Vous êtes sur la page 1sur 9

Available online at www.sciencedirect.


Aggression and Violent Behavior 13 (2008) 1 9

Examining the evidence from small-scale societies and early

prehistory and implications for modern theories
of aggression and violence
Grant S. McCall a,, Nancy Shields b
Department of Anthropology, Tulane University, 1326 Audubon St., New Orleans, LA 70118, United States
University of Missouri-St. Louis, St. Louis, MO, United States
Received 2 November 2006; received in revised form 3 April 2007; accepted 4 April 2007
Available online 12 April 2007


This paper attempts to evaluate theoretical positions concerning the causes of violence among human societies using data from
small-scale, radically non-Western societies and archaeological evidence from early hominids. The paper begins by observing that
the almost exclusive focus on violence in industrial societies misses a wide range of variability presented by non-industrial groups,
in the present as well as the past. It is argued that this narrow focus represents a significant bias within many of the social sciences.
The paper also examines evidence of an evolutionary basis for violence and aggression by looking at the early hominid
archaeological record. The paper finds significant evidence for some evolutionary basis for violence given its ubiquity in both the
present as well as the deep archaeological past. The paper closes by proposing a synthetic model combining evolutionary theory
and interactionist perspectives on the inputs leading to aggression and violence in human social groups.
2007 Elsevier Ltd. All rights reserved.

Keywords: Violence; Cross-cultural research; Evolutionary theory; Early hominids


1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2. Reviewing social psychological theories of aggression and violence . . . . . . . . . . . . . . . . . . . . . . . . . . 2
3. Aggression and violence in non-western societies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4. Archaeological evidence for interpersonal violence in early human prehistory . . . . . . . . . . . . . . . . . . . . . 6
5. Synthetic perspectives on the causes of violence in the past and present . . . . . . . . . . . . . . . . . . . . . . . . 8
6. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Corresponding author.
E-mail address: gsmccall@gmail.com (G.S. McCall).

1359-1789/$ - see front matter 2007 Elsevier Ltd. All rights reserved.
2 G.S. McCall, N. Shields / Aggression and Violent Behavior 13 (2008) 19

1. Introduction

Explaining the causes of interpersonal violence has been a problem at the forefront of research in many of the social
sciences for more than a century (Anderson & Bushman, 2002; Walker, 2001; Wrangham & Peterson, 1996). Most
attempts to examine interpersonal violence have focused on either ethnographic or experimental approaches to modern
people in Western industrial societies (Walker, 2001). This paper argues that the exclusive focus on modern Western
societies misses the substantial variability in the occurrence of interpersonal violence among human groups over space
and time. We also feel strongly that understanding the full range of variation in the manifestations of interpersonal
violence is the key to building theory concerning its origins and causes. In particular, there are two settings where
research into the causes of interpersonal violence has been underdeveloped: (1) small-scale, non-Western societies, and
(2) human groups in the past before the origins of modern industrial social and economic systems. While there is
substantial theory concerning the causes of violence in Western societies at all scales and levels of intensity developed
by all branches of the social sciences, the causes of violence outside of this rather limited context has been dramatically
Another closely related question concerns the universality of interpersonal violence to all human groups. In other
words, is interpersonal violence a universal condition of all people? We believe that this relates directly to the issues of
the recognition of wider variation in the manifestations of interpersonal violence. We also argue that this is
fundamentally an evolutionary question, because if interpersonal violence is a human universal, then it must have been
shared by our last common evolutionary ancestors. For this reason, the recognition of evidence for interpersonal
violence in fossil and archaeological records of our early hominid ancestry is extremely important to modern theoretical
perspectives on the nature of interpersonal violence. In addition, studies of aggression in modern non-human primates
have added an important point of comparison in terms of the causes interpersonal violence among our most closely
related species. This paper discusses the evidence for interpersonal violence among both our early hominid ancestors
and modern non-human primates.

2. Reviewing social psychological theories of aggression and violence

This paper rests on a fundamental definition concerning the distinction between violence and aggression. Anderson
and Bushman (2002) see the difference between aggression and violence to be a matter of degree, with aggression
defined as behavior intended to produce deliberate harm to another and violence having extreme harm as its intent
(such as murder). We consider aggression as a state of arousal manifested by various emotional communicative
strategies (e.g., shouting, gesturing, etc.). In contrast, we define violence as the physical attack of one person by another
in the context of aggressive behavior. On this basis, one can be aggressive without being violent, but not the reverse.
One important question addressed by Anderson and Bushman (2002) concerns the stimuli leading to aggressive
behavior. A more important question for this paper concerns the transition from aggression to violent behavior,
especially given the nature of archaeological and fossil evidence. Aggression without violence does not result in
archaeologically investigable remains.
Existing theories about the causes of interpersonal violence can be divided into two categories: (1) Evolutionary, and
(2) interactionist. Before proceeding to our evidence, we will briefly review these.
Evolutionary theories include the ideas of kin selection and inclusive fitness. In other words, sociobiologists have
focused on explaining how various patterns of interpersonal violence might have increased the fitness of offspring over
the long haul of evolutionary time (Wrangham & Peterson, 1996). From a sociobiological perspective, it is logical to
predict relatively low incidences of violence between related individuals. In addition, sociobiology predicts higher rates
of violence between males in the context of male/male competition. This evolutionary perspective also predicts lower
frequencies of violent behavior between females and their offspring than males and their offspring, given the certainty
of offspring belonging to females in comparison with greater uncertainty of males. These are robust ethological
principles that have been documented on all manner of organisms from eusocial insects to higher order primates, and it
would be foolish to think that they do not operate on human behavior, as well.
In fact, statistics from the U.S. Bureau of statistics (2006) support many of these predictions. In 2004, men were
more than ten times more likely than women to commit homicide, and of all homicides, 65.2% were male on male
violence. The next largest category was male on female violence, which accounted for only 22.6% of homicides.
Furthermore, in 2004 only 11% of all homicides were between intimates, which would seem to support the kin
G.S. McCall, N. Shields / Aggression and Violent Behavior 13 (2008) 19 3

selection hypothesis, with women being 3.3 times more likely than men to be victims. The statistics on child homicides
(under the age of 5) are less clear. Between 1976 and 2004, mothers were just as likely as fathers to kill a child; of all
homicides, mothers committed 31%, and 30% were committed by fathers. However, unfortunately for our current
purposes, stepparents were combined with parents, so the question of kin selection cannot be fully assessed. Finally,
reports of substantiated child abuse in 2004 (U.S. Department of Health and Human Services, 2006) were actually
significantly higher among mothers (38.8% of cases) than fathers (18.3% of cases), although these statistics do not take
into account the greater amount of time that mothers typically spend with their children.
Where evolutionary theories of conflict are underdeveloped has to do with the fitness benefits of violence in certain
situations. Phrased another way, why would it be selectively advantageous for a given individual to engage in
aggressive and/or violent behavior? Some causes are transparent, while others are not. For example, competition
between males for potential mates clearly might benefit an individual that could intimidate or defeat a competitor
through aggressive or violent behavior. A less obvious explanation has to do with competition for scarce resources. For
example, when food is scarce, aggressive or violent behavior might benefit an individual through the intimidation or
defeat of competitors, offering improved nutrition for the individual and his/her offspring. Among humans at least,
these aspects of evolutionary theory concerning aggressive and violent behavior is somewhat weak, and there is instead
a general assumption that violent behavior is a universal strategy for securing access to either reproductive or energetic
resources. Sociobiology can also be criticized as a just-so-story in the sense that it only relates large-scale phenomena
with their putative evolutionary causes, without elucidating the cognitive or behavioral dynamics in which these
phenomena occur.
Given this problem, interactionist theories of violence and aggression are useful in the sense that they provide an
analytical framework and descriptive vocabulary for studying behaviors in their dynamic social contexts. Interactionist
theories are particularly important because they make distinctions between instincts, emotions, and rational thoughts.
By treating social encounters as scripted affairs, they also consider how various small-scale situational variables bring
about certain patterns of behavior. These perspectives are instrumental in getting evolutionary theory out of its just-so-
story trap, and framing investigable problems having to do with human social and cognitive processes.
In their recent theoretical synthesis, Anderson and Bushman (2002) propose a general aggression model (GAM),
which incorporates many lower level, domain specific theories. The model is based on the concept of knowledge
structures and how they operate to produce behavior. Knowledge structures arise out of experience, influence
perception, can become more or less automatic in some cases, and are linked to affective states, beliefs and behavior. In
essence, they are used to guide responses to the environment. Knowledge structures include perceptual schemata (used
to classify objects and events), person schemata (used to classify different kinds of persons), and behavioral scripts
which define the kinds of behaviors that are appropriate in various situations.
The model focuses on characteristics of person and the situation as they relate to a person's present internal state
(affect, arousal and cognition), and ultimately appraisal and decision making processes. Appraisal and decision making
processes lead to either impulsive or thoughtful actions, which in turn cycle back to the next social encounter. Anderson
and Bushman (2002) propose the model shown in Fig. 1.

Fig. 1. General aggression model (GAM) proposed by Anderson and Bushman (2002).
4 G.S. McCall, N. Shields / Aggression and Violent Behavior 13 (2008) 19

The model begins with an analysis of the person in the situation, which are inputs. Inputs include all the
biological, environmental, psychological, and social features of the person and the situation. Person inputs include
personality traits, attitudes, values and genetic predispositions toward the use of aggression. Together they represent a
person's preparedness to aggress. Person inputs take into account evolutionary outcomes, such as sex differences in
the use of aggression. For example, Anderson and Bushman (2002) note that men are much more likely than women to
commit murder (a ratio of 10:1), although women seem as likely to aggress when provoked. This is consistent with
research on spousal violence (Brush, 1990) which has found men and women about equally likely to use violence
against their partners, but when they do, men are much more likely to inflict serious injuries. Finally, males seem more
likely to aggress in a direct fashion, while females are more likely to aggress in an indirect way.
People also possess scripts for the use of aggressive behavior. They involve the relationships between various
concepts related to aggression, and ties to different ways of responding, such as anger or retaliation. Scripts become
attached to features of the situation such as aggressive cues, provocation, frustration, pain or discomfort, and incentives
to act aggressively.
Person and situation inputs produce internal states, which Anderson and Bushman (2002) refer to as the routes to
aggressive or non-aggressive behavior. A person's internal state is the interaction between cognitions, affect, and
arousal. Cognitions include hostile thoughts and aggressive scripts, and may come to be highly salient or accessible to
an individual. Affect includes negative or hostile feelings, and anger. Arousal involves the person's physiological state
that may be the result of factors unrelated to anger, such as exercise or the use of drugs. The three routes are highly
Outcomes (what the person actually does) can be relatively automatic (immediate appraisal) or controlled
(reappraisal). The inferences a person makes about the situation are related to his or her internal state and social leaning
history. An important part of the appraisal process is the person's resources to reflect upon the situation. Resources
include available time and the person's cognitive capacities to analyze. When an outcome or course of action is
important to the person, and the immediate action is unsatisfying, reappraisal is more likely to take place.
Anderson and Bushman (2002) discuss other factors that can affect outcomes. For example, they note that most
individuals have moral standards, beliefs and attitudes that are inconsistent with the use of extreme forms of aggression,
such as murder. However, at times inhibitions can be overridden by rationalizations such as moral justification for the
use of violence (such as wars), and dehumanization of the victim (such as torturing prisoners). Extreme anger can also
override inhibitions about the use of aggression. There may also be shared motivation to aggress when a group
experiences a threat. Anderson and Bushman (2002) also note that there may be an evolutionary basis for the link
between anger and aggression, or at least the ease of learning to use aggression in response to anger.
Although noting the problems associated with the simple frustrationaggression hypothesis (frustration does not
always result in aggression, and aggression is not always the result of frustration) Anderson and Bushman (2002) do
see anger as extremely important to the use of aggressive behavior. Anger lowers inhibitions by justifying the use of
retaliation, and may interfere with higher-level reappraisal processes. It increases the probability of aggression by
enabling an individual to hold aggressive intentions over time. Anger serves as an informational cue to the individual
(i.e., he or she must be angry for a reason) and enhances hostile interpretations. Anger also activates aggressive
thoughts, behavioral scripts, and physiological responses since it is also linked to related knowledge structures. By
increasing arousal levels, anger also makes an aggressive response more likely.
The synthetic model proposed by Anderson and Bushman (2002) raises some important evolutionary questions that
are much more specific than those usually tackled in sociobiological studies. For example, several characteristics of
person inputs have universal human evolutionary underpinnings. Likewise, Anderson and Bushman (2002) see the
relationship between anger and aggression as determined by natural selection in the deep past. We believe that this
general aggression model (GAM) is very effective in dealing with aggression and violence in modern Western contexts.
However, the universal evolutionary aspects of this model warrant testing in radically non-Western ethnographic
contexts, and using archaeological evidence from our early hominid ancestors.

3. Aggression and violence in non-western societies

There is no modern human society in which both aggression and violence is unknown (Walker, 2001). Likewise,
there is no species of primate that never engages in some form of aggression (Wrangham & Peterson, 1996). On this
basis, many authors have argued that violence has deep evolutionary roots for modern humans (Anderson & Bushman,
G.S. McCall, N. Shields / Aggression and Violent Behavior 13 (2008) 19 5

2002; Walker, 2001; Wrangham & Peterson, 1996). Specifically, Walker (2001) and Wrangham and Peterson (1996)
argue that because certain forms of violence are shared by all modern human groups and closely related primates, it is
an ancestral condition possessed by our last common ancestor. In other words, certain specific capacities for violence
are unlikely to have originated independently many different times among different groups of people or even species of
primate. Still, how do patterns of aggression and violence in non-Western societies compare with those on which our
social psychological theories are based? This question is critical to investigating models such as that of Anderson and
Bushman (2002).
Modern foraging people have been a particularly frequent target of evolutionary anthropological research, and this is
no less true in terms of the study of aggression and violence. Many early anthropologists were predisposed to believe
that foragers lacked the problems of violence shared by modern Western societies, once notions of unilinear evolution
and progress were dismissed from the social sciences. Sahlins' (1968) idea of foragers as the original affluent society
typifies these beliefs in many respects. Another example is the image represented by Thomas (1959) of the Kalahari Ju/
'hoansi as the harmless people, free from the typical social strife and violence endemic to Western industrial
societies. Kelly (1995) points out that many anthropologists at mid century, during a period of intense warfare and
genocide, saw foragers as vestiges of original human innocence from which the rest of human history descended into
chaos. Indeed, views of foragers as the noble savage free from violence have been remarkably tenacious.
This perspective has proven problematic for many reasons. Numerous studies have shown modern forager societies
to have rates of assault and murder as high as modern industrial societies. Even the Ju/'hoansi in the Kalahari have
murder rates similar to (or higher than) those of modern North American cities (Lee, 1979). In addition, many stratified
forager societies, such as those of the American Northwest, not only historically had very high rates of violence, but
also had culturally sanctioned, institutionalized violence as legitimate social activity (Ferguson, 1983; Kelly, 1995).
Among stratified foragers, violence frequently takes the form of inter-group territorial conflict, involving raiding for
the capture of food resources, land for their acquisition, and slaves. Institutionalized violence also takes place in the
context of establishment of rank, gaining prestige, and the expression of political power.
One problem with this approach is that much of modern forager social and economic behavior is driven by the forces
of inclusion in global economic networks and political control of modern nation-states. For example, Lee (1979) states
that the high rate of violence among the Ju/'hoansi is largely the result of forced sedentization and population packing,
economic exploitation, resource shortage, disease, and alcohol abuse. Similarly, Ferguson (1983) argues that the
incidence of violence among American Northwest foragers was dramatically exacerbated by the pressures caused by
contact with Euro-American colonists. Around the world, foraging societies suffer from conditions caused by their
relationship with the core Western economic and political powers, and it is not surprising that rates of violence are high
since foragers are among the most marginalized peoples in the world today. Nonetheless, it is unlikely that all
interpersonal violence in modern forager societies has its roots in contact with the West, and there is also substantial
prehistoric evidence for interpersonal violence, which we will discuss more in the next section.
Among non-Western agricultural societies, violence is equally pervasive. Several studies of violence in New Guinea
exemplify this. For example, Knauft (1987) reports a homicide rate for the Gebusi of 568 per 100,000 more than one
hundred times greater than that for the United States in 2004 (Bureau of Justice Statistics, 2006). Huesmann (1986,
1998) confirms this high rate of violence, and links it with the frequent cultural rehearsal of violence scripts. The New
Guinea case clearly shows that the overall incidence of violence is remarkably high, but the specific patterns are quite
different from those seen among Western societies. Knauft (1987) shows that the victims of homicide were likely to be
related to the attacker. This study also shows essentially no revenge killing.
Knauft (1987) uses these findings to criticize high-level sociobiological theories, and this does indeed illustrate
some of the weaknesses of the evolutionary perspectives outlined in the previous section. While aggression and
violence are universal features of human behavior in all modern societies, the specific patterns of their manifestation are
actually quite variable. Returning to Anderson and Bushman's (2002) general aggression model, the differences in
patterns of violence seen in New Guinea can be explained in terms of differences in person inputs. The cultural norms
in terms of the practice of violence are much different, as well as social, political, and economic structures that underlie
them. The New Guinea case does little to undermine evolutionary perspectives on the biological aspects of person
inputs (for example, male/male violence is still much more common than any other form). It also does not challenge an
evolutionary basis concerning the connection between anger and aggression.
In examining ethnographic societies, many variables contribute to the ultimate patterns of aggression and violence
observed. Some of these variables concern the specifics of social structure, political organization, economic conditions,
6 G.S. McCall, N. Shields / Aggression and Violent Behavior 13 (2008) 19

and specific cultural norms or scripts. These are quite variable across modern societies and account for much of the
variability observed in terms of patterns of violence. Some variables have evolutionary bases, such as aggression
between males, and the link between anger and aggression. For this reason, it should be possible to see evidence of
certain patterns of violence in the archaeological record of our evolutionary past. The next section examines this

4. Archaeological evidence for interpersonal violence in early human prehistory

There is ample evidence for interpersonal violence in early hominid prehistory in terms of damage to skeletal
remains. This is particularly the case for Middle Pleistocene hominids. Many Middle Pleistocene hominid skeletons
from Africa, Asia, and Europe show evidence of skeletal trauma, particularly cranial trauma, caused by interpersonal
violence (Klein, 1999). This portion of the paper focuses on Neanderthals because of the very large numbers of fairly
complete skeletons relative to other contemporary or earlier hominids, including other contemporaneous populations of
early Homo sapiens. The evidence for this violence among Neanderthals takes several forms. Berger and Trinkaus
(1995) focus on the frequencies of skeletal lesions by anatomical part. They observe that Neanderthals in Europe and
the Near East have an unusual pattern of skeletal trauma when compared with modern hospital populations. Berger and
Trinkaus (1995) combine most known Neanderthal skeletons (representing around 200 individuals with various levels
of skeletal representation, though it is difficult to determine exactly how many individuals are represented). In
particular, the Neanderthals that Berger and Trinkaus (1995) describe have elevated rates of cranial/neck and trunk
trauma when compared with historical hospital populations from New York and London composed of all injuries
leading to skeletal fractures. Berger and Trinkaus (1995) acknowledge that interpersonal violence is one source of this
pattern. However, they see injury from animal encounters during male hunting as a more important source of these
injuries (using a clinical population of modern rodeo riders as a point of comparison).
Others have argued that the elevated rates of cranial/neck trauma are mainly the result of interpersonal violence. For
example, Gardner (2001) makes this point in discussing the fragmented skeletal remains from the Krapina cave and
Walker (2001) explores this possibility more widely for early hominids. Indeed, frequencies of skeletal lesions by
anatomical part are consistent with interpersonal violence in terms of the elevated rate of cranial/neck trauma. Fig. 1
shows the frequency of fracture by area of the body for (1) a modern New York hospital population, including 8758
individuals, reported in Berger and Trinkaus (1995), (2) a study of English assault victims, including 167 individuals,
reported by Sheperd, Shapland, Peacre, and Scully, (1990), and (3) most known Neanderthal remains showing signs of
skeletal trauma from Europe and the Near East, including 20 individuals, reported by Berger and Trinkaus (1995). It is
important to point out that, in all cases, these reports only reflect individuals injured badly enough to cause permanent
damage to bones. It is also important to point out that these reports do not reflect at all the total population from which
injured individuals were derived. In other words, it may also be significant that almost 10% of relatively complete
Neanderthal individual skeletons show signs of serious injury (Berger & Trinkaus, 1995) (Fig. 2).
It can be assumed that most of the injuries to the New York hospital population injuries were the result of accidents,
mainly caused by falling. This results in relatively low frequencies of cranial /neck trauma and relatively high
frequencies of shoulder/arm and hand trauma, resulting from individuals' attempts to catch themselves. In contrast, the
English assault victims had very high rates of cranial/neck trauma and very low rates of damage to other parts of the
body. We suggest that the high frequency of head trauma seen in Neanderthals was the result of interpersonal violence.
While there are far fewer skeletons complete enough to allow assessment of injury by anatomical part, other early
Homo sapiens populations in Africa and the Near East also show elevated rates of head trauma, suggesting that this
pattern is not unique to Neanderthals (Klein, 1999).
It is also interesting to point out that five of the Neanderthal skeletons showing cranial/neck trauma where sex could
be established were male, and only one was female (and even this assignment is somewhat doubtful; see Berger &
Trinkaus, 1995). While this sample is quite small, this supports the idea of increased levels of male/male violence seen
among modern human populations. This is an interesting pattern that might become clearer as more fossil skeletons are
In addition to the frequencies of skeletal lesion by anatomical part, there are two cases of injuries that strongly
suggest attack with weapons. Most recently, Zollikofer and colleagues (2001) report a CT scan and computer
reconstruction of the St. Cesaire 1 Neanderthal. The CT scanning revealed clear evidence for a healed stabbing wound
to the skull. Similarly, the Shanidar 3 Neanderthal shows clear evidence of a stabbing wound to the chest, through the
G.S. McCall, N. Shields / Aggression and Violent Behavior 13 (2008) 19 7

Fig. 2. Frequency bone fracture by anatomic part for New York hospitals in 1910 (Berger & Trinkaus, 1995), victims of assault in the UK in 1986
(Shepherd et al., 1990), and most known Neanderthal individuals (Berger & Trinkaus, 1995).

ninth left rib (Trinkaus, 1983). In the case of Shanidar 3, the wound shows no signs of healing and other aspects of the
archaeological context suggest that the weapon remained lodged in the individual long after his death. It is also possible
that the Shanidar 3 injury was the result of a thrown projectile. Both cases show the exertion of lethal or near-lethal
force with weapons, and both individuals were male.
Other sources of interpersonal violence are more controversial. Many physical anthropologists have seen signs of
cannibalism early in the hominid fossil record. For example, Weidenreich (1943) saw the patterns of skull breakage
among Asian Homo erectus as clear evidence for the removal and consumption of hominid brains by other hominids.
Since Weidenreich's time, it has become obvious that such bone breakage occurred on post-mortem dry bone, and was
the result of natural weathering and modification by non-human agents (Binford & Ho, 1985).
Other putative cases of early hominid cannibalism are more compelling. For example, there are cutmarks (assumed
to be the result of hominid butchery) on the various hominid skeletal remains from Sterkfontain, South Africa
(Pickering, White & Toth, 2000), Atapuerca, Spain (Fernandez-Jalvo, Diez, Caceresi, & Rosell, 1999), Bodo, Ethiopia
(White, 1986), Klasies River Mouth, South Africa (Singer & Wymer, 1982), Moula-Guercy, France (Defleur, White,
Valensi, Slimak, & Cregut-Bonnoure, 1999), and other more poorly documented sites too numerous to list here
(Walker, 2001). It is interesting that essentially every large deposit of early hominid remains from the Lower and
Middle Pleistocene show some signs of butchery by hominids. It is important to recognize that the inference of
cannibalism from these remains is often difficult, as other carnivores and natural forces can create damage on bones that
resembles that caused by butchery. It is also the case that cannibalism does not necessarily imply violence or murder.
Many have argued that Lower and Middle Pleistocene hominids frequently scavenged the remains of other large
mammals. These cases may simply represent the need-driven scavenging of hominid remains, similar to several high-
profile cases of modern cannibalism, such as the Donner Party.
One more trend in prehistory that is worth pointing out is the dramatic upsurge in the numbers of skeletons with
evidence for trauma caused by violence since the end of the Pleistocene 10,000 years ago. This time period is
dominated by graveyards filled with individuals killed during violence conflict. Of course, most of these individuals
were killed in the context of large-scale warfare, which is unique to the stratified societies of the last few millennia. This
is particularly the case for the state-level societies of Europe, the Near East, South East Asia, and Mesoamerica. The
connection between large-scale warfare and large agricultural states has been recognized by social scientists since the
Enlightenment, and this may be a special form of violence outside of the scope of this paper. Still, it is interesting to
imagine the perspective of archaeologists in the future upon examining the graveyards of soldiers from the many wars
of the 20th century and recognizing the steady increase in warfare throughout the course of the last 10,000 years.
In summary, while problems of discovery, preservation, and interpretation hinder the identification of interpersonal
violence in the past, there is clear evidence for interpersonal violence deep into our evolutionary past. This raises the
question, what were the causes of violence among our early hominid ancestors, and how different were they from
modern causes? Is the general aggression model of Anderson and Bushman (2002) applicable to these ancient cases?
While the evidence examined here is highly fragmentary, it appears that some aspects of interpersonal violence among
our early hominid ancestors were comparable to modern cases. There is clear evidence for a universal biological basis
8 G.S. McCall, N. Shields / Aggression and Violent Behavior 13 (2008) 19

for aggression as a personal input. In addition, there is emerging evidence that, in the past, this biological tendency
towards violent behavior was stronger between males than females, just as it is among modern people. What we cannot
know are the other personal inputs leading to violence, specifically those concerning certain situations and violence
scripts. Apparently, there was no time when violence was absent from human prehistory, but at the same time, there has
been no time in human prehistory with a larger scale of violence than at the present.

5. Synthetic perspectives on the causes of violence in the past and present

The causes of violence have been a prime subject matter for all branches of the social sciences. Some branches (e.g.,
history and political science) have been disposed to look for the large-scale causes of warfare, and others have focused
on the explanation of individual aggression (e.g., social psychology). However, essentially all attempts to understand
the causes of violence at any scale have examined Western industrial societies in the present or recent time periods. We
argue that this perspective misses substantial variability in patterns of interpersonal violence across differently
organized societies and over time. In a sense, modern interpersonal violence is not even the tip of the iceberg in terms
of the diversity of patterns and causes of interpersonal violence over space and time. This paper has tried to briefly
explore evidence concerning interpersonal violence in the deep evolutionary past of modern human societies.
A few immediately apparent facts have some important implications for modern theories of aggression and violence.
First, it is evident that the history of interpersonal violence is as old as the hominid evolutionary line itself. There is
abundant evidence for this in terms of trauma on early hominid fossil skeletons and the ubiquity of interpersonal
violence among all modern human groups and closely related primate species. Therefore, theories of conflict that view
interpersonal violence as purely a modern phenomenon with no evolutionary basis must be taken with a grain of salt.
Second, there is remarkable variability in the patterns and causes of interpersonal violence evident in these subjects.
The reasons for interpersonal violence and the ways in which it plays out are incredibly diverse, and the extent of
overlap is unknown. For this reason, theories of aggression and violence that see interpersonal violence as an
unchanging universal for all human societies (e.g., Wrangham & Peterson, 1996) cannot be easily substantiated.
We must begin with the premise that there is a biological basis for aggression among humans that has deep
evolutionary roots, and that it is possible to use evolutionary theoretical positions to study aggression and violence
among modern humans. Next, we must acknowledge that there are many diverse variables that influence patterns of
aggression and mediate in the transition from aggression to violence. The fact that modern patterns of violence are
highly variable does not at all undermine the proposition that certain aspects of violent behavior have an evolutionary
basis, which is the basic argument presented by Knauft (1987). Instead, in order to build theory concerning the causes
of interpersonal violence, we must seek to understand the relationships between the patterns of interpersonal violence
we observe and the many variables, including those based on evolutionary ecology, that influence them.
Does the fact that there is substantial evidence for an evolutionary basis for modern patterns of aggression and
violence mean that there are no possible pathways for their prevention and resolution? This is clearly not the case. The
model presented by Anderson and Bushman (2002) clearly shows that there are many inputs leading to violence that
offer potential loci of intervention. We would also add that an evolutionary basis for certain aspects of aggression does
not mean that certain humans are programmed to be violent. Instead, we argue that there are evolutionarily based
predispositions for individuals to respond to certain situations in specific ways. Violence is the result of interaction
between a few important biological inputs and many strong situational inputs and environmental influences. By dealing
with the situational inputs and environmental influences, such as the frequent rehearsal of violence scripts described by
Knauft (1987) in New Guinea, violence can be prevented.
It is also important to reiterate that some level of violence is apparent deep into the archaeological record of our
species' behavior. However, there is no time at which more violence has occurred than in our recent history. The many
worldwide conflicts of the twentieth century testify to this, and show the extreme flexibility of human behavior in terms
of the practice of violence. That humans can be extremely violent does not mean that they are biological determined to
do so. This represents an important problem for the applied social sciences.

6. Conclusion

This paper has reviewed evolutionary and interactionist models of aggression and violence in the light of non-
Western ethnographic and early hominid archaeological evidence. We have argued that (1) violence is ubiquitous
G.S. McCall, N. Shields / Aggression and Violent Behavior 13 (2008) 19 9

among all modern human societies, and (2) there is ample evidence for interpersonal violence among our early hominid
ancestors. On this basis, we suggest that there is clear evidence to support at least some evolutionary biological input
influencing modern patterns of violence. We have also found that there is tremendous variability and diversity in the
patterns of violence that have been observed. This implies that the occurrence of violence is a highly complex and
multivariate phenomenon. We conclude that only multi-causal models that take into account both evolutionary
biological and situational or environmental influences can explain human patterns of aggression and violence, and we
describe the general aggression model of Anderson and Bushman (2002) as an example of such a model.
Future research must focus on two central problems. The first is the refinement of archaeological methods for
recognizing evidence for interpersonal violence based on skeletal remains, and the expansion of early hominid skeletal
dataset available for studying this problem. The second involves further comparative work among diverse modern
societies in terms of documenting both patterns of violence and the many inputs that contribute to these patterns. These
directions will help build more elegant theory concerning the causes of violence among human groups, and this is the
first step towards addressing the problem.


Anderson, C. A., & Bushman, B. J. (2002). Human aggression. Annual Review of Psychology, 53, 2751.
Berger, T. D., & Trinkaus, E. (1995). Patterns of trauma among Neanderthals. Journal of Archaeological Science, 22, 841852.
Binford, L. R., & Ho, C. K. (1985). Taphonomy at a distance: Zhoukoudian, The cave home of Peking man?. Current Anthropology, 26, 413442.
Brush, L. D. (1990). Violent acts and injurious outcomes in married couples: Methodological issues in the national survey of families and households.
Gender and Society, 4, 5667.
Bureau of Justice Statistics (2006). www.ojp.usdoj.gov/bjs/. Retrieved July 12, 2006.
Defleur, A., White, T. D., Valensi, P., Slimak, L., & Cregut-Bonnoure, E. (1999). Neanderthal cannibalism at Moula Guercy, Ardeche, France.
Science, 286, 128131.
Ferguson, B. (1983). Warfare and redistributive exchange on Northwest Coast. In E. Tooker (Ed.), The Development of Political Organization in
Native North America (pp. 133147). Washington D.C.: American Ethnological Society.
Fernandez-Jalvo, Y., Diez, J. C., Caceresi, I., & Rosell, J. (1999). Human cannibalism in the Early Pleistocene of Europe (Gran Dolina, Sierra de
Atapuerca, Burgos, Spain). Journal of Human Evolution, 37, 591622.
Gardner, J. C. (2001). An analysis of the pathology of the Krapina Neanderthals.American Journal of Physical Anthropology, 35, 68 (Abstract).
Huesmann, L. R. (1986). Psychological processes promoting the relation between the exposure to media violence and aggressive behavior by the
viewer. Journal of Social Issues, 42, 125140.
Huesmann, L. R. (1998). The role of social information processing and cognitive schema in the acquisition and maintenance of habitual aggressive
behavior. In R. G. Geen & E. Donnerstein (Eds.), Aggression: Theoretical and Empirical Reviews (pp. 73109). New York: Academic Press.
Kelly, R. L. (1995). The foraging spectrum: Diversity in hunter-gatherer lifeways. Washington, D.C.: Smithsonian Institution Press.
Klein, R. G. (1999). The human career: Human biological and cultural origins, 2nd edition. Chicago: University of Chicago Press.
Knauft, B. M. (1987). Reconsidering violence in simple human societies among the Gebuis of New Guinea. Current Anthropology, 28, 457500.
Lee, R. B. (1979). The !Kung San: Men, women, and work in a foraging society. Cambridge: Cambridge University Press.
Pickering, T., White, T. D., & Toth, N. (2000). Cutmarks on a Plio-Pleistocene hominid from Sterkfontein, South Africa. American Journal of
Physical Anthropology, 111, 579564.
Sahlins, M. (1968). Notes on the Original Affluent Society. In R. B. Lee & I. Devore (Eds.), Man the Hunter (pp. 8589). Chicago: Aldine.
Singer, R., & Wymer, J. J. (1982). The Middle Stone Age at Klasies River Mouth in South Africa. Chicago: University of Chicago Press.
Sheperd, J. P., Shapland, M., Peacre, N. X., & Scully, C. (1990). Pattern, severity and aetiology of injuries in victims of assault. Journal of the Royal
Society of Medicine, 83, 7578.
Thomas, E. M. (1959). The harmless people. New York: Vintage Books.
Trinkaus, E. (1983). The Shanidar Neanderthals. New York: Academic Press.
U.S. Department of Health and Human Services. (2006). (http://www.acf.hhs.gov/programs/cb/pubs/cm04/table3_20.htm. Retrived August 28, 2006).
Walker, P. L. (2001). A bioarchaeological perspective on the history of violence. Annual Review of Anthropology, 30, 573596.
Weidenreich, F. (1943). The skull of the Sinanthropus pekinensis: A comparative study on a primitive hominid skull. Pehpei: Geological Survey of
White, T. D. (1986). Cut marks on the Bodo cranium: A case of prehistoric defleshing. American Journal of Physical Anthropology, 69, 503510.
Wrangham, R., & Peterson, D. (1996). Demonic males: Apes and the evolution of human violence. Boston: Houghton Mifflin.
Zollikofer, C. P. E., Ponce de Leon, M. S., Vandermeersch, B., & Leveque, F. (2002). Evidence for interpersonal violence in the St. Cesaire
Neanderthal. Proceedings of the National Academy of Science, 99, 64446448.