that
widespread networks facilitating diffuse access to and
transfer of resources and information have been pro-
nounced at least since the Upper Paleolithic (Gamble
1982; 1983; 1986:322-42; Whallon 1989).

Ethnographers of simple societies have consistently
emphasized the social and cultural as well as ecological
importance of sharing; diffuse distribution of major food
items (and other articles of value) is a prominent if not
preeminent cultural value (e.g., [!Kung] Marshall 1979,
Wiessner 1982; [Mbuti] Turnbull 1 961, 1965a, b; [Mbuti
archers/spearmen] Harako 1976:76-79; [Inuit] Balikci
i97o:chaps. 5 and 6; Guemple 1972; Damas 1984;
[Guayaki/Ache] Clastres 1972, Kaplan and Hill 1985a,
Kaplan et al. 1984). The functional significance of meat
sharing is obvious: game is seldom a dependable single-
family resource and yet is in some form a crucial dietary
component in virtually all simple societies (see Eaton
and Konner 1985, Hill 1982). Given that game is often
procured in large units but only sporadically obtained
by a given hunter, sharing effectively increases average
daily food consumption (see Kaplan and Hill 1985a).
Marshall (1979:363, 357) writes, 'The !Kung are quite
conscious of the value of meat-sharing and they talk
about it, especially about the benefit of the mutual obli-
gation it entails. The idea of sharing is deeply implanted
and very successfully imposes its restraints. . . . The idea
of eating alone is shocking to the !Kung. It makes them
shriek with an uneasy laughter. Lions could do that,
they say, not men.'' Testart (1985, 1987) has docu-
mented in detail the variations of food sharing among
hunter-gatherers worldwide. Food sharing in simple so-
cieties is both an index of cooperation and a key symbol
of what it is to be human.- More generally, an ethic of
communalism and equal access to resource production
is highly developed (e.g., Ingold 1988:277-85; Lee 1988,
1990; Leacock and Lee 1982).

A more complex and in some ways different ethic
emerges increasingly in middle-range human socie-
ties — food-storing and sedentary foragers, horticultur-

6. The early generalizations of Lee and DeVore (1984 [1968)) con-
cerning residential flexibility and bilateral group cooperation
among hunter-gatherers remain more applicable to simple societies
than to the much broader sample of all hunter-gatherers (contrast
Ember 1978); see, for example, [Mbuti] Turnbull (1965(3, b, 1984),
[Hadza] Woodburn (1972, 1984), [IKung] Lee (i979i>:chaps. 5 and
6), and [Eskimo] Balikci (i97o:chaps. 3-6), Damas (1984), and
Guemple (1972).

394 I c

URRENT ANTHROPOLOGY Voluuie ^2, Numbei 4, August-Octobei 1991

alists, prestate ranked societies, and chiefdoms. Though
the range of variation is enormous, group size and popu-
lation density tend to increase, and the individual's so-
cial universe becomes more differentiated (e.g., Johnson
and Earle 1987: chaps. 5-13; Upham 1990b; Carneiro
1981; Testart 1982; Haas 1990). There is increasing con-
centration and investment of labor in long-term land
and domesticated-animal resources, residential struc-
tures, and material artifacts. Correspondingly, there is
increased benefit in forcefully defending domesticated
resources and property — and in forcefully obtaining
them from other groups — while the costs of defending
them decrease. Extradomestic spheres of resource alloca-
tion become much more clearly defined. Plog (1990:193)
points to ''the development of a more restricted form
of sharing, producing a smaller, more formalized social
group composed of a limited number of households.'' As
Sahlins (1972) might put it, spheres of domestic general-
ized exchange and reciprocity are increasingly circum-
scribed and complemented by those of balanced and neg-
ative reciprocity between larger and opposed social
units. Even the giving of large gifts and the establish-
ment of alliances (Braun and Plog 1982) commonly take
on an aura of political competition, if not outright ag-
gression (Mauss 1970). Moreover, reciprocation of non-
giving and antagonism becomes institutionalized among
larger and better-defined groups of people. In short, clan
and ethnic self-interest, competition, and interrelation-
ship all become elaborated and developed, creating ''in-
creased potential for social conflict" and leading to "the
evolution of a group ideology consistent with increased
territoriality" (Plog 1990:193).

This is not to neglect the importance of intergroup
exchange and alliance in small-scale sedentary societies
(Braun and Plog 1982), nor is it to suggest that personal
or group self-interest, possessiveness, or dispute is ab-
sent in simpler ones. In simple societies, however, the
development of exclusive social identity and ethnocen-
trism tend to be opposed rather than cultivated and rein-
forced by deep-seated cultural norms and values. Norms
of generalized and diffuse reciprocity are culturally and
behaviorally foregrounded. This cultural emphasis is
particularly viable for demographic and ecological rea-
sons, since awareness of and confrontation with persons
beyond the sphere of social cooperation are limited. In
complex hunter-gatherers and small-scale sedentary so-
cieties, in contrast, negative reciprocity and violence
tend to be valued as an important dimension of in-
tergroup relationships.

Resource sharing in simple societies also contrasts in
significant ways with the patterns found among our
closest nonhuman relatives, the great apes (Pongidae) —
chimpanzee, bonobo or pygmy chimpanzee, gorilla, and
orangutan. The main sharable material resource among
these nonhuman primates is food. Foliage and fruit are
infrequently shared, except by mothers with their in-
fants. Gorillas eat predominantly foliage (86%) and con-
sume virtually no prey (Jolly 1985:47). Even gorilla in-
fants have never been seen being given solid food by
their mothers (Fossey 1979:167). Orangutan resource

sharing is also minimal (cf. Edwards and Snowdon 1980).
Large ground-foraging orangutan males have on rare oc-
casions been known to carry termite-infested logs into
trees to share with female consorts (Jolly 1985:450).
There is apparently only one report of orangutan meat
eating, and this was carried out in a strikingly individu-
alistic manner (Sugardjito and Nurhuda 1981).^

Common chimpanzees share vegetable resources and
fruit more actively than orangutans and gorillas, but the
vast majority of this sharing is between mother and in-
fant (Hiraiwa-Hasegawa 1990, Nishida 1970, McGrew
1979b). Chimpanzees are distinctive among apes for
their hunting; an estimated 10% of the time spent feed-
ing is devoted to mammalian predation (Teleki
1981:327), which is often cooperative among several
individuals, especially males (Goodall 1968; 1971;
i986:chap. ii; see also Boesch and Boesch 1989). Meat
consumption among Gombe chimpanzees is estimated
to average 27 g per individual per day, this being about
one-tenth the per capita meat intake of IKung San (Tel-
eki 1981:327). The degree to which chimpanzees will-
ingly share meat with each other has been a matter of
some contention (Teleki 1973,- McGrew i979b;453-57;
Wrangham 1975). On balance, however, there is distinct
competition and possessiveness in meat "sharing," as
is illustrated in GoodalPs (1986:299-300) compendium
volume on Gombe chimpanzees:

Meat is a highly coveted food and often there is in-
tense aggressive competition around a kill. This ag-
gression comprises [a] attacks on possessors of meat
by those who have none, {b) attacks or, more usu-
ally, displays or threats by possessors toward the in-
dividuals trying to share their prey, (c) attacks or
threats directed by those who have not managed to
acquire portions toward lower-ranking individuals
who are also trying to get some meat. Wrangham
(1975) .. . found that of the twenty attacks made on
individuals in possession of meat, the aggressor was
successful only three times. ... As Wrangham has
pointed out, an individual in possession of meat is
usually well able to defend it by crouching over it
and protecting it with arms and body (much as a
mother protects her infant when under attack).
Moreover, the possessor is usually highly motivated
to retain his prize. Even a low-ranking male may
hold onto his prey in the face of intense aggression.

Chimpanzees who successfully hunt game alone also
consume it alone unless discovered. Further, "an adoles-
cent or low-ranking captor, or a female, is likely to lose
possession of the carcass within moments of capture"
(Goodall 1986:299).

Nonetheless, as McGrew (1979a) and de Waal (1989a)
have suggested, there are some ways in which chimpan-

7. As Jolly (1985:62) describes the incident, "A female [orangutan)
completely consumed the carcass of an infant gibbon, taking 137
minutes to do so. (She may have found the gibbon dead, because
there was no blood.) An adult male she was consorting with stared
at her fixedly for the first hour, but she did not share with him — an
index of the antisocial habits of female orangs."

KNAUFT Violence and Sociality in Human Evolution \ 395

zees are preadapted to the food-sharing behavior found
in simple human societies. Dominant adult males in
particular may share food to reduce social conflict, since
such conflict can involve loss of the food item to a
scramble of others, injury, and/or disruption of impor-
tant male-male coalitions upon which male status fre-
quently depends. Recent experiments with captive
chimpanzees by de Waal (1989a) suggest that turn taking
in the exchange of social favors prevents one-sided accu-
mulation of benefits and that individuals who are reluc-
tant to share have a higher probability of encountering
aggression. Further, there is some tendency among
chimpanzees to suspend dominance relations and intro-
duce turn taking in reciprocating grooming for access to
food — at least under experimental conditions when the
concentrated food is only moderately attractive and
given to a single-male group. Even in free-ranging condi-
tions, food sharing does not always occur along lines
predicted by the dominance hierarchy (Nishida 1990:28;
Takahata, Hasegawa, and Nishida 1984).

Sharing of floral resources is more common and toler-
ated among bonobos (see Furuichi 1989; Kuroda 1984;
Kano 1983, 1990; Badrian and Badrian 1984; Badrian and
Malenky 1984). Vegetable food is commonly shared be-
tween males and females, ''and there are descriptions of
females clambering onto the shoulders of males to reach
overhead delicacies'' (Goodall 1986:485). However, ago-
nistic encounters remain most frequent ''during feeding,
directed from a high-ranking male to lower-ranking
male'' (Susman 1987:79). Though mammalian predation
appears much less frequent than among common chim-
panzees, forest bonobos do kill and share faunal re-
sources. Males' sharing of meat with females often in-
volves or leads to copulation, females being "allowed to
take parts of the carcass after copulating with the male
possessors." The distinctive genito-genital rubbing that
takes place between female bonobos is also frequently
observed during food sharing and has been interpreted
as a tension-reducing as well as a food-solicitation
mechanism in this context (Kuroda 1980, Thompson-
Handler, Malenky, and Badrian 1984, Kano 1984a, de
Waal 1987).

Even among bonobos, however, possessors of food
(except for mothers vis-a-vis their infants [Kuroda
1984:312]) never approach those without it. When in the
presence of others, bonobos possessing food shared it on
their own initiative less than 1% of the time (p. 317).
Food sharing — perhaps more accurately termed "food
transfer" — occurred particularly through active begging
as part of a "food taking bout." Bonobo food transfer
of prized fruit items parallels the common-chimpanzee
transfers of meat described above (Kuroda 1984:303-4),
and Kuroda notes (p. 306) that allowing others to share
in a plentiful food source is perhaps better described as
"co-feeding" than as "sharing."

In general, it appears that great apes do not willingly
share food, especially scarce food, with conspecifics
(other than offspring). Bonobos constitute a partial ex-
ception. Common chimpanzees may be preadapted to
food- sharing behavior through elaborate communication

concerning food, begging, tolerance of scrounging, and
ability to mitigate competitive behavior in the context
of collective eating. As Itani (1988) suggests, this may
represent an incipient ability to develop the more egali-
tarian sharing relationships found in simple human soci-
eties. There remains, however, a significant qualitative
difference between food transfer among free-ranging
great apes and the rule-governed sharing of food and
other valuable resources in simple human societies. The
strong internalization of a sharing ethic is in many re-
spects the sine qua non of culture in these societies (see
also Ingold 1987:114).

Dominance Structure

Perhaps the most striking thing about simple human
societies is how decentralized they are. Instead of indi-
viduals' striving to be "first among equals," aggressively
assertive, or powerful — striving to be big-men — there
tends to be active and assiduous devaluation of adult
male status differentiation and minimization or de-
nial of those asymmetries of ability that exist. Self-
aggrandizing behavior is disparaged and open coercion
considered highly improper (see [!Kung] Draper 1978;
Marshall 1976; Lee 1979b, 1982; [Inuit] Briggs 1970,
1978; [Mbuti] Turnbull 1961, 1965a, b, 1978; [Semai]
Dentan 1978, 1979; Robarchek 1977,- Robarchek and
Dentan 1987; [Hadza] Woodburn 1979). Leadership is ru-
dimentary and uninstitutionalized, and political life is
communal. Patriarchy and elders' authority are mini-
mal, and leadership is itself rarely a matter of assertion,
dispute, or competition. Decisions are most frequently
reached through casual consensus, in which no man has
authority over another. Major collective enterprises tend
to emerge spontaneously as the result of myriad fluid
conversations that mix stories, banter, fantasies, and
plans. As Ingold (1987: chap. 9) points out, politics in
simple societies maintains a fine balance between indi-
vidual autonomy and the collective appropriation of
nature.

Among complex hunter-gatherers and with the advent
of sedentism and horticulture/agriculture, male status
differentiation increases (see archeological evidence in
Price and Brown 1985, Zvelebil 1986, Bailey 1983; see
Price 1984). Though the range of variation is enormous
and the dynamics complex (e.g., Feinman and Neitzel
1984), the progressive development of overt status differ-
entiation and competitive leadership is apt to be influ-
enced by the functional requisites of coordinated action
among larger social groups in sedentary societies (e.g.,
Upham 1990a). Property and possessions increase with
sedentisni, and the potential for material wealth differ-
entials also increases (Johnson and Earle 1987). Opportu-
nities increase for selective control of the flow of infor-
mation and material resources and the development of
social inequality (Bender 1990). For both political and
economic reasons, competition over access to positions
of control intensifies, and various forms of leadership
and status hierarchy emerge: gerontocracy, ritual elder-

396 I CURRENT ANTHROPOLOGY Voluuie 32, Number 4, August- October 1991

ship, pronounced age-grading, headmen, war leaders,
priests, big-men, and chiefs. Such status distinctions be-
come progressively formalized in the evolution of politi-
cal society prior to the rise of the state (e.g.. Fried 1967).
With increasing socioeconomic complexity and hierar-
chy, administrative control and organization become
crucial, along with unequal control of the means of pro-
duction (e.g., Claessen and Skalnik 1978, Friedman and
Rowlands 1978). In lieu of more nuanced description, it
may be said that actively competitive male leadership
hierarchies have been common in human political evo-
lution since the advent of complex hunter-gatherers and
sedentary societies.

All of the great apes have male dominance hierarchies
(Wrangham 1987). This pattern is thrown into relief by
the relative lack of dominance hierarchies among great-
ape females (de Waal 1986). Male dominance over fe-
males is the norm, and with the notable exception of
bonobos (F. White 1989, Furuichi 1989) males are the
focal points of great-ape group cohesion.

Gorilla group structure is dominated by a single adult
male (Jolly 1985:132). Between 64% and 77% of gorilla
groups have but a single silverback (Harcourt, Fossey,
and Pi Sabater 1981), and the silverback is the focus of
a cohesive and long-lasting group of females, juveniles,
and, to a lesser extent, subadult males (blackbacks)
(Schaller 1976, Fossey 1983). As Harcourt (1979:189)
summarizes.

In groups in which there was more than one sil-
verback, one (the oldest) was clearly the leader, and
the other(s) occupied peripheral positions. The lead-
ing silverbacks were the protectors of the group,- it
was they who most frequently stopped intragroup
fights; it was they who most frequently controlled
the timing and direction of group movement; and it
was the leading silverbacks around whom most ani-
mals most often congregated, even when the groups
were undisturbed. Only silverbacks, not blackbacks,
were seen to mate with primiparous and multipa-
rous females.

Among orangutans, there is also clear domination of
younger males and females in consort by full adult
males, though there is no bounded social group over
which such males preside. Adult orangutan males are
characterized (MacKinnon 1979:265) ''by the full devel-
opment of secondary sexual adornments including long
hair and beards, full throat pouches, high fatty crowns,
expanded cheek flanges, and large size. These animals
indulge in loud, long-range calling, defending range
boundaries against other males.''

Chimpanzee males exhibit several clearly distinguish-
able dominance levels, including alpha, high, middle,
and low (Takahata 1990^, Bygott 1979). Dominance is
established ultimately through aggressive displays, en-
counters, and fights. Longitudinal studies of rank (Taka-
hata 1990^; Bygott 1979; Goodall i986:chap. 16) docu-
ment a variety of strategies for intimidating opponents
and gaining status. Particularly distinctive are chimpan-
zee male-male coalitions to challenge the dominance of

higher-ups, especially the alpha male (see also Takahata
1990b, Kawanaka 1990, de Waal 1982). In contrast to
those of many monkey species, chimpanzee adult males
rarely if ever transfer out of their home groups, even
when dethroned; transfers are limited to females and
their offspring (Nishida 1979; Goodall et al. 1979; Good-
all 1986:489). Though normal assertions of dominance
are responded to with pant-grunting or displacement and
other encounters may result in aggressive display and/or
chasing, genuine challenges at the top of the hierarchy
frequently result in severe wounds and permanent injur-
ies. Death from such wounds within free-ranging chim-
panzee groups is strongly suspected in several cases
(when the wounded male disappeared abruptly and was
never found) and has been documented definitively
among a large captive group of chimpanzees (de Waal
1982; i989b:chap. 2).

Male dominance hierarchies exist among bonobos, but
there is a lower incidence of male-male affiliative and
agonistic behavior, and male-male coalitions apparently
do not occur (Kano 1990, Badrian and Badrian 1984, Fu-
ruichi 1989, F. White 1990). In contrast to that of com-
mon chimpanzees, bonobo social organization can be
considered "female-based," with "strong affiliation
among females and between males and females, but not
among males" (F. White 1989:204; see also Furuichi
1989). This pattern poses a significant anomaly for Pan-
based models of male-collective intergroup aggression
in hominid evolution. Bonobos form quasi-stable sub-
groups within a cohesive troop with a shared home
range. Males rarely if ever transfer out of this larger
group. Patterns of bonobo aggression may be quite pro-
nounced (de Waal i989b:22i):

Although the remarkable gentleness of the bonobo
species has been noticed by other investigators, we
should also realize that, until a decade ago, the same
opinion prevailed with regard to gorillas and chim-
panzees. Now we know better. Recently, Takayoshi
Kano [1984b] published a rather shocking report on
physical abnormalities in free-living bonobos. An as-
tonishing number of them lack fingers, toes, even
entire hands. Two-thirds of the males and one-third
of the females show limb abnormalities. . . . The
higher incidence of missing or deformed body parts
among males, especially adults, supports a link with
aggression. And the tendency of bonobos to aim
bites at the extremities may explain the nature of
the defects.

Damaging fights have not, however, been witnessed (de
Waal 1 9 89b: 22 1— 22), and techniques of managing con-
flict seem highly developed; in this respect, bonobos
may exhibit important similarities with simple human
societies in harboring both strong conflict-mediation
skills and the potential for rare but ultimately extreme
violence.

In general, and again with the partial and/or possible
exception of bonobos, great apes exhibit marked male
dominance hierarchies maintained and altered by
aggressive display, fighting, and violence. Sub-alpha-

KNAUFT Violence and Sociality in Human Evolution \ 397

male-male coalitions are marked among common chim-
panzees but largely absent in other ape species. Among
gorillas and chimpanzees males are the focus of group
cohesion; bonobos are more female-based, and cohesion
among orangutan adults is limited to mating dyads.

In selected formal respects, dominance hierarchies in
middle-range human societies are more similar to those
of great apes (their other huge differences notwithstand-
ing) than to those of simpler human societies. In both
great-ape and middle-range human societies, competi-
tive dominance relations entail concrete, behaviorally
regularized relations of superordination and submission
among adult males and systematic inequalities in access
to valued resources. Hierarchical relations are ulti-
mately negotiable and subject to coalitional strategizing.
Dominance hierarchies are an important and often pre-
eminent male concern. In the simplest human societies,
by contrast, cultural emphasis is on precluding rather
than facilitating male status distinctions. The develop-
ment of the capacity for forming male hierarchies is
effectively undercut by a combination of ecological,
social, and cultural forces. Conversely, if, as de Waal
(1989^, b) has suggested, chimpanzees and bonobos have
the capacity to suspend hierarchies temporarily in shar-
ing and reconciliation, the particular conditions of their
evolution have not had much leveling effect on male
dominance hierarchies in free-ranging populations. As
Whallon (1989:448-49) notes.

If we adopt a basically primate model (more specifi-
cally ape and in particular chimpanzee) ... we can
see, among other things, that the typical pattern of
dominance relations among individuals . . . would
not be particularly adaptive in environments of low
resource density and predictability. It would prevent
an effective internal distribution of resources in the
group, which therefore could not take advantage of
the risk reduction available with division of labour
and particularly with separation of labour.

As a result, he concludes, the late Paleolithic saw the
culmination of an evolutionary trend toward the ''re-
placement of ape-like systems of interpersonal domi-
nance established through relatively constant display,
combat, and trial and error, by systems of at least rela-
tively egalitarian, stable, and reliable relations of rights
and obligations among individuals both within and be-
tween local groups.''

The available evidence thus suggests that the trajec-
tory of male status differentiation in human evolution is
U-shaped rather than linear. Correspondingly, similarity
between great apes and middle-range human societies in
terms of competitive male dominance hierarchies may
be based on analogy rather than homology, in contrast
to the assumptions of current sociobiological reasoning
(e.g., Betzig 1986, Betzig, Borgerhoff Mulder, and Turke
1988; cf. also Wrangham 1987, Manson and Wrangham
1 99 1, Tooby and DeVore 1987). The tenuousness of such
analogies is underscored by contrasting patterns among
simple human societies. It is quite possible that patterns

analogous to those of simple societies characterized a
significant portion of our evolution as H. sapiens.

As we have seen, the relative absence of male domi-
nance hierarchies in simple human societies does not
preclude the occurrence of severe male violence. The
interface between this violence and intense sociality is
all the more in need of explanation in that it is not easily
accounted for by theories often used to explain the col-
lective, reciprocating conflict common in middle-range
societies (see Knauft 1987^). To probe this question, it
is necessary to consider male-female relations and, in
particular, evolutionary contrasts in patterns of male
sexual competition.

Male Competition for Sexual Access
to Females

The conflict between sexual self-interest and coopera-
tive affiliation must have been profound during the early
evolution of human social organization. Distinctive fea-
tures of the social evolution of H. sapiens if not Homo
generally in all likelihood included high male parental
investment, a significant sexual division of labor, the
sharing of valuable food throughout a group of 25 + re-
lated and unrelated individuals, and exchange of infor-
mation and flexible access to resources among several
such dispersed but interlocking forager groups. This
form of organization provided crucial adaptive advan-
tages for humans relative to social carnivores, solitary
and herd herbivores, and omnivorous nonhuman pri-
mates. As noted above, however, the regular dispersal
of individuals to exploit patchy or unpredictable food
resources made mating exclusivity and paternity cer-
tainty problematic. The human solution to this problem
is arguably unique (contrast Foley 1987).

As Freud and the early Levi-Strauss elucidated, human
societies have developed symbolically mediated social
and psychological constraints on individuals' sexual im-
pulses,- sexual control in humans is both psychically in-
ternalized and socially and symbolically maintained. It
may be suggested that the threat posed by immediate
and disruptive self-interested sexual behavior in the evo-
lution of H. sapiens was to a significant extent, though
by no means totally, countered by cultural prescriptions
fostering intra- and intergroup cooperation.

How this process may have arisen out of processes of
standard natural selection is an intriguing question to
which the work of Boyd and Richerson (1985) provides
important insights. If imitative learning is strongly se-
lected for as a rule-of-thumb adaptive strategy — as it is
likely to have been among early humans, a species de-
pendent upon prolonged socialization — then phenotypic
traits may spread very rapidly. Added to this, and no
doubt in causal concert with it, is the highly developed
ability of humans to communicate through elaborate
symbolic communication. As has recently been sug-
gested by Goodenough (1990), Bickerton (1990), Steklis
(1985), and Marshack (1989^3), the propensity toward
elaborate symbolic communication was probably highly

398 I CURRENT ANTHROPOLOGY Voluuie 32, Nuuibei 4, August- October 1991

developed even if not fully equivalent to modem lan-
guage during the evolution of archaic H. sapiens and
possibly H. er actus. ^

8. This assessment is but little compromised by suggestions that
the fully modern repertoire of human linguistic and symbolic capa-
bilities may not have arisen until 45,000-35,000 b.p. (e.g., Mellars
and Stringer 1989, Lieberman 1984, Chase and Dibble 1987, David-
son and Noble 1989, Binford 1989). The fallacy of linking symbolic
facility narrowly with relatively elaborate archeological evidence
of ritual and art has been exposed by Lindly and Clark (1990), who
point out that Upper Paleolithic assemblages until about 20,000
B.p. typically lack evidence of elaborate symbolism (p. 239; cf. also
Dibble 1989). Gamble (1990:243) notes that even the colonization
of North America and the North European Plain after 13,000 b.p.
are "marked by very few if any" directly symbolic artifacts. Obvi-
ously, language and protolanguage were important in human evolu-
tion long before the regular encoding of symbolism in durable arti-
facts (cf. Bickerton [1990:176] on the "perils of fossilism" in the
study of language evolution). Clear evidence exists for some form
of complex symbolic capacity and corresponding behavior at least
as far back as H. erectus (Marshack 1985, i989<3), including beads,
animal-bone pendants, and other elaborate decorative items in
Mousterian assemblages (e.g., R. White 1989) and fire-hardened
hardwood digging sticks, post holes, ocher crayons, and huge
amounts of red ocher powder in association with Acheulian finds,
some dated to several hundred thousand years b.p. (Marshack
i989<3).

That some of this evidence for human symbolic capacity may
indicate supernatural belief and ritual behavior is particularly sig-
nificant in evolutionary terms (e.g., Solecki 1971, 1975; Trinkaus
1983; cf. Wreschner 1981; Marshack 1981, 1989b). As Rappaport
(i97i<3, b, 1979) has suggested, ritual reinforces and insulates from
argument shared propositions about the world and fosters deep-
seated cognitive acceptance of and behavioral compliance with
these propositions. Religious belief and ritual emphasize broad so-
cial objectives and cultural values, often advancing group goals as
opposed to more immediate individual self-interest. Self-interest is
of course in common if not constant tension with group-level goals
and constraints, but such goals and constraints have themselves
long exerted an important influence on human behavior and on
the cognition and motivation that underlie it. Among humans,
internalization of values that propagate group interests is likely to
have been important for social functioning from an early period,
as Durkheim (1965 [1912]) long ago suggested.

These issues are sometimes overshadowed by current debates
about "replacement" versus "continuity" hypotheses concerning
the preponderance of anatomically modern humans in Europe at
45,000-35,000 B.p. (see Trinkaus 1989, Smith, Falsetti, and Don-
nelly 1989). The remarkable spread and socioecological adaptation
of H. sapiens prior to this time argue for some kind of highly devel-
oped linguistic facility. Of particular present relevance is that the
faunal data "imply that sharing was a regular feature of Middle
Paleolithic economic life" (Chase 1989:333). It appears quite likely
on both conceptual and anatomical grounds that Middle Paleolithic
populations had some elaborate form of speech (Marshack 1 989*3,
Bickerton 1990; see Arensburg et al. 1989 concerning anatomical
evidence). As Whallon (1989:450) notes, "it seems most unlikely
that fully developed language capacities could have emerged from
other than an already evolving system of symbolic communication
(cf. Bickerton 1981:261, et passim]. It seems equally unlikely that
kinship, for example, as an organizing principle in cultural sys-
tems, could have emerged from an organization entirely lacking in
the definition of social roles and positions."

Bickerton (1990: chaps. 6 and 7) has recently suggested that com-
plex protolanguage was likely characteristic of H. erectus and that
fully syntactic language was characteristic of archaic H. sapiens.
The existence of some complex but not fully modern protolanguage
does appear consistent with trends characteristic of H. erectus such
as increased encephalization, increased subsistence flexibility,
greatly increased home-range size, and habitation dispersal to non-

Through socialization and symbolic communication,
behavioral traits can be learned and spread — even when
they are not maximally adaptive — faster than they can
be eradicated through biogenetic selection. This creates
the potential for the temporary spread of maladap-
tive customs in a human population — what Durham
(i99i:chap. 7) calls cultural-genetic opposition. In addi-
tion, however, it allows the spread of behaviors that can
favor the sociocultural group at the expense of a given
individual's inclusive fitness (contrast Hamilton 1964).
As a result, with extensive symbolic transmission —
itself initially favored by standard Darwinian se-
lection — group selection and genuine altruism become
possible. Put differently, biogenetic selection gives rise
to a cultural transmission process that ultimately be-
comes partially decoupled from it. This does not con-
demn humans to biogenetic dysfunctionality,- it simply
adds the cultural group to the individual as a separate
and competing unit of selection.

The assumption of many researchers that the indi-
vidual is the unit of selection in human evolution is
increasingly being challenged. From ethnography and
demography, queries are raised by the existence
of behavior patterns that systematically compromise
rather than maximize the reproductive success of indi-
vidual actors (e.g., Vining 1986; Knauft 1987^, b, 1989^;
Moore 1990). From social psychology, sociobiological as-
sumptions have been questioned on the basis of experi-
ments that document the existence of altruistic tenden-
cies toward strangers in the absence of rationally
expected payoffs (e.g., Caporeal et al. 1989, Dawes, van
de Kragt, and Orbell 1988). And from theoretical biology,
questions are raised by increasing realization that the
conditions under which self-interested reciprocal altru-
ism can explain cooperation are more restrictive than
previously thought (e.g., Boyd and Lorberbaum 1987). At
the same time, there is increasing recognition that group

tropical ecozones. Protolanguage is also consistent with the com-
plex but relatively standardized Acheulian tool kit. The increased
technological sophistication, openness, and productivity of Mous-
terian disk-core technology, by contrast, may reflect cognitive ca-
pacities consistent with grammar and syntax in archaic H. sapiens.
A further consideration of the Middle Paleolithic transition, how-
ever, might suggest that the speech of archaic H. sapiens included
syntax and productivity but lacked full duality of patterning (cf.
Tomasello 1991; Parker 1985:624-25; Parker and Gibson 1979;
see Hockett and Ascher 1964). Relative to Upper Paleolithic ones,
Mousterian assemblages were arguably undeveloped in "third-
order" objectification — blanks made for specific parts of elabo-
rately fashioned composite tools and the careful making of tools
such as burins that were used to make other tools. Conceptually
analogous, syntactic language without duality of patterning lacks
the intralexical productivity that recombines word-making sounds
to increase the number and diversity of linguistic referents.

Though specific details of language evolution are subject to de-
bate, it appears highly likely from a broader perspective — even ne-
glecting the potential for species continuity and the origin of
anatomically modern humans approaching 100,000 b.p. (Smith,
Falsetti, and Donnelly 1989; cf. Mellars 1988, i989<3) — that elabo-
rate forms of linguistic communication facilitated the remarkable
development and spread of humans from the time of archaic H.
sapiens and probably since H. erectus.

KNAUFT Violence and Sociality in Human Evolution \ 399

selection may help explain the existence of widespread
reciprocity and altruism (e.g., Boyd 1988, Boyd and Rich-
erson 1990^, Rogers 1990) and that cultural transmis-
sion may operate through selection parameters that are
nonbiogenetic, nonindividualistic, and/or superorganic
(e.g., Boyd and Richerson 1985, 1990b; Richerson and
Boyd 1989; Harpending and Rogers 1987; Nowak and
Sigmund 1989; Durham 1991; see more generally Wil-
son and Sober 1989, Wilson 1989; contrast Hamilton
1964; Trivers 1971, 1985; Axelrod and Hamilton 1981).

In the same way that, from a neo-Darwinian perspec-
tive, the individual has been suggested to be a gene's
way of producing another gene (e.g., Dawkins 1989), so
too an individual in a human cultural environment is to
a certain extent a culture's means of reproducing and
spreading a given set of symbols — propagating a collec-
tive symbolic system. What results is complex coevolu-
tion of distinct behavioral transmission and selection
systems, one symbolic, the other biogenetic (Durham
1 99 1, Boyd and Richerson 1985). The increased speed
and elaboration of cultural as opposed to biogenetic
transmission processes render symbolic effects increas-
ingly influential on behavior over time, but the tension
between them is never eliminated. Robert Paul (personal
communication; cf. 1987) has suggested that a kind of
''arms race'' emerges between cultural constraints upon
sexual impulses and biogenetic selection pressures that
encourage mating. This may well have made humans
among the sexiest and simultaneously the most sexually
repressed of species.

Cultural control of sexual impulses has never been
complete; threats to cooperation and integrity posed
by illicit sexuality and its concomitant disruption have
always been present. The unprecedented transmission
of behaviors and mental motivations through cultural
means, however, has resulted in significant group-
cooperative control over sexuality. Rule-governed cul-
tural control of sexual behavior, even if partial, could
facilitate cooperation and the sexual division of labor,
thus increasing the survival rate of the group.^

In simple human societies, cultural appropriation of
sexuality is crucially evident in (i) the institution of
marriage, (2) frequent monogamy despite collective
group living, (3) widespread classificatory extension of
the incest taboo (elaborate exogamy), and (4) marital ex-
change and alliance.

Marriage confers rights of sexual access. Though these
norms are not uncommonly broken in simple societies
(e.g., Shostak 1981), sexual norms do circumscribe sex-
ual activity for most persons to a great degree (e.g., con-
trast, for !Kung, Lee i979b:chap. 6). Simple societies of-
ten reduce the discrepancy between norm and reality by
making the establishment of marriage socially flexible
(e.g., Lee i979b:chap. 6; Balikci 1970:153-62; Wood-

9. Boyd and Richerson (1990^:124) suggest that group selection for
cooperative behavior is most likely for hard-to-learn traits, and
Rogers (1990:408) suggests that group selection via selective emi-
gration is facilitated by mobility rather than by isolation of local
groups.

burn 1972:205; TurnbuU 1961). Premarital andpostmar-
ital sexuality are often tolerated, and in some cases ex-
tramarital sexuality is itself culturally controlled and
appropriated for alliance purposes — as in Eskimo wife-
exchange partnership (Balikci 1970:140-43). Finally,
norms of sexual fidelity are underscored by the violent
repercussions of breaking them. Though many simple
societies tolerate polygyny or do not actively proscribe
it, the cultural and statistical norm for most adults in
simple societies remains monogamy (Lee 1979b: 79;
TurnbuU 196s a, b). In contrast to the situation in many
of the more complex prestate societies, polygyny is not
a prominent index or symbol of differential adult male
status (contrast Flinn and Low 1986).

One of the most important sociopolitical dimensions
of cultural sexual control in simple societies is affinal
alliance predicated on rules of extended exogamy (e.g.,
Levi-Strauss 1969 [1949]). Though incest avoidance of
various kinds has also been documented among pri-
mates, the social and political importance of extended
elaborate exogamy is distinctive and enormous among
humans. Marriage in simple societies creates enduring
links of reciprocity, and coresidence, subsistence cooper-
ation, and/or sharing among close affines are highly de-
veloped (e.g., Lee i979b:chaps. 5 and 6; TurnbuU 1965^,
b; Damas 1984). Affinal cooperation, operating among
kin who are culturally created, is in many ways a sym-
bol of cultural viability. (In addition to affinality, simple
societies are characterized by a variety of extrakinship or
fictive-kin relationships — trade partnerships, namesake
relations, ritual/totemic affiliations, and kinlike classi-
ficatory relations — that link persons in extensive social
networks [e.g., Guemple 1972; Balikci i97o:chap. 6;
Wiessner 1982; cf. Yengoyan 1984]. Extensive social ties
have been inferred from archeological evidence at least
as far back as the Upper Paleolithic [e.g.. Gamble 1982;
1983; 1986:386-91; Whallon 1989]).

Though the importance of cultural constraints upon
sexuality can hardly be overemphasized, there remains
a poignant tension in simple societies between cultural
norms and individual desires. As Collier and Rosaldo
(1981) have noted, marriage and legitimate sexual access
to a woman are predominant markers of male adult-
hood, and these rights are fervently protected. Corre-
spondingly, much of the severe violence that does occur
is ultimately related to male sexual disputes over
women (Knauft 1987^:477). Displacement of affinal or
sexual tensions also appears to underlie much of the
seemingly irrational violence over ''trivial'' issues that
occurs in many egalitarian societies (e.g., Lee 1979^:
chap. 13; van der Steenhoven 1959; Hoebel 1964). Sexual
tensions in these societies are normally kept in check
by norms of affinal harmony, group cooperation, and
personal propriety. When they are ultimately galva-
nized, however, they are frequently quite intense. This
threat is exacerbated by the absence of political leaders
or dominant individuals who might exercise control and
of institutionalized or formalized redress mechanisms.
Indeed, violence occurs in significant degree to prevent
some individuals from acquiring sexual dominance.

400 I CURRENT ANTHROPOLOGY Volume 32, Numbei 4, August- October 1991

Violence in simple societies may be aggressively self-
interested, e.g., aimed at dethroning a rival and pro-
curing a woman from him as a spouse (e.g., Balikci
i97o:chap. 7). This violence may occasionally be suc-
cessful, but it tends to be disapproved of by the group at
large and may result in compensatory violence at its
initiative. Moreover, if the aggressor is successful in kill-
ing his primary rival, he may be considered an outcast
because of his breach of norms and may be unable to
obtain support even among his closest kin. In only i of
the 22 homicides listed by Lee (1979^:383) for the !Kung
did the killer run off with the victim's wife, and this
case is itself revealing; despite the fact that the couple
had previously been lovers, the woman was frightened
by the killing and as a result soon returned alone. In
most simple societies, aggressively self-interested per-
sons may be killed with the consent or active collabora-
tion of the community at large (see Lee 1979^: chap. 13;
Balikci i97o:chap. 9). This responsive violence can in a
sense be considered a form of execution or capital pun-
ishment (see Otterbein 1986, 1987). As Otterbein em-
phasizes, it tends to uphold norms (such as egalitarian-
ism and sexual propriety) that are crucial to group sur-
vival.

Often, however, the rationale for violence in simple
societies is not so clear-cut. Norms of sexual propriety
are internalized by the individual even though they con-
flict with strong sexual desires. It is thus not surprising
that violence often erupts suddenly in a displaced, dis-
torted, or noninstrumental manner. For instance, Lee
(1979^:397) states for the !Kung, 'This is a society in
which a high proportion of 'crimes' are 'crimes of pas-
sion.' It may be useful to consider !Kung fighting as a
kind of temporary insanity or running amok rather than
as an instrumental act in a means-end framework." Lee
notes further (p. 392),

The most striking aspect of the killer-victim rela-
tion ... is that in the majority of cases the victim
was not a principal in the verbal conflict that led up
to the actual killing with arrow or spear. In only 8
of the 18 cases on which I have data was the victim
a principal in the previous argument. In 10 other
killings, the victim was struck more or less at ran-
dom: in 3 cases a man came to the aid of another
and was killed; in 4 cases a peacemaker was fatally
wounded; and 3 victims (2 of them women) were by-
standers.

That !Kung are superb marksmen with spear and arrow
underscores the blind fury of the event. As Lee writes,
"!Kung men are excellent shots when hunting game, but
are poor shots when aiming at each other." This kind of
dissociated behavior is similar to the sudden passion of
Mbuti arguments (TurnbuU 1961), strongly dissociative
violence among Inuit (Rasmussen 1932:17; Briggs 1982),
and the much-contested "blood drunkenness" among
Semai (Dentan 1979:58-59; Robarchek and Dentan
1987). Cognitively displaced and projective aggression
may also occur as violent scapegoating of persons within
the community (Knauft 1985, 1987^).

Some forms of violence in simple societies thus appear
almost more dysfunctional than functional and bear at
least a passing resemblance to the syndrome of "phylo-
genetic regression" described by Bailey (1985, 1987; cf.
Neuman 1987). Bailey argues on neurophysiological
grounds that sudden aggressive outbursts stem from
phylogenetically "primitive" parts of the brain and char-
acterize various forms of genuine brain dysfunction.
While his claims about phylogenetic regression — that it
can explain myriad types of aggression and violence —
are too sweeping, his model may apply to certain forms
of simple-society violence, particularly those galvanized
by sexual tensions (see also Konner i982:chap. 9). Given
its phylogenetic and neurophysiological underpinnings,
the human aggressive potential would have been diffi-
cult if not impossible to suppress entirely, particularly
for males in the context of mating competition (see
Hamburg and Trudeau 1981, Blanchard and Blanchard
1984; cf. Valzelli 1981). The internalization of cultural
norms conflicts powerfully with this impetus, however,
and can exacerbate psychic conflict and dysfunctional
aggression.

It may be emphasized that symbolically mediated con-
trols on sexuality that facilitate group-adaptive be-
haviors do not replace or negate individualistic sexual
strategizing; biogenetic propensities for fitness maximi-
zation continue to be acted out. However, the enormous
competing influence of cultural norms in constraining
human sexual behavior — sociologically and psychologi-
cally as well as symbolically — cannot be ignored. To ar-
gue that violence in simple societies is driven by socio-
biological selection pressures simply because it often
relates in some fashion to sexual frustration is to miss
the distinctive impact of cultural rules upon sexual be-
havior and the implications of this for sociosexual con-
flict among humans. It is arguably the tension between
group-symbolic and individual-biogenetic imperatives
that is responsible for the distinctive features of socio-
sexual conflict in simple human societies. ^^

Middle-range societies on the whole exhibit much
greater emphasis than simple ones on explicitly targeted
competition between men over female sexuality. Cor-
respondingly, these societies on the whole provide
greater and more consistent sexual rewards for men who
are successful in aggressive male status competition. In
a significant number of known "tribal" and "chiefly"
societies, for instance, polygyny is both an index and a
reflection of status or rank. For instance, Feinman and
NeitzePs (1984:58-59) tabulation of leadership status

10. It could be countered that exceptionally intelligent individuals
can enhance their reproductive success while appearing to conform
to community norms. For instance, since the distinction between
homicide and capital punishment is often murky in simple socie-
ties (Knauft i987<3:49i-92), community norms could be strategi-
cally manipulated by self-interested individuals smart enough to
subvert them and get away with it. In the long run, however, this
would stimulate increasingly intelligent countermeasures by the
society at large. Sexual drive and its cultural control are thus at
once logically antithetical and mutually reinforcing, and rule-
governed control over sexuality is not reduced but increased by its
potential contravention.

KNAUFT Violence and Sociality in Human Evolution \ 401

prerogatives in 5 1 New World prestate societies reveals
that multiple wives are a concomitant of leadership in
39.2% (20/51) of the cases. Male wealth differentials
may be translated into differential reproduction through
the use of bridewealth payments to procure new wives.
Betzig (1986) has extended this general line of reasoning
to posit a link between despotism and differential repro-
duction. Despite the many problems with her argument
and counterexamples (e.g., Moore 1990; cf. also Knauft
i9Sja, b), a connection between male prestige and repro-
ductive success appears more plausible for these socie-
ties than for simple ones. It does appear that reproduc-
tive advantage is one of the concomitants — though by
no means the exclusive cause — of victorious status ri-
valry in many prestate societies in which male status is
overtly and vigorously contested (e.g., Chagnon 1979,
1988). Carneiro (personal communication, 1 981) has fur-
ther suggested that polygyny (and corresponding repro-
ductive variance) increases greatly as a prerogative of
leadership with the appearance of chief doms. Thus, for
instance, paramount chiefs frequently use wealth and
power to arrogate unto themselves a substantial number
of mates and/or wives.

Great-ape dominance hierarchies are strongly linked
to male competition over sexual access to females. Par-
ticularly among gorillas and orangutans, this selective
pressure has led to extreme sexual dimorphism and to
distinctive physical and display characteristics in adult
males (e.g., Rodman and Mitani 1987). Females of both
species show strong sexual preference for an adult as
opposed to a mature subadult male. In part because of
injurious competition between them, fully mature
males in these species seldom interact. In those confron-
tations that have been observed, the larger, stronger
male drives off his competitor. Violent fights typically
occur when one of the competitors does not quickly give
way. The following is an account for orangutans (Galdi-
kas 1979:213):

One such [encounter between mature males] oc-
curred when the resident male, TP, was in consort-
ship with a mature female, Priscilla. A visiting male
moved toward the couple. The resident male charged
toward him and a fight ensued, which lasted 25 min-
utes. Fighting consisted mainly of chasing and furi-
ous grappling, including biting at each other's face,
shoulders, and hands. . . . Finally, they separated and
sat facing each other 10 meters apart. The resident
male threw a snag [stripped branch] and called; the
other male then disappeared while the victor went
off in search of his receptive female (and her off-
spring) who had slowly moved away foraging while
the combat was taking place.

Mitani (1985) found that most male-male orangutan in-
teractions occur in the presence of females. Large males
actively disrupt the mating of other males, with the
highest-ranking adult male consistently displacing oth-
ers. A high frequency of male orangutans are physically
disfigured by fights, and it is quite likely that some ma-
ture males are effectively excluded from reproduction

or killed outright (e.g., Galdikas 1979:216, 230). Female
orangutans are strongly drawn to mature males in their
home ranges, and the forceful attempts of subadult
males to copulate with females (when an adult male is
absent) are typically resisted fervently by females them-
selves. What may be termed rape sometimes occurs in
such contexts (Galdikas 1979; see MacKinnon 1979).

Among gorillas, the cohesive social group and harem
of the silverback male is threatened by maturing black-
backs. These young males eventually leave their group,
become solitary until more fully adult, and then attempt
through competition to gain females from existing
groups and hence form social and breeding groups of
their own (Schaller 1976, Fossey 1983, Harcourt 1979).
Though adult gorillas tend to avoid each other, confron-
tation between them ultimately tends to be quite vio-
lent (see below).

Among common chimpanzees, the relationship be-
tween dominance and sexual access is less uniform.
Many males establish temporary consort relations with
females, and these tend not to be aggressively contested.
However, nonconsort mating — when receptive females
are at the peak of estrus in a group context — is very
frequent (Goodall 1986:442), and here aggressive male
sexual competition is pronounced (Tutin 1979). To some
extent, the chimpanzee alpha male or the highest-
ranked male present can monopolize a female in a group
context when she is at the peak of estrus (Goodall
1986:452, 472-73). This pattern, evident at Gombe and
among captive chimpanzees (de Waal 1982), has been
documented definitively among Mahale Mountain
chimpanzees, with alpha-male status conferring ''a great
reproductive advantage'' in copulation as well as nutri-
tion (Nishida 1970; 1990:28). Alpha males enforce a
markedly possessive mating pattern which allows them
to copulate at an exceptionally high rate with the most
fertile females near their time of ovulation (Hasegawa
and Hiraiwa-Hasegawa 1983, 1990; Takahata 1990a).

Systematic data about the relationship between bo-
nobo dominance patterns and mating are apparently not
yet available (Wrangham 1987:65). Bonobo sexual rela-
tions are particularly promiscuous and exhibit a high
frequency of copulation outside the period of maximal
female sexual swelling (Badrian and Badrian 1984, de
Waal 1987). Bonobos form subgroups comprising two to
eight males and one to eight females and may also form
consortships. Thus, ''many, perhaps all females have
close long-term relationships with particular males
within the community" (Wrangham 1986:368; see Fu-
ruichi 1989, F. White 1989, Kano 1984, Thompson-
Handler, Malenky, and Badrian 1984). The relationship
between this process and dominance is little known. Re-
cent research by Dahl and Nadler (1989:19) suggests that
mild male sexual aggression is involved in the high fre-
quency of bonobo sexual activity and that further study
is needed ''to determine whether this is intimidation, as
defined for the other great apes (Nadler 1988), coercion,
or persuasion." Preliminary evidence suggests that copu-
lation of bonobo males with high-ranking females is in-
deed related to their period of maximal sexual swelling.

402 I CURRENT ANTHROPOLOGY Voluuie 32, Number 4, August- October 1991

whereas noncyclic heterosexuality serves as a tension-
regulation and conflict-resolution mechanism closely
related to dominance — occurring particularly between
males and immature or subordinate females and some-
times involving no intromission or even penile erection
on the part of the male (Dahl 1987, Furuichi 1987).

In summary, competition between males over mating
is evident among all the great apes, only partially excep-
ting bonobos. Resulting differentials in male reproduc-
tive success make such male aggressiveness subject to
significant natural selection (Smuts 1987). This is evi-
dent anatomically in pronounced sexual dimorphism
among single-male gorilla and orangutan breeding
groups and a quite high testes-size/body-weight ratio
among chimpanzees and bonobos, whose social organi-
zation includes promiscuous multimale mating. Hu-
mans, however, have neither of these crucial anatomical
indices of male sexual competition, and this pattern is
both highlighted and perhaps to some degree explained
by the culturally mediated ability of humans to main-
tain pair-bonded sexual relations within a diffusely co-
operative multimale group. In social terms as well, it
appears that violence systematically enhances mating
access of dominant males among great apes but much
less so in simple human societies. Among the latter, a
male challenger, even if victorious, may be a social out-
cast rather than gaining access to his rival's mate. As is
evident by the violent end that recidivist killers them-
selves ultimately meet among the !Kung and other sim-
ple societies, violent competition to gain sexual access
to women may limit rather than increase lifetime male
reproductive success.

Some sociobiologists have suggested that sexual vio-
lence among human males may be a ''high-risk'' mating
strategy for underprivileged males (e.g., Thornhill and
Thornhill 1983; see Daly and Wilson 1983, 1988). How-
ever, the highly egalitarian nature of male-male rela-
tions in simple societies — including the normative in-
stitution of marriage for adult men — makes it unlikely
that some adult males will be systematically ''under-
privileged" in sexual access (e.g.. Draper 1975, TurnbuU
1982, Lee 1982). Reproductive success is in these cases
unlikely to increase with interpersonal aggression and
may even show a negative association with it. This is
consistent with cultural values in these societies: inter-
personal aggression and violence tend to be unrewarded
if not actively devalued by men and women alike.
Among societies in which kin or territorial groups are
coherent and have exclusive interests, it could be argued
that killing of anyone on "the other side" increases the
fitness of one's group and oneself relative to one's ene-
mies (see Daly, Wilson, and Weghorst 1982). However,
simple societies tend to lack such exclusive territorial
groups. Indeed, their subsistence strategy is often highly
dependent on nonexclusive access to resources and on
diffuse territorial linkages. As a result, ethnocentric kill-
ing is rare and cannot easily be considered to increase
inclusive fitness in any event.

In significant contrast to simple human societies,
great-ape and middle-range human ones display similari-

ties, or at least analogies, between male dominance and
differential sexual access. Again, then, the known range
of variation belies simplistic assumptions of uniformity
over the course of human evolution, cautioning in par-
ticular against drawing homologous conclusions from
male status competition in great-ape and human
"tribal" societies.

Violence and Intergroup Competition

In simple human societies, exclusivity and boundedness
of social groups are largely precluded by shifting re-
source availability, fluid population movement, lack of
fixed property, and the need for intergroup alliance and
support. Territorial rights, while often formally recog-
nized, are rarely enforced when permission to hunt or
forage is requested (Balikci 1970:170; Lee i^j^b'.Sj-,
TurnbuU 1965(3; see also Myers 1986; cf. Hamilton
1982). Band membership tends to be fluid, shifting easily
to exploit available resources. Local groups also share
information about availability of subsistence resources,
and it has been suggested that the advantage of such
information sharing was a major selective force in the
evolution of human linguistic communication (Kurland
and Beckerman 1985).

With emphasis on egalitarian access to resources, co-
operation, and diffuse affiliative networks, contrary em-
phasis on intergroup rivalry and collective violence is
minimal. Even apart from cultural injunctions and affil-
iative needs, the shifting location and composition of
bands tend to undercut the demographic basis for uni-
formly opposed territorial groups. Given migratory pat-
terns and resource dispersal, resources are difficult if not
impossible to defend, and the cost of such defense typi-
cally outweighs its benefit (Netting 1986: chap. i). Like-
wise, the cost of defense and retaliation against armed
aggression is typically great; it is more expeditious to
move away. This tends to short-circuit patterns of feud
and systematic raiding or warfare (see Woodburn 1984,
TurnbuU 1984, Lee and DeVore 1984).

Some authors have projected such reasoning quite far
back in human evolution. Concerning H. erectus,
Campbell {i9Ssa:T,26) writes: "conflicts between bands,
if they occurred, must have been rare in an uncrowded
world. . . . They probably came after humans settled
down on the land, became a more numerous species,
and forged cultures that encouraged individual and
group pride in possessions, territories, and beliefs." Gell-
ner (1989) suggests that systematic use of coercive force
as an organizational feature of society was undeveloped
prior to the advent of food storage and agriculture.

The question of territoriality and armed conflict may
under certain conditions be more complicated. There is
some historical evidence of reciprocating collective con-
flict, sometimes ethnically based, among simple foragers
(Balikci 1970:182-84; Lee 1979(3:382; Clastres 1972; K.
Hill, personal communication; Griffin 1984:103-7; Ro-
barchek 1990). Feuding or warfare does not, however,
appear pronounced except where, as among the Ache,

KNAUFT Violence and Sociality in Human Evolution \ 403

Agta, and Waorani, large-scale intrusion by agricultural
societies resulted in conflict over land and internal soci-
etal reorganization.^^ Apart from the influence of state
societies, collective enmities in simple societies tend to
be minimal and to occur between groups that lie outside
the extensive networks of affiliation that link adjacent
bands and territories. Travel to and armed conflict
against persons in such areas tend to be infrequent.

In middle-range societies, social competition in-
creases with the increase in fixed, high-value resources,
sedentism, and population density. Even where popula-
tion density is low, valuable movable resources such as
large domesticated animals may become a source of in-
tergroup antagonism and systematic hostility (e.g.. Bed-
ouins and herders with respect to camels and livestock.
Plains Indians with respect to horses). With sedentism
and/or significant domestication of plants or animals
there is a proliferation of corporate groups that stress
exclusive membership and rituals of allegiance (e.g.,
Plog 1990:193; see also Smith 1975). Ethnic differentia-
tion and ethnocentrism become more pronounced, and
fraternal interest groups become increasingly important
in armed conflict to protect or extend access to valued
rights and resources (Otterbein and Otterbein 1965, Ot-
terbein 1968; see also LeVine and Campbell 1972). John-
son and Earle (1981:60) suggest that in ''increasingly
widespread and increasingly successful efforts to restrict
access to critical resources, we encounter the beginnings
of warfare.'' Paige and Paige (1981:75-78) have shown
cross-culturally that a stable and valuable resource
base — such as herds or cultivated lands — is highly corre-
lated with fraternal-interest-group strength and, indeed,
a better predictor of it than residence pattern. This un-
derscores the difficulty of using simple forager/sedentary
characterizations in a human-evolutionary context.

Some researchers, including Haas (1990), have drawn
upon the typologies of Service (1971) and Sahlins (1968)
to suggest that increased warfare in sedentary societies
is a key dimension of what is termed ''tribalization.''
Others, such as Braun (1990; Braun and Plog 1982), em-
phasize multidimensional adaptations of sedentary com-
munities, including networks of trading alliance among
villages and increasing ritualization, leadership delega-
tion, and dispute. In either case, both the potential for
and the incidence of collective armed conflict between
opposed groups would appear to increase. Feinman and
Neitzel (1984:48-5 1) found warfare to be the single most
commonly reported function of leadership in sedentary
prestate societies in South and Central America and
eastern North America. Black (1990) has suggested that
chronic vengeance as a mode of conflict management
is especially pronounced when social groups are charac-
terized by immobility, social distance, equality, and
organization — all typical of middle-range societies. In a
cross-cultural analysis of aggression, Ross (1985, 1986)

II. Dense rain-forest areas, where visibility is minimal, offer the
potential for effective surprise raiding, but the exploitation of such
habitats was relatively late because of its difficulty in the absence
of plant domestication or trade with sedentary populations (Camp-
bell i985b:chap. 2; Ichikawa 1983; Bailey et al. 15

documents, among other things, a direct association be-
tween socioeconomic complexity and external warfare
in prestate societies.

Certainly in those areas of the world where small-
scale sedentary societies have persisted until recent
times or been subject to intense archeological study,
warfare has often been shown to have been frequent and
intense, including in most parts of Melanesia (Knauft
1990b, Meggitt 1977, Berndt 1962, Koch 1974, Hallpike
1977, Clunie 1977), Polynesia (e.g., Goldman 1970,
Vayda i960), and parts of Southeast Asia (R. Rosaldo
1980, M. Rosaldo 1980, Bock 1985, Kiefer 1972, Gibson
1990), South and Central America (e.g., Chagnon 1968,
1983, 1988; Harner 1972; Ross 1984; Whitehead 1990;
Carneiro 1990; Drennan and Uribe 1987), North
America (e.g., Drago 1970, Ewers 1958, Grinnell 1956,
Wallace and Hoebel 1952), Africa (e.g., Sahlins 1961,
Kelley 1985, Fukui and Turton 1979, Kuper 1937, Van-
sina 1966), the Middle East (e.g.. Meeker 1979; Hardy
1963; Barth 1961, 1965), and Europe (e.g., Boehm 1984(3,
Moss 1979). Warfare and raiding have also been docu-
mented among a range of complex hunter-gatherers, in-
cluding equestrian or sedentary foragers and selected
prehistoric cases (see Ember 1978, Prayer n.d., Wright
1988; cf. Price and Brown 1985).

While certainly admitting an enormous range of varia-
tion in frequency and intensity of warfare, these socie-
ties exhibit a much greater tendency toward collective
armed conflict and toward culturally competitive ethno-
centrism than simple societies. The tendency toward
warfare may be yet greater when paramount chiefdoms
are created through conquest (Carneiro 1981, 1990) and
is also highly associated with — if not partly causative
of — the rise of the state (Carneiro 1970, Haas 1982, Ga-
briel 1990; cf. Claessen and Skalnik 1978).

Among the great apes, male intergroup killings have
been documented in gorillas and chimpanzees, are ex-
tremely likely in orangutans, and are uncertain, sus-
pected, or under investigation among bonobos. How-
ever, no great apes are territorial in the strict sense of
the term.^^ Among all these species and particularly
orangutans, home ranges are to some degree flexible and
overlapping. However, challenges to the dominance hi-
erarchy or to the integrity of breeding units are subject
to strong and aggressive defense. Among solitary orang-
utans and in the harem-group organization of gorillas,
intergroup violence and violence between individual
adult males are largely synonymous. Violence occurs
among orangutans as aggression between mature males
over a female in consort with one of them. Male vio-
lence in intergroup contexts among gorillas is ultimately
crucial in establishing and maintaining control of a so-
cial and breeding group. Fossey (1979) documents sev-
eral such cases, including one in which a lone silverback
exhibited aggressive displays and chest-beating, vio-

12. A definitive case for territoriality can be made for gibbons [Hy-
lobates lar), which are less closely related to humans than the
larger apes. Gibbons are arboreal and monogamous and have small,
well-defined home ranges that are continuously occupied and vig-
orously defended by males (see Carpenter 1940, Ellefson 1968).

404 I CURRENT ANTHROPOLOGY Voluuie 32, Number 4, August- October 1991

lently charged into a group, effected the killing of an
infant, and later went off with the infant's mother. The
existing information on adult male-male gorilla interac-
tions is summarized by Wrangham (1987:66):

For gorillas intergroup interactions have been seen
in the Virunga and in Kahuzi. In the Virunga about
80 percent of encounters involved violent displays,
and 50 percent involved physical aggression by adult
males (Harcourt 1978). Severe wounding (Harcourt
1978) and probably death (Baumgartel 1976) of adult
males can result. Furthermore, five out of a total of
38 infants have apparently been killed by attacks
from adult males during intergroup encounters
(Fossey 1979, 1981). In Kahuzi three interactions
were reported by Yamagiwa (1983). There was no
physical violence, but there were fierce displays in
two of them. Furthermore, an infected wound which
led to the death of a breeding male was apparently
caused by attacks from a lone male.

In short, male gorilla breeding success is significantly
related to violent confrontation with males from other
groups.

Violence among great apes both within and between
conspecific groups is best documented among chimpan-
zees (Goodall i986:chaps. 15, 17; Goodall et al. 1979;
Nishida 1979, 1983; Nishida et al. 1985; de Waal 1982;
cf. Itani 1979). Systematic confrontation and killing be-
tween troops of free-ranging chimpanzees have been
well documented at two research sites and have been
the subject of considerable professional interest. At
Gombe, a chimpanzee group that had split off from an-
other was systematically exterminated by members of
the latter over a period of five years (Goodall et al. 1979;
Goodall 1 98 6: chap. 17). During this period, all five adult
males and one female from the victim group were appar-
ently killed by the intruders. (Several of these kills were
conclusively documented.) Remaining females were ab-
sorbed into the victorious group. During this process,
the home range of the victims was systematically occu-
pied for foraging by the intruders. After the old group
was exterminated, the new larger group itself became
subject to predation — by another adjacent chimpanzee
troop to the south — and was forced to contract its home
range from 17 to 6 km^ over four years,- the deaths of at
least two infants and one adult male are suspected to
have been caused by the invaders. At Mahale, a chim-
panzee group gradually became extinct as a social, geo-
graphic, and breeding unit over a period of 1 5 years,- dur-
ing this period all seven of its adult males died without
replacement (Nishida 1979, Nishida et al. 1985). In-
tergroup fights, woundings, and the sudden disappear-
ance of several prime adult males strongly suggest that
as many of six of these deaths were brought about vio-
lently by an adjacent group. As the number of males was
reduced, females were incorporated gradually and later
en masse into the intruding group. Correspondingly, the
territory of the victim group was absorbed by the in-
vaders.

Chimpanzee males in particular monitor the periph-

eral areas of their home range. This is undertaken
through distinctive patrolling behavior, characterized by
''cautious, silent travel during which the members of
the party tend to move in a compact group'' (Goodall
1986:490). Animals on patrol carefully watch and listen
for evidence of occupation by outsiders and suppress all
normal calls among themselves. This pattern alters dra
matically if individuals from the adjacent group are en-
countered. Males and anestrous females are aggressively
intimidated and, if alone, violently attacked or effec-
tively ''hunted" by the coordinated effort of several
males in the patrol. Once seized, the individual may be
beaten and bitten until wounded or dead. According to
Goodall (1986:529), the five observed cases of lethal
conspecific attack at Gombe have several features
in common:

(a) the attacks were all long — the shortest lasted at
least ten minutes, and three continued more than
twice as long; (b) all were gang attacks, during
which the aggressors sometimes assaulted the vic-
tim one after the other, or two to five assailed the
victim simultaneously,- (c) all the victims were, at
some point, held to the ground by one or more of
the aggressors while others hit and pounded,- (d) all
the victims, in addition to being hit, stamped on,
and bitten, were dragged first in one direction, then
another . . . and [g] during each incident the observ-
ers, all thoroughly experienced in chimpanzee be-
havior, believed that the aggressors were trying to
kill their victims. . . . [This is] because the attackers
showed some of the patterns which, while com-
monly seen during the killing of large prey, have not
been seen during irztr^^ community fighting — as when
one of Goliath's legs was twisted, when a strip of
flesh was torn from De's thigh, or when Satan drank
the blood pouring from Sniff's nose.

There are several possible causes for such intergroup
killings. In the Gombe study area, there was some evi-
dence of added resource pressure — caused by land en-
croachment by humans — that may have exacerbated ag-
gression and violence (Goodall 1986:51). The degree of
land impingement, however, has not been judged severe
enough to explain such severe and persistent violence.
Goodall (198 6: chap. 17) concludes that chimpanzees are
a territorial species and that land areas are defended and
contested by males, even though these territories have
no fixed boundaries and need not be continually occu-
pied by their possessors. A further cause of chimpanzee
intergroup aggression is the enhanced sexual access and
reproductive success of males in the invading group.
Group annihilation need not greatly affect population
density for very long; while the number of males and
infants is temporarily depressed, the transfer of repro-
ductive females to the new group quickly facilitates re-
population of the area (Nishida etal. 1985:298). The ad-
ditional food resources of the newly occupied territory
aid in this process. Given these reproductive enhance-
ments, intergroup violence by chimpanzee males ap-
pears subject to strong natural selection.

KNAUFT Violence and Sociality in Human Evolution \ 405

This conclusion is supported by broadly similar rela-
tionships between social organization and intergroup vi-
olence in Gombe and Mahale. Several cases have been
observed in which infants from outsider groups have
been seized, killed, and even eaten by males of an op-
posed group. The mother in such instances not infre-
quently joins the males of the attacking group, resumes
sexual cycling, and becomes receptive for mating with
her new male companions. There is obviously strong
selective advantage for males in killing infants sired in
other groups and then reproductively appropriating the
recycling mothers for themselves (as also occurs quite
dramatically among gorillas). Sexually receptive females
are usually tolerated peacefully when encountered by
males of adjacent chimpanzee groups, and immigrant
females tend to be quite sexually active in this context
(Hasegawa 1989).

Bonobos appear to be somewhat less agonistic than
chimpanzees in irztr^^group relations, and recent evi-
dence suggests that their intergroup relations are corre-
spondingly less aggressive (Kano 1990). Nonetheless,
those few bonobo intergroup confrontations observed in-
cluded competitive vocal displays, withdrawal of the
smaller party, and one violent incident causing ''serious
injuries to several individuals'' (Kano and Mulavwa
1984:265; see Kano 1984b, Badrian and Badrian 1984).
Long-term trends of bonobo intergroup relationship have
not yet been documented. Given that many years of ob-
servation among chimpanzees were necessary before the
extent of intergroup aggression could be established, no
firm conclusion can yet be drawn concerning the fre-
quency and severity of bonobo intergroup violence
(Wrangham 1987:66; de Waal 1989:221, quoted above).
It does appear, however, that a relatively rich resource
habitat, relative lack of predation, and intragroup social
bonds facilitated by diffuse and frequent sexual encoun-
ters combine to minimize intergroup contact and in-
tergroup aggression among bonobos (Kano 1990, Blount
1990).

With the apparent and/or partial exception of bonobos,
then, the evidence suggests that intergroup violence
among great apes is both pronounced and highly related
to reproductive success. The chimpanzee pattern in par-
ticular has certain features in common with ethnocen-
tric and aggressive intergroup encounters in human sed-
entary and pastoral prestate societies (see Boehm n.d.a),
and this pattern contrasts significantly with the more
flexible and cooperative intergroup dynamics of simple
human societies. Several authors, including Wrangham
(1987; Manson and Wrangham 1991), Foley (1988:219;
cf. Foley and Lee 1989), and Ghiglieri (1987, 1989), have
attempted to draw a parallel between the patrifocal so-
cial organization of the great apes (though bonobos are a
partial exception) and the social organization of prestate
humans in general. As Barnard (1983:196-97) has
shown, however, the notion of the patrilocal band
among human foragers is ''empirically groundless,'' be-
ing based on the legacy of the early and erroneous — if
influential — typologies of Steward (1936) and Service
(1971, 1979): "New generations of scholars gave the

coup de grace to the patrilocal model [of hunter-gatherer
band organization]. All over the world, societies of small
community size were shown to be neither essentially
virilocal nor patrilineal in any sense. 'Flux,' 'flexibility,'
and 'fluidity' became the . . . words to describe their
social organization (see Lee and DeVore 1984:7-12)." Of
particular relevance here is his criticism of the Human
Relations Area Files study by Ember (1978), which sug-
gested that hunter-gatherers were typically patrilocal
and prone to warfare; he finds this suggestion untenable
given "the inaccuracy of the early ethnographies on
which her Ethnographic Atlas sample is based" (p. 197).

Manson and Wrangham's recent study (1991) is to a
certain extent an advance. On the one hand, it distin-
guishes between groups of foragers in terms of the pres-
ence or absence of significant alienable property (cf.
Testart 1982, 1985) and documents that intergroup vio-
lence more frequently occurs over resources in the for-
mer category. This finding is consistent with the present
analysis. However, Manson and Wrangham rely on Em-
ber's (1978) analysis of forager warfare and appear to
adopt her biased assessment of forager patrilocality (see,
similarly, Foley 1989:487-88). Consequently, they posit
a similarity between prestate human social organiza-
tional patterns and chimpanzee patterns of male philo-
patry (males breeding within their natal groups) and vio-
lent male attacks on outside groups (cf. also Otterbein
1985 :xxi; Ghiglieri 1987, 1989). These pongid patterns
may provide an analogy with some sedentary human
societies and complex hunter-gatherers, but they are
highly questionable as a model for simpler human
groups. ^^

To illustrate the importance of delimiting the sample
of simpler human societies from more complex hunter-
gatherers, a subset of the 563 societies in Murdock's
(1981) Atlas of World Cultures was defined in terms of
[a] absence of agriculture, significant animal domestica-
tion, and primary subsistence reliance on fishing (i.e., o
in Murdock's col. 7, pts. 4 and 5, 0-4 in col. 3, and o in
col. 39) and [b] absence of significant sociopolitical class
distinctions (o in col. 67). Of the resulting 39 societies,
71.8% (28/39) were rated by Murdock as having an "ab-
sence of any patrilineal kin groups and also of patrilineal
exogamy" (o in col. 20). In contrast, 59.0% had cognatic
kin groups recognizing ambilineal or bilateral descent (B
or K in col. 24). Only 25.6% (10/39) were rated as exhib-
iting patrilocal residence (P in col. 16), the remainder
being characterized as practicing various combinations
of ambilocal and uxorilocal residence or virilocal resi-
dence "confined to instances where the husband's pat-
rikin are not aggregated in patrilocal and patrilineal kin
groups." These trends would probably have been still
more pronounced if cases based on somewhat question-
able older data had been excluded and if a less rigid ty-

13. Stanford and Allen (1991:59) cogently criticize several recently
proposed models of human behavioral evolution as being chimp-
anzee-referent models couched in Darwinian terms to give the ap-
pearance of a broader evolutionary perspective.

4o6 I CURRENT ANTHROPOLOGY Voluuie 32, Number 4, August- October 1991

pology of residential types had been used.^"^ These find-
ings are consistent with ethnographic suggestions above
that social organization and residence tend to be shift-
ing, open, and flexible among nonintensive foragers.
Available evidence thus supports the conclusion that
fraternal interest groups and the kind of violence associ-
ated with them are seldom of importance in simple hu-
man societies.

Overall Patterns of Sociality and Violence

In simple human societies, strong emphasis is placed
on cooperative sociality both within and between flexi-
bly constituted groups. Cultural values and behavior
strongly indicate sharing of valuable food items and ma-
terial property, flexible access to resources, and exten-
sion of sociality through diffuse networks of classi-
ficatory kinship, namesake, totemic, and/or trade
partnerships. Both within and between groups, rates of
aggression are low. The aggression that does occur, how-
ever, has a relatively high probability of resulting in
homicide, with dispute resolution measures other than
dispersal undeveloped. Violence in simple human socie-
ties is generally unrelated to disputes over material prop-
erty, territory, competitive leadership interests, status
hierarchy, or opposition between corporate or ethnic
groups; it is often if indirectly related to sexual disputes.
While adult male competition over sexual access to
women is to a significant extent controlled by cultural
rules of pair-bonded sexual union within extended
multimale/multifemale social groups, disputes over sex-
uality, when they do erupt, often result in severe vio-
lence within the band or larger community. However,
the killings that result are often not an effective way to
gain a new spouse or lover.

In middle-range societies, by contrast, violence is re-
lated to frequent disputes over male dominance and po-
litical status hierarchy and sometimes also to conflicts
over property and territorial resources. Aggressively
achieved male status domination is more likely to result
in increased access to mates and/or polygyny. Middle-
range societies also tend toward a stronger ethic of com-
petitive virility linked to positive valuation of male as-
sertiveness and aggression in gender relations, politics,
and warfare. Violence between corporate or ethnic
groups is more frequently collective, more often occurs
between fraternal interest groups, and is more frequently
reciprocated in raiding or warfare than in simple socie-
ties. Ongoing blood feuds are relatively common. All of
these features vary in complex ways and correlate with
various social structural and socialization factors (M.
Ross 1985, 1986; Otterbein 1980, 1985; Knauft 1990b).

Known socialization practices are consistent with

14. If Australian Aboriginal societies (which, as we have seen, are
exceptional among nonintensive foragers in a variety of ways) are
excluded, the Murdock-ranked percentage of simple societies lack-
ing patrilineal kin groups increases to 83.9% (26/31), the percent-
age having cognatic kin groups increases to 74.2% (23/31), and
the percentage lacking patrilocal residence increases to 87.1%
(27/31).

these broad contrasts. Simple societies have been consis-
tently noted as particularly indulgent and nonpunitive
in their child-rearing and adolescent socialization prac-
tices, including a high degree of supportive paternal as
well as maternal contact with young children (see gener-
ally Leacock and Lee 1982:8; West and Konner 1976;
Konner 1981; [Mbuti] TurnbuU 1978, Tronick, Morella,
and Winn 1987, Hewlett 1990; [IKung] Draper 1978,
Marshall 1976; [Semai] Dentan 1978, Robarchek 1977;
[Inuit] Briggs 1970, 1978, 1982). In contrast, child rearing
and adolescent socialization in small-scale sedentary so-
cieties tend on the whole to be relatively harsher and
more punitive, including a greater frequency of trau-
matic male initiations (cf. Rohner 1980, Lambert 1971,
Whiting and Whiting 1975). Paige and Paige (1981:77)
conclude on the basis of detailed cross-cultural study
that ''in societies in which there are stable and valuable
economic resources, strong fraternal interest groups are
likely to develop, and the dilemmas posed by the major
events of the human reproductive cycle are responded
to by the form of ritual bargaining characterized . . .
as surveillance rituals.'' These conditions maximize a
group's ''ability to make explicit bargains concerning
women and children and to protect such bargains
through military action of a large and loyal political fac-
tion" (pp. 76-77)'

These trends are influenced by the socialization capac-
ities afforded by the larger size of residential age cohorts
in aggregated sedentary communities. Children in sim-
ple societies have few age mates and tend to play in
multiage and cross-sex groups, with few rigid boundaries
among them. In larger sedentary communities, there are
greater numbers of same-sex age cohorts and greater
potentials for age-grading and formation of interest
groups. Age-cohort differentiation can occur not only
in adult male leadership but in child-child social-
ization, with age-cohorts of older boys socializing and
not infrequently dominating their juniors and with sex-
segregated play and socialization becoming more com-
mon. Not surprisingly, age-sex status differentiation and
accompanying relations of social inequality are frequent
mechanisms for developing strong warriors among the
male initiate class (e.g., Langness 1974, Allen 1967).

Among great apes the most severe forms of violence
appear related to enhanced sexual access and reproduc-
tive success for male victors. In all great-ape species for
which data are available, defeating male competitors
systematically increases the winner's sexual access to
receptive females. Killing the female's offspring sired by
another male brings the female more quickly into es-
trus. Females commonly gravitate to such successful
males of their own volition, and in this sense male vio-
lence is actively selected for through females as well as
through males. Among chimpanzees, extermination or
dispersal of competitive males may also increase the
home range of the victorious males, thus increasing
their access to food resources and the nutritional fitness
of themselves, their mates, and their progeny. Coali-
tional male violence among chimpanzees benefits coali-
tion members by intimidating and/or eliipinating attack
targets both within and between groups. Among orang-

KNAUFT Violence and Sociality in Human Evolution \ 407

utans and gorillas, violence is pursued by adult males
only as individuals, and selection thus operates directly
in promoting large male size. Whereas sexual dimor-
phism is pronounced among single-male gorilla and
orangutan breeding groups and a high testes -size/body-
weight ratio occurs among chimpanzees (whose social
organization includes promiscuous multimale mating),
humans have neither of these crucial indices of male
sexual competition. In the context of multimale mating
groups, this suggests that cultural rules may have re-
duced male sexual competition in human evolution.

Among bonobos, the incidence of nonaggressive food
transfer is relatively high, male dominance and sexual
competition appear relatively undeveloped, sexual di-
morphism is low (Zihlman and Cramer 1978), and in-
tergroup aggression, though not yet thoroughly investi-
gated, may be somewhat lower than among other great
apes. These and other aspects of bonobo organization
have been considered suggestive in modeling the behav-
ior patterns of early humans (see Susman 1984, 1987;
Furuichi 1989). The present analysis supports this view
but points to major qualitative differences between bo-
nobos and the simplest human societies (see also Blount
1990). Bonobo patterns of diffuse sexuality and use of
sexual contact as an appeasement strategy contrast
sharply with the rule-governed and restricted patterns of
mating in simple human societies. Bonobo food sharing
is tacit rather than active, and relative lack of aggression
is due as much to a plentiful environment and relative
freedom from predation as to a propensity for coopera-
tive social interaction. The actively cooperative social-
ity of early humans in a predator-rich savannah environ-
ment was probably significantly different from and
much more strongly selected for than the patterns of
affiliative sociality found among bonobos. As has often
been noted, no single species is apt to provide an ade-
quate model of early human social organization.

Conclusions

Patterns of sociality and violence are in certain formal
respects similar among great apes and middle-range hu-
man societies but systematically different among simple
human societies. Male competition in resource control,
dominance hierarchy, overt competition over sexual ac-
cess to females (including polygynous mating), and com-
petitive defense of the social group are all arguably
greater among both great apes and middle-range socie-
ties than among the simplest human societies.

This comparison is based on superficial analogy; in-
deed, the divergence of simple societies from these
trends underscores that correspondences between great-
ape and middle-range human societies are not homolo-
gies based on phylogenetic continuity through human
evolution (contrast Wrangham 1987; Ghiglieri 1987,
1989; Foley and Lee 1989; Tooby and DeVore 1987).
Thus, no assumption can be made of uninterrupted evo-
lutionary progression from primate patterns to those
found in human ''tribal,'' much less state, societies (con-
trast Otterbein i985:xxii, Bigelow 1975,- Eibl-Eibesfeldt

1979; Popp and DeVore 1979; Chagnon 1979, 1988).
That the evolutionary period characterized by simple
human societies may have been many times longer than
that characterized by middle-range ones suggests that
findings about violence and sociality based on selected
case studies of the latter may be limited in their evolu-
tionary implications (e.g., Chagnon 1988).

The present analysis recontextualizes Wrangham's
(1987:68) assessment, based in part on a lumping of sim-
ple with more complex human societies, that the com-
mon ancestor of humans had ''closed social networks,
hostile and male-dominated intergroup relationships
with stalk-and-attack interactions, female exogamy and
no alliances between females, and males having sexual
relationships with more than one female.'' This and
similar arguments by Ghiglieri (1987, 1989), by Foley
(1988:217-19) for late Pleistocene H. sapiens sapiens,
and by Manson and Wrangham (i 991) for human evolu-
tion generally are here countered by distinguishing be-
tween simple and more complex human societies as
points of primatological comparison and by reasoning
"backward" rather than "forward" in evolutionary
time — projecting simple human society traits into the
evolution of H. sapiens and perhaps Homo generally
rather than projecting nonhuman primate tendencies
forward into hominid and then equally into human evo-
lution.

Nonhuman primate models of human social evolution
are likely more applicable to early hominids than to hu-
mans. Conversely, models based on simple societies are
likely to have greater relevance for H. sapiens and per-
haps for Homo generally. Rather than reduce these to
competing species-archetype models, however, it is im-
portant to address the question how and over what
period of time the distinctive features of human cul-
tural transmission and corresponding social alteration
emerged. From the present perspective, the evolution of
Homo is likely to have proceeded in large part among
groups that had relatively open social networks, nonhos-
tile intergroup interactions, and a significant degree of
institutionalized if not monogamous pair bonding. More
generally, the distinctive "cooperative niche" discern-
ible in simple human societies needs to be considered
alongside the "social carnivore niche" and the "cogni-
tive niche" that are so often attributed to them in mod-
els of hominid evolution (Knauft 1988, 1989(3; Ingold
1987; Carrithers 1990; cf. Foley 1982, 1984; Hill 1982;
Tooby and DeVore 1987). That cultural rules of coopera-
tion are pronounced in such societies does not preclude
violence, but it does facilitate patterns of sociality and
diffuse sharing that may have been characteristic of a
large portion of our genus's evolution. Generalizations
about human societal evolution are easily biased by
HRAF samples weighted heavily with middle-range so-
cieties, which are far more numerous in the ethno-
graphic record than simple ones though they have per-
sisted for a much shorter period of evolutionary time.

The present model is not intended as a definitive or
an exclusive alternative to others, and it is itself self-
defeating in that it uses static types such as "simple
societies" to argue for developmental changes in pat-

4o8 I CURRENT ANTHROPOLOGY Voluuie 32, Number 4, August-October 1991

terns of sociality and violence in the course of human
evolution. If these patterns have indeed changed, it is
hardly to be expected that they were static within the
category of ''simple societies'' itself. It could be, for in-
stance, that patterns of violence and sociality among late
Pleistocene hunters of periglacial Eurasia, who were
highly dependent on very large game (offering the poten-
tial for large aggregations of consuming foragers and siz-
able frozen food stores), were somewhat more like those
among complex hunter-gatherers than were those of
foragers relying more on dispersed floral resources
and smaller game (cf. Foley 1988:217-19; Mellars
i989b:356-57). More generally, the difference between
simple- and complex-forager patterns of violence and so-
ciality may parallel differences in resource concentra-
tion or population aggregation that potentiate them.
Highly decentralized, nonintensive foraging adaptations
are, on a global scale, likely to be both underrepresented
in the archeological record and subject to less scholarly
interest than the relatively dramatic material assem-
blages of more socioeconomically complex prehistoric
groups, but the bulk of our genus's evolution was spent
as simple foragers.

The range of variation within the category of nonin-
tensive foragers has in fact been underemphasized here,
in part for the sake of brevity and in part to highlight
larger, more pronounced, and more neglected aggregate
differences. That the delayed-return reciprocity systems,
male status distinctions, patrilocality, polygyny, and
conflict between fraternal interest groups found in some
Australian Aboriginal societies are rare or absent among
nonintensive hunter-gatherers elsewhere suggests that
this cluster of traits may among simple societies be an
Australian isolate (cf. Testart 1988). The range of varia-
tion in great-ape and middle-range societies, if anything
even more evident from observational data, is likewise
collapsed here for the sake of illuminating larger evolu-
tionary trends. Like any other evolutionary typology,
this one is useful as a heuristic device to stimulate dis-
confirmation and promote more sophisticated and re-
fined hypotheses (see Upham 1990(3:88-91).

Despite inevitable problems, the comparative assess-
ment of empirical data from great-ape and simple- and
middle-range human societies is crucial. Lately, the rele-
vance of these findings has typically been viewed in
terms of neo-Darwinian assumptions of individualis-
tic competition. It has been suggested here that the
strengths of evolutionary biology and behavioral ecology
should be complemented by a fuller and more sensitive
consideration of symbolic transmission and cultural
rules of cooperation in human evolution. In this regard,
humans may indeed be unique.

Postscript

Rodseth et al.'s (1991) suggestions about human social
organization, derived from extensive cross-species com-
parison with nonhuman primates, are in important ways
both convergent with and divergent from the findings

of the present analysis. The similarities and differences
reveal important possibilities for further research and
theory building.

This paper's findings are consistent with Rodseth et
al.'s statement that ''humans are unique in the extent
to which both males and females form affiliative rela-
tionships, with nonkin as well as kin, both within and
between groups" (p. 232) and their assessment that hu-
mans are unique in forging alliances through systematic
exchange of mates, e.g., that intergroup affinity among
humans should be "appreciated as a unique and revolu-
tionary primate pattern" (p. 237). They differ, however,
with regard to the suggestion that human social organi-
zation consists of closed or semiclosed groups that tend
toward male philopatry (males breeding within their na-
tal groups) and fraternal interest groups. As we have
seen, such generalizations [a] collapse middle-range or
"tribal" society patterns with those of simple societies,
[b] rely on Ember's (1978) skewed cross-cultural analy-
sis, and (c) tend to project nonhuman primate patterns of
violence and social organization onto humans. Alliances
among local groups in simple human societies tend to
be highly flexible and shifting, importantly including
coresidence among adult male affines for significant pe-
riods of time.^^

In the social organization of the common ancestor of
apes and humans, a tension could plausibly have existed
between chimpanzee-like patterns of male philopatry
and selective pressures for more open, dispersed, and
flexible alliances, and such tension may have been
played out and resolved differentially in the several com-
peting australopithecine species (Foley 1989:487-89). By
the time of Homo's development and the progressive
elaboration of human cultural transmission, however,
development may have turned decisively away from
male philopatry and closed fraternal interest groupings
and toward more open and more flexible social net-
works.

Elaborate symbolic communication and cultural
transmission were likely crucial in maintaining the pat-
tern of diffuse alliance in the absence of ongoing spatial
proximity that Rodseth et al. rightly emphasize. This
dovetails with the elaboration of cultural constraints
upon domestic behavior to maintain exclusive sexual
bonds in the face of diurnal dispersion/reunion and a
sexual division of labor. From Rodseth et al.'s perspec-
tive, "the problem . . would not be how early hominids
came to avoid incest or how one sex came to breed in
other groups but how exclusive sexual bonds evolved
from a chimpanzee-like pattern of promiscuity" (p. 237).

1 5 . This conclusion is nicely developed through detailed empirical
analysis for the Mbuti by Terashima (1985; cf. also Pedersen and
Woehle 1988) in contrast to the more simplistic residential catego-
rization of Bailey and Aunger (1990). The characterization of early
human social organization as largely male philopatric, closed or
semiclosed, and divided into fraternal interest groups rests uneasily
with the other features of Rodseth et al.'s own analysis, which
would appear to emphasize the distinctive openness of human so-
cial networks. This implicit tension appears to force an expansion
of the notion of fraternal interest to accommodate even matrilocal
alliances, making the "interest group" hardly "fjratemal" at all.

KNAUFT Violence and Sociality in Human Evolution \ 409

My intuition is that, given the geometric decrease
of genetic relatedness with genealogical distance,
inclusive-fitness benefits are insufficient to outweigh
the threats to pair-bonded sexual exclusivity and to male
parental investment posed by sexual cheating in a regu-
larly dispersing forager group of 25-50 persons — even if
the males of the group are all agnatically related.^^ In
other words, an independent force of cultural as opposed
to biogenetic selection is needed. Wrangham's perspec-
tive could be construed as supporting such an assess-
ment insofar as it suggests that language was likely cru-
cial as an aspect of absentee mate-guarding — i.e., that
nonsymbolic behavior-selection variables were by them-
selves insufficient to maintain such a mating system.
This raises the question how a uniquely human rule-
influenced mating system could have evolved.

It will be important to develop competing analyses of
the likely timing and extent of cultural development in
humans. My own analysis, which emphasizes the poten-
tial for group selection that cultural transmission af-
fords, predicts that from an early period in human evolu-
tion there is a crucial and driving opposition between
cultural and biogenetic selection processes. Durham's
( 1 991: chap. 8) fruitfully competing hypothesis is that
cultural and biogenetic selection processes, while oppos-
able in principle, tend in the main to support each other
in fact. To me this seems to give too little attention
to the significant conflict between group-level and
individual-level selection and the increasing gap be-
tween fast rates of symbolic dissemination and slower
rates of differential biogenetic reproduction. This is ob-
viously an area for further study and refinement. An at-
tempt has been made here to characterize the opening
stages of this relationship, addressing issues that Rod-
seth et al. independently raise by implication but do not
directly consider.

Cross-species analyses can indeed be valuable, and
Rodseth et al. are to be congratulated for illuminating
key overall differences in the social organization of hu-
mans and nonhuman primates. With humans in particu-
lar, however, one must be wary of assuming continuity
of social organization across the evolving spectrum of
socioeconomic intensification. As Rodseth et al. recog-
nize (fig. 3), their own analysis can in principle be re-
fined to include patterns of intraspecific organizational
variation. There are, however, some intrinsic limits to
such a procedure. Considering humans only in sociobe-
havioral terms makes it difficult to give adequate con-
sideration to the distinctive dynamics of human sym-
bolic transmission and the selective effects of this
transmission upon subsequent evolutionary develop-
ment. An analysis of symbolic dynamics as well as so-
cioecological patterns is therefore crucial in the study of
human evolution.

16. This issue can perhaps be illuminated through mathematical
modeling.

Comments

THOMAS S. ABLER

Department of Anthropology, University of Waterloo,
Waterloo, Ont, Canada NiL 3G1. 10 iv 91

Knauf t has made a significant contribution to our under-
standing of violence and warfare in an evolutionary con-
text. The breadth of his paper is such that I lack confi-
dence to comment on all its arguments,- these
comments will focus on his interpretation of warfare
and violence in non-state human societies. Incidentally,
contrary to an oft-expressed opinion (as in Foster and
Rubinstein i986:xii), the bibliography of the anthropol-
ogy of warfare is lengthy and contains numerous works
of high quality (Ferguson and Farragher 1988; see also
Ferguson 1984b). While I agree with the general tone of
Knauf t's paper and with his conclusions, there are sev-
eral major areas where I would question his analysis.

I applaud Knauft's analytical category of ''simple hu-
man societies.'' Too many fish and fowl (or, rather,
fishers and fowlers) have been lumped together as ''for-
agers'' or "band societies." Knauf t, however, contrasts
these with "middle-range societies," and surely the di-
versity among societies in this group is far greater than
that among foraging peoples. His emphasis on fraternal
interest groups in this category ignores the many such
societies that are matrilineal and matrilocal (Divale
1974, 1984). It also ignores the egalitarian ideology, the
absence of statuses possessing authority, and the nature
of leadership in many of these societies (see Trigger
1990). Knauft's characterization of "middle-range" soci-
eties emphasizes those with social ranking (those we
used to and occasionally still call "chiefdoms"). To be
fair to him, he phrases these characteristics in evolu-
tionary terms, seeing these institutions as developing.
However, as Mann (1986:67) has observed, "movement
toward rank and political authority seems endemic but
reversible" (my emphasis); ambitious elites that attempt
to impose authority frequently find themselves deposed
or simply with no subjects, the egalitarian-minded popu-
lation having moved elsewhere.

Also troubling is a failure to consider the effect of state
or imperial expansion (both military and economic) on
"middle-range societies" (see Wolf 1982, Ferguson and
Whitehead n.d.). Frequency of violence and warfare is
doubtless greatly increased by desire for trade, displace-
ment of populations, ecological disruption, employment
as allies by neighboring states, etc.; one can point to
analyses of Iroquois history by Hunt (1940) and Trelease
(1962) and to explanations of warfare on the North
American grasslands (Newcomb 1950). Any state expan-
sion has an impact on its neighbors, but European impe-
rialism has had a strong effect on our ethnographic sam-
ple. Parker's (1988:115) observation on the early stages
of European expansion is telling: "The principal export
of pre-industrial Europe to the rest of the world was vio-
lence."

This ignoring of history also leaves open the question.

4IO

CURRENT ANTHROPOLOGY Volume 32, Number 4, August- October 1991

as Donald (1987) has noted in a critique of an earlier
paper by Knauft, whether Knauft's ''simple human soci-
eties'' are simply products of the marginal environments
they exploit and the resulting low population densities.
While ''simple human societies'' may be simpler than
any others known to ethnography, they also may be
simpler than those of earliest H. sapiens (or even
earlier Homo) known only from the archaeological
record.

I also feel it is incorrect to see, as Knauft does, "simi-
larities, or at least analogies," between polygyny and the
mating patterns of a silverback male gorilla. Marriage is
much more than simple coresidence and sexual access.
The multiple spouses of elites may relate much more to
politics than to sex. Marriage also does not guarantee
exclusive sexual access to one's spouse. The Tiwi (Hart
and Pilling i960) are a classic example of a society in
which multiple marriages are linked to the high status
accorded elderly males whose prime days for fathering
children are long past.

LAURA BETZIG

Evolution and Human Behavior Program, University
of Michigan, looH Rackham, Ann Arbor, Mich.
48109, U.S.A. 15 IV 91

Culture is keeping the Gebusi from behaving adaptively.
For the second time in four years, Bruce Knauft has
made that case in CA. For the second time, I'm not con-
vinced (Knauft 1987, Betzig 1988).

Knauft draws parallels among great-ape societies,
"simple" human societies, and "middle-range" human
societies. He looks at conflict — dominance within and
violence between groups. He looks at cooperation — food
sharing and other forms of affiliation. And he looks at
reproduction — which males get more females and, prob-
ably, father more young. And he draws this conclusion:
Among apes and in middle-range human societies, dom-
inance and violence among males are positively related
to reproductive success, but among simple people like
the Gebusi, neither dominance nor violence increases
fitness. In Gebusi and other simple societies, culture tri-
umphs over nature; an ethic of cooperation wins out
over genes for conflict.

What's wrong with this argument? Two things, at
least. First, it's not at all clear that dominance and vio-
lence in simple human societies are not related to repro-
duction. In fact, what work has been done suggests that
they are. Hewlett (1988), for example, found that Aka
Pygmy leaders are more than twice as polygynous as
other men and tend to father more children. Qualitative
statements are consistent. Osgood (1958:200, 203), for
instance, noted that "powerful" Ingalik men might ac-
quire second wives and added that "a long time ago there
was a fellow who had five wives at one time and seven
at another. This man was a great fighter and had ob-
tained his women by raiding." As Lee's interviews of
the !Kung, which Knauft reviews, suggest, a killer may

not always get the girl — that day — but even if he
doesn't he may gain the resources or respect that help
him get another woman on another occasion. Darwin
(1871:896) wrote long ago that "polygamy ... is almost
universally followed by the leading men." As Knauft
rather grudgingly admits, this seems to be true in
middle-range societies (Betzig 1986); it seems too to be
true in ancient (Betzig n.d.a) and medieval (Betzig n.d.b)
societies, and it may even be true among "simples" (e.g.,
Hewlett 1988) and moderns (e.g.. Low n.d.).

Second, I think that the importance of cooperation
among most people and primates is underplayed. As
Knauft writes, and as others have written, food sharing
can be essential in simple human societies faced with
environmental uncertainty and risk. But food sharing is
ubiquitously human,- redistribution is everywhere, from
the smallest band to the biggest state. And other kinds
of cooperation are crucial among other primates. As do
humans, apes cooperate, at least, in order to compete. If
food sharing or some other kind of cooperation helps
keep sharers alive, helps them get mates, and helps them
raise children, then cooperation should be as "natural"
as the kinds of conflicts that can, under different condi-
tions, be a means to reproduction. Again, the evidence
that I'm aware of suggests that it is. Kaplan and Hill
(1985b) found that better Ache hunters attract more ex-
tramarital mates, father more children, and raise more
of them than poorer hunters (see also Hill and Kaplan
1988). And again, qualitative conclusions are consistent.
Marshall (1959:346), for instance, wrote that among the
!Kung "the number of wives a man may have is not
regulated by social rules but by his ability to obtain and
support them." And Jenness (1922:161) pointed out that
a Copper Eskimo polygynist "must be a man of great
energy and skill in hunting, bold and unscrupulous, al-
ways ready to assert himself and uphold his position by
an appeal to force." There should, in short, be genes for
cooperation as well as for competition; both should be
naturally selected when they increase their bearers' re-
production.

One bone that Knauft picks, that correspondences be-
tween ape and human societies may reflect "homologies
based on phylogenetic continuity through human evolu-
tion," seems to me unlikely to stand up to much if any
evidence or theory. Species are, by definition, set apart
by enough evolutionary time to allow for significant
change. It seems reasonable, then, to suspect that simi-
larities between them exist because similar behaviors
have been adaptive under similar conditions rather than
because phylogenetic legacies made them persist against
selective force. And it seems ironic that the idea that
"culture" has to win over "nature" is more guilty of
this kind of homology than the "assumptions of current
sociobiological reasoning" often are.

Having begun with an attack, I'll end with a little
reconciliation: Few problems are as interesting or as im-
portant as lethal conflict. Both Knauft's study of Gebusi
homicide and this review of killing, etc., in apes and
humans are packed with interesting — and potentially
very important — information. I appreciate that.

KNAUFT Violence and Sociality in Human Evolution \ 411

CHRISTOPHER BOEHM

Department of Anthropology, Northern Kentucky
University, Highland Heights, Ky. 41076, U.S.A.
16 IV 91

This is an astute and wide-ranging synthesis. A signifi-
cant contribution is the emphasis placed on the evolu-
tionary anomaly that one faces if one begins with the
great apes and ends with modern humans, concentrating
on what might be called agonistic style. The glaring dis-
continuity comes, of course, with egalitarian societies,
that is, societies in which leadership is unassuming and
hierarchy among adult males is weakly developed. The
discontinuity is as follows: African great apes and hu-
man societies at the chiefdom level and higher exhibit
pronounced male dominance hierarchies, while egalitar-
ian societies exhibit the opposite (see Boehm 1982,
1984b). This well-known ''egalitarian'' type has come to
include foraging societies but also many sedentary ones,
among them segmentary societies with rather high pop-
ulation densities and warfare.

Knauft confronts anthropologists interested in social
evolution and warfare with an absolutely critical dis-
tinction within this rather diffuse class, that between
simple and complex foragers. In pondering the apparent
evolutionary riddle of the contrast between the simple
foragers who long typified human political organization
and both African great-ape and centralized human socie-
ties, anthropologists must, I think, look much more
closely at the leveling factors that make small-scale hu-
man societies ''egalitarian.'' So far, causes of leveling
among males have been found mostly in ecological fac-
tors or in the dynamics of social or economic systems
(e.g.. Fried 1967); to my knowledge only Lee (1979),
Woodburn (1982), and I (Boehm 1982, 1984(3, b) have
argued that egalitarian society is a direct result of hu-
man intentions.

One way to consider the role of intentions in simple
foraging and other egalitarian societies is to start with
the genotype: I assume in a common ape-ancestor a sig-
nificant disposition to form hierarchies, since this is
present in African great apes and modern humans. One
then asks what cultural behaviors might account for the
absence of hierarchy in simple foragers, assuming that
they too share that disposition. The answer would ap-
pear to be that humans deliberately create and carefully
maintain "reverse dominance hierarchies" (Boehm
1984b, n.d.) whereby the potentially subordinate adult
males — the rank and file — band together assertively to
limit the dominance of more aggressive or otherwise
outstanding individuals. I have surveyed a large number
of egalitarian societies (simple foragers, complex forag-
ers, and many others) from various cultural areas and
found evidence of such behavior. The result is far from
an absence of male rivalry or inequality, but there is a
sharply negative response if someone behaves too asser-
tively or tries to control other males. This syndrome
relates to personal autonomy: males will not counte-
nance being "bossed around" (e.g., Woodburn 1979).
Among Knauft's simple foragers it also involves the eq-

uitable distribution of foodstuffs and, sometimes,
women within a group that believes in cooperation and
equality.

An evolutionary hypothesis would be that dominant
egalitarian control of leaders arrived with the moral
community (Boehm 1982), a socially manipulative group
that is strongly (and often consensually) "judgmental."
Such an argument has obvious relevance for coevolu-
tionary theory. In effect, if a cultural behavior can sup-
press selected aspects of a marked behavioral disposition
such as the tendency to form dominance hierarchies,
without even physically eliminating individuals who in-
herit large doses of that disposition, then at the pheno-
typic level a force set in motion by genetic selection is
being powerfully and radically — but not permanently —
transformed by cultural selection,- this is implemented
through group decisions (see Boehm 1978). It is possible
that this same reverse-dominance-hierarchy cultural be-
havior is also modifying the gene pool, since males who
overstep sometimes are executed.

To explain the evolution of social behavior, we must
choose wisely among extant models for prehistoric so-
cial life, all of which have been problematic. Knauft has
facilitated a major stride in this direction. Once the best
possible model is chosen, evolutionary "triangulation"
can be improved. For example, judging from the behavior
of the African great apes, it sems likely that the common
ancestor exhibited a decisive and forceful triadic-
intervention style of conflict management within the
group (Boehm n.d., Goodall 1986; see also de Waal 1982).
By contrast, in Knauft's simple nomadic foragers very
weak egalitarian leadership combined with an absence
of effective institutions to mediate conflict can result in
higher rates of intragroup homicide compared with
those inferable for chimpanzees. However, Knauft over-
emphasizes the fact that simple societies appear to take
little interest in mediating conflicts. I think that a close
reading of certain ethnographic accounts (e.g., von
Fiirer-Haimendorf 1967) would suggest that some seri-
ous attempts are made in this direction but that, since
for a nomad to pull up stakes and move to another open
group involves no very heavy investment loss, usually
one protagonist pursues a strategy of avoidance rather
than immediate reconciliation. In the more complex
egalitarian societies, including clan-based territorial
ones that stimulate male lethal aggressiveness by war-
rior socialization and training, forceful control of inter-
nal conflict by high-ranking males is still preempted by
egalitarian behavior. But in many of these feuding socie-
ties (e.g., von Fiirer-Haimendorf 1967, Boehm 1984^3) in-
stitutionalized types of noncoercive conflict manage-
ment can become quite effective, even in the absence of
any really authoritative human agency.

Thus it would appear that when Knauft's prehistoric
simple foraging nomads became settled, an already high
group-internal homicide rate came under more immedi-
ate control. This took place as territorial behavior be-
came more developed, groups became more bounded,
and adjacent groups came to have more to fight about
and therefore began to train warriors. When inevitably

412 I CURRENT ANTHROPOLOGY Volume 32, Number 4, August- October 1991

the warriors quarreled among themselves and threat-
ened to divide the group and make it vulnerable, people
coped with this perceived threat. Their mode of inter-
vention was not a strong, chimpanzee-gorilla style of
leadership (even though we know from what happened
later with kingdoms and states that the potential for
that was still strong in the species) but the creation of
cultural rules and institutions that regulated but realis-
tically did not try to suppress entirely the homicidal
self-expression of individuals.

A fascinating problem for future exploration is how
the vigilantly egalitarian dominance behavior of follow-
ers described above was ''disarmed'' during the transi-
tion to kingdoms and states. State-formation theory will
be incomplete until this important political element, in-
volving both genotypic dispositions and purposive cul-
tural selection, is taken directly into account.

Knauft has taken a very important first step in differ-
entiating simple from complex foragers and has con-
vinced me that the simple foraging society, with its sig-
nificant political differences from other egalitarian
societies, is the appropriate model for earlier prehistoric
''triangulation.'' Another necessary ingredient of a more
effective approach would be a move away from the
overly mechanistic functional frameworks that have
served anthropology rather well over the past century
and a more aggressive search for documented instances
in which the deliberate kind of ''strategizing'' shapes
evolutionary outcomes (see Boehm 1978, Vayda 1989).
I believe that one of the first such instances was the
emergence of control of leaders by followers instead of
vice versa. This was the beginning of Knauft's simple
foraging society — surely, as he says, the longest-lasting
social form invented by humans.

ROBERT KNOX DENTAN

Departments of American Studies and
Anthropology, State University of New York/
University at Buffalo, Amherst, N.Y. 14260, U.S.A.
20 III 91

As usual, Knauft's theories and speculations are clear,
cogent, and stimulating. His argument that ''culture''
potentiates an evolutionary mechanism of group selec-
tion is attractive. His assertion that competitive male
dominance hierarchies among people are analogous but
not homologous to those found among other anthro-
poids is precise and, given his evidence, inarguable.

Native Australians constitute an exception that tests
his characterizations of "simple human societies." I am
completely ignorant of the prehistory or ethnography of
the area. Still, along the path that the ancestors of native
Australians probably took lies New Guinea, a huge is-
land so extraordinarily rich (e.g., in carbohydrates from
taro and sago) that native foragers using simple stone
tools (Townsend 1969) routinely formed settled commu-
nities of considerable complexity (e.g., Ellen 1988;
Townsend 1974, 1977, 1990; cf. Langub 1988). The in-
fluence of social evolutionary theory has obscured the
fact that its founders saved a space, albeit secondary, for

"degeneration theory," the idea that some societies are
in some ways less "advanced" than their ancestors in
social evolutionary terms (e.g., Tylor 1958 [i87i]:4i-
48). Foragers moving from a relatively rich into a rela-
tively harsh habitat might change their foraging strate-
gies faster than their ideologies, particularly those
ideologies involved with religion, which is often subject
to "cultural lag." Shifts in foraging strategy due to new
circumstances are not rare cross-culturally (e.g.. Head-
land 1988, 1990; Townsend 1990). Perhaps part of the
complexity of native Australian culture grew out of a
more complex form of social organization adapted to
richer surroundings.

It is important to remember that, in talking of social
evolution, Knauft is particularly interested in a supposed
sequence of adaptations. His article would be stronger,
as he indicates in a couple of places, if he had focussed
more on culture as an adaptation and less on culture
as a factitiously static stage in an implicitly unilinear
scheme. The foregoing speculation about native Austra-
lians, whatever its worth, exemplifies a focus on adapta-
tion rather than evolution. My concern with this matter
stems from my belief that patterns similar to those
Knauft attributes to simple human societies may de-
velop as adaptations to a number of other environments,
for example, slaving (Dentan n.d.a] or religious en-
clavement (Dentan n.d.b). There are differences: "nega-
tive peace" rather than the "positive peace" that Knauft
discusses tends to be salient among people who flee slav-
ers (for this distinction, see Sponsel 1989:29-30; Ste-
phenson 1990:5), and child rearing by enclaved religious
communities tends to be harsh. But Knauft's supposed
evolutionary sequence is difficult to demonstrate arche-
ologically, and the emphasis on social evolution (i) ob-
scures similarities in adaptation that might be enlight-
ening, (2) could foster erroneous despair about the
possibility of recreating the admirable qualities of "sim-
ple human societies," such as egalitarianism and love of
peace (e.g., Denich 1987), and (3) can lead to the stereo-
typing of peoples as "harmless" or "fierce" on the basis
of observations limited to particular times and places.

Knauft's admirable philosophical achievement in this
article is to free Darwinism from the Social Darwinist
scientism that has recently cloaked it (e.g., Daly and
Wilson 1991). Since human egalitarianism and peace-
ability are social facts as real as human hierarchies and
violence, any theory that purports to explain the latter
should explain the former as well, not just dismiss them
as epiphenomenal or "reaction formation" (cf. Robar-
chek 1991). A politically neutral Darwinism that ac-
cepts the fact of cultural adaptation should prove a pow-
erful intellectual tool.

THOMAS M. KIEFER

18 Worthington St., Boston, Mass. 02120, U.S.A.
15 IV 91

Humans evolved in groups of about 25 + or so, although
in very favorable situations considerably larger groups
might have been possible. Notwithstanding the cultural

KNAUFT Violence and Sociality in Human Evolution \ 413

changes of the last 12,000 years, the optimal number of
fellow humans with whom we can be really intimate
has not changed. I am generally in agreement with
Knauft's argument; unfortunately, our really good exam-
ples of simple foragers (!Kung, Shoshone, Eskimo, Aus-
tralian Aborigines, et al.) are ecological oddities, but one
works with the data one has!

As Knauft implies, the evolution of the interface be-
tween human sociality and the capacity for violence (in-
cluding the reptilian-mammalian-primate baggage
which we still carry) must be analytically separated
from what humans fight about. Simple foragers (and all
humans) contend over access to sexual partners,
breaches of norms of reciprocity, and possibly some
types of incorporeal property. Conflict over corporeal
property, territory, and political office is very recent.

The term ''violence,'' in both academic and common
usage, can be a trap for the unwary. At least five mean-
ings may be identified in English usage: (i) any agonistic
conduct, (2) any agonistic conduct resulting in injury, (3)
any conduct intended to cause injury, (4) in the classic
Marxian sense, the use of tools as a means of social dom-
ination, and (5) culturally disapproved conduct, as when
we describe disapproved police conduct as "violent" but
not routine police activities (see Wolff 1971).

This last use of the term is particularly troublesome
in the analysis of simple societies in that Westerners
since the Renaissance have increasingly regarded indi-
vidual violence as "irrational," leaving "rational" vio-
lence to the mysteries of the state. But in prestate socie-
ties, an opposition of "reason" to "violence" surely
makes no sense. Humans can, and very often do, ratio-
nally use violence to achieve individual and group goals.
Indeed, the rational use of violence in any simple face-
to-face society more often than not involves feigning the
running-amok "reptilian" conduct that Knauft describes
(see Kiefer [1972] for one of many ethnographic exam-
ples). For most mammals and all higher primates, ago-
nistic behavior can be playful: humans are certainly the
most playful primates of all. But to feign agonistic con-
duct playfully presupposes that others recognize the
"real thing" as a capacity in themselves.

Knauft argues — correctly, I think — that violence in
simple human societies was generally dysfunctional for
both the society and the individual: population densities
were small, and "picking up one's marbles and leaving"
was the efficient way of dealing with hostility. However,
the capacity for "reptilian" violence, whether maladap-
tive, adaptive, or whatever, did in fact persist. I suspect
that one of the reasons was precisely that the playful
feigning of agonistic conduct makes such great social
drama and, provided that it is subject to self-control,
indeed contributes to social solidarity in the long run.

Self-control is absolutely crucial, and it presupposes a
self. In saying that each human has a "self" I mean only
what George Mead (1934) meant: that one can act to-
ward oneself as if one were an other. The evolution of
this capacity is critical to the cortical control of vio-
lence. Being able to act toward oneself (and imagination
is an act) has at least two significant consequences for
human socialitv. First, it enables humans to project

long-range intentional self-interested strategies. Second,
without it no "morality" — at least as we would recog-
nize it — would be possible. Knauft rightly emphasizes
that in simple societies there is a tension between the
force of cultural norms and individual desire to break
them. I would argue that this tension is intrinsic to the
human condition as a self-conscious animal and is pecu-
liar to simple foraging societies only in the sense that it
evolved there. Mead (1934) stressed the importance of
fully developed human language in the evolution of the
capacity for self, but all recent studies of apes have
shown that this is outdated. Some degree of "proto-self-
control" and a high degree of ability to predict others'
(and presumably one's own) intentions is deeply rooted
in our primate past (Jolly 1985:400).

In my view, the relationship between the evolution of
human violence and sociality is mediated by the gradual
perfection of the capacity for self-control, which was
highly adaptive in many ways. As Knauft implies, it is
probable that, in the very long period between the emer-
gence of (normative?) food sharing and 12,000 b.p. or so,
most interpersonal violence (and feigning thereof) in-
volved sexual conflict, and access to women may have
been one important factor. Ambivalence over sexual
self-interest and cooperative group life was plausibly one
adaptive function of self-control, particularly if losing
self-control did not win the woman and might easily
lead to exile or death.

KEITH F. OTTERBEIN

Department of Anthropology, State University of New
York at Buffalo, Buffalo, N.Y. 14261, U.S.A. 27 iii 91

Knauft's "typological" mode of presentation creates an
apparent homogeneity that is not supported by the data.
One receives the impression that most intragroup vio-
lence is related to "male sexual disputes over women,"
that "feuding or warfare does not appear pronounced,"
and that "strong emphasis is placed on cooperative soci-
ality." Rather than draw his support from only a few
examples, usually the IKung, Mbuti, or Inuit, it would
have been methodologically preferable for him to have
utilized the subsample of 39 societies drawn from the
Atlas of World Cultures — coding these societies for key
variables such as brawling, homicide, capital punish-
ment, feuding, and warfare.

I believe that the alleged similarity of simple human
societies would disappear if Knauft made a genuine ef-
fort to look for differences. In my comment upon his
previous article I pointed out (1987:485) that the fre-
quencies of each form of violence vary from society to
society and offered a theory to explain the variation (see
also Otterbein 1988).

Variation also exists in the frequency of warfare.
Knauft's 39-society subsample greatly overlaps a sample
of hunter-gatherers utilized by Ember (1975, 1978). The
similarity of selection criteria insures that many socie-
ties must appear in both samples. Ember obtained
frequency-of-warfare ratings for many of the societies,
in her sample. Using this sample and her codes in a

414 I c

URRENT ANTHROPOLOGY VoluHie 32, Nuuihei 4, August—Octobei 199 1

cross-cultural study relating dependence on hunting to
warfare frequency (Otterbein 1989), I have found that
the greater the dependence upon hunting, the greater the
frequency of warfare (i = +.45, p < .01). I was able to
perform this analysis because of the great variability in
warfare frequency; in a sample of 3 1 societies, the num-
ber of societies for each point on the variable is 6, 4, 6,
4, 2, 2, 4, and 2.

JOHN PADDOCK

Apartado 464, 68000 Oaxaca, Oax,, Mexico, 5 iv 91

Visiting Mexico to explain the journal he was to launch
the following year (current anthropology), Sol Tax
mentioned in 1959 while chatting with Ignacio Bernal
that the paradise myth probably referred to vague mem-
ories of preagricultural times, before the appearance of
war. The idea stayed with me. In 1968, with violence
breaking out in France, Mexico, and elsewhere, I was in
a Valley of Oaxaca Zapotec community whose people
boasted of not having, in contrast with their neighbors,
a local problem of interpersonal violence. Skeptical, I
checked and later with the aid of others documented an
astonishing difference. Homicide rates were high in
most Oaxaca villages but almost incredibly low in some,
though they lacked police, courts, or prisons (Paddock
1974, 1975, 1976a, b, 1979^, b, 1980). Evolution does
not begin with the great apes. The antiviolent tendency
we documented with statistics and observations in vil-
lages apparently like their violent neighbors might re-
late, then, to the near absence of violence in pre-primate
species, even more ancestral than the violence among
the great apes.

Among the several hundred references Knauft lists
there is no citation of the journal Aggressive Behavior,
organ of the International Society for Research on Ag-
gression, though works by several members of the Soci-
ety do appear. This might be because that journal tends
to publish on extremes — laboratory animals (psycholo-
gists) and urban-industrial communities (sociologists,
anthropologists) — while Knauft's area is intermediate.

Having personal experience only with Zapotec Indian
villages in far southern Mexico, I am stimulated by the
observations that Knauft makes about simple and
middle-range societies, for traits of both classes are sig-
nificant in these villages (which of course do not fit into
either). In pre-Hispanic times they were much like
''peasant'' communities, and the Spanish conquest in-
troduced some European ideas without changing their
status as rural producers for urban consumers. However,
Mexican independence and modernization have invali-
dated the "peasant" status even for many almost purely
American Indian communities when they participate in
urban life to the degree that Valley of Oaxaca villagers
now do.

Our research showed nearly incredible differences in
homicide rates between villages that are closely similar
with respect to the commonly ascribed "causes" of vio-
lence such as alcohol, sexual access, boundary quarrels.

crowding, poverty, and political ambition. The crucial
difference, we found, was local consensus on the desir-
ability of living without interpersonal violence. While
we did not locate the origin of the phenomenon, a recent
case may help. A Oaxaca village far from the central
valleys we studied once had one of the highest internal
homicide rates in the world, but the women rebelled
and banned alcohol, arms, and violence (Greenberg
1989:230-34; Paddock 1990). Our antiviolent villages
do not forbid alcohol and in fact often use it to excess,
but they have traditions of antiviolence that are thor-
oughly established, perhaps centuries old. (The term
"antiviolent" was applied here because these commu-
nities are not totally successful in preventing inter-
personal violence, hence "nonviolent" would be too
strong.)

Ranking does exist in our antiviolent villages, but it
is controlled and diffused. Official posts are assigned to
all adult males by turns, though assignment of power to
elders is separate and there is also unofficial ranking.
Such villages are part of a larger society and are assigned
low ranks within it, but they are also social units, com-
munities with individual characters of which they are
proud. More remote communities, and groupings of
them, are nearly complete societies in themselves, al-
most lacking participation in state and national soci-
eties.

Hawaiian ho'oponopono (Shook 1985) also seems not
to fit either of Knauft's classes or in some ways to fit
both, though it shares some traits with Valley Zapotec
villages practicing antiviolence. It is an effective device
for conflict resolution but in a society marked by male
ranking and, apparently, intercommunity violence as
well as intracommunity antiviolence. (Interestingly, ur-
ban social workers are finding it effective among non-
Hawaiians.)

The humanity of scientists is often ignored; Knauft is
to be commended for finding ways to compensate for it.
Born and socialized in large societies with high rates
of external homicide (i.e., war), archeologists commonly
attribute similar orientations to ancient societies while
ignoring evidence of other sorts, on the apparent as-
sumption that war making is an inescapable human
trait. Shocked by personal experience in violent Oaxaca
societies, ethnographers may simply deny the reality of
Oaxaca antiviolence, abundantly documented though it
is (Paddock 1988, 1990).

The vast majority of humanity now lives neither in
simple nor in middle-range societies. Therefore it would
be helpful to see a proposal for classifying the societies
Knauft has prudently avoided, even though that will be
difficult and controversial.

LARS RODSETH

Department of Anthropology, University of Michigan,
Ann Arbor, Mich. 48109, U.S.A. 18 iv 91

This paper demonstrates what anthropology can do in
the 1990s if the resources of its various subdisciplines

KNAUFT Violence and Sociality in Human Evolution | 415

are brought to bear on the one problem that unites the
field: human social evolution. In synthesizing findings
of primatology, archeology, and ethnology, Knauft
brings new respectability to general or philosophical
anthropology — the only kind of anthropology that may
yet save the field from its own centrifugal forces.

While I am sympathetic, then, with Knauft 's general
aims and admire the range of data he has marshalled to
press his case, there seems to me another interpretation
of those data that he does not adequately consider. His
own view is that the distinctive social patterns of ethno-
graphically known simple foragers demonstrate that hu-
man social evolution is ''U-shaped,'' with great-ape soci-
eties ironically resembling groups of complex foragers
and food producers more than they do simple foraging
societies. While such a U-shaped trajectory is of course
possible, the irony itself points to the alternative inter-
pretation: simple foraging societies as known from the
ethnographic record may not be representative of such
societies in the Pleistocene and may in fact be radically
different, precisely because they have adapted to mar-
ginal areas outside the main currents of human social
evolution. Although Knauft notes that "highly decen-
tralized" foragers are likely to be "underrepresented in
the archeological record," he does not mention that they
are also probably oyerrepresented in the ethnographic
record compared with less marginal foragers who pre-
sumably evolved into or were assimilated by food pro-
ducers. This familiar caveat about extant hun-
ter-gatherers is perhaps best illustrated by Gellner's
(1988:36) striking analogy:

Let us suppose that in the twenty-second century
the world is fully industrialized, but that somewhere
near Yasnaya Polyana, or in the English shires, a few
communities of Tolstoyans or William Morris en-
thusiasts survive, firmly rejecting the values and
practices of the surrounding world, and perpetuating
the lives of muzhiks or of English medieval crafts-
men, or what they fondly imagine to be such. How
justified would a twenty-second-century anthropolo-
gist be in studying such communities, and on this
basis reconstructing a general model of the agrarian
world as it truly was?

If modern foragers are indeed "fringe groups from the
Stone Age," perhaps the best we can do is try to recon-
struct the central habitats of that age so that, by combin-
ing this information with what we know about modern
foragers (and perhaps nonhuman primates as well), we
can retrodict how typical Pleistocene foragers might
have lived.

Knauft's own reasoning suggests a way to begin such
a reconstruction. Following Foley (1989), he notes that
"in the social organization of the common ancestor of
apes and humans, a tension could plausibly have existed
between chimpanzee-like patterns of male philopatry
and selective pressures for more open, dispersed, and
flexible alliances." With the evolution of Homo and the
elaboration of cultural transmission, Knauft argues, the
pattern of social organization "turned decisively away

from male philopatry and closed fraternal interest group-
ings and toward more open and more flexible social net-
works." Then, with intensified production and the es-
tablishment of "stable and valuable resource bases,"
human social organization reverted, in his view, to the
chimpanzee-like pattern.

Now, there are two points to be made here. First, even
if we accept simple foragers in the ethnographic record
as models for those of the Pleistocene, the very fact that
humans all over the world consistently shifted, with the
transition to more intensive foraging and food produc-
tion, from their open flexible groupings to the atavistic
pattern of relatively closed male-bonded groups suggests
that the tension between these two poles was not deci-
sively resolved with the evolution of Homo. It tribal and
chiefly societies all over the world spontaneously "rein-
vented" a pattern of social organization resembling that
of chimpanzees, the most reasonable explanation may
be that this pattern survived in the human social reper-
toire from the common ancestor of humans and African
apes and was only dormant through the long period of
simple foraging bands.

Second, this pattern may not have been dormant at
all. Relatively closed fraternal interest groups are pre-
sumably formed by both chimpanzees and humans to
defend concentrations of resources which generally do
not exist in extant populations of simple foragers. In
the human case, such concentrations are now artificially
generated through food production, but in the Pleisto-
cene they must have occurred naturally in at least some
human habitats, just as they do in chimpanzee habitats.
And if humans consistently jump at the opportunity to
defend artificial concentrations of resources by forming
fraternal interest groups, there seems little doubt that
they would have done the same thing wherever natural
concentrations occurred. The picture that emerges has
the common ancestor of humans and African apes living
in relatively closed, male-philopatric groups, then
spreading out from central areas where resources were
concentrated into marginal habitats where resources
were dispersed. In these marginal areas, the open, flexi-
ble groupings characteristic of extant hunter-gatherers
would have emerged, but in the central areas even sim-
ple foragers would have remained in relatively closed,
male-philopatric communities. Eventually these would
have been transformed smoothly into patrilocal tribal
societies, with no intervening stage of flexible social or-
ganization. The U -shape of human social evolution pro-
posed by Knauft would be an illusion created by casting
an adaptation to extreme conditions as a global evolu-
tionary stage.

While such a scenario deserves more attention than
Knauft gives it, human residence patterns may be beside
the point anyway if one's ultimate aim is to understand
patterns of social relationships. In contrast to nonhu-
man primates, in which philopatry, consanguinity, and
cooperation tend to coincide, humans display dispersal
patterns that predict patterns of kinship and cooperation
only weakly, if at all (Rodseth et al. 1991). In this sense,
my colleagues and I are stressing "the distinctive open-

4i6 I CURRENT ANTHROPOLOGY VoluHie 32, Nuuihei 4, August-Octobei 1991

ness of human social networks/' but this openness re-
fers to patterns of social relationships independent of
spatial proximity, making it conceptually distinct from
the ''open" residence patterns Knauft attributes to sim-
ple foragers.

Thus, leaving aside the issue of flexible versus patrilo-
cal residence, we might ask with regard to Knauft's sim-
ple foragers whether men are permanently cut off from
their close male kin in the way that most male primates
are when they migrate to other groups. The answer, to
my knowledge, is no. In fact, men usually maintain cru-
cial alliances with their natal kin whether they reside
with them or not. The fact that these alliances do not
always constitute ''fraternal interest groups'' in the con-
ventional sense is irrelevant, since such groups are con-
ventionally defined in part by patrilocal residence.
Whatever their residence patterns, it may be argued, hu-
mans are very generally if not universally "male-
bonded." (Women of course also tend to maintain rela-
tionships with natal kin but are not "female-bonded"
in that they seldom form alliances with each other for
purposes of aggression against members of their own sex
[Rodseth et al. 1991].) Even matrilocality, which dis-
perses male kin at the level of the household or "neigh-
borhood," tends to reassemble them at the level of the
community or tribe, especially for purposes of long-
distance warfare against unrelated groups (Rodseth et al.
1 991). This is why we describe matrilocality as a case of
"agnatic sprawl," with the whole network of intermar-
rying households often acting as one large "fraternal in-
terest group" — not because brothers reside together, of
course, but because the community or tribe as a whole
constitutes a "power group of related males" (Otterbein
and Otterbein 1965:1472).

Reply

BRUCE M. KNAUFT

Atlanta, Ga., U.S.A. 2 i 91

There is a great need for testable models of the relation-
ship between cultural and biogenetic selection in spe-
cific periods of human evolution. My paper is but a
rough initial effort in this direction, and even then it
approaches human evolutionary theory through the top-
ical "back door" of violence and sociality. I hope that
further commentary on both my paper and those of Rod-
seth et al. (1991) and Manson and Wrangham (1991) will
stimulate a richer consideration of human evolutionary
theory as well as of the developmental trajectories of
human violence and sociality per se.

Acknowledgment should be made of Boehm's inde-
pendent prior assessment (1984:12) that "social and po-
litical hierarchy is strong particularly in the other terres-
trial primates with which we like to compare ourselves,
and it is strong with us, too, after we centralized our
polities and economies to form states,- however, it is

virtually absent in the critical transitional form of hunt-
ers or hunters and gatherers" and his further assessment
(p. 13) that "culture acquires . . . force precisely at the
point that a moral community is formed — a face-to-face
community of individually self-conscious individuals
who, through verbal symbolic communication, reach
the point that they understand their common identity
in terms of a common self-interest." As have I, Boehm
has suggested the plausible association of this process
with the development of elaborate protolanguage, cul-
tural transmission, and rule-influenced behavior in hu-
man evolution as early as H. erectus (e.g., Boehm
1982:417). He has also given more attention than I to
the parasitic selection (and even group selection) that
might operate prior to the development of widespread
cultural transmission. In his view, these patterns pre-
adapt primate conflict interference for subsequent devel-
opment into human patterns of culturally mediated con-
flict management (Boehm 1981).

Boehm is correct to criticize my paper for not giving
sufficient attention to informal conflict mediation tech-
niques in simple human societies (e.g., Briggs 1970;
Marshall 1976; TurnbuU 1961, 1982; see Knauft 1987^3:
376). The question of how these societies maintain "re-
verse dominance hierarchies" is correspondingly criti-
cal. Boehm's query goes well beyond the misplaced criti-
cism of sociobiologists that since at least some
inequalities of skill, leadership ability, and number of
offspring exist in simple societies, these must function
collectively to foster the reproductive success of domi-
nant, aggressive leaders. Such inequities are both cultur-
ally and self-consciously minimized in these societies.
Perhaps more important, they tend not to "line up" in
a mutually supportive way: aggression, material ag-
grandizement, dominant leadership, and reproductive
success do not coalesce as a Darwinian set. As Boehm
suggests, it will be crucial to explore how various ine-
qualities are culturally precluded from coalescing and
under what conditions they begin to coalesce as more
formalized and centralized leadership forms develop.

Boehm (cf. particularly 1981) alludes to the possibility
of culturally mediated group selection,- my own analysis
places much greater emphasis on this factor. It seems to
me that the high costs of altruistic behavior in both sex-
ual restraint and conflict mediation render it unlikely
to be selected for systematically in humans through par-
asitic selection or noncultural interdemic group selec-
tion alone (cf. Wilson 1980). That cultural transmission
competes with biogenetic transmission in influencing
behavior dovetails with Boehm's (1989) suggestion, im-
plicit also in Kief er's comment here, that human nature
is better analyzed as a series of "universal dilemmas" — a
distinctive set of competing behavioral impetuses —
than as a series of single universal traits or motivational
continuities. The primary competition is, in my view,
between Darwinian propensities toward maximal bioge-
netic propagation and cultural propensities toward ide-
ational transmission and dissemination.

Rodseth's suggestion that I am not sensitive enough
to the contextual factors influencing qurrent hunter-

KNAUFT Violence and Sociality in Human Evolution \ 417

gatherer variants is arguably more applicable to his own
article than to mine: while developmental variation of
social organization in the course of human evolution
is foregrounded in my model, his ultimately collapses
human variation into a single aggregate type. There have
certainly been decisive changes in social organization,
as in fertility and mortality, in the course of human evo-
lution, viz., with the Neolithic and industrial revolu-
tions. I suggest that the human revolution, with the
emergence of symbolic culture in Homo, is at least as
momentous as the advent of agriculture or industry in
changing patterns of social organization. If assessed crit-
ically in the context of known historical trends, the eth-
nographic record of simple societies can help illuminate
the selective parameters of this earlier revolution.

Human socioecology as it developed in the course of
human evolution is in many ways consistent with that
of simple societies documented in this century: both are
characterized by large home ranges, dispersed, patchy
resources, omnivory, and high mobility. The striking
human trajectory from H. eiectus until quite recently
has been, as Rodseth would appear to concede, one of
expansion into environments that were at the outset
marginal. As Sahlins (1972) suggests, the prevailing eco-
nomic tendency in simple societies is to minimize both
labor and food surplus and to keep yields well below
environmental limits. The low-intensity human ecology
and the patterns of sociality and violence associated
with simple societies are likely to have been quite com-
mon in human evolutionary history. Much has been
made of archeological evidence of complex hunter-
gatherer societies (particularly in Europe) in the late Pa-
leolithic and Mesolithic, and a few resource-rich areas
could have spawned complex hunter-gatherers yet ear-
lier. When the evolution of Homo is considered as a
whole and on a global scale, however, a nonintensive
forager model is far more appropriate than one based on
complex hunter-gatherers.

Rodseth's apparent recognition that male residence,
bonding, and alliance in simple societies are frequently
affinal and even Active as well as patrifiliative tends to
dissolve his lingering sociobiological assumption that
human males coalesce predominantly in fraternal inter-
est groups, i.e., in ''power groups of related males." As
I have shown, the cross-cultural evidence strongly dis-
confirms this suggestion. In most simple societies the
effective social group is a bilateral band and not a dis-
persed group of closely related agnates.

Betzig's comments contains several serious misunder-
standings and echo the kind of unsophisticated sociobi-
ology so effectively critiqued by Kitcher (1985). Most
curious is her opening claim that my article is predi-
cated on showing that ''culture is keeping Gebusi from
behaving adaptively'' when the word "Gebusi'' never
even appears in it. I do not suggest that culture in simple
societies is generally maladaptive or that "culture tri-
umphs over nature." I stress that culture is group-
adaptive, that the tension between cultural and bioge-
netic selection is never resolved, and that biogenetic
propensities for fitness maximization continue to be

acted out. At the same time I argue that these propensi-
ties are crosscut by symbolically externalized and psy-
chologically internalized cultural rules. ^ By failing to
give sufficient attention to this dimension of my work,
Betzig reads my suggestions about middle-range socie-
ties as an admission that polygyny is almost universal
among their leaders and that dominance and violence
are almost invariably related to reproductive success.
These are claims I do not make. Despite the increased
incidence of polygyny as a leadership prerogative in New
World middle-range societies, a majority of these socie-
ties (31/5 1; 61%) still lack this prerogative (Feinman and
Neitzel 1984:58-59). Aggression and reproductive suc-
cess are not invariably associated in these societies (see,
e.g., Moore 1990), and any such association as exists
may represent sociocultural correlation rather than so-
ciobiological causation. Finally, that Betzig is able to
cite a few cases of polygyny in simple societies does not
contravene the empirical trend to the contrary: "Nim-
koff and Middleton (i960) demonstrated from HRAF
data long ago that there is a highly significant correla-
tion between type of family system and subsistence pat-
tern,- the independent, or monogamous, family is the
most common in hunting and gathering cultures" (Kin-
zey 1987:111).^

The inappropriateness of asserting homogeneity for
human societies as a whole and then drawing connec-
tions between aggregate primate patterns is only in-
creased by the variations among great apes. A male-
coalitional model of violence does not fit bonobos,
orangutans, and gorillas, being characteristic primarily
of chimpanzees. The persistence of the notion of
fraternal-interest-group violence in human evolution
may reflect the disproportionate attention once given to
models of baboon dominance and aggression as a proto-
type for human violence. Strum and Mitchell (1987)
have pointed to the abundant counterevidence about ba-
boon behavior and social organization that was available
even in the 1960s.

Recent considerations of human and ape sexual mor-
phology (e.g., Nadler and Phoenix 1991) suggest that

1. It is important to realize that Darwin (1981 [1871]) also suggested
that collective moral traits, such as sympathy, give human socie-
ties an important adaptive advantage over groups composed en-
tirely of self-interested individuals. The possibility of group-level
adaptation raised by Darwin himself (see especially 1981 [i87i]:72,
98-104, 164-66) has been curiously neglected by neo-Darwinians.

2. Betzig's interpretation of Ache hunting and reproductive success
may also be slanted. Kaplan and Hill (1985:133) found no significant
difference in the number of certain or estimated total children born
to good versus poor hunters; they found a difference only in the
number of illegitimate or "possible" children. I have been informed
that Ache believe multiple male inseminators to be co-conceivers
of a single illegitimate child. In light of this, the pressure to desig-
nate only one of these persons as "the father" to satisfy the etic
perspective of the ethnographic investigator could introduce sig-
nificant bias. As Kaplan and Hill themselves point out, the docu-
mented greater survivorship of children of good hunters may be a
product of genetically mediated physical robusticity that increases
both men's hunting skill and the birth weight and survivorship of
their children.

4i8 I CURRENT ANTHROPOLOGY Voluuie 32, Nuuihei 4, August-Octobei 1991

humans had neither a polygynous mating system simi-
lar to that of gorillas nor a multimale mating system
similar to that of chimpanzees. Given the comments of
Abler concerning my supposed collapsing of human and
ape ''polygyny/' I reemphasize that the formal similarity
between some dimensions of ape and middle-range hu-
man societies is an analogy only. The point is that what
appear to some behavioral ecologists and sociobiologists
to be cross -species similarities are in crucial ways quite
different, and this fact is highlighted by the contrasting
simple-society patterns. The evolutionary trajectory of
violence that I propose does not suggest that patterns of
leadership and mating in middle-range societies can be
explained in any simple way by sociobiological princi-
ples, and indeed it argues against that notion.

The cogent comments of Abler, Dentan, Paddock, and
Kiefer lead me to reemphasize that the enormous varia-
tion in middle-range societies is collapsed here only for
the sake of broader comparison. Post-state development
is outside my purview, which is dauntingly broad as it is
(cf. Paddock, Abler). I have confronted variation among
middle-range societies in the Melanesian context (e.g.,
Knauft 1985:339-48; 1989b, 1990c) and with respect
to warfare (i99ofo). My awareness of this ethnographic
complexity is one reason I make little attempt to differ-
entiate sociodevelopmental patterns within this cate-
gory. Comparative patterns of socioeconomic and politi-
cal variation in small-scale sedentary societies are well
addressed by Feinman and Neitzel (1984) and Upham
(1990^), and systematic studies of prestate warfare have
iDeen made by Otterbein (1985) and Ross (1985, 1986).
I have no argument at all with Abler's and Paddock's
assessments that relatively nonviolent middle-range so-
cieties exist and that matriliny may be present and fra-
ternal interest groups absent. My point is simply that
middle-range societies in the aggregate exhibit a much
greater propensity toward fraternal interest grouping,
ethnocentrism, and collective armed conflict than sim-
pler ones.

The purpose of my broader comparison was (a) to inte-
grate data from diverse fields into a conceptual frame-
work for the evolutionary study of human violence and
sociality, (b) to counter evolutionary and biobehavioral
models that uncritically project nonhuman primate
tendencies onto humans, and (c) to suggest alternative
and more sophisticated models that consider cultural as
well as biosocial dimensions of human evolution. Evolu-
tionary typologies are by their nature Popperian in func-
tion, serving as a lightning rod for counterexamples and
refutation. An analysis on a broad scale elicits counter-
examples at a more detailed level.

One of the drawbacks of evolutionary typologies such
as mine is their neglect of complex evolutionary coun-
tercurrents, including devolution. I admit that a number
of sedentary societies may exhibit egalitarian forms of
social organization, particularly under conditions of ex-
ternal oppression (Abler, Dentan). This point has been
made particularly forcefully by Jayawardena (1963, 1968;
see Bern 1987). My argument encompasses this develop-
ment not in evolutionary terms but in structural ones:

whatever the socioecological or historical pattern of
their development, simple societies are likely to evi-
dence a relative lack of fraternal interest groups, warfare,
blood feuds, and raiding. Conversely, hunter-gatherers
exhibiting significant male political-status differentia-
tion, such as selected Australian Aborigine groups, may
be expected to have a greater incidence of collective vio-
lence and warfare than more decentralized societies in-
habiting similar environments. Thus developmental
correspondences are not uniform in each case but are
evident in aggregate evolutionary trends: most societies
of the period between 1,500,000 b.p. and 13,000 b.p. are
likely to have been closer to the ''simple'' end of the
continuum. By contrast, most sedentary societies are,
through both endogenous and exogenous processes,
likely to be closer to the "complex" end. Isolated cases
to the contrary do not invalidate the model; what would
falsify it is many cases of politically decentralized sim-
ple societies with a high incidence of warfare, blood
feuds, and raiding.

Abler's suggestion that warfare in rniddle-range socie-
ties can in large part be accounted for as a function of
Western encroachment deserves particular consider-
ation. This argument (which has been strongly made
as well by Blick [1988], Ferguson [1990], and Ferguson
and Whitehead [n.d.]) is the first step toward a more his-
torically sensitive understanding of the relationship be-
tween colonial and precolonial patterns of violence and
warfare. There is no question but that Western en-
croachment has transformed, redirected, and in some
cases intensified indigenous conflict; the introduction
of guns and the coopting of indigenous forces in the ser-
vice of colonial military objectives are well documented
from numerous parts of the world. It is erroneous, how-
ever, to assume that intense collective violence was ab-
sent aboriginally or that the homicides involved in in-
digenous warfare were demographically insignificant.
Precolonial warfare among middle-range societies was,
as I have pointed out, strongly evident in native North
America, South America, Africa, and Melanesia. The
Melanesian cases are particularly dramatic and well doc-
umented (see review in Knauft 1990b and also Shank-
man 1 991), and archeological evidence suggests similar
patterns in the New World (e.g., Haas 1990).

Otterbein's finding that extreme dependence on hunt-
ing may be associated with warfare is interesting and,
in terms of my analysis, expectable. A large proportion
of the societies in Ember's (1978) hunter-gatherer sam-
ple, upon which Otterbein's analysis is based, are eques-
trian or fishing societies, both of which may be predicted
to have greater frequencies of warfare than simple socie-
ties. Equestrianism in particular is often linked with
strong dependence on hunting. Ember's sample is heavi-
ly weighted toward North America, where the Western
introduction of the horse and of fur trading had a pro-
found impact; this geographic pattern may itself be sig-
nificant in explaining the range of variation that Ot-
terbein finds. Finally, as I have noted with respect to
intensive big-game hunters of prehistoric periglacial
Eurasia, socioecological factors that inqrease the con-

KNAUFT Violence and Sociality in Human Evolution \ 419

centration of large-scale resources and make them easily
exploitable with available technology can be expected
to potentiate sociopolitical complexity. When such fac-
tors are not pronounced — and I suggest that they have
not been for most of human evolution — the simple-
society pattern is likely to be particularly applicable.

I agree with Otterbein that a cross-cultural study of
the 3 9 -society sample I employed would be an important
complement to my argument. The nature of state-
society influences, the historical context, and the types
of bias must, however, be very carefully weighed in each
individual case,- the data cannot simply be taken at face
value. Moreover, the constellation of factors that is most
useful for evolutionary projection must be carefully de-
lineated; if one is to avoid the charge that observed sim-
ple societies are not representative of the past, the simi-
larities and differences between present and past
societies on key socioeconomic, ecological, and even an-
atomical factors must be assessed. In contrast to many
past approaches, my perspective emphasizes the evolu-
tionary importance of cultural transmission and group
selection, suggesting that we cannot adequately under-
stand social development since the emergence of Homo
without taking this dynamic into account.

References Cited

ALLEN, MICHAEL R. 1967. Male cults and secret initiations in
Melanesia. Melbourne: Melbourne University Press.

ARENSBURG, B., A. M. B. VANDERMEERSCH, H. DUDAY,

L. A. scHEPARTZ, AND Y. RAK. 1989. A Middle Paleolithic
human hyoid bone. Nature 338:758-60.

AXELROD, R., AND W. D. HAMILTON. I981. The evolution of

cooperation. Science 211:1390-96.
BADRiAN, ALISON, AND NOEL BADRiAN. 1984. "Social Orga-
nization of Pan paniscus in the Lomako Forest, Zaire," in The
pygmy chimpanzee: Evolutionary biology and behavior. Edited
by R. L. Susman, pp. 325-46. New York: Plenum Press.

BADRIAN, NOEL, AND RICHARD K. MALENKY. I984. "Feeding

ecology of Pan paniscus in the Lomako Forest, Zaire," in The
pygmy chimpanzee: Evolutionary biology and behavior. Edited
by R. L. Susman, pp. 275-99. New York: Plenum Press.

BAILEY, GEOFF. Editor. 1983. Hunter-gatherer economy in pre-
history: A European perspective. Cambridge: Cambridge Uni-
versity Press.

BAILEY, KENT G. 198$. Human paleopsychology: Applications
for aggression and pathological processes. Hillsdale: Lawrence
Erlbaum Associates.

.1987. "Human paleopsychology: Roots of pathological ag-
gression," in Origins of human aggression: Dynamics and eti-
ology. Edited by G. G. Neuman, pp. 50-63. New York: Human
Sciences Press.

BAILEY, ROBERT C, AND ROBERT AUNGER. I99O. HumanS aS

primates: The social relationships of Efe Pygmy men in com-
parative perspective. International Journal of Primatology
11:127-46.

BAILEY, ROBERT C, GENEVIEVE HEAD, MARK JENIKE,

BRUCE OWEN, ROBERT RECHTMAN, AND ELZBIETA ZECH-

ENTER. 1989. Hunting and gathering in tropical rain forest: Is
it possible? American Anthropologist 91:59-82.

BALiKci, ASEN. 1970. The NetsiUk Eskimo. Garden City: Natu-
ral History Press.

BARNARD, ALAN. 1983. Contemporary hunter-gatherers: Cur-
rent issues in ecology and social organization. Annual Review
of Anthropology 12:193-214.

BARTH, FREDRiK. 1961. Nomads of South Persia. New York:
Humanities Press.

. 1965. Political leadership among Swat Pathans. London:

Athlone Press.

BAUMGARTEL, w. 1976. Up among the mountain gorillas. New
York: Hawthorn.

BENDER, BARBARA. 1990. "The dynamics of nonhicrarchical
societies," in The evolution of political systems: Sociopolitics
of small-scale sedentary societies. Edited by Steadman Upham,
pp. 247-63. Cambridge: Cambridge University Press.

BERN, JOHN. 1987. Is the premise of egalitarianism inequality?
Mankind 17:212-23.

BERNDT, RONALD M. 1962. Excess and restraint' Social control
among a New Guinea mountain people. Chicago: University
of Chicago Press.

. 1965. "Law and order in Aboriginal Australia," in Ab-
original man in Australia: Essays in honour of emeritus pro-
fessor A. P. Elkin. Edited by R. M. Berndt, pp. 167-206. Lon-
don: Angus and Robertson.

BETZiG, LAURA L. 1986. Despotism and differential reproduc-
tion: A Darwinian view of history. Chicago: Aldine.

. 1988. On reconsidering violence in simple human socie-
ties. CURRENT ANTHROPOLOGY 29:264-65. [lb]

-. n.d.a. Why monogamy? Behavioral and Brain Sciences

14. In press, [lb]
. n.d.b. Medieval monogamy. MS. [lb]

BETZIG, LAURA L., MONIQUE BORGERHOFF MULDER, AND

PAUL TURKE. Editors. 1988. Human reproductive behaviour:
A Darwinian perspective. Cambridge: Cambridge University
Press.

BicKERTON, DEREK. 1981. Roots of language. Ann Arbor:
Karoma.

. 1990. Language and species. Chicago: University of Chi-
cago Press.

BiGELOW, ROBERT. 1 97 5. "The rolc of competition and coopera-
tion in human evolution," in War: Its causes and correlates.
Edited by M. A. Nettleship, R. Dale Givens, and A. Nettleship,
pp. 235-63. Chicago: Aldine.

BiNFORD, LEWIS R. 1981. Bones: Ancient men and modern
myths. New York: Academic Press.

. 1983. In pursuit of the past: Decoding the archaeological

record. London: Thames and Hudson.

-. 1989. Debating archaeology. New York: Academic Press.

BLACK, DONALD. 1990. "The elementary forms of conflict
management," in New directions in the study of justice, law,
and social control. Edited by the School of Justice Studies,
Arizona State University. New York: Plenum Press.

BLANCHARD, D. CAROLINE, AND ROBERT J. BLANCHARD.

1984. "Affect and aggression: An animal model applied to hu-
man behavior," in Advances in the study of aggression. Edited
by R. J. Blanchard and D. C. Blanchard, p. 2-62. New York:
Academic Press.

BLicK, JEFFREY P. 1988. Gcnocidal warfare in tribal societies as
a result of European-induced culture conflict. Man 23:654-70.

BLOUNT, BEN G. 1990. Issucs in bonobo (Pan paniscus) sexual
behavior. American Anthropologist 91:702-14.

BOCK, CARL. 1985. Headhunters of Borneo. New York: Oxford
University Press.

BOEHM, c. 1978. Rational preselection from hamadryas to
Homo sapiens: The place of decisions in adaptive process.
American Anthropologist 80:265-96. [cb]

. 1 98 1. "Parasitic selection and group selection: A study

of conflict interference in rhesus and Japanese macaque mon-
keys," in Primate behavior and sociobiology. Edited by A. B.
Chiarelli .and R. S. Corruccini, pp. 161-82. Berlin: Springer-
Verlag.

1982. The evolutionary development of morality as an ef-

fect of dominance behavior and conflict interference. Journal of
Social and Biological Structures 5:413-22. [cb]

-. i984<3. Blood revenge: The anthropology of feuding in

Montenegro and other nonliterate societies. Lawrence: Univer-
sity of Kansas Press, [cb]
. 1984b. Can social hierarchy and egalitarianism both be

420 I CURRENT ANTHROPOLOGY Volume 32, Nuuibei 4, August-October 1991

ascribed to the same causal forces? Politics and the Life Sci-
ences 1:12-14. [cb[

1989. Ambivalence and compromise in human nature.

American Anthropologist 91:921-39. [cb]

n.d. a. "Segmentary 'warfare' and the management of

conflict: Comparison of East African chimpanzees and
patrilineal-patriloeal humans," in Cooperation and competi-
tion in animals and humans. Edited by A. Harcourt and Frans
de Waal. London: Oxford University Press. In press.

-. n.d. b. Why do egalitarian societies stay egalitarian? The

case for human intentions. MS. [cb]

BOESCH, CHRISTOPHE, AND HEDWIGE BOESCH. I989. Hunt-
ing behavior of wild chimpanzees in the Tai National Park.
American Journal of Physical Anthropology 78:547-73.

BOYD, ROBERT. 1988. The cvolutiou of reciprocity in sizable
groups. Journal of Theoretical Biology 132:337-56.

BOYD, ROBERT, AND J. P. LORBERBAUM. 1 987. No purC Strat-
egy is evolutionarily stable in the repeated prisoner's dilemma
game. Nature 327:58-59.

BOYD, ROBERT, AND PETER J. RICHERSON. I985. Culture and

the evolutionary process. Chicago: University of Chicago
Press.

. 1990^. Group selection among alternative evolutionarily

stable strategies. Journal of Theoretical Biology 145:331-42.
1990b. "Culture and cooperation," in Beyond self-

interest. Edited by Jane J. Mansbridge, pp. 11 1-32. Chicago:
University of Chicago Press.
BRAUN, DAVID P. 1990. "Selection and evolution in nonhierar-
chical organization," in The evolution of political systems:
Sociopolitics in small-scale sedentary societies. Edited by
Steadman Upham, pp. 62-86. Cambridge: Cambridge Univer-
sity Press.

BRAUN, DAVID P., AND STEPHEN PLOG. I982. Evolution of

"tribal" social networks: Theory and prehistoric North Ameri-
can evidence. American Antiquity 47:504-25.

BRiGGS, JEAN L. 1970. Never in anger: Portrait of an Eskimo
family. Cambridge: Harvard University Press.

. 1978. "The origins of nonviolence: Inuit management of

aggression (Canadian Arctic)," in Learning non-aggression: The
experience of non-literate societies. Edited by Ashley Montagu,
pp. 54-93. New York: Oxford University Press.

1982. "Living dangerously: The contradictory foundations

of value in Canadian Inuit society," in Politics and history in
band societies. Edited by Eleanor Leacock and Richard Lee,
pp. 109-31. Cambridge: Cambridge University Press.

BYGOTT, J. DAVID. 1979. "Agonistic bchavior, dominance, and
social structure in wild chimpanzees of the Gombe National
Park," in The great apes. Edited by David A. Hamburg and
Elizabeth R. McCown, pp. 405-27. Menlo Park, Calif.:
Cummings.

CAMPBELL, BERNARD G. 1985^. 4th edition. Humankind
emerging. Boston: Little, Brown.

. 1985b. Human ecology. New York: Aldine.

CAPOREAL, LINDA R., ROBYN M. DAWES, JOHN M. ORBELL,

AND ALPHONS J. c. VAN DE KRAGT. 1989. Sclfishncss ex-
amined: Cooperation in the absence of egoistic incentives.
Behavioral and Brain Sciences 12:683-739.

CARNEiRO, ROBERT L. 1970. A theory of the origin of the state.
Science 69:733-38.

. 198 1. "The chief dom: Precursor of the state," in The

transition to statehood in the New World. Edited by Grant D.
Jones and Robert R. Kautz, p. 37-79. Cambridge: Cambridge
University Press.

1990. "Chiefdom-level warfare as exemplified in Fiji and

the Cauca Valley," in The anthropology of war. Edited by Jon-
athan Haas, pp. 190-21 1. Cambridge: Cambridge University
Press.

CARPENTER, c. R. 1940. A field study in Siam of the behavior
and social relations of the gibbon (Hylobates lar). Comparative
Psychology Monographs i6.

CARRiTHERS, MICHAEL. 1990. Why humans have cultures.
Man 25:189-206.

CASHDAN, ELIZABETH A. 1980. Egalitarianism among huntcrs
and gatherers. American Anthropologist 82:116-20.

. 1983. Territoriality among human foragers: Ecological

models and an application to four Bushman groups, current

ANTHROPOLOGY 24:47-66.

-. 1985. Coping with risk: Reciprocity among the Basarwa

of northern Botswana. Man 20:454-74.

-. 1986. Competition between foragers and food-producers

on the Botletli River, Botswana. Africa 56:299-318.

Editor. 1990. Risk and uncertainty in tribal and peasant

economies. Boulder: Westview Press.

CHAGNON, NAPOLEON A. 1968. "Yanomamo social organiza-
tion and warfare," in War: The anthropology of armed conflict
and aggression. Edited by Morton Fried, Marvin Harris, and
Robert Murphy, pp. 109-59. Garden City: Natural History
Press.

. 1979. "Is reproductive success equal in egalitarian socie-
ties?" in Evolutionary biology and human social behavior:
An anthropological perspective. Edited by N. A. Chagnon and
W. Irons, pp. 374-401. North Scituate: Duxbury Press.
-. 1983. 3d edition. Yanomamo: The fierce people. New

York: Holt, Rinehart and Winston.

-. 1988. Life histories, blood revenge, and warfare in a tribal

population. Science 239:985-92.

CHASE, PHILIP G. 1989. "How different was Middle Palaeo-
lithic subsistence? A zooarchaeological perspective on the Mid-
dle to Upper Palaeolithic transition," in The human revolu-
tion: Behavioural and biological perspectives on the origins of
modern humans, vol i. Edited by Paul Mellars and Chris
Stringer, pp. 321-37. Princeton: Princeton University Press.

CHASE, PHILIP, AND HAROLD DIBBLE. 1987. Middle Paleo-
lithic symbolism: A review of current evidence and interpreta-
tions. Journal of Anthropological Archaeology 6:263-96.

CLAESSEN, HENRI J. M., AND PETER SKALNIK. EditOrS. I978.

The early state. The Hague: Mouton.

CLASTRES, PIERRE. 1972. "The Guayaki," in Hunters and gath-
erers today: A socioeconomic study of eleven such cultures in
the twentieth century. Edited by M. G. Bicchieri, pp. 138-73.
New York: Holt, Rinehart and Winston.

CLUNiE, FERGUS. 1977. Fijian weapons and warfare. Bulletin of
the Fiji Museum 2.

COLLIER, JANE F., AND MICHELLE Z. ROSALDO. I981. "Poli-
tics and gender in simple societies," in Sexual meanings: The
cultural construction of gender and sexuality. Edited by Sherry
B. Ortner and Harriet Whitehead, pp. 275-329. New York:
Cambridge University Press.

DAHL, JEREMY F. 1987. Scxual initiation in a captive group of
pygmy chimpanzees (Pan paniscus). Primate Report 16:43-53.

DAHL, JEREMY F., AND RONALD D. NADLER. I989. ScXUal ag-

gression and extended sexual activity in pygmy chimpanzees
(Pan paniscus). Primate Report 23:13-21.

DALY, MARTIN, AND MARGO I. WILSON. I983. 2d edition.

Sex, evolution, and behavior. Boston: Duxbury Press.

. 1988. Homicide. New York: Aldine de Gruyter.

. 1 99 1. Anti-science and the pre-Darwinian image of man-
kind. American Anthropologist 93:162-64. [rkd]

DALY, MARTIN, MARGO I. WILSON, AND SUZANNE J.

WEGHORST. 1982. Male sexual jealousy. Ethology and Socio-

biology 3:69-78.
DAMAS, DAVID. 1984. "The diversity of Eskimo societies," in

Man the hunter. Edited by Richard B. Lee and Irven DeVore,

pp. 1 1 1- 1 7. New York: Aldine.
DARWIN, CHARLES. 1871. The descent of man and selection in

relation to sex. New York: Modern Library, [lb]
. 198 1 (1871). The descent of man and selection in rela-
tion to sex. Princeton: Princeton University Press.
DAVIDSON, IAN, AND WILLIAM NOBLE. 1989. The archacol-

ogy of perception: Traces of depiction and language, current

ANTHROPOLOGY 30:125-55.
DAWES, ROBYN M., ALPHONS J. C. VAN DE KRAGT, AND

JOHN M. ORBELL. 1988. Not me or thee but we: The impor-
tance of group identity in eliciting cooperation in dilemma
situations: Experimental manipulations. Acta Psychologica

68:83-97.
DAWKiNS, RICHARD. 1989. Ncw edition. The ^elfish gene.
Oxford: Oxford University Press.

KNAUFT Violence and Sociality in Human Evolution \ 421

DENiCH, BETTE s. 1987. As wdi has evolved, can also peacel
Commission on the Study of Peace, International Union of
Anthropological and Ethnological Sciences, Newsletter 5 (2),
suppl. [rkd]

DENTAN, ROBERT K. 1978. "Notcs on childhood in a nonvio-
lent context: The Semai case (Malaysia)," in Learning non-
aggiession: The experience of non-literate societies. Edited by
Ashley Montagu, pp. 94-143. New York: Oxford University
Press.

. 1979. Fieldwork edition. The Semai: A nonviolent people

of Malaya. New York: Holt, Rinehart and Winston.

-. n.d.a. "The rise, maintenance, and destruction of peace-

able polity," in To fight or not to fight: Violence and peace-
fulness in humans and other primates. Edited by James Sil-
verberg and Patrick Gray. New York: Oxford University Press.
In press, [rkd]

n.d.b. " 'Surrendered men': Peaceable enclaves in the

post-Enlightenment West," in Nonviolence and peace: An-
thropological insights. Edited by Leslie Sponsel and Thomas
Gregor. In preparation, [rkd]

DE WAAL, FRANS B. M. 1982. Chimpanzee poUtics: Power and
sex among apes. New York: Harper and Row.

. 1986. Integration of dominance and social bonding in pri-
mates. Quarterly Review of Biology 61:459-79.

-. 1987. Tension regulation and nonreproductive functions

of sex in captive bonobos (Pan paniscus). National Geographic
Research 3:318-35.

1989^. Food sharing and reciprocal obligations among

chimpanzees . fourn al of Hum an Evolu tion 18:433-59.

1989b. Peacemaking among primates. Cambridge: Har-

vard University Press.

DIBBLE, HAROLD L. 1989. "The implications of stone tool
types for the presence of language during the Lower and Mid-
dle Palaeolithic," in The human revolution: Behavioural and
biological perspectives on the origins of modern humans, vol.
I. Edited by Paul Mellars and Chris Stringer, pp. 415-32.
Princeton: Princeton University Press.

DivALE, WILLIAM. 1974- Migration, external warfare, and mat-
rilocal residence. Behaviour Science Research 9:75-133. [tsa]

. 1984. Matrilocal residence in pre-literate society. Ann

Arbor: UMI Research Press, [tsa]

DONALD, LELAND. 1987. Comment on: Reconsidering violence
in simple human societies, by Bruce M. Knauft. current an-
thropology 28:483. [tsa]

DRAGO, HARRY. 1970. The great range wars: Violence on the
grasslands. Lincoln: University of Nebraska Press.

DRAPER, PATRICIA. 1975- "!Kung women: Contrasts in sexual
egalitarianism in foraging and sedentary contexts," in Toward
an anthropology of women. Edited by Rayna R. Reiter, pp.
77-109. New York: Monthly Review Press.

. 1978. "The learning environment for aggression and

anti-social behavior among the !Kung (Kalahari Desert, Bots-
wana, Africa)," in Learning non-aggression: The experience of
non-literate societies. Edited by Ashley Montagu, pp. 31-53-
New York: Oxford University Press.

DRENNAN, ROBERT D., AND CARLOS A. URIBE. I987. Chief-

doms in the Americas. Lanham: University Press of America.
DURHAM, WILLIAM H. 1991. Coevolution: Genes, culture, and

human diversity. Stanford University Press.
DURKHEiM, EMiLE. 1965 (1912). The elementary forms of the

religious life. New York: Free Press.

EATON, S. BOYD, AND MELVIN J. KONNER. I985. Paleolithic

nutrition: A consideration of its nature and current implica-
tion. New England Journal of Medicine 312:283-89.

EDWARDS, s., AND c. SNOWDON. 1980. Social bchavior of
captive, group-living orang-utans. International Journal of
Primatology 1:39-62.

EiBL-EiBESFELDT, iRENAus. 1979- The hiology of peace and
war: Men, animals, and aggression. New York: Viking Press.

ELLEFSON, J. o. 1968. "Territorial bchavior in the common
white-handed gibbon [Hylohates lar Linn)," in Primates:
Studies in adaptation and variability. Edited by P. C. Jay,
pp. 180-99. New York: Holt, Rinehart and Winston.

ELLEN, ROY. 1988. "Foraging, starch extraction, and the seden-

tary lifestyle in the lowland rainforest of central Seram," in
Hunters and gatherers, vol. i. History, evolution, and social
change. Edited by Tim Ingold, David Riches, and James
Woodburn, pp. 117-34. Oxford: Berg, [rkd]

EMBER, CAROL R. 1 97 5. Residential Variation among huntcr-
gatherers. Behavior Science Research 10:199-227. [kfo]

. 1978. Myths about hunter-gatherers. Ethnology

17:439-48-
EWERS, JOHN. 1958. Blackfeet: Raiders of the northwestern
Plains. Norman: University of Oklahoma Press.

FEINMAN, GARY, AND JILL NEITZEL. 1984. ToO many typCS:

An overview of sedentary prestate societies in the Americas.
Advances in Archeological Method and Theory 7:39-102.

FERGUSON, R. BRIAN. Editor. 1984a. Warfare, culture, and en-
vironment. Orlando: Academic Press.

. 1984b. "Introduction: Studying war," in Warfare, culture,

and environment. Edited by R. Brian Ferguson, pp. 1-8 1. Or-
lando: Academic Press, [tsa]

1990. Blood of the Leviathan: Western contact and war-

fare in Amazonia. American Ethnologist 17:237-57.

FERGUSON, R. BRIAN, AND LESLIE E. FARRAGHER. I988. The

anthropology of war: A bibliography. New York: Harry Frank
Guggenheim Foundation, [tsa]

FERGUSON, R. BRIAN, AND NEIL L. WHITEHEAD. Editors, n.d.

War in the tribal zone: The impact of expanding states on in-
digenous warfare. Santa Fe: School of American Research. In
press, [tsa]

FLiNN, MARK V., AND BOBBi s. LOW. 1986. "Rcsourcc distri-
bution, social competition, and mating patterns in human soci-
eties," in Ecological aspects of social evolution. Edited by D. I.
Rubenstein and R. W. Wrangham, pp. 217-43. Princeton:
Princeton University Press.

FOLEY, ROBERT A. 1982. A rcconsidcration of the role of preda-
tion on large mammals in tropical hunter-gatherer adaptation.
Man 17:393-402.

. 1984. "Early man and the red queen: Tropical African

community evolution and hominid adaptation," in Hominid
evolution and community ecology. Edited by R. A. Foley,
pp. 85-110. New York: Academic Press.

1987. Another unique species: Patterns in human evolu-

tionary ecology. New York: Longman.

-. 1988. "Hominids, humans, and hunter-gatherers: An evo-

lutionary perspective," in Hunters and gatherers, vol. i. His-
tory, evolution, and social change. Edited by Tim Ingold, Da-
vid Riches, and James Woodburn, pp. 207-21. Oxford: Berg.
1989. "The evolution of hominid social behaviour," in

Comparative socioecology: The behavioural ecology of hu-
mans and other mammals. Edited by V. Standen and R. A.
Foley, pp. 473-94. London: Blackwell Scientific Publications.

FOLEY, ROBERT A., AND c. P. LEE. 1989. Finite social spacc,
evolutionary pathways, and reconstructing hominid behavior.
Science 243:901-6.

FOSSE Y, DiAN. 1979. "Development of the mountain gorilla
(Gorilla gorilla beringei): The first thirty-six months," in The
great apes. Edited by David A. Hamburg and Elizabeth R.
McCown, pp. 139-84. Menlo Park, Calif.: Cummings.

. 1 98 1. The imperial mountain gorilla. National Geo-
graphic 159:501-23.

1983. Gorillas in the mist. Boston: Houghton Mifflin.

FOSTER, MARY LECRON, AND ROBERT A. RUBINSTEIN. Edi-
tors. 1986. Peace and war: Cross-cultural perspectives. New
Brunswick: Transaction Books, [tsa]

FRAYER, DAVID w. n.d. "Ofnct: Evidence for violence in the
Central European Mesolithic." Proceedings of the ^d Sympo-
sium on Upper Paleolithic, Mesolithic, and Neolithic Popula-
tions of Europe and the Mediterranean, Budapest.

FRIED, MORTON H. 1967. The evolution of political society.
New York: Random House.

. 1975- Tl^^ notion of tribe. Menlo Park, Calif.: Cum-
mings.

FRIEDMAN, JONATHAN, AND M. J. ROWLANDS. EditOrS. I978.

The evolution of social systems. Pittsburgh: University of
Pittsburgh Press.

FUKUI, KATSUYOSHI, AND DAVID TURTON. EditOrS. I979.

422 I C

URRENT ANTHROPOLOGY Volume 32, Number 4, August- October 1991

Warfare among East African herders. Senri Ethnological Stud-
ies 3.
FURUiCHi, TAKESHI. 1987. Scxual swcUing, receptivity, and

grouping of wild pygmy chimpanzee females at Wamba, Zaire.

Primates 28:309-18.
. 1989. Social interactions and the life history of female

Pan paniscus in Wamba, Zaire. International Journal of Prima-

tology 10:173-97.
GABRIEL, RICHARD A. 1990. The culture of war: Intention and

early development. New York: Greenwood Press.
GALDiKAS, BiRUTE M. F. 1979. "Orangutan adaptation at Tan-

jung Puting Reserve: Mating and ecology," in The great apes.

Edited by David A. Hamburg and Elizabeth R. McCown, pp.

195-233. Menlo Park, Calif.: Cummings.
GAMBLE, CLivE. 1982. Interaction and alliance in Palaeolithic

society. Man 17:92-107.
. 1983. "Culture and society in the Upper Palaeolithic of

Europe," in Hunter-gatherer ecology in prehistory: A European

perspective. Edited by Geoff Bailey, pp. 201-11. Cambridge:

Cambridge University Press.

1986. The Palaeolithic settlement of Europe. New York:

. 1990. "Warfare and the evolution of tribal polities in the

prehistoric Southwest," in The anthropology of war. Edited by
Jonathan Haas, pp. 171-89. Cambridge: Cambridge University
Press.

-. Editor. 1990. The anthropology of war. Cambridge: Cam-

Cambridge University Press.

-. 1990. Comment on: Symbolism and modem human ori-

gins, by J. Lindley and G. Clark, current anthropology
31:243-44.

GELLNER, E. 1 988. Plough, sword, and book: The structure of
human history. Chicago: University of Chicago Press, [lr]

. 1989. "Culture, constraint, and community: Semantic

and coercive compensations for the genetic under-determina-
tion of Homo sapiens sapiens," in The human revolution: Be-
havioural and biological perspectives on the origins of modern
humans, vol. i. Edited by Paul Mellars and Chris Stringer, pp.
514-25. Princeton: Princeton University Press.

GHiGLiERi, MICHAEL P. 1987. Sociobiology of the great apes
and the hominid ancestor. Journal of Human Evolution

16:319-57-

. 1989. "Hominid sociobiology and hominid social evolu-
tion," in Understanding chimpanzees. Edited by Paul G.
Heltne and Linda A. Marquardt, pp. 370-79. Cambridge:
Harvard University Press.

GIBSON, THOMAS. 1990. "Raiding, trading, and tribal auton-
omy in insular Southeast Asia," in The anthropology of war.
Edited by Jonathan Haas, pp. 125-45. Cambridge: Cambridge
University Press.

GOLDMAN, IRVING. 1970. Ancicnt Polynesian societies. Chi-
cago: University of Chicago Press.

GOODALL, JANE. 1968. Bchavior of free-living chimpanzees of
the Gombe Stream area. Animal Behavior Monographs
1:163-311.

. 1 97 1. In the shadow of man. Boston: Houghton Mifflin.

. 1986. The chimpanzees of Gombe: Patterns of behavior.

Cambridge: Belknap/Harvard University Press.

GOODALL, JANE, ADRIANO BANDORA, EMILIE BERMANN,
CURT BUSSE, HIALI MATAMA, ESILOM MPONGO, ANN

PIERCE, AND DAVID Riss. 1979- "Intercommunity interac-
tions in the chimpanzee population of the Gombe National
Park," in The great apes. Edited by David A. Hamburg and
Elizabeth R. McCown, pp. 13-53. Menlo Park, Calif.: Cum-
mings.

GOODENOUGH, WARD H. 1990. Evolution of the human capac-
ity for belief. American Anthropologist 91:597-612.

GREENBERG, J. B. 1989. Blood tics. Tucson: University of Ari-
zona Press, [jp]

GRIFFIN, p. BioN. 1984. "Foragcr resource and land use in the
humid tropics: The Agta of northeastern Luzon, the Philip-
pines," in Past and present in hunter-gatherer studies. Edited
by Carmel Schrire, pp. 95-121. New York: Academic Press.

GRiNNELL, GEORGE. 1956. The fighting Chcyennes. Norman:
University of Oklahoma Press.

GUEMPLE, LEE. Editor. 1972. Alliance in Eskimo society. Pro-
ceedings of the American Ethnological Society, 1971, suppl.

HAAS, JONATHAN, 1982. The cvolution of the prehistoric state.
New York: Columbia University Press.

bridge University Press.
HALLPiKE, CHRISTOPHER R. 1977. Bloodshcd and vcngcancc
in the Papuan mountains: The generation of conflict in Tuade
society. Oxford: Clarendon Press.

HAMBURG, DAVID A., AND MICHELE B. TRUDEAU. EditOrS.

1 98 1. Biobehavioral aspects of aggression. New York: Liss.

HAMILTON, ANNETTE. 1982. "Descended from father, belong-
ing to country: Rights to land in the Australian Western De-
sert," in Bands and history in band societies. Edited by Elea-
nor Leacock and Richard B. Lee, pp. 85-108. Cambridge:
Cambridge University Press.

HAMILTON, w. D. 1964. The gcnctical evolution of social be-
havior. Journal of Theoretical Biology 7:1-52.

HARAKo, REizo. 1976. The Mbuti as hunters: A study of eco-
logical anthropology of the Mbuti Pygmies, pt. I. Kyoto Uni-
versity African Studies 10:37-99.

HARCOURT, A. H. 1978. Strategies of emigration and transfer by
primates, with particular reference to gorilla. Zeitschrift far
Tierpsychologie 48:401-20.

. 1979. "The social relations and group structure of wild

mountain gorilla," in The great apes. Edited by David A. Ham-
burg and Elizabeth R. McCown, pp. 187-92. Menlo Park,
Calif.: Cummings.

HARCOURT, A. H., D. FOSSEY, AND J. PI SABATER. 1981.

Demography of Gorilla gorilla. Journal of Zoology (London)
195:215-33.

HARDY, M. J. L. 1963. Blood fcuds and the payment of blood
money in the Middle East. Leiden: Brill.

HARNER, MICHAEL J. 1972, Jivaro: People of the sacred water-
falls. Berkeley: University of California Press.

HARPENDING, HENRY, AND ALAN ROGERS. 1987. On

Wright's mechanism for intergroup selection. Journal of Theo-
retical Biology 127:51-61.

HART, C. W. M., AND ARNOLD R. PILLING. I960. The Tiwi of

North Australia. New York: Holt, Rinehart and Winston.
HASEGAWA, TOSHiKAzu. 1989. "Scxual bchavior of immigrant
and resident female chimpanzees at Mahale," in Understand-
ing chimpanzees. Edited by Paul G. Heltne and Linda A. Mar-
quardt, pp. 90-103. Cambridge: Harvard University Press.

HASEGAWA, TOSHIKAZU, AND MARIKO HIRAIWA-

HASEGAWA. 1983. Opportunistic and restrictive matings
among wild chimpanzees in the Mahale Mountains, Tanzania.
Journal of Ethology 1:75-85.

. 1990. "Sperm competitiion and mating behavior," in The

chimpanzees of the Mahale Mountains: Sexual and life history
strategies. Edited by Toshisada Nishida, pp. 115-32. Tokyo:
University of Tokyo Press.

HEADLAND, THOMAS. 1988. Ecosystcmic change in a Philip-
pine tropical rainforest and its effect on a Negrito foraging so-
ciety. Tropical Ecology 29:121-35. [rkd]

. 1990. Time allocation, demography, and original afflu-
ence in a Philippine Negrito hunter-gatherer society. Paper pre-
sented at the 6th International Conference on Hunting and
Gathering Societies, University of Alaska, Fairbanks, May
2 8- June I. [rkd]

HEADLAND, THOMAS N., AND LAWRENCE A. REID. I989.

Hunter-gatherers and their neighbors from prehistory to the
present, current anthropology 30:43-66.

HENRY, DONALD o. 1985. "Prcagricultural intensification: The
Natufian example," in Prehistoric hunter-gatherers: The emer-
gence of cultural complexity. Edited by T. Douglas Price and
James A. Brown, pp. 365-84. Orlando: Academic Press.

HEWLETT, BARRY. 1988. "Scxual Selection and paternal invest-
ment among Aka pygmies," in Human reproductive behavior:
A Darwinian perspective. Edited by L. Betzig, M. Borgerhoff
Mulder, and P. Turke, pp. 263-76. Cambridge: Cambridge Uni-
versity Press, [lb]

HEWLETT, BARRY s. 1990. Intimate fathers:. The nature and

KNAUFT Violence and Sociality in Human Evolution \ 423

context of Aka Pygmy paternal care. Ann Arbor: University of
Michigan Press.

HiATT, L. R. 1965. Kinship and conflict: A study of an Aborigi-
nal community in northern Arnhem Land. Canberra: Austra-
lian National University Press.

HILL, KIM. 1982. Hunting and human evolution. Journal of Hu-
man Evolution 11:521-44.

HILL, KIM, AND HiLLARD KAPLAN. 1988. "Tradeoffs in male
and female reproductive strategies among the Ache, pt. i," in
Human reproductive behavior: A Darwinian perspective. Ed-
ited by L. Betzig, M. Borgerhoff Mulder, and P. Turke, pp.
277-89. Cambridge: Cambridge University Press, [lb]

HiRAiwA-HASEGAWA, MARiKO. 1990. "The rolc of food shar-
ing between mother and infant in the ontogeny of feeding be-
havior," in The chimpanzees of the Mahale Mountains: Sexual
and life history strategies. Edited by Toshisada Nishida, pp.
267-83. Tokyo: University of Tokyo Press.

HITCHCOCK, ROBERT K. 1987. Sociocconomic change among
the Basarwa in Botswana: An ethnohistorical analysis. Ethno-
history 34:219-55.

HOCKETT, CHARLES F., AND ROBERT ASCHER. 1964. The hu-
man revolution, current anthropology 5:135-68.

HOEBEL, E. ADAMSON. 1964. The law of primitive man: A
study in comparative legal dynamics. Cambridge: Harvard
University Press.

HUNT, GEORGE T. 1940. Wars of the Iroquois: A study in in-
tertribal trade relations. Madison: University of Wisconsin
Press. [tsaJ

icHiKAWA, MiTSuo. 1983. An examination of the hunting-
dependent life of the Mbuti Pygmies, eastern Zaire. African
Study Monographs 4:44-76.

INGOLD, TIM. 1987. The appropriation of nature: Essays on hu-
man ecology and social relations. Iowa City: University of
Iowa Press.

. 1988. "Notes on the foraging mode of production," in

Hunters and gatherers, vol. i. History, evolution, and social
change. Edited by Tim Ingold, David Riches, and James
Woodburn, pp. 269-85. Oxford: Berg.

-. 1989. "The social and environmental relations of human

beings and other animals," in Comparative socioecology: The
behavioural ecology of humans and other mammals. Edited by
V. Standen and R. A. Foley, pp. 495-512. London: Blackwell
Scientific Publications.

ISAAC, GLYNN. 1978. The food-sharing behavior of protohuman
hominids. Scientific American 238:90-108.

. 1984. The archaeology of human origins: Studies of the

Lower Pleistocene in East Africa 1971-1981. Advances in
World Archaeology 3:1-89.

iTANi, JUNICHIRO. 1979. "Distribution and adaptation of
chimpanzees in an arid area," in The great apes. Edited by Da-
vid A. Hamburg and Elizabeth R. McCown, pp. 55-71. Menlo
Park, Calif.: Cummings.

. 1988. "The origin of human equality," in Social fabrics

of the mind. Edited by M. R. A. Chance, pp. 137-56. London:
Lawrence Erlbaum Associates.

jAYAWARDENA, CHANDRA. 1963. Conflict and Solidarity in a
Guianese plantation. London: Athlone.

. 1968. Ideology and conflict in lower-class communities.

Comparative Studies in Society and History 10:423-46.

JENNESS, D. 1922. The life of the Copper Eskimo. Ottawa: Ack-
land. [lb]

JOHNSON, ALLEN W., AND TIMOTHY EARLE. I987. The CVO-

lution of human societies: From foraging group to agrarian
state. Stanford: Stanford University Press.

JOLLY, ALISON. 1985. The evolution of primate behavior. New
York: Macmillan.

KANO, TAKAYOSHi. 1983. An ccological study of pygmy chim-
panzees (Pan paniscus) of Yalosidi, Republic of Zaire. Interna-
tional Journal of Primatology 4:1-31.

. 1984^. "Reproductive behavior of the pygmy chimpanzee.

Pan paniscus, of Wamba, Republic of Zaire," in Sexuality of
the primates. Edited by T. Maple and R. D. Nadler. New York:
Van Nostrand Reinhold.

. 1984b. Observations of physical abnormalities among the

wild bonobos of Wamba, Zaire. American Journal of Physical
An thropology 63:1-11.

-. 1990. The bonobos' peaceable kingdom. Natural History,

no. 11/90, pp. 62-71.

KANO, TAKAYOSHI, AND MBANGI MULAVWA. 1984. "Feeding

ecology of pygmy chimpanzees (Pan paniscus) of Wamba,"
in The pygmy chimpanzee: Evolutionary biology and
behavior. Edited by R. L. Susman, pp. 233-74. New York:
Plenum Press.

KAPLAN, HILLARD, AND KIM HILL. 1985^. Food sharing

among Ache foragers: Tests of explanatory hypotheses, cur-
rent ANTHROPOLOGY 26:223-45.

. 1985b. Hunting ability and reproductive success among

male Ache foragers, current anthropology 26:131-33. [lb]

KAPLAN, HILLARD, KIM HILL, KRISTEN HAWKES, AND MAG-

DALENA HURTADO. 1984. Food sharing among Ache hunter-
gatherers of eastern Paraguay, current anthropology
25:113-15.

KAWANAKA, KENji. 1990, "Alpha malcs' interactions and social
skills," in The chimpanzees of the Mahale Mountains: Sexual
and life history strategies. Edited by Toshisada Nishida, pp.
171-87. Tokyo: University of Tokyo Press.

KELLY, RAYMOND c. 1985. The Nucr conqucsV. A case study in
the structure of non- equilibrium systems. Ann Arbor: Univer-
sity of Michigan Press.

KENT, SUSAN. 1989. And justice for all: The development of po-
litical centralization among newly sedentary foragers. Ameri-
can Anthropologist 91:703-12.

KiEFER, THOMAS. 1972. The Tausug: Violence and law in a
Philippine Moslem society. New York: Holt, Rinehart and
Winston.

KiNZEY, WARREN G. 1987. "Monogamous primates: A primate
model for human mating systems," in The evolution of human
behavior: Primate models. Edited by W. G. Kinzey, pp.
105-14. Albany: State University of New York Press.

KiTCHER, PHILIP. 1985. Vaulting ambition: Sociobiology and
the quest for human culture. Cambridge: M.I.T. Press.

KNAUFT, BRUCE M. 1985. Good company and violence: Sorcery
and social action in a lowland New Guinea society. Berkeley:
University of California Press.

. 1987^. Reconsidering violence in simple human socie-
ties: Homicide among the Gebusi of New Guinea, current

ANTHROPOLOGY 28:457-500.

1987b. Divergence between cultural success and repro-

ductive fitness in preindustrial cities. Cultural Anthropology
2:94-114.

-. 1988. Reply [to Betzig and Dentan]. current anthropol-

ogy 29:629-33.

-. 1989^. Sociality versus self-interest in human evolution.

Behavioral and Brain Sciences 12:712-13.

1989b. "Bodily images in Melanesia: Cultural substances

and natural metaphors," in Fragments for a history of the hu-
man body, pt. 3. Edited by Michel Feher, Ramona Naddaff, and
Nadia Tazi. New York: Urzone.

-. 1990^. Violence among newly sedentary foragers. Ameri-

can Anthropologist 92:1013-15.

-. 1990b. Melanesian warfare: A theoretical history. Ocea-

nia 60:250-311.

-. 1990C. The question of ritualized homosexuality among

the Kiwai of south New Guinea. Journal of Pacific History

25:188-210.
KOCH, KLAUS-FRiEDRiCH. 1974. War and pcacc in Jalcmo.

Cambridge: Harvard University Press.
KONNER, MELViN J. 1981. "Evolution of human behavior devel-
opment," in Handbook of cross-cultural human development.

Edited by Ruth H. Munroe, Robert L. Munroe, and Beatrice B.

Whiting, pp. 3-52. New York: Garland STPM Press.
. 1982. The tangled wing: Biological constraints on the

human spirit. New York: Harper and Row.
KUPER, HILDA. 1937. The development of military organization

in Swaziland. Africa 10:55-74, 176-205.

KURLAND, JEFFREY A., AND STEPHEN J. BECKERMAN. I985.

424 I CURRENT ANTHROPOLOGY Volume 32, Numbei 4, August-October 1991

Optimal foraging and hominid evolution: Labor and reciproc-
ity. American Anthropologist 86:73-93.

KURODA, suEHiSA. 1980. Social bchavior of the pygmy chim-
panzees. Primates 21:181-97.

. 1984. "Interaction over food among pygmy chimpan-
zees/' in The pygmy chimpanzee: Evolutionary biology and
behavior. Edited by R. L. Susman, pp. 301-24. New York: Ple-
num Press.

LAMBERT, w. w. 1971. "Personality development and socializa-
tion of aggression/' in Comparative perspectives on social psy-
chology. Edited by W. W. Lambert and R. Weisbrod. Boston:
Little, Brown.

LANGNESS, L. L. 1974. Ritual powcr and male domination in
the New Guinea Highlands. Ethos 2:189-212.

LANGUB, jAYL. 1988. "The Penan strategy," in People of the
tropical rain forest. Edited by Julie Sloan Denslow and Chris-
tine Padoch, pp. 207-8. Berkeley: University of California
Press, [rkd]

LAYTON, ROBERT. 1986. Political and territorial structures
among hunter-gatherers. Man 21:18-33.

LEACOCK, ELEANOR, AND RICHARD B. LEE. 1982. "Introduc-
tion," in Politics and history in band societies. Edited by Elea-
nor Leacock and Richard B. Lee, pp. 1-20. Cambridge: Cam-
bridge University Press.

LEE, RICHARD B. 1979^. The IKung San: Men, women, and
work in a foraging society. New York: Cambridge University
Press.

. 1979b. The Dobe IKung. New York: Holt, Rinehart and

Winston.

-. 1982. "Politics, sexual and non-sexual, in an egalitarian

society," in Politics and history in band societies. Edited by
Eleanor Leacock and Richard B. Lee, pp. 37-59- New York:
Cambridge University Press.

— . 1988. "Reflections on primitive communism," in Hunters
and gatherers, vol. i. History, evolution, and social change.
Edited by Tim Ingold, David Riches, and James Woodbum, pp.
252-68. Oxford: Berg.

. 1990. "Primitive communism and the origin of social in-

equality," in The evolution of political systems: Sociopolitics
of small-scale sedentary societies. Edited by Steadman Upham,
pp. 225-46. Cambridge: Cambridge University Press.

LEE, RICHARD B., AND IRVEN DE VORE. I984 (1968). "Prob-
lems in the study of hunters and gatherers," in Man the
hunter. Edited Richard B. Lee and Irven DeVore, pp. 3-12.
New York: Aldine.

LEVi-STRAUSS, CLAUDE. 1969 (1949). The elementary Struc-
tures of kinship. Boston: Beacon Press.

LEVINE, ROBERT A., AND DONALD T. CAMPBELL. 197^. Eth-

nocentrism: Theories of conflict, ethnic attitudes, and group
behavior. New York: Wiley.

LEWiN, R. 1988. New views emerge on hunters and gatherers.
Science 240:1146-48.

LiEBERMAN, P. 1984. The biology and evolution of language.
Cambridge: Harvard University Press.

LiNDLY, J. M., AND G. A. CLARK. 1990. Symbolism and mod-
ern human origins, current anthropology 31:233-61.

LOVEjOY, c. OWEN. 1981. The origin of man. Science
211:341-50.

-. 1982. Models of human evolution. Science 217:304-5.

LOW, BOBBi. n.d. Reproductive life in nineteenth-century Swe-
den: An evolutionary perspective of demographic phenomena.
Ethology and Sociobiology 12. In press, [lb]

MC BURNEY, c. B. 1976. Early man in the Soviet Union. Lon-
don: British Academy.

MC GREW, WILLIAM c. 1979 fl. "Evolutionary implications of
sex differences in chimpanzee predation and tool use," in The
great apes. Edited by David A. Hamburg and Elizabeth R.
McCown, pp. 441-63. Menlo Park, Calif.: Cummings.

. 1979b. "Patterns of plant food sharing by wild chimpan-
zees." Proceeding of the sth Congress of the International Pri-
matological Society, Nagoya, Japan, pp. 304-9- Basel: Karger.

MAC KiNNON, JOHN. 1979. "Reproductive behavior in wild
orangutan populations," in The great apes. Edited by David A.

Hamburg and Elizabeth R. McCown, pp. 257-73. Menlo Park,
Calif.: Cummings.
MANN, MICHAEL. 1986. The sourccs of social power. Vol. i. A
history of power from the beginning to A.D. ij6o. Cambridge:
Cambridge University Press, [tsa]

MANSON, JOSEPH H., AND RICHARD W. WRANGHAM. I99I.

Intergroup aggression in chimpanzees and humans, current

ANTHROPOLOGY 32:369-90.

MARSHACK, ALEXANDER. 1981. On Paleolithic ochre and the
early uses of color and symbol, current anthropology
22:188-91.

. 1985. Hierarchical evolution of the human capacity: The

Paleolithic evidence. (Fifty-Fourth James Arthur Lecture on the
Evolution of the Human Brain, 1984.) New York: American
Museum of Natural History.

-. 1989^. Evolution of the human symbolic capacity: The

symbolic evidence. Yearbook of Physical Anthropology

32:1-34.
. 1989b. On depiction and language, current anthropol-

ogy 30:332-35-
MARSHALL, LORNA. 1959. Marriage among IKung Bushmen.

Africa 29:335-64. [lb]
. 1976. The IKung of Nyae Nyae. Cambridge: Harvard

University Press.

1979. "Sharing, talking, and giving: Relief of social ten-

sions among IKung Bushmen," in Kalahari hunter-gatherers:
Studies of the IKung San and their neighbors. Edited by Rich-
ard B. Lee and Irven DeVore, pp. 349-72. Cambridge: Cam-
bridge University Press.

MAUss, MARCEL. 1970. The gift. New York: Norton and
Norton.

MEAD, GEORGE HERBERT. 1934. Mind, Self, and socicty.
Chicago: University of Chicago Press, [tmk]

MEEKER, MICHAEL. 1979- Literature and violence in North
Arabia. New York: Cambridge University Press.

MEGGiTT, MERVYN J, 1962. Desert people: A study of the
Walbiri Aborigines of Central Australia. London: Angus and
Robertson.

. 1977. Blood is their argument: Warfare among the Mae

Enga tribesmen of the New Guinea Highlands. Palo Alto:
M^yfield.

MELLARS, PAUL A. 1985. "The ccological basis of social com-
plexity in the Upper Paleolithic of southwestern France," in
Prehistoric hunter-gatherers: The emergence of cultural com-
plexity. Edited by T. Douglas Price and James A. Brown, pp.
271-97. Orlando: Academic Press.

. 1988. The origin and dispersal of modern humans, cur-
rent ANTHROPOLOGY 29:186-88.

-. 1989^. Major issues in the emergence of modern humans.

CURRENT ANTHROPOLOGY 30:349-85.

1989b. "Technological changes across the Middle-Upper

Palaeolithic transition: Economic, social, and cognitive per-
spectives," in The human revolution: Behavioural and biologi-
cal perspectives on the origins of modern humans, vol. i. Ed-
ited by Paul Mellars and Chris Stringer, pp. 338-65. Princeton:
Princeton University Press.

MELLARS, PAUL, AND c. B. STRINGER. Editors. 1989. The hu-
man revolution: Behavioural and biological perspectives on
the origins of modern humans. Vol. i. Princeton: Princeton
University Press.

MiTANi, JOHN c. 1985. Mating behavior of male orangutans in
the Kutai Reserve, East Kalimantan, Indonesia. Animal Behav-
ior 33:392-402.

MOORE, JOHN M. 1990. The reproductive success of Cheyenne
war chiefs: A contrary case to Chagnon's Yanomamo. current

ANTHROPOLOGY 31:322-30.

MOSS, DAVID. 1979. Bandits and boundaries in Sardinia. Man

14:477-96.
MURDOCK, GEORGE PETER. 1981. Atlas of world cultures.

Pittsburgh: University of Pittsburgh Press.

MURPHY, ROBERT F., AND JULIAN H. STEWARD. I955. Tap-
pers and trappers: Parallel process in acculturation. Economic
Development and Cultural Change 4:335-53.

KNAUFT Violence and Sociality in Human Evolution \ 425

MYERS, FRED R. 1986. Pintupi country, Pintupi self: Sentiment,
place, and politics among Western Desert Aborigines. Wash-
ington, D.C.: Smithsonian Institution Press.

. 1988. Critical trends in the study of hunter-gatherers.

Annual Reviews in Anthropology 17:261-82.

NADLER, RONALD D. 1988. "Sexual aggression in the great
apes," in Human sexual aggression: Current perspectives. Ed-
ited by R. A. Prentky and V. L. Quinsey, pp. 154-62. Annals of
the New York Academy of Sciences 528.

NADLER, RONALD D., AND CHARLES H. PHOENIX. I99I.

"Male sexual behavior: Monkeys, men, and apes," in Under-
standing behavior: What primate studies tell us about human
behavior. Edited by James D. Loy and Calvin B. Peters, pp.
152-89. New York: Oxford University Press.

NETTING, ROBERT MC c. 1986. "Huntcr-gatherers, " in Cul-
tural ecology, id edition. Prospect Heights: Waveland.

. 1990. "Population, permanent agriculture, and polities:

Unpacking the evolutionary portmanteau," in The evolution of
political systems: Sociopolitics of small-scale sedentary socie-
ties. Edited by Steadman Upham, pp. 21-61. Cambridge: Cam-
bridge University Press.

NEUMAN, GERARD G. 1 987. "How wc became (in)human," in
Origins of human aggression: Dynamics and etiology. Edited
by G. G. Neuman, pp. 78-104. New York: Human Sciences
Press.

NEwcoMB, w. w. 1950. A re-examination of the causes of
Plains warfare. American Anthropologist 52:317-30. [tsa]

NIMKOFF, M. F., AND RUSSELL MIDDLETON. I960. TypCS of

family and types of economy. American Journal of Sociology

66:215-25.
NiSHiDA, TOSHisADA. 1970. Social bchavior and relationship

among wild chimpanzees of the Mahale Mountains. Primates

9:167-224.
. 1979. "The social structure of chimpanzees of the

Mahale Mountains," in The great apes. Edited by David A.

Hamburg and Elizabeth R. McCown, pp. 73-121. Menlo Park,

Calif.: Cummings.

-. 1983. Alpha status and agonistic alliance in wild chim-

panzees (Pan troglodytes schweinfurthii). Primates 24:318-36.
1990. "A quarter- century of research in the Mahale

Mountains: An overview," in The chimpanzees of the Mahale
Mountains: Sexual and life history strategies. Edited by Toshi-
sada Nishida, pp. 3-35. Tokyo: University of Tokyo Press.

NISHIDA, TOSHISADA, MARIKO HIRAIWA-HASEGAWA,
TOSHIKAZU HASEGAWA, AND YUKIO TAKAHATA. I985.

Group extinction and female transfer in wild chimpanzees in
the Mahale National Park, Tanzania. Zeitschrift fiir Tierpsy-
chologie 67:284-301.

NITECKI, MATTHEW W., AND DORIS H. NITECKI. EditOrS.

1987. The evolution of human hunting. New York: Plenum
Press.
NOWAK, M., AND K. siGMUND. 1989. OsciUations in the evo-
lution of reciprocity. Journal of Theoretical Biology 137:21-26.

O'CONNELL, JAMES F., KRISTEN HAWKES, AND NICHOLAS

BLURTON JONES. 1988^. Hadza scavenging: Implications for
Plio/Pleistocene hominid subsistence, current anthropol-
ogy 29:356-63.

. 1988b. Hadza hunting, butchering, and bone transport

and their archaeological implications. Journal of Anthropologi-
cal Research 44:113-61.

OSGOOD, CORNELIUS. 1958. Ingalik social culture. New
Haven: Yale University Press, [lb]

OTTERBEiN, KEITH F. 1968. Internal war: A cross-cultural
study. American Anthropologist 70:277-89.

. 1985. 2d edition. The evolution of war: A cross-cultural

study. New Haven: HRAF Press.

-. 1986. The ultimate coercive sanction: A cross-cultural

study of capital punishment. New Haven: HRAF Press.
-. 1987. Comment on: Reconsidering violence in simple

human societies, by B. Knauft. current anthropology
28:484-85.

-. 1988. Reply to: On Semai homicide, by Robert K.

. 1989. Socialization for war: A study of the influence of

hunting upon warfare. Paper read at the 88th Annual Meeting
of the American Anthropological Association, Washington,
D.C., November 19. [kfo]

OTTERBEIN, KEITH F., AND CHARLOTTE S. OTTERBEIN.

1965. An eye for an eye, a tooth for a tooth: A cross-cultural
study of feuding. American Anthropologist 67:1470-82.

PAIGE, KAREN E., AND JEFFERY M. PAIGE. I981. The poUtiCS

of reproductive ritual. Berkeley: University of California Press.

PADDOCK, J. (with SANTIAGO GENOVEs). 1974- Violence and be-
havior: A symposium on the origins of man's inhumanity to
man. American Journal of Physical Anthropology 41. [jp]

. 1975. Studies on antiviolent and "normal" communities.

Aggressive Behavior 1:217-33. [jp]

-. 1976^. "Pueblos antiviolentos: Notas sobre un estado de

salud social," in Comportamiento y violencia. Edited by Santi-
ago Genoves and Jacques F. Passy. Mexico: Editorial Diana, [jp]
1976b. Values in an antiviolent community. Humanitas

12:183-94. [jp]

-. 1979^. "A new look at the problem of human violence,"

in Social, political, and economic life in contemporary
Oaxaca. Edited by Aubrey Williams, Jr., pp. 1-41. Vanderbilt
University Publications in Anthropology 24. [jp]

-. 1979b. Symposium: Faces of antiviolence (abstracts).

Aggressive Behavior 5:210-14. [jp]

-. 1980. Review of: Learning non- aggression: The experi-

ence of non-literate societies, edited by Ashley Montagu
(Oxford: Oxford University Press, 1978). Aggressive Behavior
6:180-84. [jp]

-. 1988. Review of: Intervillage conflict in Oaxaca, by

Philip Dennis (New Brunswick: Rutgers University Press,
1988). Man 23:571. [jp]

1990. Review of: Blood ties, by J. B. Greenberg (Tucson:

Dentan. current anthropology 29:633-36. [kfo]

University of Arizona Press, 1990). Aggressive Behavior
16:399-401. [jp]

PARKER, SUE T. 1985.A socio- technological model for the evo-
lution of language, current anthropology 26:617-39.

PARKER, SUE T., AND KATHLEEN R. GIBSON. 1 979. A devel-
opmental model for the evolution of language and intelligence
in early hominids. Behavioral and Brain Sciences 2:367-408.

PARKER, GEOFFREY. 1988. The military revolution: Military in-
novation and the rise of the West, is 50-1800. Cambridge:
Cambridge University Press, [tsa]

PAUL, ROBERT A. 1987. The individual and society in biological
and cultural anthropology. Cultural Anthropology 2:80-93.

PEDERSEN, JON, AND ESPEN woEHLE. 1988. "The Complexi-
ties of residential organization among the Efe (Mbuti) and
Bambombi (Baka): A critical view of the notion of flux in
hunter-gatherer societies," in Hunters and gatherers, vol. i.
History, evolution, and social change. Edited by Tim Ingold,
David Riches, and James Woodburn, pp. 75-90. Oxford: Berg.

PLOG, STEPHEN. 1990. "Agriculture, sedentism, and environ-
ment in the evolution of political systems," in The evolution
of political systems: Sociopolitics of small-scale sedentary so-
cieties. Edited by Steadman Upham, pp. 177-99. Cambridge:
Cambridge University Press.

POPP, JOSEPH L., AND IRVEN DE VORE. 1979- "AggrCSSivC

competition and social dominance theory," in The great apes.
Edited by David A. Hamburg and Elizabeth R. McCown, pp.
317-38. New York: Holt, Rinehart and Winston.

POTTS, RICHARD. 1987. "Reconstructions of early hominid so-
cioecology: A critique of primate models," in The evolution of
human behavior: Primate models. Edited by Warren G. Kin-
zey, pp. 28-47. Albany: State University of New York Press.

. 1988. Early hominid activities at Olduvai. New York:

Aldine de Gruyter.

PRICE, BARBARA J. 1984. "Competition, productive intensifi-
cation, and ranked society: Speculations from evolutionary
theory," in Warfare, culture, and environment. Edited by
R. Brian Ferguson, pp. 209-40. Orlando: Academic Press.

PRICE, T. DOUGLAS, AND JAMES A. BROWN. Editors. I985.

Hunter- gatherers: The emergence of cultural complexity.
Orlando: Academic Press.

426 I CURRENT ANTHROPOLOGY Voluuie 32, Nuuibei 4, August-Octobci 1991

RAPPAPORT, ROY A. i^jid. Ritual, sanctity, and cybernetics.

American Anthropologist 73:5 9-76 .
. 19 J lb. The sacred in human evolution. Annual Review

of Ecology and Systematics 2:23-44.

-. 1979. "Obvious aspects of ritual/' in Ecology, meaning,

and religion. Richmond, Calif.: North Atlantic Books.
RASMussEN, KNUD. 1932. Intellectual culture of the Copper
Eskimo. Report of the Fifth Thule Expedition, vol. 9.

RICHERSON, PETER J., AND ROBERT BOYD. I989. The rolc of

evolved predispositions in cultural evolution. Ethology and

Sociobiology 10:195-219.
ROBARCHEK, CLAYTON A. 1977. Frustration, aggression, and

the nonviolent Semai. American Ethnologist 4:762-79.
. 1990. The Aucas, the cannibals, and the missionaries:

From warfare to peace among the Waorani. Paper presented at

H. F. Guggenheim Conference, Charleston, S.C.

1 99 1. "Agnostic about consciousness" — science, anti-

science, and ratomorphic psychology: A reply to Daly and
Wilson. American Anthropologist 93:165-66. [rkd]

ROBARCHEK, CLAYTON A., AND ROBERT K. DENTAN. I987.

Blood drunkenness and the bloodthirsty Semai. American
An thropologist 89:356-65.
RODMAN, PETER s., AND JOHN c. MiTANi. 1987. "Orang-
utans: Sexual dimorphism in a solitary species," in Primate so-
cieties. Edited by B. B. Smuts, D. L. Cheney, R. M. Seyfarth,
R. W. Wrangham, and T. T. Struhsaker, pp. 146-54. Chicago:
University of Chicago Press.

RODSETH, LARS, RICHARD W. WRANGHAM, ALISA HARRI-

GAN, AND BARBARA B. SMUTS. 1991. The humau Commu-
nity as a primate society, current anthropology 32:221-55.

ROGERS, ALAN R. 1990. Group Selection by selective emigra-
tion: The effects of migration and kin structure. American
Naturalist 135:398-413.

ROHNER, R. P. 1980. Worldwide tests of parental acceptance-
rejection theory: An overview. Behavior Science Research
15:1-21.

ROSALDO, MICHELLE z. 1980. Knowledge and passion: Ilongot
notions of self and social life. New York: Cambridge Univer-
sity Press.

ROSALDO, RENATO. 1980. Ilongot headhunting, 1883-1974: A
study in society and history. Stanford: Stanford University Press.

ROSS, JANE BENNETT, 1984. "Effccts of contact ou rcvcngc
hostilities among the Achuara Jivaro," in Warfare, culture,
and environment. Edited by R. Brian Ferguson, pp. 83-109.
Orlando: Academic Press.

ROSS, MARC HOWARD. 1985. Internal and external conflict and
violence: Cross-cultural evidence and a new analysis. Journal
of Conflict Resolution 29:547-79.

. 1986. A cross-cultural theory of political conflict and

violence. Political Psychology 7:427-69.

SAHLiNS, MARSHALL D. 1961. The Segmentary lineage: An or-
ganization of predatory expansion. American Anthropologist
63:322-45.

. 1968. Tribesmen. Englewood Cliffs: Prentice-Hall.

. 1972. Stone Age economics. Chicago: Aldine.

scHALLER, G. B, 1 976. The mountain gorilla: Ecology and
behavior. Chicago: University of Chicago Press.

scHRiRE, CARMEL. Editor. 1984. Past and present in hunter-
gatherer studies. Orlando: Academic Press.

SERVICE, ELMAN R. 1971. 2d edition. Primitive social organiza-
tion: An evolutionary perspective. New York: Random House.

. 1979. 2d edition. The hunters. Englewood Cliffs:

Prentice-Hall.

SHANKMAN, PAUL. 1991. Culturc coutact, cultural ecology, and
Dani warfare. Man 26:602-19.

SHiPMAN, PAT. 1983. "Early hominid lifestyle: Hunting and
gathering or foraging and scavenging?" in Animals and archae-
ology, vol. I. Edited by J. Clutton-Brock and C. Grigson, pp.
31-49. British Archaeology Reports 163.

. 1986. Scavenging or hunting in early hominids?: Theoret-
ical framework and tests. American Anthropologist 88:27-43.

SHOOK, E. V. 1985. Ho'oponopono. Honolulu: University of
Hawaii Press, [jp]

SHOSTAK, MARjORiE. 1981. Nisa: The life and words of a
IKung woman. Cambridge: Harvard University Press.

SMITH, ERIC A. 1988. "Risk and uncertainty in the 'original af-
fluent society': Evolutionary ecology of resource-sharing and
land tenure," in Hunters and gatherers, vol. i. History, evolu-
tion, and social change. Edited by Tim Ingold, David Riches,
and James Woodbum, pp. 222-68. Oxford: Berg.

SMITH, FRED J., ANTHONY B. FALSETTI, AND STEVEN M.

DONNELLY. 1989. Modcm humau origins. Yearbook of Physi-
cal Anthropology 32:35-68.

SMITH, M. G. 1975. Corporations and society. New York:
Aldine.

SMUTS, BARBARA B. 1987. "Scxual Competition and mate
choice," in Primate societies. Edited by B. B. Smuts, D. L.
Cheney, R. M. Seyfarth, R. W. Wrangham, and T.T. Struh-
saker, pp. 385-99. Chicago: University of Chicago Press.

soFFER, OLGA. 1985. "Patterns of intensification as seen from
the Upper Paleolithic of the Central Russian Plain," in Prehis-
toric hunter-gatherers: The emergence of cultural complexity.
Edited by T. Douglas Price and James A. Brown, pp. 235-70.
Orlando: Academic Press.

soLECKi, RALPH. 1971. Shanidar: The first flower people. New
York: Knopf.

. 1975. Shanidar IV, A Neanderthal flower burial in north-
ern Iraq. Science 190:880-81.

SOLWAY, JACQUELINE S., AND RICHARD B. LEE. I99O. Forag-

ers, genuine or spurious?: Situating the Kalahari San in history.

CURRENT ANTHROPOLOGY 31:109-46.

SPONSEL, LESLIE. 1989. "An authropologist's perspective on
peace and quality of life," in Peace and development: An in-
terdisciplinary perspective. Edited by Daniel S. Sanders and
Jon K. Matsuoka, pp. 29-48. Honolulu: University of Hawaii
School of Social Work, [rkd]

STANFORD, CRAIG B., AND JOHN S. ALLEN. I99I. On Strate-
gic storytelling: Current models of behavioral evolution, cur-
rent ANTHROPOLOGY 32:58-61.

STEKLis, HORST D. 1985. Primate communication, comparative
neurology, and the origin of language re-examined. Journal of
Human Evolution 14:157-73.

STEPHENSON, CAROLYN. 1990. Pcacc studics: The evolution of
peace research and peace education. University of Hawaii In-
stitute for Peace Occasional Paper i . [rkd]

STEWARD, JULIAN H. 1936. "The ccouomic and social basis of
primitive bands," in Essays in anthropology in honor of Alfred
Louis Kroeber, pp. 331-50. Berkeley: University of California
Press.

STREHLOW, T. G. H. 1947. Aranda traditions. Melbourne: Mel-
bourne University Press.

. 1965. "Culture, social structure, and environment," in

Aboriginal man in Australia. Sydney: Angus and Robertson.

STRUM, SHIRLEY C, AND WILLIAM MITCHELL. I987. "Ba-

boon models and muddles," in The evolution of human behav-
ior: Primate models. Edited by Warren G. Kinzey, pp. 87-104.
Albany: State University of New York Press.

suGARDjiTO, J., AND N. NURHUDA. 1981. Mcat-catiug bchav-
ior in wild orangutans, Pongo pygmaeus. Primates 22:414-16.

susMAN, RANDALL L. 1984. The pygmy chimpanzcc: Evolu-
tionary biology and behavior. New York: Plenum Press.

. 1987. "Pygmy chimpanzees and common chimpanzees:

Models for the behavioral ecology of the earliest hominids," in
The evolution of human behavior: Primate models. Edited by
Warren G. Kinzey, pp. 72-86. Albany: State University of New
York Press.

TAKAHATA, YUKio. 1990^. "Adult malcs' social relations with
adult females," in The chimpanzees of the Mahale Mountains:
Sexual and life history strategies. Edited by Toshisada Nishida,
pp. 133-48. Tokyo: University of Tokyo Press.

. 1990b. "Social relationships among adult males," in The

chimpanzees of the Mahale Mountains: Sexual and life history
strategies. Edited by Toshisada Nishida, pp. 149-70. Tokyo:
University of Tokyo Press.

TAKAHATA, YUKIO, TOSHIKAZU HASEGAWA, AND TOSHI-
SADA NISHIDA. 1984. Chimpanzee predation in the Mahale

KNAUFT Violence and Sociality in Human Evolution \ 427

Mountains from August 1979 to May 1982. International Jour-
nal of Primatology 5:213-33.

TANNER, NANCY M. 1987. "The chimpanzec model revisited
and the gathering hypothesis/' in The evolution of human
behavior: Primate models. Edited by Warren G. Kinzey, pp.
3-27. Albany: State University of New York Press.

TELEKi, GEZA. 1973. The predatory behavior of wild chimpan-
zees. Lewisburg: Bucknell University Press.

. 1 98 1. "The omnivorous diet and eclectic feeding habits

of chimpanzees in Gombe National Park, Tanzania," in Om-
nivorous primates: Gathering and hunting in human evolu-
tion. Edited by R. S. O. Harding and G. Teleki, pp. 303-43.
New York: Columbia University Press.

TERASHiMA, HiDEAKi. 1985. Variation and composition princi-
ples of the residence group (band) of the Mbuti Pygmies: Be-
yond a typical/atypical dichotomy. African Study Monographs
4:103-20.

TESTART, ALAIN. 1982. The Significance of food storage among
hunter-gatherers: Residence patterns, population densities, and
social inequalities, current anthropology 23:523-37.

. 1985. Le communisme primitif: Economic et ideologic.

Paris: Maison des Sciences de PHomme.

1987. Game sharing systems and kinship systems among

hunter-gatherers. Man 22:287-304.

-. 1988. Some major problems in the social anthropology of

hunter-gatherers, current anthropology 29:1-31.

THODEN VAN VELZEN, H. U. E., AND W. VAN WETERING.

i960. Residence, power groups, and intra-societal aggression:
An enquiry into the conditions leading to peacefulness within
non-stratified societies. International Archives of Ethnography
49:169-200.

THOMPSON-HANDLER, NANCY, RICHARD K. MALENKY, AND

NOEL BADRiAN. 1984. "Scxual bchavior of Pan paniscus un-
der natural conditions in the Lomako Forest, Zaire," in, The
pygmy chimpanzee: Evolutionary biology and behavior. Edited
by R. L. Susman, pp. 347-68. New York: Plenum Press.

THORNHILL, RANDY, AND NANCY W. THORNHILL. I983. Hu-
man rape: An evolutionary analysis. Ethology and Sociobiol-
0^74:137-73.

TOMASELLO, MICHAEL. 1991. "Proccsscs of Communication in
the origins of language," in Studies in language origins, vol. 2.
Edited by W. Roffler-Ergle. Philadelphia: Benjamins. In press.

TOOBY, JOHN, AND iRVEN DE voRE. 1987. "The reconstruc-
tion of hominid behavioral evolution through strategic model-
ing," in The evolution of human behavior: Primate models.
Edited by Warren G. Kinzey, pp. 183-237. Albany: State Uni-
versity of New York Press.

TOWNSEND, PATRICIA K. w. 1974. Sago production in a New
Guinea economy. Human Ecology 2:217-36. [rkd]

. 1977. "The cultural ecology of sago in New Guinea," in

Sago-76: Papers of the First International Sago Symposium.
Edited by Tan Koonlin, pp. 91-95. Kuala Lumpur, [rkd]

1990. On the possibility /impossibility of tropical forest

hunting and gathering. American Anthropologist 92:745-47.
[rkd]

TOWNSEND, WILLIAM H. 1969. Stouc and Steel tool use in a
New Guinea society. Ethnology 8:199-205. [rkd]

TRELEASE, ALLEN w. 1962. The Iroquois and the western fur
trade: A problem in interpretation. Mississippi Valley Histori-
cal Review 49:32-51. [tsa]

TRIGGER, BRUCE G. 1990. "Maintaining cconomic equality in
opposition to complexity: An Iroquoian case study," in The
evolution of political systems: Sociopolitics in small-scale so-
cieties. Edited by Steadman Upham, pp. 119-45. Cambridge:
Cambridge University Press, [tsa]

TRiNKAus, ERIK. 1983. The Shanidar Ncandcrtals. New York:
Academic Press.

. Editor. 1989. The emergence of modern humans: Bio-
cultural adaptations in the later Pleistocene. Cambridge:
Cambridge University Press.

TRivERS, ROBERT. 1971. "The cvolutiou of reciprocal altru-
ism," in Readings in sociobiology. Edited by T. Clutton-Brock
and P. Harvey, pp. 189-226. San Francisco: Freeman.

. 1985. Social evolution. Menlo Park, Calif.: Cummings.

TRONICK, EDWARD Z., GILDA A. MORELLA, AND STEVE

WINN. 1987. Multiple caretaking of Efe (Pygmy) infants.
American Anthropologist 89:96-106.

TURNBULL, COLIN M. 1961. The forcst pcoplc. Garden City:
Natural History Press.

. 1965^. The Mbuti Pygmies: An ethnographic survey. An-
thropological Papers of the American Museum of Natural His-
tory 50:139-282.

1965b. Wayward servants: The two worlds of the African

Pygmies. Garden City: Natural History Press.

1978. "The politics of non-aggression (Zaire)," in Learn-

ing non- aggression: The experience of non-literate societies.
Edited by Ashley Montagu, pp. 1 61-221. New York: Oxford
University Press.

1982. "The ritualization of potential conflict between the

sexes among the Mbuti," in Politics and history in band socie-
ties. Edited by Eleanor Leacock and Richard B. Lee, pp.
133-55- Cambridge: Cambridge University Press.

1984. "The importance of flux in two hunting societies,"

in Man the hunter. Edited by Richard B. Lee and Irven DeVore,
pp. 132-37. New York: Aldine.

TUTiN, CAROLINE E. G. 1979- Mating patterns and reproduc-
tive strategies in a community of wild chimpanzees. Behav-
ioral Ecology and Sociobiology 6:29-38.

TYLOR, SIR EDWARD BURNETT. 1958 (1871). Primitive culture.
Pt. I. The origins of culture. New York: Harper Torchbooks.
[rkd]

UPHAM, STEADMAN. 1990^. "Analog or digital: Toward a ge-
neric framework for explaining the development of emergent
political systems," in The evolution of political systems:
Sociopolitics of small-scale sedentary societies. Edited by
Steadman Upham, pp. 87-115. Cambridge: Cambridge Univer-
sity Press.

. Editor. 1990b. The evolution of political systems: Socio-
politics of small-scale sedentary societies. Cambridge: Cam-
bridge University Press.

vALZELLi, LuiGi. 1981. The psychobiology of aggrcssion and
violence. New York: Raven Press.

VAN DER STEENHOVEN, GEERT. 1959. Legal conccpts among
the Netsilik Eskimos of Pelly Bay, N.W.T Northern Co-ordi-
nation and Research Centre, Department of Northern Affairs
and National Resources, NCRC-59-3.

VANSINA, TAN. 1966. Kingdoms of the savanna. Madison:
University of Wisconsin Press.

VAYDA, A. p. i960. Maori warfare. Wellington: Polynesian
Society.

. 1989. Explaining why Maring fought. Journal of Anthro-
pological Research 45:159-77. [cb]

viNiNG, DANIEL R., JR- 1986. Social vcrsus reproductive suc-
cess: The central theoretical problem of human sociobiology.
Behavioral and Brain Sciences 9:167-216.

VON FURER-HAiMENDORF, c. 1967. Morals and merit. Chi-
cago: University of Chicago Press. [cb[

WALLACE, ERNEST, AND E. ADAMSON HOEBEL. I952. The

Comanches: Lords of the south Plains. Norman: University of
Oklahoma Press.

WARNER, w. LLOYD. 1930. Mumgin warfare. Oceania
1:457-94.

. 1969 (1937). Black civilization: A social study of an Aus-
tralian tribe. New York: Peter Smith.

WEST, M. M., AND M. J- KONNER. 1976. "The rolc of the fa-
ther: An anthropological perspective," in The role of the father
in child development. Edited by M. E. Lamb. New York: Wiley.

WHALLON) ROBERT. 1989. "Elements of cultural change in the
Later Palaeolithic," in The human revolution: Behavioural and
biological perspectives on the origins of modern humans, vol.
I. Edited by Paul Mellars and Chris Stringer, pp. 433-54.
Princeton: Princeton University Press.

WHITE, FRANCES J- 1989. "Social Organization of pygmy chim-
panzees," in Understanding chimpanzees. Edited by Paul G.
Heltne and Linda A. Marquardt, pp. 194-207. Cambridge:
Harvard University Press.

428 I CURRENT ANTHROPOLOGY Voluuie 32, Nuuiber 4, August-October 1991

WHITE, RANDALL. 1989. 'Troduction complexity and standard-
ization in early Aurignacian bead and pendant manufacture:
Evolutionary implications/' in The human revolution: Behav-
ioural and biological perspectives on the origins of modern
humans, vol. i. Edited by Paul Mellars and Chris Stringer,
pp. 366-90. Princeton: Princeton University Press.

WHITEHEAD, NEIL. 1990. 'The suakc warriors — sons of the ti-
ger's teeth: A descriptive analysis of Carib warfare ca. 1500-
1820," in The anthropology of war. Edited by Jonathan Haas,
pp. 146-70. Cambridge: Cambridge University Press.

WHITING, JOHN W., AND BEATRICE B. WHITING. I975.

Aloofness and intimacy between husbands and wives. Ethos
3:183-207.

wiESSNER, POLLY. 1982. "Risk, reciprocity, and social influ-
ences on IKung San economics," in Politics and history in
band societies. Edited by Eleanor Leacock and Richard B. Lee,
pp. 61-84. Cambridge: Cambridge University Press.

wiLMSEN, EDWIN N. 1989. Land filled with flies: A political
economy of the Kalahari. Chicago: University of Chicago
Press.

WILSON, DAVID SLOAN. 1989. Lcvcls of sclcction: An alterna-
tive to individualism in biology and the human sciences. So-
cial Networks 11:257-72.

. 1980. The natural selection of populations and commu-
nities. Menlo Park, Calif.: Cummings.

WILSON, D. s., AND E. SOBER. 1989. Reviving the supcrorgau-
ism. Journal of Theoretical Biology 136:337-56.

WOBST, H. M. 1978. The archaeo-ethnology of hunter-gatherers,
or The tyranny of the ethnographic record in archaeology.
American Antiquity 43:303-9.

WOLF, ERIC R. 1982. Europe and the people without history.
Berkeley: University of California Press, [tsa]

WOLFF, ROBERT PAUL. 1971. "Violence and the law," in The
rule of law. Edited by Robert Paul Wolff, pp. 54-75. New
York: Simon and Schuster, [tmk]

wooDBURN, JAMES c. 1972. "Ecology, uomadic movcmcut,
and the composition of the local group among hunters and
gatherers: An East African example and its implications," in
Man, settlement, and urbanism. Edited by Peter J. Ucko, Ruth
Tringham, and G. W. Dimbleby, pp. 293-06. Hertfordshire:
Duckworth/Garden City Press.

. 1979. "Minimal politics: The political organization of the

Hadza oif North Tanzania," in Politics in leadership: A com-

parative perspective. Edited by William A. Shack and Percy S.
Cohen, pp. 244-66. Oxford: Clarendon Press.

1980. "Hunters and gatherers today and reconstruction of

the past," in Soviet and Western anthropology. Edited by Er-
nest Gellner, pp. 95-117. New York: Columbia University
Press.

. 1982. Egalitarian societies. Man 17:431-51.

-. 1984. "Stability and flexibility in Hadza residential

groupings," in Man the hunter. Edited by Richard B. Lee and
Irven DeVore, pp. 103-10. New York: Aldine.

-. 1988. "African hunter-gatherer social organization: Is it

best understood as a product of encapsulation?," in Hunters

and gatherers, vol. i. History, evolution, and social change.

Edited by Tim Ingold, David Riches, and James Woodburn,

pp. 31-64. Oxford: Berg,
WRANGHAM, RICHARD w. 1975- The bchavioral ecology of

chimpanzees in Gombe National Park, Tanzania. Ph.D. diss.,

Cambridge University, Cambridge, England.
. 1986. "Ecology and social relationships in two species of

chimpanzee," in Ecological aspects of social evolution. Edited

by D. L Rubenstein and R. W. Wrangham, pp. 353-78.

Princeton: Princeton University Press.

-. 1987. "The significance of African apes for reconstructing

human social evolution," in The evolution of human behavior:
Primate models. Edited by Warren G. Kinzey, pp. 51-71.
Albany: State University of New York Press.

WRESCHNER, ERNEST E, 1981. Red ochrc and human evolution:
A case for discussion, current anthropology 21:631-44.

WRIGHT, G. R. H. 1988. The scvcrcd head in earliest Neolithic
times. Journal of Prehistoric Religion 2:51-56.

YAMAGiwA, J. 1983. Diachronic changes in two eastern low-
land gorilla groups (Gorilla gorilla graueri) in the Mt. Kahuzi
region, Zaire. Primates 24:174-83.

YENGOYAN, ARAM A. 1984. "Demographic and ecological in-
fluences on Aboriginal Australian marriage sections," in Man
the hunter. Edited Richard B. Lee and Irven DeVore, pp.
185-99. New York: Aldine.

ziLHMAN, A. L., AND D. L. CRAMER. 1978. Skeletal differ-
ences between pygmy (Pan paniscus) and common (Pan troglo-
dytes) chimpanzees. Folia Primatologica 29:86-94.

zvELEBiL, MAREK. Editor. 1986. Hunters in transition: Meso-
lithic societies of temperate Eurasia and their transition to
farming. Cambridge: Cambridge University Press.

Wanted

For a two-volume annotated bibliography of world the-
oretical archaeology, reprints and books on theory and
methods published in the last decade. This decade is
represented hardly at all not only in my personal col-
lection but also in the libraries of Leningrad. The
kindness of colleagues helped me to write my 'Tan-

orama of Theoretical Archaeology/' published in CA
in 1977 and updated in Fennoscandia Archaeologica 7
(1990), and I hope that it will assist me in compiling
the bibliography that every archaeologist needs. Please
write: Leo Klejn, Zheleznovodskaja 27, kv. 27, Lenin-
grad 199 155, U.S.S.R.