Is Adolescent Bullying an Evolutionary Adaptation?

AGGRESSIVE BEHAVIOR
Volume 00, pages 1-17 (2012)
Is Adolescent Bullying an Evolutionary Adaptation?
Anthony A. Volk1, Joseph A. Camilleri2, Andrew V. Dane3, and Zopito A. Marini1
1Department of Child and Youth Studies, Brock University, Ontario, Canada
2Department of Psychology, Westfield State University, Westfield, Massachusetts
3Department of Psychology, Brock University, Ontario, Canada
: : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : :
Bullying appears to be ubiquitous across cultures, involving hundreds of millions of adolescents worldwide, and has potentially serious
negative consequences for its participants (particularly victims). We challenge the traditionally held belief that bullying results from
maladaptive development by reviewing evidence that bullying may be, in part, an evolved, facultative, adaptive strategy that offers
some benefits to its practitioners. In support of this view, we draw from research that suggests bullying serves to promote adolescent
bullies' evolutionarily-relevant somatic, sexual, and dominance goals, has a genetic basis, and is widespread among nonhuman
animals. We identify and explain differences in the bullying behavior of the two sexes, as well as when and why bullying is adaptive
and when it may not be. We offer commentary on both the failures and successes of current anti-bullying interventions from an
evolutionary perspective and suggest future directions for both research and anti-bullying interventions. Aggr. Behav. 00:1-17, 2012.
C 2012 Wiley Periodicals, Inc.
: : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : :
Keywords: bullying; evolutionary psychology; evolution; adaptation; adolescence
INTRODUCTION
This then begs the question--why are so many ado-
lescents in so many cultures involved in a potentially
Bullying is a specific form of aggression that is
harmful and/or antisocial behavior?
characterized by an imbalance of power whereby a
Traditionally, bullying researchers have viewed bul-
more powerful individual repeatedly and intention-
lying as the result of maladaptive development [e.g.,
ally causes harm to a weaker individual [Olweus,
Laslett, 1980]. Bullying is believed to be what happens
1993, pp. 8-10]. In the last two decades, a considerable
when something goes wrong with the developmen-
amount of research has focused on bullying [Berger,
tal process [Smokowski and Kopasz, 2005; Walden
2007]. Bullying has been observed in every society in
and Beran, 2010]. However, using the above global
which it has been examined [Smith, 1999] where it
prevalence rates, this means that something has gone
peaks around the age of 14 and has prevalence rates
"wrong" with the development of 100-600 million
ranging between 10% and 60% of adolescents depend-
adolescents each year. What is more, data from non-
ing on how one defines "repeated" [Volk et al., 2006].
industrialized societies suggest that this same mal-
Translated to a global scale, this yields a total of 100-
adaptive developmental influence would have to en-
600 million adolescents who are directly involved in
gender bullying over the vastly different cultural and
bullying each year [Volk et al., 2006]. Moreover, bully-
environmental contexts inhabited by geographically
ing is not limited to modern nation-states. Significant
diverse groups such as Arctic Utku Eskimo [Briggs,
levels of bullying have been described in numerous
1970], Amazonian Yanomam o [Chagnon, 1983], and
modern hunter-gatherer [e.g., Briggs, 1970] and his-
African mountain Ik [Turnbull, 1972]. The same
torical civilizations [e.g., Hsiung, 2005]. The ubiquity
could be said for bullying in historical civilizations.
of bullying is matched by the frequency in which bul-
lying is associated with a wide range of serious neg-
ative outcomes. While bullies may suffer from some
Correspondence to: Tony Volk, PhD, 500 Glenridge Avenue, St.
negative consequences [Volk et al., 2006], this is par-
Catharines, ON, Canada L2S 3A1. E-mail: tvolk@brocku.ca
ticularly true for victims, who suffer from higher levels
Received 1 June 2011; Accepted 16 November 2011
of depression, loneliness, generalized and social anxi-
Published online in Wiley Online Library (wileyonlinelibrary.com).
ety, and lower self-worth [Hawker and Boulton, 2000].
DOI: 10.1002/ab.21418
C 2012 Wiley Periodicals, Inc.

---

2
Volk et al.
Evidence for bullying can be seen in historical records
trait or behavior, tied at least probabilistically to (if
from cultures as diverse as ancient Greece [Golden,
not the direct result of) genes, that increases the likeli-
1990], ancient Rome [Rawson, 2003], medieval China
hood that an organism would successfully reproduce
[Hsiung, 2005], medieval Europe [Orme, 2001], and
and pass on those genes down to future generations
renaissance Europe [Cunningham, 2005]. Bullying in
[Dawkins, 1989]. Thus, for bullying to be an evolved
these historical societies took the form of hazing,
adaptation, it must display two necessary components
physical and verbal bullying, as well as indirect social
[Williams, 1966]. First, it must solve adolescent prob-
bullying resulting in the exclusion and isolation of in-
lems ultimately related to reproductive success (in a
dividuals. Interpreted from a traditional point of view,
way that is plausibly linked to past environments).
these data could suggest that bullying is the result of
Second, it must be heritable. That is, it must be reli-
maladaptive development occurring on a staggering
ably related to the specific expression of one or (almost
cross-cultural, environmental, geographical, and his-
certainly) more genes.
torical scale. But perhaps bullying is not solely the
result of maladaptive, problematic development. The
ADAPTIVE NATURE OF BULLYING
prevalence and ubiquity of adolescent bullying sug-
gest an alternate explanation: adolescent bullying may
We therefore begin by examining whether bully-
serve an adaptive purpose for some adolescents.
ing meets the first prerequisite--that it is adaptive
Unfortunately, influential, traditional theories of
toward problems faced in the ancestral environment,
aggression and bullying generally view this behav-
which is where any evolution related to bullying would
ior solely as maladaptive and thus do not address
have occurred [Alcock, 2001]. This Environment of
the possibility that bullying may be adaptive [Berger,
Evolutionary Adaptation (EEA) is not a single time
2007]. For example, traditional theories have been
or place, but rather is the combination of environ-
shaped by theory and research suggesting that ag-
ments in which adolescents evolved [Irons, 1998].
gressive behavior is a function of imitating aggressive
While we cannot make definitive observations of these
role models [e.g., Bandura, 1978] and/or due to de-
past environments, we can infer some basic prop-
ficient social skills and social information processing
erties of the EEA. We believe that adolescents had
[Crick and Dodge, 1994, 1999]. However, researchers
two important goals from an evolutionary perspec-
have begun to question the assumption that bullying
tive: growth/health/survival and securing appropri-
is a consequence of impairments in social skills, in
ate mating opportunities [Konner, 2010]. The ultimate
part because theory-of-mind research has shown that
biological/evolutionary meaning of life is passing on
children who bully do not seem to lack social under-
genes [Dawkins, 1989]. This requires organisms to sur-
standing [Shakoor et al., 2011; Sutton et al., 1999].
vive long enough to successfully reproduce. Lacking
Furthermore, bullying does not appear to be strongly
the ability to directly test the goals of survival and
associated with other environmental factors that of-
reproduction in the EEA, we instead rely on data
ten engender maladaptation, including cultural vari-
from modern bullying supplemented by data from
ables such as low socioeconomic status [Berger, 2007]
hunter-gatherers and historical records to determine
and poor parent-child relationships [Veenstra et al.,
the adaptiveness of bullying [Marlowe, 2010; Volk and
2005; Volk et al., 2006]. This leaves an unsatisfac-
Atkinson, 2008]. It should be noted that adaptations
tory gap between theory and empirical data, insofar
do not require that the individual consciously and
as research findings do not adequately illuminate the
efficiently pursues the goal in question [Geary and
dysfunction underlying bullying, or delineate the de-
Bjorklund, 2000]. Instead, evolutionary theory dic-
velopmental pathways leading to its emergence. For-
tates that she must simply behave in a way that does,
tunately, evolutionary psychology may offer a more
on average, improve her reproductive fitness, whether
parsimonious explanation of the data. Specifically,
or not she is consciously and/or efficiently address-
evolutionary psychology offers the hypothesis that
ing that problem [Daly and Wilson, 1988; Ellis and
some adolescent bullying may not be maladaptive.
Bjorklund, 2005].
Instead, it may be the result of an evolved adaptation
That said, we would like to address three impor-
toward bullying [Kolbert and Crothers, 2003; Volk
tant semantic issues related to the evidence for adap-
et al., in press].
tiveness. First, we would like to note that bullying is
Evolutionary developmental psychology suggests
not a homogenous behavior, and that different forms
that some behaviors develop because they are at least
of bullying are likely to have arisen in response to
in part due to evolved mental adaptations that are a re-
different selection pressures. Volk et al. [2006] iden-
sponse to past evolutionary pressures [Bjorklund and
tified five forms of bullying, including racial/ethnic,
Pellegrini, 2000]. An evolved adaptation is a physical
verbal, physical, sexual, and indirect/social, each of
Aggr. Behav.

---

Is Adolescent Bullying an Evolutionary Adaptation
3
which may serve a particular function. New forms
level of social groups will be discussed separately in
of bullying such as cyberbullying [Raskauskas and
the third section on adaptive functions related to dom-
Stoltz, 2007; Wade and Beran, 2011] further compli-
inance.
cate the matter. While we explicitly explain specific
Individual-Based Somatic Functions of
forms, functions, and evolutionary origins of some
Bullying
bullying subtypes later in the current manuscript (e.g.,
sexual bullying), a lack of data on specific forms of
A key potential function of bullying is whether it
bullying means that we often refer to bullying in gen-
is associated with positive outcomes for growth and
eral. While these generalizations remain true at the
survival (i.e., somatic resources). These positive out-
aggregate level, it is important to note that they may
comes could be the result of bullying itself and/or
not always hold for all the individual forms and func-
they could be inherent characteristics of bullies that
tions of bullying.
are correlated with, but not necessarily caused by,
Second, we would like to draw attention to the fact
bullying. Somatic resources do not directly transmit
that bullying is not the same thing as general aggres-
more genes into future generations, but larger, health-
sion. All bullying is aggression, but not all aggres-
ier, longer-lived individuals are likely to obtain better
sion is bullying [Hawley et al., 2010]. For example,
reproductive success [Allal et al., 2002; Archer and
the establishment of dominance hierarchies among
Thanzami, 2009; Gallup et al., 2007]. There is no di-
many mammals involve brief contests to determine a
rect evidence that bullying causes greater or faster
hierarchy that results in low levels of future aggres-
physical growth, but there is evidence that bullies (par-
sion other than between evenly-matched aggressors
ticularly male bullies) tend to be larger and stronger
[Alcock, 1988, pp. 402-406]. The lack of serious harm
than non-bullies [Gallup et al., 2007; Olweus, 1993].
and repetition means these incidents do not qualify
Hunter-gatherers (e.g., Aboriginal Australians,
as bullying. Nor do more serious and harmful domi-
Hadza, !Kung San) may offer a more salient test of
nance conflicts between two evenly-matched individu-
the somatic benefits of bullying as they face greater
als, as there is a lack of clear power imbalances. Thus,
survival obstacles compared to adolescents in devel-
when we refer to evidence based on general aggression
oped nations [Lee and Daly, 1999; Volk and Atkin-
readers should consider it to be less conclusive than
son, 2008]. Mortality rates are high among youth,
evidence based specifically on bullying.
approaching an average of 50% mortality to age 15 as
Third, bullying done by "pure" bullies may differ in
compared to 1% among youth in developed nations
important ways from bullying done by bully-victims
[Volk and Atkinson, 2008]. We suggest that if bullying
who are both bullies and victims. Thus, unless oth-
is related to somatic benefits, bullying should be more
erwise stated, we use "bullying" to refer to bullying
valuable (and thus prominent) when resources are
done only by "pure" bullies. In particular, as we dis-
scarce. There is indeed evidence of this among hunter-
cuss later in the current manuscript, there is reason
gatherers. The Ik were displaced hunter-gatherers
to believe that the bullying conducted by individuals
who experienced extreme resource deprivation, and
who are both bullies as well as victims of bullying
whose adolescents were highly involved in bullying
has more of a reactive function and may be a prod-
over life-and-death physical resources such as obtain-
uct of dysregulation, and therefore may be less objec-
ing food from others vs. starving to death [Turnbull,
tively adaptive than the bullying performed by "pure"
1972]. Within the brief time that they had experienced
bullies.
this deprivation, the youngest generation participated
With those caveats in mind, we begin by exploring
more actively in bullying than their parents, who in
two functions of adolescent bullying directly related
turn were more involved than the grandparents, who
to evolutionary survival and success: health/survival
fondly recalled past times of plenty and group cohe-
and sex. Both of these functions operate on the dyadic
sion [Turnbull, 1972]. In a similar vein, an incident
level of two individuals (i.e., they can operate exclu-
of food scarcity prompted greater levels of indirect
sively between a single bully and a single victim). They
bullying between families of Utku Eskimo [Briggs,
are thus independent of the social group, unlike other
1970]. Thus bullying for "lunch money," or more di-
forms of bullying that depend upon the actions of
rectly, lunch itself, may be an important function of
other individuals (e.g., bystanders) or the gaining of
bullying.
social dominance/status to obtain their effects [Kol-
Another important ancestral factor for bullying is
bert and Crothers, 2003]. We refer to this latter form
competition not for material resources, but for the
of bullying as occurring at the level of social groups
jobs that lead to those material resources. While these
rather than dyads. The benefits for an individual's
kinds of opportunities were sharply limited in egali-
health/survival and sex resulting from bullying at the
tarian hunter-gatherer societies [Lee and Daly, 1999],
Aggr. Behav.

---

4
Volk et al.
they became very important as civilization introduced
suggests that adolescents involved in bullying have at
increasingly specific and competitive job markets.
least equal and in some cases better physical and men-
Bullies in historical societies did indeed often bully
tal health than victims and uninvolved youth. Fur-
and compete for limited, zero-sum jobs or appren-
thermore, it seems that these positive outcomes may
ticeships [Cunningham, 2005; Golden, 1990; Hsiung,
be partly attributable to bullying providing greater
2005; Orme, 2001; Rawson, 2003]. The winners of
access to somatic (especially in hunter-gatherer so-
these contests would gain the necessary resources to
cieties where there is food scarcity) and material re-
promote their survival and reproductive success. This
sources, affording protection from aggressive attacks,
kind of bullying is still seen in modern society, par-
and, more generally, resulting in lower levels of stress.
ticularly in lucrative professions that possess limited
Consequently, we believe the evidence supports the
enrolment, such as law schools [Flanagan, 2007].
adaptive functioning of bullying in potentially en-
Besides obtaining material resources, bullies may
hancing prospects for health and survival, which in
also acquire physical protection for themselves and
turn increases, opportunities for current and future
their resources by building a tough reputation, in-
reproductive success.
creasing their position in a dominance hierarchy,
Individual-Based Sexual Benefits of Bullying
and/or gaining allies, all of which would prevent
further conflicts and minimize certain costs, such as
A second key potential adaptive function of bully-
physical injuries or loss of tangible resources. Ado-
ing at the individual level is increasing opportunities
lescent boys who are perceived as being tough also
for sex. The survival of the fittest is a biological term
enjoy protection from aggression as their reputation
that refers not to survival of the strongest, fastest, or
makes other boys less likely to physically aggress
healthiest, but of the best reproducer [Alcock, 2001;
against them even if provoked by insult [Archer and
Dawkins, 1989]. We therefore consider the evidence
Benson, 2008]. These benefits may extend to their gen-
that bullying is associated with increased reproduc-
eral health too. There appears to be a link between
tion, or at least mating opportunities, to be a key test
stress (as measured by hormonal cortisol levels),
of our hypothesis that bullying is an adaptation. The
health, and social dominance, suggesting general
first line of evidence for this adaptive function is that
health benefits to those who win social competitions
bullies may in fact engage in more sex (as evidenced
such as those involved in bullying [Flinn, 2006; van
by increased dating/mating). Supporting the sexual
Honk et al., 2004]. Individuals at the top of a dom-
adaptiveness of bullying for both sexes are data show-
inance hierarchy and/or those with sufficient physi-
ing that bullies of both sexes appear to enter puberty
cal resources (e.g., socioeconomic status) may expe-
and start dating at a younger age, are more active
rience less stress, and thus fewer health issues associ-
with members of the opposite sex, report greater dat-
ated with chronic stress as compared to those without
ing/mating opportunities, and are more likely to be
dominance and/or physical resources [Newman et al.,
in a dating relationship [Connolly et al., 2000].
2005; Sapolsky, 2004]. This is consistent with findings
How does this occur? One reason may be that
that show bullies tend to have some health benefits
bullies display traits that are evolutionarily attrac-
such as less frequent sore throats, coughs, colds, and
tive to members of the opposite sex. While there
breathing problems than victims, and bully-victims,
are similar long-term mate preferences for both sexes
and no more total health problems than neutral, un-
[e.g., kindness, social skills, intelligence; Buss 1988a,
involved children [Fekkes et al., 2004; Wolke et al.,
1988b], there are also important differences [Archer
2001]. This evidence regarding health and survival is
and Thanzami, 2009; Geary, 2010; Moller and Alat-
contrary to the prediction of standard theories of bul-
alo, 1999]. For boys, this means exhibiting primary
lying as a maladaptive response to poor environments
traits such as physical strength, dominance, material
[Rigby, 2003]. Furthermore, several studies note that
resources as well as secondary traits such as physi-
pure bullies (excluding bully-victims) report equal or
cal attractiveness. These evolutionarily relevant traits
better mental health than uninvolved adolescents and
advertise a boy's future ability to provide and protect
victims [Berger, 2007; Ireland, 2005; Juvonen et al.,
for a mate, as well as to provide her with good genes.
2003; Volk et al., 2006; Wolke, et al., 2001]. Bully-
For a girl, this means displaying traits of attractive-
ing is also positively linked with other positive mental
ness and resources that signal her future fertility and
traits such as theory of mind ability, cognitive empa-
potential to care for his children.
thy, leadership, social competence, and self-efficacy
These male and female traits are in line with the
[Caravati et al., 2009, 2010; Vaillancourt et al., 2003].
evolutionarily-derived hypothesis that boys and men
So although the directionality of some of the rela-
are more tolerant toward risk due to greater aver-
tions cannot be determined, the evidence as a whole
age female investment in offspring combined with a
Aggr. Behav.

---

Is Adolescent Bullying an Evolutionary Adaptation
5
greater variance in male reproductive success [Daly
friends with perpetrators [Cunningham et al., 2010].
and Wilson, 1988; Geary, 2010; Hrdy, 1999; Trivers,
Similarly, female adolescents who rated themselves as
1972]. While the average number of children is obvi-
attractive had higher odds of being victimized rela-
ously equal between the sexes, men are much more
tionally through the spreading of rumors or social
likely to have either no children, or many more
exclusion [Leenaars et al., 2008]. These lines of evi-
children than average, with dominant/powerful men
dence suggest a plausible connection between sexual
having many more than subordinate/powerless men
or relational bullying, physical attractiveness and op-
[Daly and Wilson, 1988]. For example, 69 is the
portunities for sex. Attractive victims of sexual bul-
highest number of children born to a single (unfor-
lying, which includes making sexual jokes, comments
tunate) woman, wife of Feodor Vassilyev (b.1707-
or gestures [Volk et al., 2006], may be targeted by
c.1782) who was a peasant from Shuya, Russia. In 27
bullies trying to signal their sexual interest, and thus
pregnancies she gave birth to 16 pairs of twins, seven
to increase sexual opportunities. Consistent with this
sets of triplets, and four sets of quadruplets. In con-
idea, targets of sexual bullying are more likely to be
trast, the most prolific father of all time is believed to
reproductively viable [i.e., early puberty vs. younger
be the (highly dominant) last Sharifian Emperor of
children; Craig et al., 2001].
Morocco, Mulai Ismail (1646-1727). In 1703, he had
In a related vein, both sexes may use indirect inter-
at least 342 daughters and 525 sons and by 1721 he was
sexual bullying, particularly "pushing and poking"
reputed to have 700 male descendents. Yet even this
harassing courtship behaviors, as public or private
impressive number is far lower than the total number
efforts intended to gauge a potential partner's sexual
of children believed to be sired by Genghis Khan and
interest at the same time as minimizing the costs of re-
his descendents. This family dynasty, which included
jection (humiliation) or retaliation [from the target or
some of histories' most dominant and violent men, in
their friends/family; Pellegrini, 2001]. For example,
the world, ruled an empire with a larger territory than
this might entail making unwanted but noncommit-
any other before or since. Yet they were not only so-
tal jokes or comments about the attractiveness of, or
cially dominant, they also translated that dominance
desire to date, a particular member of the opposite
to the global gene pool as they are believed to be
sex. In contrast, relational bullying such as spreading
the paternal ancestors of at least 8% of Asian men,
rumors or excluding individuals from social groups
or 0.5% of the world's population, which is roughly
is likely used to denigrate characteristics of attractive
35 million people [Zerjal et al., 2003]! Thus, like vir-
opposite-sex adolescents who may be seen as rivals
tually all mammals, human men are literally phys-
in intrasexual competition for mates, as is discussed
ically and mentally built toward accepting riskier
in greater detail below in regard to the group-based
competition because the relative gains and losses are
benefits of bullying [e.g., Benenson, 2009].
much higher, justifying the increased willingness to
Additional research indicates that another impor-
engage in risky competition [Geary, 2010]. Just as the
tant individual-based function of sexual bullying may
physical adaptations of increasing muscle mass and
be to coerce members of the opposite sex to estab-
testosterone levels appear to prepare boys for suc-
lish dating/mating opportunities, [Cunningham et al.,
cess at intraspecific competition [Geary, 2010], so may
2010; Pellegrini and Long, 2003]. Given that adoles-
their heightened propensity for bullying [Kolbert and
cence is a period of sexual maturation, it is not sur-
Crothers, 2003].
prising that the number of adolescents involved in
As previously mentioned, male bullies appear to
romantic relationships increases with age and sexual
generally display the primary traits of strength, dom-
maturity [Collins, 2003]. Coercive sexual behaviors,
inance, and material resources [Gallup et al., 2007;
including bullying, typically emerge during this age
Olweus, 1993; Turnbull, 1972]. Boys who bullied other
period as well [McMaster et al., 2002]. Furthermore,
boys were also more accepted by girls [Veenstra et al.,
sexual harassment is correlated with the onset of pu-
2010], suggesting that the traits they display are in-
berty for cross-sex harassment, but not same-sex ha-
deed attractive to the opposite sex. Given that physi-
rassment [McMaster et al., 2002; Pepler et al., 2006],
cally attractive individuals are seen as desirable mates,
underlining its reproductive correlation.
further evidence that bullying may be a tool used to in-
Evolutionary theory predicts that due to the lower
crease sexual opportunities would come from the data
potential costs of boys' reproduction, and the greater
showing that attractive individuals are more likely to
costs incurred by girls, boys should generally be more
be targeted by bullies. In fact, adolescents of both
aggressive and less selective in choosing mates [Geary,
sexes who rated themselves as highly attractive were
2010; Trivers, 1972]. In agreement with this pro-
not only more likely to be victims of sexual bully-
posed evolutionary reproductive strategy of boys, it
ing, but also to have been perpetrators, observers, or
is adolescent boys who are the most common
Aggr. Behav.

---

6
Volk et al.
perpetrators
of
sexually
coer-
increased opportunities for dating, mating, and mate-
cive
or
harassing
intersexual
control. These outcomes are again not predicted by
behaviors [McMaster et al., 2002; Pepler et al., 2006].
traditional models emphasizing the maladaptive ori-
While there is no doubt that such tactics can be offen-
gins of bullying.
sive toward their target, it is clear that similar forms of
Group-Based Benefits of Bullying: Dominance
sexual aggression can lead to sexual access in adults
[Lalumiere et al., 2005; Walker, 1997]. Although not
A third function of bullying is that it may be used
well-studied among adolescents, we presume this
to obtain benefits in both somatic and sexual do-
is also the case for adolescent intersexual bullying.
mains through attainment of dominance and/or so-
Cross-cultural research suggests that there exists
cial status [Kolbert and Crothers, 2003]. Bullying for
similar sexual coercion among traditional cultures
dominance and status (we use the terms interchange-
such as Samoa or the Yanomam o [Chagnon, 1983;
ably) is really bullying for social resources that can
Nardi, 1984].
in turn be translated into current or future adaptive
Intersexual bullying may also be used to maintain
benefits in somatic, sexual, and/or parental domains.
relationships by operating as a form of jealous con-
Specifically, the peer relationships of higher-status in-
trol. Connolly et al. [2000] note that bullies' aggressive
dividuals (especially when grouped with other high-
tendencies tended to carry over into their romantic
status individuals) provide social resources that give
relationships, as did Pellegrini [2001]. There is pre-
them the power to influence, persuade, or compel oth-
liminary evidence that general sexual aggression in
ers, and to enhance their own access to resources.
relationships is preceded more often by cues of in-
In large part then, dominance achieved by bullying
fidelity than general physical aggression [Camilleri
would serve the same functions as individual-level
and Quinsey, 2009]. Land [2003] reported that vi-
bullying by securing somatic resources and mating op-
gnettes of sexual bullying among adolescent included
portunities. So are bullies in fact more dominant than
themes of jealousy. Among adolescent females, inter-
non-bullies?
sexual bullying of partners appears to be a way of
Dominance has been found to be positively asso-
obtaining control over one's partner [Capaldi et al.,
ciated with both bullying and peer nominations of
2004]. For these girls, aggression and bullying toward
dating popularity among adolescents [Pellegrini and
their partner help ensure both their partner's fidelity,
Long, 2003]. Bullying is also positively correlated with
as well as their continued provisioning of resources
peer nominations of power [Vaillancourt et al., 2003],
[Capaldi et al., 2004]--a key evolutionary consider-
social prominence [Zimmer-Gembeck et al., 2005],
ation for future mothers [Geary, 2010; Hrdy, 1999].
student and teacher ratings of perceived popularity
Young women's intersexual aggression and bullying
[de Bruyn et al., 2010; Estell et al., 2007; Juvonen et al.,
differs from men's in that their aggression is typically
2003; Rodkin and Berger, 2008], and peer leadership
of a lower intensity [reflecting women's lower toler-
[Estell et al., 2007], all of which relate to dominance
ance for risk, Geary, 2010; Hrdy, 1999, 2009], and
and control over resources. Indeed, Hawley [1999] has
it occurs more frequently in the context of mutual
posited that aggression in general is frequently used
partner aggression [perhaps reflecting a suboptimal
as a means to control resources [i.e., Resource Con-
situation; Capaldi et al., 2007]. Thus, intersexual bul-
trol Theory; Hawley, 2007]. While not tested directly,
lying appears to be used as a mechanism to increase
we believe it is likely that adolescent bullies are sim-
a partner's fidelity, which in turn would have ben-
ilarly able to utilize social dominance to obtain and
efits with regard to enhancing reproductive fitness.
control valued resources. It is clear that adolescents
Male intersexual bullying in response to the threat
are very sensitive to issues of dominance and social
of a female partner's infidelity would increase the
power as witnessed by bullies targeting only those
probability of the male investing parental resources
who were not protected by people who could harm
in protecting and providing for his offspring (as op-
the bully [Veenstra et al., 2010]. However, as in sex-
posed to unknowingly raising the child of another
ual bullying, there again are important evolutionary
man), and thus ensuring that his child survives to
sex differences between boys' and girls' motives and
pass on her genes. On the other hand, female inter-
methods for dominance-oriented bullying. We start
sexual bullying directed toward an unfaithful male
by examining girls' bullying for dominance and social
partner would heighten the chance of the offspring
power.
being protected and provisioned by the father, thus
Girls who bully not only show a greater desire for
boosting the likelihood that the child survives to pass
male acceptance, they also show higher levels of male
on his genes. In summary, adolescent bullying appears
acceptance [Dijkstra et al., 2008]. However, this in-
to be often related to sexual attractiveness and offers
trasexual competition comes at a cost as they are less
Aggr. Behav.

---

Is Adolescent Bullying an Evolutionary Adaptation
7
positively accepted by other girls with whom they are
to exert social control, as powerful individuals, and
presumably competing [Olthoff and Goosens, 2008].
as members of popular groups. Consistent with this
Evolutionary theory suggests that girls should have
contention, adolescents with high levels of perceived
a lower tolerance toward risk given the less variable
popularity, social preference, and social prominence
reproductive outcomes (see further discussion below)
have been found to be more likely to engage in future
and their greater likelihood in assuming a future role
relational aggression [Cillessen and Mayeux, 2004;
as the primary caregiver to a child [Geary, 2010]. Con-
Zimmer-Gembeck et al., 2005]. Thus, for adolescent
sistent with this theory, and with data on girls' general
girls, dominance plays an important role in obtain-
aggression and risk tolerance [Archer, 2009; Vaillan-
ing and controlling access to boys [Benenson, 2009;
court, 2005], adolescent girls' bullying for dominance
Geary, 2010; Hrdy, 1999], demonstrating that indirect
relies more on verbal and indirect social bullying than
bullying for female dominance can be a valuable tool
on riskier direct and physical bullying [Volk et al.,
for girls.
2006]. As employed by boys and girls, indirect bully-
Direct bullying may be particularly salient for boys,
ing is particularly difficult for adults to observe [Craig
who more than girls, are believed to channel much of
and Pepler, 1998]. From an evolutionary perspective,
their aggression and bullying toward members out-
adolescents' sophisticated Theory of Mind [Wellman,
side of one's friendship group [Maccoby, 2004]. This
1992] would have allowed them to engage in covert,
includes physical bullying, which is likely to carry
indirect bullying [Caravita et al., 2010] that offered
higher costs of general physical aggression such as re-
many of the same benefits as overt, direct bullying
taliation, punishment, and likelihood of being caught
while reducing the likelihood of adult punishments
[Rivers and Smith, 1994]. As previously mentioned,
and victim retaliation. Even among the most out-
stronger adolescent boys are more likely to engage in
wardly peaceful cultures, adult hunter-gatherers ap-
bullying [Gallup et al., 2007], as are more aggressive
pear to employ gossip and social exclusion as effective
adolescent boys [Bollmer et al., 2006]. Both of these
aggressive or bullying strategies [Briggs, 1970; Ingold,
factors may enhance the appeal of the risky physi-
2004; Lee, 1979; Marlowe, 2010; Thomas, 1989; Turn-
cal bullying, leading to the higher levels of physical
bull, 1972]. Thus, while it may be less effective in ob-
bullying used by boys [Volk et al., 2006].
taining immediate short-term goals due to the lack of
Besides the previously discussed individual reputa-
immediate leverage over one's target, indirect bully-
tions for toughness, bullying may generate a pow-
ing likely evolved as an effective long-term tactic that
erful incentive for boys to be part of the bullying
allowed adolescents (particularly girls) to apply the
in-group [Sherif and Sherif, 1970] in order to avoid
(albeit diminished compared to direct) social power
being members of an out-group who are more likely
of bullying over long periods of time without incur-
to be targeted [Cillessen and Mayeaux, 2007]. As fa-
ring negative sanctions in return.
mously demonstrated in the Stanford Prison Exper-
With regards to the content of their bullying, much
iment, the creation of a scape-goat(s) can generate
of girls' intrasexual bullying appears to consist of in-
powerful forces that amplify both in-group cohesive-
sults that emphasize other girls' sexual promiscuity or
ness and out-group aggression [Zimbardo, 2008]. Bul-
perceived flaws in physical appearance [Shute et al.,
lying someone from an out-group who is not a friend
2008]. As discussed above, these are the traits that are
and who you do not need to rely upon may further al-
believed to hold strong evolutionary value for men
low boys to engage in costlier, more damaging, forms
who have evolved general preferences for sexual fi-
of bullying. Thus, bullying for adolescent boys may be
delity and attractiveness in women desired as long-
a means of increasing in-group power and cohesive-
term mates [Buss, 1988a, 1988b]. So when these traits
ness. Research has shown that boys are in fact more
are attacked by bullies publicly, these insults appear
likely than girls to engage in solidarity in the face of
to be a way of reducing the appeal of a competitor to
conflict [Benenson, 2009], highlighting both the po-
possible mates, thereby raising the bully's relative ap-
tential importance of coalitions among boys and their
peal and status [Timmerman, 2003]. Adolescent girls
willingness to use riskier forms of bullying to cement
can also use indirect bullying tactics such as social
those coalitions [Archer, 2009; Geary, 2010]. This may
exclusion or rumors to compete over potential sexual
have been particularly important from an evolution-
partners by not only damaging others' reputations,
ary perspective if being a victim of severe physical
but by attempting to socially limit competitors' access
bullying lead to severe injury or death among boys.
to potential partners [Benenson, 2009; Owens et al.,
While anthropologists have not directly measured
2000]. Having high social status is likely to enable
bullying among pre-industrialized adolescents, many
adolescent girls to bully more effectively using indi-
cultural ethnographies often refer to highly aggres-
rect or relational means, as it puts them in a position
sive, bullying boys, in warrior agriculturalist cultures
Aggr. Behav.

---

8
Volk et al.
such as the Ache [Hill and Hurtado, 1996], Huron
than the potential risks among boys as compared to
[Trigger, 1969], New Guinea Highlanders [Heider,
girls.
1991], and Yanomamo [Changon, 1983]. In contrast
to modern hunter-gatherers, in these cultures, men
GENETIC BASIS OF BULLYING
actively band together to protect their women and
territory from raids by bands of neighboring men
Having examined the adaptiveness of bullying, we
[similar to chimpanzees; Goodall, 1986]. This lifestyle
turn to the second component of evolved adaptations.
places a premium on male aggression, cooperation,
Recall that evolutionary adaptations require a genetic
and status [Chagnon, 1983]. In particular, the greater
linkage that allows natural or sexual selection to alter
reproductive variance of men in more stratified
the ratio of genes in future generations based upon the
societies increases the potential benefits of bullying
effects of the adaptation [Dawkins, 1989; Williams,
for dominance/status [Daly and Wilson, 1988], while
1966]. Without this linkage, evolution is unable to se-
at the same time greater social acceptance of vio-
lect for adaptive behaviors. Thus, it is a necessary, but
lence lowers the social costs of bullying, particular
not sufficient, prerequisite for bullying to have some
toward members of an out-group. The fact that an
kind of genetic linkage that would have allowed evolu-
estimated 90% of "pre-civilized" human cultures en-
tion to act upon it as an adaptation. In humans, com-
gaged in war/raiding at least twice per year [Keeley,
plex behaviors are almost certainly polygenic [Hern-
1996] suggests that historically there have been ample
derson, 1982]. This means that their expression de-
opportunities for male adolescent bullying to serve
pends upon the effects of multiple genes. As such, we
an adaptive function in protecting and promoting the
do not expect there to be one, or even only a few, genes
welfare, power, and reproductive success of its practi-
that control bullying. Instead, we expect bullying to
tioners. Nor did the potential evolutionary utility of
be related to a host of genes, including those that are
bullying and violence end with the dawn of civiliza-
already known to control for a wide range of devel-
tion, as witnessed by the bloody politics of Ancient
opmental factors including: temperament [Goldsmith
Greece [Golden, 1990], Rome [Rawson, 2003], and
et al., 1999], personality [Eysenck, 2006], and general
China [Hsiung, 2005], as well as the previous exam-
aggression [DiLalla, 2002]. This is an important point,
ples of Mulai Ismail and the Khan family.
because it means that the developmental factors be-
What is more, members of these coalitions may be
lieved to increase the likelihood of bullying [e.g., per-
desired as companions not only by other boys who
sonality; Book et al., in press] were already linked to
want to be part of the "in-group," but also by girls who
genes. Thus, an adaptation for bullying would not re-
are attracted by the member's ability, as a high-status,
quire the more complicated and statistically unlikely
powerful individual, and as a member of a cohesive
event of evolving an entirely new suite of traits. In-
and high-status group, to protect and/or provision
stead, to the extent that bullying increases success in
them and any future offspring [Buss, 1988a, 1988b).
natural or sexual selection, the frequency of this be-
Consistent with this contention, dominant male ado-
havior in the population may increase as a function of
lescents have a higher level of dating popularity [Pel-
a heightened prevalence of the genes that predispose
legrini and Long, 2003] and perceived popularity was
individuals to greater involvement in bullying. Links
associated with increased adolescent sexual activity at
in the literature between bullying and aspects of per-
a 2-year follow-up assessment [Mayeux et al., 2008].
sonality such as lower agreeableness and higher levels
Bullying boys who achieve dominance and/or per-
of aggression [Bollmer et al., 2006; Olweus, 1994; Tani
ceived popularity may therefore enjoy increased mat-
et al., 2003] provide a plausible mechanism by which
ing success as a result. Boys' bullying of boys was
this could occur. Genetic-based individual differences
in fact specifically related to greater acceptance by
in temperament may also play a role, because research
girls in grades 5 to 8 [Veenstra et al., 2010]. Thus,
has shown that bullies exhibit a higher level of nega-
besides promoting in-group solidarity, boys' bully-
tive emotionality than do uninvolved peers, reacting
ing for dominance and popularity facilitates intrasex-
to stressful situations or provocations with stronger
ual competition for dating and mating opportunities,
emotions [Pellegrini and Bartini, 2000]. Furthermore,
which provides a powerful social incentive for male
their ability to inhibit impulses toward aggressive be-
adolescents to engage in riskier, more severe, overt
havior is limited by deficits in behavioral regulation
forms of bullying than girls, thereby explaining one
[Marini et al., 2006], and a relative lack of fearful-
of the most common findings in the bullying litera-
ness [Bacchini et al., 2008]. Thus, the existing genetic
ture. Boys bully more often, more directly, and more
links with these individual traits mean that the evo-
intensely than girls [Berger, 2007; Olweus, 1993; Volk
lution of bullying did not necessarily require the evo-
et al., 2006] because the potential rewards are greater
lution of new mental adaptations. Instead, it could
Aggr. Behav.

---

Is Adolescent Bullying an Evolutionary Adaptation
9
have required the much simpler and more probable
Female alpha wolves, African wild dogs, and banded
event of modifying the existing genotypes for temper-
mongooses will employ bullying in an effort to pre-
ament, personality, and/or aggression, to magnify the
vent subordinate females from mating and producing
probability that bullying occurs [a process known as
pups in order to commit the pack's resources solely
exaptation; Gould and Verba, 1982].
to the alpha's pups [Bell et al., 2011; Creel and Creel,
Indeed, a recent behavioral genetics study of the
2002; Mech, 1970, Scott, 1991]. Dominant animals of
heritability of bullying provides some evidence that
all these social carnivores will also frequently bully
genetic factors play a role in the development of bul-
the weakest member of the pack, presumably over re-
lying. The investigators found that in a cohort of over
sources, often resulting in its death or its dispersal
1,000 10-year-old twins genetic differences accounted
from the pack [Cant et al., 2010; Creel and Creel,
for 73% of the variation in victimization and 61% of
2002; Mech, 1970]. Dominant female spotted hye-
the variation in bullying, with environmental factors
nas put their abnormally high levels of testosterone
accounting for the rest of the variation [Ball et al.,
to work by bullying weaker females and males from
2008]. This does not mean that 61% of bullying is
group kills [Stewart, 1987]. Dominant hyena mothers
due to genetic factors, or that bullying is primarily
also pass on their dominance status to their offspring,
determined by genetic rather than environmental fac-
who can then bully other youngsters, and sometimes
tors. Rather, we simply take it to mean that there is
even adults, to gain privileged access to resources
sufficient genetic linkage for evolution to have acted
such as food [Stewart, 1987; van Lewick and Goodall,
upon bullying. Without any genetic linkages, evolu-
1978].
tion could not have influenced traits related to bul-
Among primates, similar behaviors are seen in a
lying. Therefore, we only argue that bullying satisfies
wide range of species whereby dominant females gain
the second necessary criterion for an adaptation--it
greater access to food resources and male protec-
is heritable/has genetic links and is thus selectable by
tion, and both privileges get passed on to their off-
natural or sexual selection.
spring [Smith, 2005, pp. 184-187; Wrangham, 1980].
Further evidence for the genetic plausibility of bul-
Dominant male primates often get greater access to
lying as an adaptation comes from viewing the phylo-
mating opportunities either directly by bullying fe-
genetic (large-scale evolutionary) history of the be-
males or indirectly by bullying away male competitors
havior, as repeated phylogenetic adaptations offer
[Hrdy, 1999; Thompson, 2010]. Male chimpanzees
strong evidence for the plausibility of a genetic link-
have been observed to bully, either as individuals or
age [Carroll, 2005]. From a phylogenetic perspective,
as a coalition, subordinate males in order to pre-
bullying appears to have a long evolutionary history
vent them from mating, while female chimps bully
among a wide range of social animals other than hu-
other females over access to food resources [Goodall,
mans. Bullying as a means of obtaining social dom-
1986].
inance is in fact a relatively common social adapta-
Combined, research on genetic linkages to traits
tion in the animal world [Alcock, 1988; Archer, 1988;
associated with bullying and evidence for bullying
Lorenz, 1966]. Indeed, dominance achieved through
among nonhuman animals strongly support the pos-
bullying is often signified by the expression "estab-
sibility of a genetic basis for bullying in humans. Ad-
lishing a pecking order," which is drawn from data
mittedly, this is an area of research that has not been
on chickens showing that they literally rely on re-
thoroughly studied, but the research done to date sup-
peated aggressive pecking to establish social hierar-
ports a human genetic linkage with bullying.
chies [Masure and Allee, 1934]. In other animals,
dominance hierarchies may have evolved to serve an
WHY ARE NOT ALL ADOLESCENTS BULLIES?
anti-bullying function by minimizing aggressive in-
teractions over limited resources by reducing the like-
Clearly, if bullying was purely adaptive and was
lihood that more powerful (i.e., dominant) animals
biased by strong genetic predispositions, one would
will make potentially costly and likely unnecessary
predict that all adolescents should engage in it all the
aggressive challenges against weaker (submissive) an-
time. However, statistics clearly indicate that they do
imals that also pay some of the same costs during
not. A large number of adolescents do not engage
such a fight [Alcock, 1988, p. 239]. This appears to
in bullying [Berger, 2007]. Why? As with many other
be most common among animals that do not social-
adaptations [e.g., an adaptive preference for fat and
ize intensively for long periods of time (e.g., deer),
sugar; Birch, 1992], bullying comes with costs as well
whereas bullying appears to exist among social ani-
as benefits. We emphasize that human evolutionary
mals who maintain intensive social contact for pro-
behavioral strategies are not believed to be fixed, un-
longed periods of time (e.g., many social carnivores).
responsive, genetically pre-determined programs. For
Aggr. Behav.

---

10
Volk et al.
example, given the potential long-term costs of vio-
nation is not incompatible with traditional aggression
lent conflicts over food, the Ik bullying behaviors pre-
theories of social learning [Bandura, 1978] and/or so-
viously discussed may reflect a facultative adaptation
cial information processing [Crick and Dodge, 1994].
that presents itself in response to the difficult envi-
Rather, an evolutionary explanation explains when
ronments and/or during times of drought or famine
and why adolescents should be sensitive to environ-
that are believed to have frequently plagued hunter-
mental cues and/or particular forms of social infor-
gatherers [Lee and Daly, 1999].
mation.
A facultative adaptation is an adaptation that is
An alternative (and complementary) explanation
expressed only under certain environmental circum-
for bullying sometimes not being adaptive is that bul-
stances [Underwood, 1954]. Other adolescent behav-
lying is a heterogeneous behavior [Marini et al., 2010;
iors believed to be facultative adaptations include ear-
Volk et al., 2006], and that the type of bullying in-
lier sexual maturation and behaviors as a result of
volvement likely affects its adaptiveness. Specifically,
father absence [Belsky et al., 1991; Ellis, 2004]. These
an important theoretical and empirical distinction has
sexual behaviors and the heightened levels of bullying
been made between bullies and bully-victims [individ-
(particularly risky physical bullying) may both share
uals who are both bullies and victims, see Marini et al.,
the common cause of adolescents engaging in behav-
2009]. Research has shown that bully-victims are less
iors that discount future investments (because they are
adept at prosocial behavior, less athletic and attrac-
less likely to pay off) in favor of present investments
tive, lower in social competence and self-control, and
[even if they carry future costs as those costs are less
higher in activity level than bullies, victims, or un-
likely to occur; Del Giudice and Belsky, 2010]. So un-
involved children [Estell et al., 2007; Haynie et al.,
der harsh conditions, it may make more sense to invest
2001; Marini et al., 2006; Rodkin et al., 2008]. These
all your resources in whatever option helps you repro-
attributes fit well with both popular stereotypes of
duce now, even if that means sacrificing better long-
bullying as well as the general maladaptive view that
term options. If bullying is a facultative adaptation
bullying is the result of individual and/or environ-
(and we believe it is), this would partly explain why
mental developmental insults rather than an adap-
not all adolescents engage in bullying because they
tive response to one's environment [see Berger, 2007,
do not experience the necessary environmental pre-
for a review]. Bullying performed by bully-victims
cursors to trigger bullying. Instead, bullying should
is not linked to positive outcomes in social domi-
be more prevalent when adolescents receive cues that
nance and intrasexual competition. Like bullies, they
motivate them to embrace their present at the expense
were lower in social preference and peer liking, and
of their future.
higher in peer disliking [Estell et al., 2007; Pellegrini,
Facultative adaptations are designed to be sensitive
Bartini, & Brooks, 1999; Rodkin et al., 2008]. How-
to the costs and benefits of engaging in a particular
ever, in contrast to bullies, bully-victims were ranked
behavior in a particular environment. Thus prospec-
lower than uninvolved participants, bullies, or vic-
tive bullies should consciously or unconsciously weigh
tims, in teacher-rated popularity and peer nomina-
the personal and environmental factors that influence
tions of perceived popularity [Estell et al., 2007; Rod-
their odds of success (e.g., body size, positive peer sup-
kin et al., 2008]. Furthermore, bully-victim status
port, target's weakness, lax supervision) against those
was not correlated with dominance, was associated
factors that influence their odds of failure (e.g., per-
with fewer reciprocated friendship nominations, and
sonal weaknesses, likely adult punishment, negative
these individuals had more peer relationship prob-
peer responses, target's defenses) prior to engaging in
lems than did bullies, victims, or uninvolved peers
bullying behavior. Archer and Southall [2009] have
[Marini et al., 2006; Pellegrini et al., 1999]. Notably,
argued that male prisoners engage in just this sort of
these peer relationship problems include both a lack
cost vs. benefit analysis to determine their bullying be-
of friendships and dating opportunities [Marini et al.,
haviors. Individuals who possess the appropriate tem-
2006].
perament, personality, and requisite physical and/or
Several other lines of research suggest that the bully-
social power would be more likely to positively eval-
ing practiced by bully-victims is less likely than that of
uate their chances of success at bullying. But if envi-
"pure" bullies to be selective with respect to potential
ronmental factors are not conducive to success (e.g.,
targets and to the social context, which would likely
strong adult sanctions, cultural taboos against vio-
reduce its adaptiveness. First, as indicated, bully-
lence, etc.), these predispositions should not lead to
victims are characterized by a lack of social com-
bullying [Volk et al., in press]. This would explain why
petence and self-control, and their bullying is seen as
not all children, even those with the "right" predispos-
a product of emotional dysregulation [Estell et al.,
ing factors, engage in bullying. Note that this expla-
2007; Haynie et al., 2001; Marini, et al., 2006; Rodkin
Aggr. Behav.

---