AGGRESSION

AGGRESSION
a social learning analysis
Albert Bandura
© 1973 Albert Bandura
e-Book 2015 International Psychotherapy

Institute
All Rights Reserved
This e-book contains material protected under

International and Federal Copyright Laws and
Treaties. This e-book is intended for personal
use only. Any unauthorized reprint or use of this
material is prohibited. No part of this book may
be used in any commercial manner without
express permission of the author. Scholarly use
of quotations must have proper attribution to
the published work. This work may not be
deconstructed, reverse engineered or
reproduced in any other format.
Created in the United States of America

to Ginny, Mary, and Carol
www.freepsychotherapybooks.org 4
Table of Contents
preface
Theories of Aggression
LABELING OF AGGRESSION
THEORETICAL ANALYSES OF AGGRESSION
DIFFERENTIAL IMPLICATIONS OF DRIVE AND

SOCIAL LEARNING THEORIES
Origins of Aggression
LEARNING THROUGH MODELING
LEARNING THROUGH PRACTICE
LEARNING OF AGGRESSION UNDER NATURALLY

OCCURRING CONDITIONS
Instigators of Aggression
ESTABLISHMENT OF STIMULUS CONTROL OF
AGGRESSION
MODELING INFLUENCES
AVERSIVE TREATMENT
INSTRUCTIONAL CONTROL OF AGGRESSION
NONRECURRENT AGGRESSORS
Maintaining Conditions
EXTERNAL REINFORCEMENT
VICARIOUS REINFORCEMENT
SELF-REINFORCEMENT
PUNISHMENT CONTROL OF AGGRESSION
REINFORCEMENT INFLUENCES IN COLLECTIVE

AGGRESSION
Modification and Control of Agression

MODELING PRINCIPLES
DIFFERENTIAL REINFORCEMENT
MODIFICATION OF ANTISOCIAL AGGRESSION
CHANGES IN SOCIAL SYSTEMS
REFERENCES
preface
This book is concerned with why man

aggresses. There are several reasons for
addressing this issue, despite the great deal of
attention that has already been devoted to it.
Although aggression pervades our lives, few

concerted efforts have been made to substantiate
its causes or to devise constructive ways of
reducing the level of societal violence. Man’s
technological capacity for massive destruction has
now developed to the point where he can no
longer continue to settle conflicts by destructive
means. In addition to the threats posed to survival,

violence increasingly encroaches on daily human
affairs to impair the quality of life. These
developments make it all the more necessary to
understand the determinants of injurious conduct.
Theories of human behavior have been
undergoing drastic change in the last decade.
Interest in traditional approaches that depict

behavior as either instinctively determined or
impelled by drive forces declined as their
deficiencies became apparent. New perspectives

based on social learning principles have emerged
that substantially advance understanding of
human behavior. In this book I have attempted to
formulate a social learning theory of aggression
sufficiently broad in scope to integrate evidence
on all facets of aggression, whether individual or

collective, personal or institutionally sanctioned.
The aim of this undertaking is not only to offer a
better basis for explaining, predicting, and

modifying aggression, but also to provide impetus
for new lines of research likely to augment the
explanatory power of social learning theory.
The worth of a psychological theory should be

partly measured in terms of the effectiveness of
the modification procedures that it produces. A
sizable portion of this book is therefore devoted to
demonstrating how social learning principles can
be applied individually and at the social systems

level to reduce deleterious forms of aggression.
There will always be social conflicts, but they need
not require violent solutions. Some forceful
measures have functional value in promoting

constructive social changes which lessen the
likelihood that people will resort to violent action.
The exercise of social power as an instrument of

change receives detailed consideration.
No account of aggression can be complete
without discussing the social labeling and ethics of
aggressive action. Findings bearing on these
issues, along with evidence that aggressive
conduct can be rapidly and drastically altered by

moral sanctions, strongly indicate that the
determinants of human aggression are best sought
in man’s social practices.
Many people have contributed to this book in

various stages of its growth, and I am pleased to
take this opportunity to acknowledge my

indebtedness to them. Fellowship grants from the
National Institute of Mental Health and the John
Simon Guggenheim Foundation greatly aided my
work. This project was begun during my residence
as a Fellow at the Center for Advanced Study in the

Behavioral Sciences. The opportunity to pursue
one’s studies in the company of distinguished

scholars free from intrusive routines is an
experience that will be long treasured and for

which I am deeply grateful to the staff of the
Center. Violence was more than a topic for
intellectual discussion at the Center that year. It
personally touched the lives of the Fellows and the

staff. In the predawn hours on April 24, 1970 ten
studies were gutted by firebombs in a calculated
attempt to burn the Center to the ground. It was
tragic to see a facility devoted to the
enlightenment and betterment of man selected as

a target for destruction.
I am thankful to Ted Rosenthal and Leonard
Berkowitz who took time from their crowded
schedules to comment on preliminary versions of

the manuscript. To Jane Crane, whose diligent
efforts in preparing the manuscript eased my task
considerably, I owe a large debt of gratitude. My

daughter, Mary, helped me along the way with
typing and proofreading. I was most fortunate to
have the able services of Julia Baskett in preparing
the final version of the manuscript for publication.
Writing a book at times takes possession of the
household as well as the author. I am thankful to
my family for their forbearance. It is to them that I

dedicate this work.
Albert Bandura
chapter one
Theories of Aggression
Among the various human activities that are
the subject of attention, none has aroused deeper
concern than man’s aggressiveness. Though
aggression has always been an important social
problem, developments during the past few
decades have fully justified increased concern.
With the progressive growth of instruments of

destruction, simple aggressive acts can produce
widespread disastrous consequences. The hazards

of ill-judged actions have thus become enormously
magnified. Man’s aggressive potential has also
been increased, independently of expanding
destructive accouterments, by changes in the
social conditions of life. When populations were
widely dispersed, the consequences of any given

aggressive act were principally confined to
persons toward whom the behavior was directed.

Under conditions of urbanized life, in which the
welfare of hordes of residents depends upon
smooth functioning of intricate interdependent
systems, destructive behavior that can be easily

performed without requiring elaborate apparatus
instantaneously harms vast numbers of people.
Concern over the adverse consequences of
aggression obscures the fact that such behavior
often has functional value for the user. Indeed,

there is a property unique to aggression that
generally creates conditions fostering its
occurrence. Unlike other social behaviors that

cannot be effective without some reciprocity
acceptable to the participants, aggression does not
require willing responsiveness from others for its
success. One can injure and destroy to self-

advantage regardless of whether the victim likes it
or not. By aggressive behavior, or dominance
through physical and verbal force, individuals can
obtain valued resources, change rules to fit their
own wishes, gain control over and extract

subservience from others, eliminate conditions
that adversely affect their well-being, and remove
barriers that block or delay attainment of desired
goals. Thus, behavior that is punishing for the
victim can, at least on a short-term basis, be

rewarding for the aggressor. Although, as we shall
see later, aggression has many different causes, its

utilitarian value undoubtedly contributes heavily
to the prevalence of such behavior in the
interactions of everyday life.
Over the years a number of theories have been
proposed to explain why people behave
aggressively. This chapter is focused on a
comparison of the tenets of instinct and drive

theories, which historically attributed aggression
to internal aggressive forces, with the explanatory
system based upon social learning theory. The
basic principles of social learning are then applied
in succeeding chapters to the explanation of both

individual and collective aggression. This
formulation differs in several important respects
from the alternative lines of theorizing not only in
the factors considered to be the causes of
aggression, but also in its implications for the

management and control of human aggression.
LABELING OF AGGRESSION
Attempts to define a concept essentially
represent an invitation for a stroll through a
semantic jungle. The journey, however, is
instructive because it reveals important issues
about the phenomena selected for analysis.
Aggression, like most other concepts, has been

characterized in many different ways. Variations
in defining features arise mainly because some
authors describe aggression solely in terms of
attributes of the behavior, while others include
assumptions about the instigators, the emotional

concomitants, or the intent of potentially injurious
actions.
In the theory originally advanced by Dollard,
Doob, Miller, Mowrer, and Sears (1939), which

occupies a prominent historical position,
aggression is defined as “any sequence of
behavior, the goal response to which is the injury

of the person toward whom it is directed.” Most
subsequent theorizing and research have adopted
injurious intent as an essential aspect of
aggression (Berkowitz, 1962; Feshbach, 1970;
Sears, Maccoby and Levin, 1957). A major
limitation of such a definition is that it assumes
that aggression serves only a single purpose,

namely, to inflict injury.
If aggression is restricted to behavior that is
performed solely for the purpose of injuring
others, then a wide range of activities that are
commonly judged as aggressive, including some of

the most violent forms of interpersonal assault,
would be excluded from consideration. Some
writers (Berkowitz, 1965a; Feshbach, 1970) have
attempted to deal with this problem by proposing

different types of aggression. Instrumental
aggression, which is aimed at securing extraneous
rewards other than the victims' suffering, is

distinguished from hostile aggression, the sole aim
of which is presumably to inflict injury on others.
Since hurtful and destructive actions were largely
attributed to aggressive drive forces, so-called
instrumental aggression received only passing
notice. The differentiation generally conveys the
impression that aggressive behavior performed

for rewarding outcomes represents a form of
pseudoaggression relegated to the subsidiary
status of a means to other ends. According to this
valuation, the holocaust in Hiroshima, which was
ordered to force a quick end to war, would

represent a mere instrumental act. So would any
act of war, for that matter.
In point of fact, so-called hostile aggression is
equally instrumental except that the actions are

used to produce injurious outcomes rather than to
gain status, power, resources, or some other types
of results. Whatever its merits, the distinction

reflects differences in desired outcomes, not in
instrumentality. It would therefore be more
accurate to differentiate aggressive actions in
terms of their functional value rather than in
terms of whether or not they are instrumental.
Most aggressive acts serve ends other than solely
to produce injury. Hence, when researchers report

that instrumental aggression is excluded from
their theoretical analysis, it is unclear whether
they are mislabeling what is being studied or
whether, in fact, they are concerned with only a
small aspect of aggression.
As a rule, aggressive acts not only hurt the

victim, but create a variety of results for the
aggressor. It is therefore difficult to determine, in
any given instance, whether the aggressive

behavior is primarily reinforced by signs of injury,
by associated noninjurious outcomes, or by both
sets of consequences. A given action is most apt to

be misjudged as rewarded by the suffering of
others when the purposes served by aggression
are not fully understood and when it produces no
evident material gains. In many social groups, for
example, aggression has powerful status-
conferring value. Under these types of
contingencies homicidal assaults are sometimes

performed mainly to gain the approval and
admiration of peers and to maintain one’s status in
the social hierarchy of a deviant reference group
(Yablonsky, 1962). The way in which threat of loss
of “rep" may compel a person to engage in a

homicidal assault is graphically illustrated in the
following excerpt from an interview with one of
the boys studied by Yablonsky:
Momentarily I started to thinking about it

inside; I have my mind made up I’m not
going to be in no gang. Then I go on inside.
Something comes up, then here comes all my
friends coming to me. Like I said before, I’m
intelligent and so forth. They be coming to
me—then they talk to me about what they
gonna do. Like, “Man, we’ll go out there and
kill this cat.” I say, “Yeah.” They kept on
talkin’. I said, “Man, I just gotta go with you.”
Myself, I don’t want to go, but when they
start talkin’ about what they gonna do, I say,
“So, he isn’t gonna take over my rep. I ain’t
gonna let him be known more than me.” And
I go ahead, (p. vii)
The pull of status rewards rather than the push
of an aggressive drive also appears to account for
the behavior of another youth involved in a gang

killing:
If I would of got the knife, I would have

stabbed him. That would have gave me more
of a build-up. People would have respected
me for what I’ve done and things like that.
They would say, “There goes a cold killer.”
(p. 8)
The operation of similar reinforcement

contingencies is illustrated by a gang that used
attacks upon strangers without provocation as its

main admissions requirement (San Francisco
Chronicle, 1964). Each physical assault, which had

to be observed by a club member to be valid, was
valued at 10 points, a total of 100 points was
required for full-fledged membership.
A large body of research will be cited later to
demonstrate that aggressive behavior declines
when social reactions to it are withheld. Such

evidence further supports the view that much
interpersonal aggression that appears to have no
utilitarian value may be extensively regulated by
the social consequences it produces. Physical
aggression is also often employed to gain control

over other people, to secure material resources,
and to force changes in social practices that
adversely affect one’s everyday life. In other
words, people frequently resort to aggressive acts
because they produce desired results that cannot

as readily be achieved through nonaggressive
means, rather than simply to generate expressions

of suffering. A comprehensive theory of aggression
must therefore account not only for aggressive

actions that are primarily reinforced by
satisfactions derived from hurting others, but also
for much broader classes of aggressive behavior in
which the infliction of suffering is essentially

irrelevant or serves, at best, a secondary purpose.
For purposes of the present discussion,
aggression is defined as behavior that results in
personal injury and in destruction of property. The
injury may be psychological (in the form of

devaluation or degradation) as well as physical.
Although this formulation delimits the
phenomenon in a meaningful way, it should be
made clear that aversive effects cannot serve as

the sole defining characteristic of aggression.
Individuals who hurt others while performing a
socially sanctioned function (for example, dentists

repairing teeth or surgeons making painful
incisions) would not be considered as acting in an
aggressive manner. Nor would bulldozer
operators destroying condemned buildings to
make way for new construction be charged with
committing aggressive acts. Conversely, some
forms of conduct would be judged aggressive even

though no personal injury or property damage
occurred. A person who attempted to hurt another
individual by firing a gun at him or by striking him
with a lethal object, but who happened to miss the
unsuspecting victim, would be judged as behaving

violently. As the preceding examples show,
additional criteria for distinguishing between
accidental and intended injury must be used both
to exclude numerous pain-producing responses
from the category of aggression and to include

others that do not injure anyone because they are
poorly executed.
In social learning theory, aggression is treated
as a complex event including behavior that
produces injurious and destructive effects as well
as social labeling processes. According to this
view, a full explanation of aggression must
consider both injurious behavior and social
judgments that determine which injurious acts are

labeled as aggressive. The behavioral component
has been studied in considerable detail.
Subsequent chapters describe fully the
mechanisms by which potentially injurious and
destructive behaviors are acquired and the

conditions regulating their expression. The
judgmental component, on the other hand, has
received relatively little attention, so that the
factors that lead people to attach aggressive labels
to social behavior are less well understood.
Social Labeling Processes
People come to judge certain actions as

aggressive on the basis of a variety of criteria,
some of which are grounded in the behavior, while
others are extraneous to it and quite subjective.
The characteristics of the behavior itself
undoubtedly exert a strong influence on how it
will be judged by others. As is evident in virtually

all formal definitions of aggression, behavior that
is likely to produce aversive consequences, such as
physical assaults, humiliation, social rebuffs, and
property destruction, is generally designated as
aggressive, quite apart from the actual effects it

may have on recipients.
The intensity of responses also often influences
the labeling of behavior as aggressive. Behaviors
of high magnitude, even relatively acceptable ones,

such as addressing a person loudly, performing
activities that exceed the tolerance levels of
others, and almost any actions that are executed

impetuously, tend to be interpreted as aggressive.
Since the intensity of the response is an important
factor in determining whether or not it will be
painful or damaging, intensity is often a major
distinguishing feature of aggressive responses that
otherwise differ widely in character.
It follows from the conceptualization of

aggression in terms of high-magnitude responses
(Bandura and Walters, 1963) that teaching people
to behave forcefully on a variety of nonaggressive
tasks would, other factors being constant, increase
the likelihood of behavior judged to be aggressive.

In a test of this hypothesis, Walters and Brown
(1963) found that children who were rewarded
for responding vigorously either in hitting an
automated Bobo doll or in depressing a lever later

displayed more physically aggressive behavior in
an interpersonal situation than children trained to
respond in a less intense manner.
Expressions of pain and injury by recipients is
another important determinant of how the
associated behavior will be evaluated. Scuffles
between children are likely to be labeled
aggressive if a member cries or otherwise conveys
the impression that he has been injured, whereas

the same behavior is apt to be viewed as youthful
exuberance if the participants display no pain
reactions. This feature of an aggressive event
reflects the ruggedness and endurance of the

target person rather than properties of the actions.
At the broader social level, coercive behavior of
enforcement agents that is denounced by the
victims is more apt to be viewed as unwarranted
brutality than treatment to which they acquiesce.
The discussion thus far has been concerned
with observable aspects of behavior and its

evident consequences. Whether or not a given
action is construed as aggressive is also strongly

affected by factors extraneous to the behavior
being evaluated. The intentions attributed to the

performer will alter the way his actions are
categorized. If they are judged to be unintended,
then his behavior is not thought to be aggressive.
The same actions would be interpreted as

aggressive, even though no injury was in fact
inflicted, by observers who assumed that the
person wanted to hurt others.
The task of assessing a person's intentions is
no simple matter. Intent is typically inferred from,

among other factors, the social context of the act,
the role status of the perpetrator of the act, and
recent or more remote antecedent conditions.
Injuries produced through physical assaults on a

football field, for example, are less likely to be
ascribed to pernicious intent than identical
wounds caused by similar actions on city streets.

The painful behavior performed by agents in
socially sanctioned roles, as in the case of
physicians, veterinarians, trial lawyers, and
policemen, and the destructive activities of
lumberjacks and house-wreckers, are rarely
attributed to evil purposes. In these instances,
general standards of acceptable conduct are

delineated and grievance procedures are created
to discourage legitimate agents from exhibiting
unwarranted punitiveness. Disputes, however,
often arise over whether hurtful practices reflect
vigorous enactment of agents’ roles or thinly

veiled aggression.
Injurious intentions are also inferred from
knowledge of earlier provocations. A baseball
pitcher who strikes, with a supposedly wild throw,

an opposing pitcher who had “beaned” a star
performer in a preceding inning will be judged by
most fans as exhibiting retaliatory aggression

rather than momentary loss of control. Although
high consensus may be achieved in inferring
hostile intent from an obvious immediate
provocation, agreement in labeling behavior as
aggressive is relatively low if the instigation for
vengeful actions is more subtle, far removed in
time, or generalized from other sources of

provocation. Under these types of conditions,
which predominate in interpersonal interactions,
the alleged causes of injurious actions are difficult
to verify. The validity of causal attribution is
especially debatable when hurtful consequences

are presumed to be intended unconsciously, since
confirmable criteria are even more elusive. In
many instances the attribution of injurious
intentions is, in fact, determined more by liking for
the performer than by information about the

instigators of his actions. People tend to ascribe
evil designs to those whom they dislike and well-

meaning intentions to those they like.
As alluded to in the preceding comments,
characteristics of the labelers may significantly
affect the way in which particular patterns of
behavior are interpreted. It has been repeatedly
shown that individuals tend to ascribe to others
attributes that they themselves are known to

possess (Holmes, 1968). Whether the assumed
similarity represents a form of imputation or
predictions about others made on the basis of
one’s own response tendencies remains unclear.
The empirical findings nevertheless suggest that

the individuals who are strongly disposed to
behave aggressively would be most inclined to
attribute hostile intent to others and to perceive
their actions as aggressive. One might also expect
persons who have a low threshold for aversive

stimulation to classify a wider variety of activities
as aggressive than those whose pain tolerance is

appreciably higher. The manner in which social
training for attending to violent activities may
enhance responsiveness to aggressive cues is

illustrated in a study by Toch and Schulte (1961),
who presented subjects with violent pictures to
one eye and matched neutral pictures to the other
in a binocular rivalry procedure. Advanced police

administration students, who had completed a
three-year course in law enforcement, reported

seeing more violent activities than novices in the
same police training program.
It is evident from informal observation that

people who differ in socioeconomic level, sex,
ethnic background, and in educational and
occupational status often vary in the way they
view the same behavior with respect to its
aggressive qualities. When social learning

experiences associated with these diverse
backgrounds give rise to different normative
standards for judging behavior, it is not
uncommon for activities promoted as suitable
forms of conduct by members of some groups to
be considered antisocially aggressive by people
who adhere to different codes.
An additional source of influence on labeling
behavior is the characteristics of the aggressor. On

the basis of direct experiences and knowledge of
cultural norms, people develop stereotypes of
behaviors that are considered appropriate for
given ages and sexes, and for members of different
religious, ethnic, occupational, and socioeconomic

groups. Evaluation of particular forms of behavior
will vary depending on the normative standard
against which they are compared. Thus, physical
assertiveness is more likely to be defined as

aggression if performed by a female than by a
male because such behavior departs more widely
from common expectations of appropriate female

conduct. Conversely, similar assertiveness by boys
in a delinquent gang would in all probability be
underrated with respect to aggressiveness.
Definitions of aggression generally convey the
impression that they are concerned solely with
factors that reside in the performer, namely, his

behavior and his intentions. It is apparent from
the preceding discussion that social labeling

processes are also a fundamental feature of the
phenomenon. For this reason, aggression is
characterized as injurious and destructive behavior
that is socially defined as aggressive on the basis of

a variety of factors, some of which reside in the
evaluator rather than in the performer. The
characteristics of the performer’s behavior, carry a

great deal of weight in the interpretation, and the
basis for excluding certain injurious behaviors is

commonly accepted so that, in practice, people
generally agree on what forms of behavior should
be designated as aggressive.
Judgmental controversies are most likely to
arise when injurious actions take subtle or
indirect forms. People ordinarily refrain from
direct personal assaults because such obvious
actions carry high risk of retaliation. Rather, they

favor disguised modes of aggression that, being
difficult to interpret or to consider blameworthy,
afford protection against counterattack. For
similar reasons, people often hurt others
indirectly by setting in motion a series of

detrimental events or by fostering certain
environmental conditions that eventually produce
injurious consequences for others. Here the causal
link between a person’s behavior and the harm
experienced by others is remote, circuitous, and

impersonal. When aversive conditions are
collectively promoted, as is typically the case, no

one individual bears responsibility for causing the
injurious effects. Indeed, decision-making systems
are specifically designed so as to diffuse or

obscure responsibility for detrimental actions. To
the victims, however, people who endorse
prejudicial social practices, who oppose reforms
that would alleviate suffering in disadvantaged

members, and who press for military attacks on
foreign populations are acting in a destructive

manner. Thus, an individual who votes for social
practices that have injurious physical and
psychological effects on others is, in his view,
engaging in democratic action, but to the victims

who must endure the harmful consequences, he is
behaving aggressively. Social scientists have
examined direct assaultive behavior in minute

detail, whereas remote circuitous acts, which
produce widespread devastating consequences,

are rarely given consideration.
Disputes over the labeling of aggressive acts
assume special significance in the case of
collective behavior. What one calls it entails more
than semantic issues. Because the way in which
civil disturbances are labeled partly determines
the countermeasures employed, such labeling can
have far-reaching consequences. As Grimshaw

(1970) has noted, civil strife characterized as
lawlessness or insurrection arouses widespread
fears and demands for suppressive controls. On
the other hand, if the same activities are
designated as legitimate protest against

deleterious conditions, people are more disposed
to seek remedies rather than coercive
countercontrols. The established leadership and
their constituencies who come under attack are
understandably quick to apply criminal labels to

collective protest.
Social labeling of forcible tactics is not entirely

an arbitrary matter. Turner (1969) identifies
several factors that are likely to influence public
interpretations of collective disruptions. When
justifiable grievances are combined with
restrained, selective pressure and principled
rejection of violent tactics, protest activities gain
legitimacy that makes them less amenable to

labeling as antisocial outbursts. The balance
between appeal and coercive threats in the
protests affects whether or not the actions of
participants are viewed as transgressive.
Challengers appear more villainous when their

professed aims are radical change in the social
system or its leadership than when they seek to
ameliorate inequitable conditions. Dissident
leaders are prone to exaggerate the magnitude
and gravity of their impending actions, either

through miscalculation or in efforts to gain
compliance by threatening rhetoric with

revolutionary overtones. To the extent that such
alarming self-labeling leads others to view
subsequent coercive behavior as violence rather

than protest, it is more apt to usher in the police
than social reforms.
Social scientists, like other people, are not

immune to the influence of their own ideological
biases in their interpretations of collective
aggression. When social scientists theorized about

the aggression of the disadvantaged, they
characterized group aggression as an impulsive,
emotional, pathological manifestation. More

recently, as social scientists and their close
associates themselves began to participate in
coercive protest and to take aggressive action
against deleterious social practices, a different
perspective of aggression and its determinants

emerged. What appears irrational and
pathological to onlookers or to those attacked
often serves the aggressor as a method of getting
what he wants when other options fail or remain
unavailable. Group aggression, which had long
been regarded in sociological theory as an
irrational “mob” phenomenon, came to be viewed
by some theorists as a highly principled form of
collective bargaining intended to sway an

irrational social system (Currie and Skolnick,
1970). Less partisan approaches treat collective
aggression as multiply determined like other
forms of social conduct, without attributing to
participants either psychic aberrations or selfless

morality (Marx, 1970a). As in designations of
disturbances in terms of crime or protest, people
respond differently depending upon whether
pathological or problem-solving labels are
attached to collective acts of aggression.
A closely related issue concerns the use of a
double standard in judging dissident and

institutionally sanctioned aggression. Agencies of
government are entrusted with considerable
coercive power designed to protect the citizenry.
When countries wage wars in the name of national
security, their supporters view military combat as
an admired act of patriotism, whereas opponents
consider it an unauthorized use of government

authority to kill human beings. They insist that
violence per se should be deplored, regardless of
whether it is under governmental or private
sanction.
Similar controversies arise about how people

evaluate domestic use of governmental force.
Those whose social and economic interests are
well served by the system applaud coercive law
enforcement practices as acceptable means of

keeping lawlessness under control. By contrast,
dissident factions tend to regard military and
police forces as control agents of political

authority more intent on preserving the status quo
than impartially protecting the welfare of all
segments of society. Few would deny that
governmental agencies are dispensers of
aggression. Authoritarian states rule by force. Nor
is institutionally sanctioned violence foreign to
democratic systems in checking oppositional

behavior that challenges the existing order
(Iglitzin, 1970; Marx, 1970b). When the dispenser
of aggression is a sanctioned authority, his
injurious behavior is minimized as vigorous
pursuit of duty, but if a free-lancing individual

does it, he is judged to be acting violently. The
rhetoric of recent years demonstrates how one
man’s violence is another man’s self-defined
altruism.
Adoption of a definition of aggression

primarily serves to delimit the range of
phenomena that a given theory is designed to

explain. But it does not necessarily aid in
identifying causal relationships, because one
selects for study only specific classes of behavior
and not an abstract “aggression.” That is, research
workers investigate the factors that cause people
to commit murder, to destroy certain property, to
engage in prejudicial practices that have harmful

effects, or other specific forms of injurious
behavior. A high degree of specificity is required at
the investigatory level because there is little
reason to believe that the diverse activities
subsumed under the omnibus label “aggression,”

though sharing some ingredients in common, have
the same determinants. In fact, apparent
inconsistencies in findings relating to aggression
arise in part from the fact that writers frequently
use the generic term rather than specifying the

particular injurious behaviors they have in mind.
The preceding comments apply equally to other

areas of functioning. Creativity, for example, is a
heterogeneous phenomenon that subsumes a vast
array of diverse activities. To advance

understanding of the determinants of
inventiveness, one would select specific endeavors
such as painting, sculpting, composing, or writing,
rather than omnibus artistry. It matters little what

the activity is called as long as it is clearly
designated.
THEORETICAL ANALYSES OF AGGRESSION
Early psychological theories sought to explain
human behavior principally in terms of instinctual

forces. During the period when such
interpretations were in vogue, many writers
advanced the view that man is by nature

aggressive. Although eventually the instinct
doctrine fell into disrepute, the belief that man is
innately endowed with an aggressive drive still
enjoys a sizable following. Proponents of this view
usually draw heavily on the doctrine of biological
determinism in accounting for aggressive actions.
The reasoning typically proceeds along the
following lines: Under primitive conditions of life
aggression had high survival value. Consequently,
survivors of the process of natural selection were

those who possessed strong aggressive
tendencies. Over the years, environmental
conditions have been radically altered, but
unfortunately man's instinctual aggressive drive
and his innate responses, though no longer

serviceable in contemporary society, remain
unchanged. As man’s destructive powers are
markedly augmented by technological advances,
his evolutionary heritage continuously threatens
his very survival. Other versions of this thesis

convey the impression that the neuroendocrine
system with which man is endowed by heredity

generates energy under stress in quantities more
fitting for fight or flight under primitive than
under contemporary conditions. These

characterizations rarely acknowledge that without
the energy-mobilizing system, man would lack the
capacity for emotional experience or for coping
with the common challenges of everyday life.
Instinctual explanations of aggression assumed
many different forms, depending on whether they

posited inborn motivational mechanisms, innately
organized responses, or innate stimulus functions
wherein external cues evoke aggressive actions

supposedly without any prior learning. These
theories ranged from McDougall’s (1931) view
that virtually all social behavior is under
instinctual control, to Freud, for whom aggression
represented overt expressions of the death

instinct, to some present-day ethologists who
regard aggression as innately organized response
patterns that are released in each species by
specific external stimuli.
Psychoanalytic Instinctual Theory
Freud (1920) initially believed that aggression
was a “primary response” to the thwarting of
pleasure-seeking or pain-avoiding behavior. His

explanation of aggression changed markedly,
however, as he modified his instinct theory of

motivation. Freud originally assumed that human
behavior was regulated by two opposing sets of
instincts—the sexual instincts and the self-
preservative ego instincts that altered, deferred,

or inhibited pleasure-striving in the service of the
reality principle. There were certain behavioral
phenomena, however, such as compulsive

repetitions of unpleasant experiences, sadism, and
especially self-destructive actions, that could not

be adequately understood in terms of the dual
instinctual impulses. Freud (1922, 1933) therefore
adopted a new instinctual system of motivation,

with the opposition being between life instincts
(Eros) aimed at enhancing and prolonging life and
death instincts (Thanatos) that continuously
strove for destruction of life within the organism.
Specific actions are presumably generated by
various admixtures of the two instinctual
impulses. In this conceptual revision, aggression

became an inborn drive rather than a by-product
of thwarting libidinal strivings. Sadism and other

forms of interpersonal aggression represented the
death instinct discharged outward; self-injurious
actions were considered manifestations of the self-
destructive instinct directed inward.
The social implications of this theory for the
modification of human aggression provide little

basis for optimism. In an exchange of letters with
Einstein on how mankind might be spared future
wars, Freud (1950) reiterated his belief that
destruction satisfies an instinctual inclination and
that it is therefore fruitless to attempt to eliminate
aggressiveness. Neither satisfaction of material
needs, establishment of equality, nor other
improvements in the conditions of life can alter
the level of aggression. Though in Freud’s view

aggression is unavoidable, the intensity and the
form of its expression are modifiable. Aggression
could presumably be attenuated by the
antagonistic influence of the life instincts.
Development of emotional ties between people

was suggested by Freud as one indirect means of
curtailing extreme destructiveness. The provision
of opportunities for outward discharge of the
innate aggressive impulse was also proposed as a
regulatory device. For Freud (1933) maintained

that when expression of aggression is impeded,
people are forced to behave destructively in order

to protect themselves from self-destruction.
Freud won few adherents to his position even
among enthusiastic advocates of psychoanalytic
theory. Apparently the notion that people
harbored an inborn impulse constantly striving to
kill them exceeded the bounds of credibility. In
recent years much more parsimonious—and

empirically verifiable—explanations of why
people repeatedly behave in self-injurious ways
have been advanced by learning theorists. It is
amply documented in psychological research that
immediate reinforcement outweighs the effects of

delayed punishment in controlling behavior.
Through appropriate temporal arrangement of
positive and negative outcomes, one can cause any
organism, be it man, chimpanzee, or flatworm, to
engage in self-hurtful behavior. Most detrimental

actions such as self-defeating defensive behavior,
excessive drinking, overeating, and other addictive

behaviors are powerfully maintained by their
immediate and automatic reinforcing effects,
whereas the aversive consequences are often

slowly cumulative and not experienced for some
time. Evidence that self-injurious behavior in
humans can be turned off and on by
environmental cues and reduced in frequency by

varying its immediate consequences provides
further support for the view that such behavior is

primarily under social rather than instinctual
control.
As Gillespie (1971) has noted at a recent

Vienna Congress devoted to aggression, most
psychoanalysts have compromised with Freud’s
theory, treating aggression as an instinctual drive
but rejecting the self-directed death instinct.
Aggression presumably originates internally like

other endogenous drives. The lack of evidence for
a physiological drive mechanism in aggression is
either dismissed as unimportant because the
instinctual drive is only a “construct,” or it is
believed that an internal drive will be discovered
by future research.
It is doubtful that the instinctual drive theories
of aggression are capable of empirical verification.
Most of them are formulated in such broad terms

that they do not generate specific predictions that
could be put to experimental tests. When a

nonmeasurable instinctual force is combined with
many qualifying factors that are also somewhat
elusive, the theory can explain any variety of
events that have already happened, though it

cannot predict them. The postdictions, of course,
are compatible with alternative theories that do
not invoke the operation of an innate aggressive

drive. The conceptual status of an instinctual drive
is especially dubious if it is presented as an

autonomous energy-generating system, as in the
case of the aggressive instincts posited by Freud
and by Lorenz. Innate drives usually have an

identifiable biological source such as food
deprivation in hunger, water deprivation in thirst,
and gonadal hormones and evocative external
stimuli in sexual urges. Their strength is therefore
externally modifiable and at least indirectly

measurable.
Ethological Theories
The writings of some ethologists and their

popularizers renewed interest in explanations of
aggressive behavior in terms of instinctual
mechanisms. The view that man possesses a

fighting instinct was engagingly presented in an
anecdotal blend of ethology, anthropology, and
moral philosophy by the distinguished ethologist
Lorenz (1966) in his publicly acclaimed book On
Aggression. Certain aspects of this position were

further popularized by a former playwright,
Ardrey (1966), who contended that humans are

driven to fight each other by a property instinct
designated as a territorial imperative. Other
books, published under catchy titles, similarly
proclaim that man is instinctively aggressive.
For Lorenz, aggression involves an instinctual

system that generates its own source of aggressive
energy independently of external stimulation; this
fighting urge gradually builds up until relieved by
an appropriate releasing stimulus. The property of
being self-generating rather than reactive to

external conditions accounts for its danger and its
unmodifiability. Intraspecies aggression in
subhumans, however, serves a number of positive
functions. Fighting disperses populations over a

habitable area and thus ensures optimal
utilization of food resources. It produces selective
breeding of strong members of the species. And

the aggressive urge can provide the driving force
for activities that outwardly do not appear
aggressive. The potential benefits of aggression
are realized by animals because, through the
evolutionary process, they have developed
aggression inhibitions that prevent them from
destroying members of their own species. For

example, animals typically ward off dangerous
combat by threatening displays that presumably
convey information innately to the protagonists.
When animals aggress toward each other, they do
so in harmless ritualized ways that establish the

victor more through endurance than serious
physical injury. And as an additional protection,
overpowered combatants can promptly terminate
further attacks on themselves by submission
signals that instinctively inhibit the victors’

aggressive behavior.
Lorenz contends that man is endowed with the

same fighting instinct as lower animals, but that it
is poorly controlled because he lacks inborn
inhibitions against severely injuring and killing his
fellow human beings. The following evolutionary
explanation is offered for this tragic state of
affairs: Natural selection ensured that animals
with powerful destructive potential evolved

strong aggression-inhibiting mechanisms to
prevent self-extermination of the species. Since
man is “a basically harmless, omnivorous creature,
lacking in natural weapons with which to kill big
prey,” he never developed built-in inhibitors

against aggression as did the dangerously armed
carnivores. Although man’s natural fighting
equipment may be puny, his intelligence gave him
lethal weaponry for which he has no innate
inhibitors. Paradoxically, man’s highest qualities

—his capacity for thought and verbal
communication—drive him out of the “paradise in

which he could follow his instincts with impunity.”
The proverbial Martian, viewing the skeletons of
dinosaurs in natural history museums and the

masses of lively people inhabiting the earth, would
undoubtedly leave unconvinced that intelligence
and the capacity to learn through experience are
necessarily life-shortening attributes.

Interestingly, Montagu (1968) argues the opposite
position, that man has been favored in the process

of natural selection because most of his behavior
is not under instinctual control and thus can be
adjusted to changing environmental
circumstances.
In point of fact, animals do not possess innate
signals for stopping attacks, and the stereotyped
signals they do use have variable effects on the
responses of their foes (Barnett, 1967). Lorenz’s

widely cited account of defeated wolves turning
off their assailants by exposing the jugular vein
turned out to represent faulty observation. What
Lorenz interpreted as vulnerable exposure by the
defeated that automatically blocks aggression was
actually a challenging posture by the victor
preparatory to dangerous attack (Schenkel, 1967).
Defeated rivals can best forestall further assaults
by surrender, by relinquishing claims, or by

escape. Those that are unskilled in defense or
escape may meet with serious injury or even
death. Under conditions of confinement animals
are known to kill members of their own species, a
phenomenon that is not readily explainable in

terms of innate signal control of lethal responses.
It is true that man does not rely heavily on
auditory, postural, and olfactory signals for
conveying aggressive intent or appeasement. He
has a much more intricate system of

communication—namely, language—for
controlling aggression. National leaders can

therefore better safeguard against catastrophic
violence by verbal communiqués than by snapping
their teeth or erecting their hair, especially in view

of the prevalence of baldness among the higher
echelons.
In the closing chapter of his book, Lorenz

offers some suggestions for how a society might
cope with a poorly controlled aggressive drive. He
cautions against either social prohibitions or

genetic engineering on the grounds that negative
sanctions dam up aggressive impulses to
dangerous levels and that breeding out the

fighting instinct could remove a generalized
source of motivation. Believing firmly in cathartic
drainage, Lorenz recommends that aggression
should be encouraged toward substitute targets
and in sublimated forms. International

competitive sports are considered especially well
suited for this purpose, though in fact such
contests often breed discord. Occupying privileged
positions in society, theorists of different
persuasions readily prescribe gladiatorial combat
for mass viewing and other recreational activities
supposedly conducive to the release of pent-up
aggressive impulses. But to the less advantaged
who rely on aggression to improve their life

circumstances, the recreational prescriptions
would be ineffectual remedies for misdiagnosed
conditions.
As additional preventive measures, Lorenz
advises people to laugh more and to embrace all

humanity with love and friendship. Finally,
militant enthusiasm widely used by demagogues
to breed discord can be enlisted to promote
international friendships by channeling the

devotion to transnational values of art, science,
and medicine. The latter humanitarian practices,
which stand in sharp contrast to the fatalistic view

of human aggression, seem grafted on rather than
derivable from Lorenz’s theory of behavior.
According to the hydraulic motivational system,
no amount of laughter or friendship can stop the
self-generating aggressive drive from welling up
and discharging periodically. Prescriptions of
benevolent remedies increase the appeal of

instinctual doctrines.
In professional circles, the works of Lorenz and
Ardrey were admired for their literary quality, but
severely criticized for their weak scholarship. A
volume edited by Montagu (1968) contains

critiques by a number of scientists who evaluated
the validity of the views advanced in these books.
Ardrey’s case for an innate territorial drive, which
is disputed by a sizable body of evidence, was

dismissed as grossly misleading. Lorenz was
similarly taken to task, though with less enmity, on
a variety of grounds; many of his generalizations

about animal behavior and cultural patterns
contain errors of fact; questionable causal
interpretations are assuredly proffered without
substantiating data; and evidence that fails to
support the instinctual theory is either slighted or
disregarded. Nowhere does Lorenz provide
adequate criteria for differentiating inborn

patterns of behavior from those developed and
maintained by experiential influences. Since he
attributes most social behaviors to instinctual
sources, usually by analogy to subhuman patterns,
the disregard of learning influences comes as no

particular surprise. Thus, for instance, defense of
one’s own beliefs and customs is depicted as a
“phylogenetically programmed behavior
mechanism” and likened to the instinctual
“triumph ceremony” of geese. Instinctual

attribution constitutes explanation in this scheme.
In a scholarly critique of the principles later

used to explain aggression, Lehrman (1953)
marshaled considerable evidence to dispute many
of Lorenz’s nativistic interpretations. Complex
behavior does not emerge as a unitary pattern, but
is formed through integration of many component
activities of differing origins. Thorough causal
analyses of behavior reveal that instinctual

explanations are often not only misleading, but
discourage examination of developmental
processes. Animals are sometimes raised in
isolation from members of their species to show
that responses can appear without learning.

However, practice of component activities, which
is never completely controlled in isolation
experiments, may make a considerable
contribution to the development of supposedly
innate patterns.
Although some researchers have sought the
nature of animals by observing how they behave

under impoverished or limited natural
circumstances, others have explored how their
nature may vary depending upon their social
experiences. The classic experiments by Kuo
(1930) illustrate the latter approach. Kittens were
raised by themselves, with rat-killing mothers, or
with rat companions. Within each of these

conditions half were brought up as vegetarians
and the other half as carnivores. The different
rearing conditions produced cats with pacifistic
and vicious natures. Those raised with rat-killing
mothers became avid rat killers (85 percent); less

than half of those reared in isolation (45 percent)
ever killed any rats; whereas those that grew up
with rats developed a strong attachment to them
and rarely killed any members of their species (17
percent). Vegetarianism reduced rat-eating but

not rat-killing. Kittens that failed to attack any rats
after several months of testing under hungry or

satiated conditions were then exposed to
modeling influences in which they observed adult
cats kill rats. Aggressive modeling converted 82

percent of the pacifistic kittens with the isolation
background into vigorous rat killers, but even the
power of example and severe hunger could not
induce rat-raised kittens to attack rats (only 7

percent did so).
Neutral contact with customary prey also

drastically reduces attack behavior and
susceptibility to aggressive modeling, although
early positive contact serves as a greater

aggression-neutralizing influence (Kuo, 1938).
Given appropriate reinforcement conditions,
animals can adapt to a noncompetitive social
system within which they function amicably, or
they can organize themselves into dominance

hierarchies based on fighting that maintain
orderly relations by threat and submission (Kuo,
1960). Examining the habits of humans raised
alone or in a particular environment would
likewise yield a different view of man’s nature
than if one studies how man behaves under
diverse cultural practices. The higher the
evolutionary development of a species, the greater
is its plasticity.
Lorenz’s motivational model, which assumes
special importance in the explanation of
aggression, is open to even more serious criticism
than his speculations about the origins of
behavior. As a number of researchers (Hinde,

1960; Lehrman, 1953; Scott, 1972) have pointed
out, there exists no neurophysiological evidence
that functional activities generate their own
motivating energy which accumulates with time in

the absence of appropriate releasing stimuli. Nor
is there any means by which energy can become
“dammed up” in the nervous system, forcefully

discharging without external elicitation or spilling
over to brain centers controlling other activities.
In other words, the neurophysiological
mechanisms that mediate aggression do not
themselves create stimulation for fighting.
Researchers working in this field not only
questioned the partial and casual treatment of

empirical evidence, but they also objected to
uncritical extrapolations from a dubious
hereditarian model and to analogical reasoning
based on superficial similarities in cross-species
behavior. Considering that the factors regulating

behavior often tend to be specific to a given
species, most theorists understandably rejected
the notion that mechanisms governing aggressive
behavior in greylag geese and sticklebacks apply

equally to human suburbanites. Even if this
assumption were valid, how could alleged
instinctual mechanisms of aggression that do not

apply even to all lower animals explain the
behavior of man? Territorial fighting, which
figures prominently in the writings of both Ardrey
and Lorenz, is a good case in point. Many herd
animals, including some of man’s closest
forebears, do not stake out territories, nor do they
jealously defend areas they inhabit at any given

time. Even in animals that exhibit territorial
aggression, such behavior tends to be quite
variable depending upon environmental
characteristics, population densities, sex hormonal
states, scarcity of food, and other external factors

(Crook, 1968; Hinde, 1956). The specificity and
variable nature of territorial fighting in lower
animals casts grave doubt on the notion that man
is impelled by a territorial instinct to wage
warfare against trespassers. Indeed, as Carrighar

(1968) has noted, nations’ need for resort to
military conscription suggests that their citizens

are far from moved by an overpowering urge to
fight for territory.
Unlike coral fish that are quick to repulse
approaches by members of their own species,
trespassers on a person’s property do not as a rule
evoke territorial fighting. Under most conditions
interlopers are more likely to elicit curiosity than

combat. On the occasions when people do fight
over territorial intrusions, the phenomena can be
much better explained in terms of learning
processes than instinctual mechanisms. Some
types of intruders have no particular effect on the

proprietor’s well-being; the appearance of others
may produce positive outcomes; and still others
appropriate or destroy valued resources, creating
adverse effects. People may thus incline to be
initially wary of a stranger, but they base their

reactions toward him on judgments of probable
consequences. They will welcome a visitor of good

will and threaten a menacing one. Differential
experiences, either direct or vicarious, with
different types of intruders produce

discriminative responses that cannot be readily
explained by an indiscriminate property instinct
that automatically compels an organism to chase
whatever approaches its territorial boundaries.
Carefully conducted ethological studies
(Carthy and Ebling, 1964; van Lawick-Goodall,

1971) have produced a large body of descriptive
data on the way animals of different species
behave in their natural habitats. Some of this

information is specifically concerned with
environmental situations that stimulate fighting.
In certain species aggressive responses are
regularly elicited by specific environmental
stimuli. Stickleback fish, for example, vigorously

attack male members of their species during
periods when their bellies have a red coloration.
Their assault on red-bellied facsimiles that
otherwise differ markedly in shape from actual
sticklebacks shows that their aggressiveness is
controlled by color cues (Tinbergen, 1951).
Territorial encroachment, as previously noted,
is another event that evokes fighting in some
animal species and in many fish and birds. After

proprietors establish rights over a given locality,
they generally attack rivals who intrude on their

domain. A number of other factors have been
identified as determinants of aggressive behavior,
though as in territorial fighting, the degree of
behavioral control exerted by the evocative

stimuli is significantly affected by a variety of
conditions. Males of some species fight over
possession of females, especially during mating

periods. Animals also fight over food, over rank in
a dominance hierarchy, in defense of their young

offspring, in warding off attacks from predators on
members of their own species, and they often
behave aggressively when subjected to painful

stimulation (Hamburg, 1970).
Field studies have been principally concerned
with environmental elicitors of aggression,
perhaps because the cues that evoke combat can
be easily observed and documented. On the other

hand, identification of social learning
determinants of aggression under natural
conditions of life requires painstaking analyses of
interaction processes and therefore tends to be
neglected. Few researchers are willing to follow

selected animals continuously as they wander
through their natural habitats in order to record
incidents in which aggressive responses are
modeled and are affected by the consequences of
fighting various adversaries. Detailed observations

would also be needed to trace the manner in
which social stimuli become endowed with fight-

eliciting properties through associations with
combat experiences. When elicitors are favored
over causal processes in field observations, the

selective data tend to convey the impression that
aggressiveness is a preestablished phenomenon
triggered by more or less fixed stimulus events.
Lower species usually come equipped with
rudimentary preformed habits and high initial
susceptibility to modeling influences that enable

them to respond adaptively from the outset to
their environmental circumstances. By contrast,
man is furnished with few inborn habits, but with

vast potentialities for learning. Not only are people
more malleable than biologically simpler
organisms, but advanced information-processing
capacities render human behavior more subject to
social and cognitive control. Rather than

responding instinctively to releasing sign stimuli,
humans regulate their behavior in accordance
with judgments of anticipated consequences for
prospective actions. In other words, innate
releasing and inhibiting mechanisms have been
largely replaced by cortical control. For these
reasons, causal relationships established in lower
species may be misleading when applied to man
without empirical confirmation of the equivalence.

Interspecies studies of the mechanisms governing
sexual behavior provide an excellent illustration of
this point.
Hormonal control of sexual behavior decreases
with advancing evolutionary status (Beach, 1969;

Ford and Beach, 1951). In lower mammalian
species, for instance, sexual activities are closely
regulated by gonadal hormones; among primates,
sexual responsiveness is partially independent of

hormone secretions; human sexual arousal is
exceedingly variable and relatively independent of
hormonal conditions. Thus, to produce a rodent

Don Juan would require repeated administration
of testosterone, whereas showing him lascivious
pictures of well-endowed female mice would have
no stimulating effects. One would, on the other
hand, rely on sexually valenced displays rather
than on hormonal injections to produce erotic
arousal in human males.
Cross-cultural studies of sexual behavior (Ford
and Beach, 1951) disclose further that the regions
of the body, the physical characteristics, the social
acts, and the extrinsic sensory events that function
as sexual arousers all vary considerably from

society to society. What is erotically stimulating in
one culture—corpulence or skinniness; upright
hemispherical breasts or long pendulous ones;
shiny white teeth or black pointed ones; deformed

ears, nose, and lips or naturally shaped ones;
broad or slim hips; light or dark skin color—may
be neutral or repulsive to members of another

social group. A similar diversity exists in the age at
which sexual interest first emerges, in the forms
that sexual behavior takes, and in the choice of
sexual objects. Considering that human sexual
responsiveness is, in large part, socially rather
than hormonally determined, findings based on
studies of subhuman species concerning

regulatory mechanisms have limited
generalizability to humans. There is every reason
to believe that the evolutionary decrease in innate
determinants of behavior holds true for
aggression as well.
Studies of hormonal influences have shown

that androgen fosters aggressiveness in animals
(Conner and Levine, 1969; Rothballer, 1967). In
addition to their effects on brain organization, sex

hormones can influence combative behavior
indirectly by stimulating physical growth and
muscular development. Moreover, during mating

periods when sex hormone levels are high, male
species are especially combative. Animals must
rely upon their biological equipment for whatever
success they can achieve in combat. Hence, those
with large canine teeth, powerful jaw muscles, and
imposing size become belligerent fighters through
victories; the physically less well-endowed

become submissive through defeats. By contrast,
man’s capacity to use destructive weapons greatly
reduces His dependence upon structural
equipment to aggress successfully. A puny man
with a gun can easily triumph over powerfully

built opponents who are unarmed. In violent
human encounters other than fisticuffs, weapons
will almost always beat physical build. Man’s
capacity for social organization likewise reduces
the importance of structural characteristics in

aggressive attainments. At the social level,
aggressive power derives from organized

collective action rather than from biological
advantage. Victory is thus assured by the force of
numbers acting in concert, and the physical

stature of individual challengers does not matter
much, if at all. Organized fighting between groups,
as in international warfare, domestic rebellion,
and intergang combat, is a human creation that

does not exist except in the most rudimentary
form among animals.
Animal studies have shown that testosterone
treatment of fetal or newborn females
masculinizes them and increases their

aggressiveness. In humans, excess fetal androgens,
produced by either enzyme defect or
administration of progestin to pregnant mothers,
causes masculinization of external genitalia in
female offspring and fosters male orientations.

These females are, however, no more physically
aggressive than nonandrogenized girls (Ehrhardt,
Epstein, and Money, 1968). Interpretation of the
differences in interest patterns is complicated by
the fact that approximately half of the
androgenized girls were raised as males during
the early months of life before their condition was
correctly diagnosed. If excessive prenatal
testosterone is highly associated with later

tomboyishness, further research would be needed
to determine whether hormones exert their effect
through neurosexual organization, physical
structure, or by some other means (Hamburg,
1969). Androgenized girls who selectively

associate with boys because of their physical
characteristics might be expected to develop male
orientations through peer modeling and
participation in masculine play activities.
Structural characteristics related to

aggressiveness also have somewhat different
evolutionary and survival consequences for

animals and man. In some animal species, physical
strength determines which males do the mating.
Combat victors gain possession of females so that
the most dominant males have the highest
reproduction rates. With humans, most of whom
practice monogamy, mate selection is based more
on such qualities as good looks, winsome style,

intelligence, and financial standing than on
fighting prowess. Within such social
arrangements, differential reproduction rates are
primarily determined by religious and ecological
beliefs, as well as by access to birth control

devices. For these reasons, variations in fertility
do not necessarily favor aggressiveness.
Questions of generality likewise arise in
evaluating the influence of genetic factors on

aggressiveness. The propensity of laboratory
animals to behave aggressively can be significantly
altered by selective breeding. Descendants of

aggressive strains tend to be increasingly more
combative across succeeding generations than
offspring of docile stocks. The matter is more
complicated, however, than it might appear from
limited assessments. Contrasting strains can vary
widely in their relative combativeness depending
upon the circumstances under which they are

tested and the nature of their targets. Thus,
members of docile strains may behave
submissively in one setting but aggressively in a
different situation; similarly, fighting breeds may
be vicious toward certain targets and pacific in

relation to others (Fredericson and Birnbaum,
1954; Ginsburg and Allee, 1942).
Corning and Corning (1972) put it well when
they noted that genes do not produce behavior;

rather, they generate enzymes that affect
biochemical processes and structural
development which, in turn, can influence

response potentialities. Therefore, to understand
the means by which genetic variations affect
aggressive responsiveness requires knowledge
about the links between genes, neurophysiological
mechanisms, and behavioral capabilities to act
aggressively. When animals are bred with respect
to an outstanding characteristic, they may display

other features that largely account for observed
differences between strains. This is especially true
of activities that require a variety of functions (for
example, perceptual, cognitive, locomotor,
emotional, and motivational), as is usually the

case. By selective breeding, for example,
researchers produced populations of animals
bright and dull in maze learning. Further analysis
revealed that not only did the relative superiority
of the groups change on different learning tasks,

but that differences in motivation and
emotionality were largely responsible for the

variation in learning performances (Searle, 1949).
With regard to fighting, combative strains have
been found to be more active motorically and less

emotionally upset by stress (Lagerspetz, 1969).
Such factors might be expected to determine
success in physical combat.
To complicate the picture further, some of the
variability attributed to genetic factors may stem
from other sources. When animals are reared by

their own mothers, postnatal maternal influences
can contribute to interstrain differences in
behavior (Denenberg, 1970). Offspring from

aggressive strains nursed by their own mothers
are likely to turn out to be more combative than
those reared by docile foster mothers. Behavioral
differences can result from maternal handling
during the nursing period and from maternal

modeling of aggression in later developmental
phases, as convincingly demonstrated by Kuo
(1930).
Some characteristics are highly heritable and
difficult to modify, whereas others, which are only
indirectly affected by genetic factors, may be
highly susceptible to environmental influences.
The importance attached to genetically mediated

effects on aggressiveness therefore depends upon
their modifiability through experience. Several
investigators have shown that aggression in
combative strains can be markedly reduced by

repeated defeats, whereas submissive animals can
be made somewhat more aggressive through
combat victories (Ginsburg and Allee, 1942;
Lagerspetz, 1964, 1969). Since even victors are
likely to receive painful bruises while fighting, the

experienced punishment may detract from the
value of eventual success. To take a human

example, a person who triumphed over an
opponent but in the process lost his teeth would

be somewhat hesitant about violent encounters
even if assured victory. The relative influence of

reinforcement and genetic determinants is best
tested under conditions in which conquests gain
victors access to food, mates, territories, and social
rank. These are the types of purposes that

aggression serves for animals in their natural
habitat. In the case of man, although some may

derive self-esteem rewards from disposing of
opponents, few go around picking hurtful fights
merely to win them. They too risk the costs of
aggression to obtain a variety of outcomes

extrinsic to the act itself.
On the assumption that the male-determining
chromosome is related to aggressiveness,
researchers explored whether a genetic

abnormality involving an extra male chromosome
increases propensity to violent conduct. (Normal
males possess an XY pair of sex chromosomes,
whereas females have an XX pair; the Y is
considered to be the chromosome determining
maleness.) Surveys of newborn children reveal
that approximately 1 in 500 males possesses an
extra Y chromosome (XYY).
Little interest was shown in the XYY anomaly

until Jacobs and her colleagues (Jacobs, Brunton,
and Melville, 1965) reported that mentally
defective men institutionalized for various forms
of criminal behavior show a higher prevalence of
the extra Y chromosome (2.9 percent of all those

tested) than might be expected in the general
public (0.2 percent). Since the XYY condition is
comparatively rare but goes with tallness, many of
the subsequent studies used biased sampling

procedures by selecting only tall imprisoned men
for chromosomal analysis. On the basis of further
reports of higher prevalence rates, some

investigators concluded that XYY males are
genetically predisposed to violent and antisocial
conduct. The stereotype of the XYY males as
dangerously assaultive was quickly established by
publicized murder trials in which legal
responsibility for homicidal acts was disclaimed
on the grounds of chromosomal defect. Richard

Speck, a ruthless mass murderer, was reported in
the press and in professional journals as an XYY
aberration, although in fact he had no genetic
defect. Some medical practitioners even advocated
genetic counseling for parents about the antisocial

propensities of XYY children. Such a stigmatizing
practice could very well lead apprehensive adults
to overreact to misconduct in ways likely to
confirm the prophecy. Lost in the publicity was the
notable feature that XYY inmates in the original

survey rarely aggressed against people and, of
their total convictions, 88 percent were for

offenses against property. The XYY prisoners in
fact have a lower incidence of physical and sexual
assaults than suitably matched XYs (Price and

Whatmore, 1967). Perhaps the most striking
aspect of the chromosome story is the ready
proclivity to blame crime on “bad genes.”
The alleged relationship between the double Y
chromosome and aggressiveness was strongly
disputed by some researchers on a number of

grounds (Kessler and Moos, 1970; Owen, 1972;
Shah, 1970). To begin with, the prevalence studies
confounded, among other things, chromosomal

anomaly with height and social influences. Given
evidence that the XYY karyotype is found more
often among tall men, its prevalence among tall
inmates should be compared not against the
general population, but against a subsample of tall

noninstitutionalized individuals matched with the
prisoners on social factors conducive to antisocial
aggression. The misleading results of biased
sampling are shown by Clark et al. (1972), who
report that when only tall prisoners are selected
for chromosomal screening, the prevalence of XYY
is higher (2.7 percent) than when prisoners under
6 feet are also included (1.8 percent). Although
XYYs who run afoul of the law are not especially

assaultive, they are nevertheless arrested at an
earlier age and more often. This differential arrest
pattern, which can have a decisive impact on the
course of future behavior, may conceivably be
related to physical stature. Apart from the

psychological strains and older companionship
fostered by conspicuous tallness, large offenders
are likely to be treated by arresting officers as
older and more dangerous than are smaller ones.
Early commitment to a reformatory populated

with delinquents is apt to launch one on an
antisocial career, regardless of genetic makeup.
An extra male chromosome was believed to
enhance aggressiveness by producing
hypermasculinity. Results of a comparative study
by Clark et al. (1972) cast serious doubt on this
line of reasoning. They conducted a chromosomal
survey of men in several penal institutions for the
rate of XYY anomaly and the Klinefelter syndrome

(XXY), which is presumed by many to be
associated with nonaggressive tendencies. The
XYYs are no more prevalent (1.8 percent) than
XXYs (2.6 percent), nor is there any discernible
difference in their criminal records. It might be

noted in passing that results of a large number of
chromosomal surveys yield a lower prevalence of
XYY prisoners (2.3 percent) than originally
reported, and even this figure overestimates the
rate for prison populations because, with few

exceptions, only tall males are selected for analysis
(Owen, 1972).
What effects, if any, an extra Y chromosome
has on behavioral development can be
satisfactorily answered only by a double-blind
prospective study in which psychological
functioning is periodically assessed in children
with the XYY karyotype and in suitably matched
controls who have no chromosomal anomaly. A

longitudinal investigation of this type would most
likely demonstrate that the vast majority of XYY
males, like others, are engaged in prosocial
pursuits. The data already reviewed suggest that
the small numbers who may get into legal

difficulty are, if anything, apt to be less
interpersonally assaultive than transgressors of
normal genetic makeup.
It is interesting that people vigorously disclaim

that instincts or genes cause them to behave the
way they do, but are quick to believe that man is
instinctively aggressive. The wide popular appeal

of such interpretations cannot be attributed to the
attraction of simple explanations, nor to the
prevalence of aggressive behavior, because other
social practices found to be widespread or
universal, such as bartering behaviors, are rarely
ascribed to an economic instinct. One possible
explanation is that most people find it exceedingly

difficult to understand how socialized human
beings can repeatedly commit atrocities, often at
the risk of self-extermination, unless driven by an
inherent viciousness. Also, by attributing
aggression to inherited tendencies, people are

absolved from the responsibility of changing social
conditions that benefit their self-interests at the
risk of provoking aggression in disadvantaged
members of society.
The dichotomous view that behavior is either

learned or innate has dwindling support as
knowledge about human functioning increases.

Though radical hereditarians and
environmentalists still exist, most theorists
acknowledge that social and physiological
influences are not that easily separable, since both
sets of factors interact in subtle ways in
determining behavior. Where certain biological
equipment is needed to perform manual

aggressive acts, structural factors, which have a
genetic basis, may partly determine whether
initial aggressiveness proves successful and is
further developed, or whether it fails and is
discarded. Possession of a brawny physique, for

example, increases the probability that physically
aggressive modes of behavior will prove effective.
Physical characteristics can also indirectly
affect the development of aggressive behavior

through their influence on associational
preferences. The social acquaintances and
activities of robust boys, for instance, are likely to

differ substantially from those of frail boys or of
petite girls. Association preferences, in turn,
determine to a large degree the types of models
repeatedly observed and, consequently, the
patterns of behavior that will be most thoroughly
learned. The groups with which children choose to
affiliate also define the forms of behavior that are

normatively sanctioned, rewarded, and selectively
perfected.
Even in instances in which new patterned
responses are formed entirely on the basis of
learning experiences, physiological factors still

serve as contributing influences. Learning capacity
is affected by genetic characteristics. And while
the organization of behavioral components into
novel intricate patterns is a product of experience,

the rudimentary elements are present as part of
the natural endowment. Thus, for example,
children begin with a set of elementary sounds

that they subsequently learn to combine into an
infinite variety of larger verbal units to form
grammatical speech. The basic phonetic elements
may appear trivial compared to the learned
intricate compounds, but they are nevertheless
essential. But it would be misleading to designate
social behavior as instinctual simply because it

employed a few innate elements. The fact that a
child is born with the capacity to strike objects
with his hand does not mean that the elaborate
fighting skills of a boxer reflect an inherited
tendency. One of the basic issues in dispute with

regard to instinctual interpretations of aggression
is whether complex patterned responses
themselves come preformed or are fashioned
through experiential influences.
Many so-called instinctual behaviors may

contain a large learning component even in the
common patterns displayed by members of a

species. Observational learning is a principal
means of acquiring new response patterns in
animals and humans alike (Bandura, 1971d).
Developmental studies of modeling processes
disclose that animals readily learn adaptive
responses soon after their eyes become functional
by observing the behavior of skilled models (Adler

and Adler, 1968). In the natural environment
young offspring are provided with countless
opportunities to observe and to learn the behavior
exhibited by members of the same species.
Observational learning may play an especially

important role in species that are highly
susceptible to imprinting. This is a process
wherein young offspring develop a strong
attachment to, and rapidly learn general

characteristics of the model to which they were
first exposed during a developmentally sensitive
period (Hess, 1959). Although some response

patterns are transmitted during the
impressionable period itself, the formation of a
close social attachment to a model greatly
enhances observational exposure and thereby
assures further learning of the modeled behavior.
Not all species are susceptible to imprinting.
Apparently those whose locomotor abilities

mature rapidly are most prone to display the
phenomenon, whereas higher mammalian species
that are born biologically incompetent and hence
require extended dependence on caretakers do
not (Klopfer, 1962). These notable evolutionary

differences further underscore the need for
caution in generalizing across species. Moreover,
modeling and reinforcement influences, which are
universally present, have powerful effects on the

behavior of species that are influenceable
throughout life rather than mainly at an early
optimal period of development. For this reason,

informal ethological observations cannot provide
full understanding of the origins of behavior.
When the effects of social experiences are not
systematically measured, behaviors that result
from observational and reinforcement learning
are readily ascribed to biological inheritance.
Denenberg (1970) has shown that neonatal

rearing practices not only have a profound effect
on aggressiveness, but produce lasting changes in
the animals’ physiology as reflected in
adrenocortical activity. Evidence that behavioral
potentialities can be significantly affected by fetal

environment further complicates identification of
instinctually determined behavior. Finally, given
some indications of fetal conditioning (Spelt,
1948), the possibility of prenatal learning must be
considered in evaluating simple response patterns

present at birth.
The coordination of aggressive behavior, like

other forms of visceral and motor responsiveness,
depends on neurophysiological mechanisms.
Substantial research has been conducted on
animals to determine whether the mechanisms
mediating aggression are localized in specific
brain regions. The results show that subcortical
structures, principally the hypothalamus and the

limbic system, are involved in facilitating and
inhibiting aggressive behavior. Lesioning these
areas or stimulating them chemically or
electrically can modify the expression of threat
displays and stereotyped attack responses. The

effects induced by such methods are not entirely
specific, however. Not only are behaviors other
than aggression similarly affected, but activation
of any one of several regions can produce similar
behavioral effects (Clemente and Lindsley, 1967;

Garattini and Sigg, 1969). While the hypothalamus
is a prominent mediator of attack responses,

animals can be prompted to aggress by external
events even after the hypothalamus has been
surgically isolated from the rest of the brain

(Ellison and Flynn, 1968). The same neural
structure can thus serve a variety of behavioral
functions, and different structures can mediate the
same social behavior.
Although aggressive actions are partly
coordinated at subcortical levels, these neural

systems are selectively activated and controlled by
higher cortical functions. Consequently,
responding is regulated to a large extent by

sensory inputs that are centrally processed. An
ingenious study by Delgado (1967) shows how
experiential influences can govern response
selection during electrical activation of the same
neural center. It has been shown that electrical

stimulation of the hypothalamus in the vicinity of
the ventromedial nucleus generally evokes attack-
like behavior in animals. Delgado added a social
dimension to the research that threw new light on
thalamic control of aggression. He recorded the
social behavior of a small colony of monkeys at
normal times and at periodic intervals when
selected members with brain-implanted
electrodes were electrically stimulated through

radio transmission.
Hypothalamic stimulation of a monkey who
assumed a dominant role in the colony instigated
him to attack subordinate male members, but the
stimulated boss monkey did not assault the

females. Nor did he attack his immediate
subordinate, with whom he was on friendly terms.
By contrast, hypothalamic stimulation elicited
cowering and submissive behavior in a monkey of

low social rank. Even more impressive is evidence
that electrical stimulation of the same cerebral
mechanism can evoke markedly different behavior

in the same animal as his social rank is modified
by changing the membership of the colony. Thus,
hypothalamic stimulation elicited submissiveness
in the animal when he occupied a low hierarchical
position, but marked aggressiveness when he was
the dominant member in the group.
Other investigators (Panksepp, 1971;

Panksepp and Trowill, 1969; Roberts, Steinberg,
and Means, 1967) also report that electrical
stimulation of the same brain region produces no
attacks in some animals, that attacks are more
readily elicited in normally aggressive animals

than in more docile members of the same species,
that hypothalamically stimulated aggressive
responses are inhibited under threat of
counterattack, and that assaultive responses

decline and even disappear with repeated
electrical stimulation. Aggressive responses are
not, in fact, directly and automatically elicited, but

vary depending upon the types of targets
available. Such evidence indicates considerable
environmental control of centrally stimulated
effects.
For the most part, researchers studying neural
regulation of attack responses directly stimulate
subcortical areas with an artificial electrical input

in a relatively barren social environment without
varying prior learning experiences. Experimental

arrangements of this sort are advantageous for
localizing lower brain functions. However,
Delgado’s finding that social influences can
determine the nature of the cerebrally evoked

behavior indicates that such procedures may
exaggerate the degree of neural control under
natural conditions, in which environmental

activators transmitted by informative sensory
messages are modulated through higher memory

and learning mechanisms. The functional
organization of these higher control systems is
largely determined by learning experiences. It is

therefore only by varying neural and experiential
influences that a full understanding can be
obtained of how the various brain structures
interact in regulating aggressive behavior.
It is valuable to know how biological
equipment works internally, but from the
standpoint of explaining aggression it is even

more important to understand how it is socially
activated for different courses of action. In
everyday life, biological systems are roused in

humans by provocative sensory inputs from the
environment and by anticipated response
outcomes and other types of ideational activation.
Hence, a remark interpreted as an insult will
doubtless generate activity in the hypothalamus,

whereas the same comment viewed innocuously
will leave the hypothalamus undisturbed. Given a
negative interpretation, social factors are likely to
determine how the recipient will respond.
The social learning theory of human
aggression adopts the position that man is
endowed with neurophysiological mechanisms
that enable him to behave aggressively, but the

activation of these mechanisms depends upon
appropriate stimulation and is subject to cortical
control. Therefore, the specific forms that
aggressive behavior takes, the frequency with

which it is expressed, the situations in which it is
displayed, and the specific targets selected for
attack are largely determined by social experience.
A problem of clinical interest that has
implications for neural regulatory functions is
concerned with whether brain injuries affect
expression of aggressive behavior. Literature on

the neural correlates of human aggression is
sparse. In exploratory studies of assaultive

patients with brain pathologies, Delgado and his
associates (1968) have occasionally been able to

elicit aggressive responses by telemetered
stimulation of the amygdala, but as a rule, the
stimulation mainly evokes strange sensations and
feelings.
Mark and Ervin (1970) report a high incidence
of neurological dysfunctions in both self-referred
violent patients and repetitively assaultive
prisoners. According to these authors, the
aggressiveness may reflect impairment of cortical

control that leaves the person vulnerable to
excitation by stimuli that would ordinarily be
ineffective. Alternatively, external stimulation of a
malfunctioning area of the brain may trigger

neural discharges that prompt actions, sometimes
at reduced levels of consciousness.
In judging how much importance to attach to
neural dysfunctions, one also needs to know in
how many cases brain damage is present but
physical aggression does not occur. Incidence
rates are less subject to misinterpretation when
individuals are selected because they possess the
condition supposedly conducive to aggression

rather than because they display a history of
aggressive conduct. Let us assume, for the sake of
illustration, a 30 percent incidence of brain
damage in the criminally assaultive, whereas in

the total population of brain-damaged people 3
percent are physically aggressive. Knowing only
the proportion of incarcerated offenders who
suffer neurological impairment, it would be easy
to overrate the influence of neural factors.
Cerebral disorder can accompany aggression
either as cause, as effect, or as a joint product of

other conditions. Abnormal
electroencephalograms (EEGs) have been

reported by some investigators to be more
common among assaultive persons than in the

general population (Mark and Ervin, 1970). The
meaning of differential incidence rates can change,
however, depending upon what groups are
compared. In a study reported by Ostrow and

Ostrow (1946), for example, abnormal EEGs were
just as prevalent among conscientious objectors
who eschew violence and other inmates jailed for
nonaggressive offenses as among assaultive
prisoners. It is conceivable that deviant EEG

patterns may, in many cases, be merely
accompaniments of aggression and not its cause.
Considering the impoverished backgrounds from
which most prisoners come, they probably show a
higher incidence of physiological deficiencies of

almost any kind. Moreover, assaultive people are
repeatedly involved in fights, in auto accidents,

and in forcible arrests that might easily result in
head injuries (Bach-y-Rita et al., 1971; Bandura
and Walters, 1959). Thus, some cerebral damage

would not be a surprising product of an aggressive
style of life. A longitudinal study of boxers would
attest to progressive cerebral damage with
continued fighting. Finally, a person who suffers

brain injury that impairs his level of functioning
may be treated by others in a rejecting,

disparaging manner productive of aggression. If a
neurologically intact person were dealt with in a
similar fashion, he too might be prompted on
occasion to respond assaultively. If the effects

operate principally through social consequences,
brain damage with maltreatment may increase the
likelihood of aggression, but brain damage with

considerate treatment may not.
Theories appear comprehensive when they

contain an all-embracing analysis to the effect that
aggressive behavior is determined by biological,
psychological, and sociological factors. Such
general propositions, which apply equally to
almost anything people may do, hardly advance
our understanding of aggression. It is also easy to
misjudge learning approaches on the grounds that,
as their name seems to imply, they must ignore

biological determinants. The critical issue is not
whether aggressive behavior is multiply
determined, but to what extent do biological,
psychological, and social influences contribute to
variations in aggression between different people

and in the same individual at different times and
under different circumstances. It is evident from
the preceding discussion and material presented
later that the proportion of variance accounted for
by these determinants differs across species, types

of aggressive behavior, and social conditions.
Aggressive Drive Theories
Until recent years most theorists subscribed to

an energy model of aggression, but they rejected
instincts as the driving forces. In an effort to
resolve the motivational problem, drives
eventually replaced instincts as internal impellers
of action. According to these theories, man is

motivated to behave aggressively by a frustration-
produced drive rather than by an innate
aggressive force. This view, which gained
widespread popular acceptance, was originally
expounded in a set of testable propositions by

Dollard and his co-authors (1939) and elaborated
further in later publications (Feshbach, 1964,
1970; Sears, Whiting, Nowlis, and Sears, 1953;
Whiting and Child, 1953).
The frustration-aggression hypothesis

contended that interference with goal-directed
activity induces an aggressive drive which, in turn,

motivates behavior designed to injure the person
toward whom it is directed. Infliction of injury was
assumed to reduce the aggressive drive. In the
form in which it was originally presented, the
hypothesis presupposed that frustration always
produces aggression. In later modifications of the
hypothesis, aggression was still considered the

naturally dominant response to frustration, but
nonaggressive reactions could occur if aggressive
behavior had previously been eliminated through
punishment or nonreward. Although some
members of the Yale group (Miller, 1941; Sears,

1941) discarded the notion that aggression is the
only unlearned response to frustration, frustration
nevertheless continued to be regarded as an
inevitable cause of aggression; in other words,
whenever an aggressive act occurred, frustration

was assumed to be always present as the
instigating condition. For years the standard

explanation of aggression was in terms of
frustration.
A number of factors were identified by Dollard
and his associates as major determinants of
aggressive behavior and the specific forms it takes.
Since this formulation drew heavily on
psychoanalytic theory, the variables singled out

for attention are similar to those with which
psychoanalysts concerned themselves at the time,
rather than to those that have subsequently been
found to be powerful controllers of aggressive
behavior. One set of conditions that was proposed

concerned antecedent factors governing degree of
frustration, from which strength of the aggressive
drive is inferred. The strength of instigation to
aggression was assumed to vary directly with how
strongly the thwarted response was motivated,

the extent to which the thwarted response was
interfered with, and the number of actions that

were frustrated.
Criticism of the frustration-aggression
hypothesis focused at first on the nature of
responses to frustration. Anthropologists
(Bateson, 1941) pointed out that in some cultures
aggression was by no means a typical response to
frustration. Barker, Dembo, and Lewin (1941) and

Wright (1942, 1943) demonstrated that young
children were inclined to regress rather than to
aggress when frustrated. Other critics argued that
only some kinds of frustration evoke aggressive
behavior and that other forms do not. Maslow

(1941), Rosenzweig (1944), and more recently
Buss (1961) noted that personal insult and threat
are more likely to evoke aggression than blocking
of ongoing behavior, an issue that will be
examined more fully later. And finally, Pastore

(1952) emphasized the role of justifiability of the
frustration in determining whether or not an

aggressive response will occur. For example, a
person who is passed by a municipal bus returning
to a garage for repairs will display much less

aggression than if the bus he has been waiting for
drives by without justification, even though the
goal-directed behavior of the ignored passenger
has been equally thwarted in both instances. The

effects of justifiability of frustration have not been
tested behaviorally, but people typically report

more aggression to hypothetical thwartings that
appear unwarranted than to those for which
excusable reasons exist (Cohen, 1955; Pastore,
1952).
Frustration subsumes such a diverse set of
conditions that it no longer has a specific meaning.
As new instigators of aggression were identified,
the definition of frustration was stretched to

accommodate them. In experimental studies, for
example, people are frustrated by being personally
insulted, subjected to physically painful treatment,
deprived of valued rewards, blocked from
reaching desired goals, and by experiencing
failure. A large body of evidence exists showing
that punishment, extinction, delay of reward, and
response obstruction do not have uniform
behavioral effects (Bandura, 1969a). Even the

same treatment can elicit markedly different
responses at different intensities and with
different learning histories.
Not only is there great heterogeneity on the
antecedent side of the relationship, but the

consequent part of the formula—the aggressive
behavior—also subsumes a vast array of activities
colored by value judgments. As we already have
seen, whether or not a given action is considered

aggressive depends on a variety of factors, many
of which reside in the definer rather than in the
performer. The same injurious behavior can be

labeled aggressive or not depending on subjective
judgments of whether it was intentional or
accidental; whether or not it is socially sanctioned;
and among other factors, on the age, sex, role, and
socioeconomic level of the performer. Considering
the mixed items that are included under both
frustration and aggression, it is questionable

whether any general statements about the
relationship between the two events have much
meaning or validity.
The widespread acceptance of the frustration-
aggression notion is perhaps attributable more to

its simplicity than to its demonstrated predictive
power. In point of fact, the formula that frustration
breeds aggression does not hold up well under
empirical scrutiny in laboratory studies in which

conditions regarded as frustrative are
systematically varied. They may sometimes
increase aggression (Berkowitz, 1965a; Hartmann,

1969; Ulrich, Hutchinson, and Azrin, 1965); they
may have no effect on aggressive behavior (Buss,
1966a; Jegard and Walters, 1960; Walters and
Brown, 1963); or they may reduce aggressive
responding (Kuhn, Madsen, and Becker, 1967). In
experiments reporting positive results, frustration
usually exerts an influence only in conjunction

with prior training in aggression or exposure to
aggressive modeling influences. The fact that
negative findings occurred in studies in which
social learning variables were highly influential
lends support to the view that frustration, as

commonly defined, is only one—and not
necessarily the most important—factor affecting
the expression of aggression.
There are two possible approaches to the

problem of frustration. One may define it in terms
broad enough to encompass heterogeneous
conditions that could conceivably be judged to be

frustrating. This strategy provides the
convenience of a single shorthand label, but for
reasons given earlier, the diverse operations
subsumed under the omnibus term will yield
inconsistent and contradictory empirical
relationships to aggressive behavior. Alternatively,
one may distinguish between dissimilar

antecedent conditions and study how each type
affects different classes of aggressive behavior.
The latter approach is selected in social learning
analyses of aggression.
Since frustration does not always produce

aggression, inhibitory factors were introduced by
drive proponents as additional controlling
influences. A frustrated person will tend to
suppress overt aggressive actions that are likely to

be punished. The strength of inhibition was
postulated, in this theory, to vary positively with
the severity of punishment anticipated for

behaving in an aggressive manner. There is
substantial evidence, to be reviewed later, that
punishment does indeed exert a regulatory
function over aggressive behavior. However, the
relationship between punishment and aggressive
behavior proved to be more complex than was
originally believed. The degree of behavioral

control exerted by punishment is, in addition to its
intensity, determined by the type and distribution
of aversive consequences; their temporal relation
to the behavior to be modified; the strength with
which punished responses are concurrently

reinforced; the availability of alternative modes of
behavior for securing rewards; and the
characteristics of punishing agents (Bandura,
1969a). Depending on its nature and interaction
with other determinants, punishment can thus

increase, reduce, or have no appreciable effect on
aggressive behavior.
The relative strengths of the competing
instigative and inhibitory tendencies served as the
basis for explaining two other features of
aggressive behavior, namely, displacement of
aggression to substitute targets and change in the
form of the aggressive behavior. According to
Miller’s (1959) conflict model, the objects and

strength of displaced responses can be predicted
from knowledge of three variables—the strength
of aggressive instigation, the severity of
punishment of aggressive behavior, and the
similarity of alternative targets to the original

frustrator. In empirical tests of this paradigm it
was additionally assumed that both aggressive
and inhibitory tendencies developed toward the
original instigator were generalized to other
people as well, the extent of the generalization

being a function of the degree of similarity; and
that inhibitory responses were generalized less

than aggressive ones. It follows from this model
that the stronger the inhibitions, the further out on
the dissimilarity continuum the aggressive

response is likely to be displaced. Thus, aggression
will be diverted to other persons who are similar
to the original instigator when fear of punishment
is relatively weak, but to dissimilar people when it

is strong.
The process of response displacement is

depicted in the same way as victim displacement,
except that the generalization dimension
represents a continuum of directness of response

(for example, physical aggression, verbal
aggression, covert fantasy aggression) rather than
differing characteristics of people. The effect of
increasing the inhibitory tendency is to shift the
mode of aggression to more indirect forms.
Deductions from this model were evaluated on
the basis of findings derived from field studies of
aggression and from controlled laboratory
situations. It is difficult to assess the theoretical
implications of the field studies (Bandura and
Walters, 1959; Sears, Whiting, Nowlis, and Sears,
1953; Whiting and Child, 1953; Wright, 1954)
because the ordering and locating of targets of

aggression on the similarity dimension are
somewhat arbitrary (and usually post hoc), and
additional assumptions are often introduced to
explain discrepancies between findings and

predictions based on Miller's conflict model.
Moreover, selective aggressiveness cannot be fully
understood without knowledge of the patterns of
reinforcement associated with expression of
aggression toward different target persons.

Antisocially aggressive boys, for example, have
been found to have parents who strongly

disapprove of and punish aggression in the home
(Bandura, 1960; Bandura and Walters, 1959;

Glueck and Glueck, 1950), a finding that has
frequently been interpreted as an instance of

displacement. However, since parents of such boys
encourage and reward aggression outside the
home, the apparently displaced aggression may be
primarily an outcome of discrimination training

(Bandura, 1960; Bandura and Walters, 1959).
The conflict model was tested most

systematically in experimental studies of the
scapegoat theory of prejudice. Applied in this
context, the theory predicted that when a

frustrating agent is feared, aggression will be
displaced to people who differ from the social
group to which the frustrator belongs. Evidence
for this formulation is, at best, equivocal (Bandura
and Walters, 1963; Berkowitz, 1958, 1962). Nor

do informal observations of prejudicial behavior
appear to provide much support for derivations
from the conflict paradigm—namely, people will
be most prejudiced toward highly dissimilar
minorities when inhibitions are strong, but they
will divert their hostility to minorities that are
similar to the original frustrator if their inhibitions
are weak. Holding inhibitory tendencies constant,
the theory would also predict that the more

strongly a person is angered, the more likely he is
to displace his aggression to someone who is
similar to the instigator.
The displacement model has limited predictive
power when applied to the problem of human

aggression because it adopts a basically nonsocial
approach to a phenomenon that involves intricate
social learning. Prejudices, like other aggressive
acts, are developed through modeling and social

reinforcement, and they make their appearance
relatively early in a child’s life (Stevenson, 1967;
Radke-Yarrow, Trager, and Miller, 1952). Parents

and other influential people, through precept,
example, and their approving reactions, teach
children whom they should hate and for what
reasons, and how they should express their
aggression toward the stigmatized targets.
Reports of Ku Klux Klan rallies (San Francisco
Chronicle, 1963) provide illustrations of how fear

and hatred can be transmitted through modeling
and social sanction to the Klansmen’s observant
offspring. People similarly come to abuse
Irishmen, Italians, Poles, Jews, and other minority
groups as a result of the prevailing sanctions in

different communities, rather than on the basis of
their resemblance to frustrators. Indeed,
frustrators come in too many diverse forms to aid
much in victim selection.
Displaced aggression is further modified by the

reactions it elicits from the people who are
aggressed against. Acquiescent targets will be

victimized more often than retaliative ones.
Moreover, the theory makes no provision for self-
imposed restraints arising from self-devaluative
reactions for hurting others, which may be more
influential in controlling aggressive behavior than
inhibitions motivated by fear of external
punishment. One might expect the conflict

paradigm to prove most successful under
circumstances in which individuals have had no
prior experiences, either direct or vicarious, with
potential scapegoats who receive attacks but
otherwise do not counterreact, since this factor

would significantly alter the generalization
gradients. Such conditions rarely, if ever, obtain in
everyday life.
The preceding comments in no way question

the existence of aggression displacement. One can
readily cite examples in which an individual who
inhibits aggressing toward a feared antagonist

later mistreats another person without adequate
justification. A social learning interpretation of the
phenomenon, however, takes into account a
number of factors that are ignored in Miller’s
paradigm. The degree of provocation and
punitiveness of the frustrator is, of course,
relevant. In addition, however, accurate prediction

requires knowledge about the individual’s learned
responses to thwarting; about the types of
reactions modeled by influential figures to
potential victims; about the social sanctions for
aggressing toward different classes of people;

about the likelihood of counteraggression by
different victims; and about the self-evaluative
reactions the aggressor experiences whenever he
hurts people possessing certain characteristics.
Catharsis of aggression is another

hypothesized process that is assigned a prominent
role in drive theories. After the aggressive drive

has been aroused, it presumably remains active as
a motivating force until discharged by some form
of aggressive behavior. In accord with the original
formulation, theorists working within this
tradition assume that expression of aggression
reduces the urge or drive to behave aggressively.
This energy discharge is designated catharsis. The

theory further postulates that all forms of
aggression are functionally equivalent; that is, one
type of aggressive response can serve as a
substitute for another in reducing the aggressive
drive. Thus, for example, a person who was

blocked from striking someone who annoyed him
could drain the aroused drive by fantasizing
physical assaults or by observing others acting
violently.
Of the various propositions advanced by drive

theorists, those pertaining to catharsis have been
studied most extensively, perhaps because of their

important social implications for the management
and control of human aggression. The catharsis
issue was debated by the illustrious Greek savants
long before the advent of the frustration-
aggression hypothesis. Aristotle contended that
emotional displays purged emotions, whereas
Plato maintained that they aroused them. As in

most of their disputes over the nature of man,
Plato turned out to be the better psychologist. The
catharsis hypothesis has fared even less well than
other aspects of the drive theory. A large body of
research evidence shows that direct or vicarious

participation in aggressive activities tends to
maintain the behavior at its original level or to
increase the likelihood of subsequent aggression.
The relevant findings and their theoretical and
practical implications are fully discussed in a later

chapter.
The problem of catharsis has been further

reformulated (Feshbach, 1970) so that neither
increases, decreases, nor absence of changes in
aggressive behavior necessarily refute the
occurrence of cathartic discharges of an aggressive
drive. According to this revised position, direct or
vicarious participation in aggression can have
three separable effects that work in different

directions: It can reduce the aggressive drive; it
can reinforce aggressive responses; and it can
alter the strength of inhibitions. Under conditions
in which the aggression-augmenting effects of
reinforcement and disinhibition outweigh the

reductive effects of drive discharge, it could be
argued that catharsis occurred even though
aggression increased. Similarly, if these
hypothesized counteracting influences are of
equal strength, catharsis could be claimed on the

basis of no evident changes in the level of
aggressive behavior. Since there exist no adequate

measures of these subprocesses that are
independent of changes in aggressive behavior,
the catharsis hypothesis cannot be subjected to a

decisive empirical test.
In the most recent version of the drive theory,
Feshbach (1970) retains the notion that
frustrating events elicit an aggressive drive that

causes people to aggress. However, the properties
ascribed to the underlying aggressive drive differ

in several respects from traditional conceptions of
how an energetic drive system operates. After
defining hostile aggression as behavior for which
the goal is hurting a disliked person, Feshbach

states that infliction of injury is not really the
principal goal of most hostile aggression; rather,
producing pain in others serves to restore the

aggressor's self-esteem and sense of power. It
follows from this redefined purpose that esteem-

enhancing experiences could discharge the
aggressive drive without any aggressive behavior
ever being performed. Indeed, bolstering self-

esteem through positive accomplishments may
serve as a more direct and effective means of
reducing aggression than having people try to
derive esteem rewards from the suffering of
others. What remains to be explained by drive
theorists is how a nonaggressive activity

discharges an aggressive source of energy. In
theories employing drive concepts, drives are
characteristically altered by drive-related

activities. Conceptual ambiguities arise when
aggressive acts are attributed to a drive that is
claimed to be satisfied only by infliction of injury
—and then the actual purpose of aggression is

redefined as restoration of self-esteem.
A similar interpretive problem arises in

Feshbach’s observation that aggression is
frequently eliminated when people reinterpret
situations they had erroneously perceived as
insulting. Redefinition of a social situation can
eliminate further aggressive instigation, but what
happens to the aggressive drive that has already
been induced? Dollard and his collaborators
assumed that after the original instigation is
removed, the residual drive still remains and must

be reduced through some form of aggressive
behavior. The questions posed here do not dispute
the notion that esteem-enhancing experiences and
favorable redefinition of social events can
diminish aggressive behavior, the issue is whether

it is necessary to invoke an aggressive drive to
explain human aggression. Doing so creates more
conceptual problems than it explains.
A given theory prescribes the factors that will

be regarded as determinants of behavior and
those that will be excluded from consideration.
Since drive theories viewed aggression as

impelled by the action of an aggressive drive, a
number of important processes, including the
conditions governing acquisition of injurious
modes of behavior, the powerful influence of
response consequences in regulating behavior,
and symbolic activities that enable people to hurt
others without experiencing self-contempt, were

essentially ignored. Obviously, people are not
innately equipped with military combat
repertoires or with proficient means of insulting
others. A complete theory of aggression must
therefore explain how intricate behaviors that are

potentially injurious and destructive are learned
in the first place. Inasmuch as the frustration-
produced drive accounted for the occurrence of
aggressive behavior, the main function of response
consequences was a negative, inhibitory one.

Punishment was singled out as a factor that
inhibits, attenuates, and displaces aggression, but

the positive value of aggressive behavior in
attaining desired outcomes was overlooked.
Feshbach gives somewhat greater recognition to

learning variables than did his theoretical
forerunners. In his view, the frustration-
aggressive-drive relationship is innately furnished,
but the frustration-aggressive-behavior linkage

can be increased through reinforcement and
modeling. Evidence will be presented later to

show that aggression is generally better explained
in terms of its rewarding consequences than on
the basis of frustrative conditions and the
punishments it incurs. In effect, despite their

avowed relationship to learning principles, drive
theories failed to provide an adequate starting
point for a social learning approach to aggression.
Although the motivation forces for aggression

differ in instinctual and drive theories in terms of
whether they are innately supplied or externally
stimulated, they are much alike in their
implications for the regulation of aggressive
behavior. Since frustration is ever present in one
form or another, in both approaches man is
burdened with a continuous source of aggressive
energy that requires periodic release. And both
assume that aggression is reduced by behaving

aggressively.
Social Learning Theory
Until recently, most personality theories
depicted behavior as impelled by inner forces in
the form of needs, drives, and impulses, often

operating below the level of consciousness. Since
the principal causes of behavior resided in forces
within the individual, that is where one looked for

explanations of man’s actions. Although this view
enjoyed widespread professional and popular
acceptance, it did not go unchallenged.
Theories of this sort were criticized on both
conceptual and empirical grounds. Because the

inner determinants were typically inferred from
the behavior they supposedly caused, the result
was pseudoexplanations. Thus, for example, a
hostile impulse was deduced from a person’s
irascible behavior, which was then attributed to

the action of the inferred impulse. In a similar
manner, various traits and dynamics, which
represent the descriptive constructs of the
assessor, frequently become entities within the

individual that supposedly cause his behavior.
Different personality theories proposed diverse
lists of motivators, some containing a few all-
purpose drives, others embracing a varied
assortment of specific drives.
The conceptual structure of psychodynamic
theories was further criticized for disregarding the

tremendous complexity of human responsiveness.
An internal motivator cannot possibly account for

the marked variation in the incidence of a given
behavior in different situations, toward different

persons, at different times, and in different social
roles. One can predict with much greater accuracy
the expression of aggressive behavior from
knowledge of the social contexts (for example,

church, school, ghetto sidewalk, athletic
gymnasium), the targets (for example, parent,
priest, teacher, or peer), the role occupied by the
performer (for example, policeman, soldier,
teacher, sales clerk), and other cues that reliably

signify potential consequences for aggressive
actions, than from assessment of the performer
(Bandura, 1960; Bandura and Walters, 1959).
When diverse social influences produce
correspondingly diverse behaviors, the inner

cause implicated in the relationship cannot be less
complex than its effects.
It should be emphasized here that it is not the
existence of motivated behavior that is being
questioned, but rather whether such behavior is at
all explained by ascribing it to the action of drives
or other inner forces. The deficiencies of this type
of simplistic analysis can be illustrated by
considering a common activity, such as reading

printed matter, which has the qualities of a highly
motivated behavior. People spend large sums of
money purchasing reading material; they expend
effort obtaining it from libraries; they engage in
reading for prolonged periods; and they become

emotionally upset over being deprived of reading
material (as when their daily newspaper is not
delivered through an oversight or when a library
is unable to locate a desired book in its collection).
Following the common practice of inferring drives

from prepotent behaviors, one could ascribe the
activated reading to the force of a reading drive.

However, in predicting what a person reads, when
he chooses to read it, and the order in which
different contents are read, one would appeal not

to drives, but rather to the stimulus inducements
and consequences of reading, and to a variety of
specific factors that influence reading behavior. On
the stimulus side, one would want to know the

person’s reading assignments, their deadlines, the
type of information he requires to deal effectively

with projects he has undertaken, and the presence
of other reading instigators; on the consequences
side, knowledge about the contents the person
finds rewarding and those he dislikes and the

effects of reading or ignoring certain materials
would constitute important controlling influences.
There is a marked difference between ascribing

motivating properties to social inducements and
acquired incentives, a quality that is easily

demonstrable, and invoking acquired drives, a
concept that has been found lacking in explanatory
power (Bolles, 1967).
Although the conceptual adequacy of
psychodynamic formulations was debatable, their
empirical limitations could not be ignored
indefinitely. They provided intriguing
interpretations of events that had already

happened, but they lacked power to predict how
people would behave in given situations (Mischel,
1968). Moreover, it was difficult to demonstrate
that persons who had undergone
psychodynamically oriented treatment benefited

more than untreated cases (Bandura, 1969a;
Rachman, 1971). Acquisition of insight into the
underlying impulses, through which behavioral
changes were supposedly achieved, turned out to
represent more of a social conversion than a self-

discovery process. As Marmor (1962), among
others, pointed out, each psychodynamic approach

had its own favored set of inner causes and its
own preferred brand of insight. The presence of
these determinants could easily be confirmed

through suggestive proving and selective
reinforcement of clients’ verbal reports in self-
validating interviews. For these reasons,
advocates of differing theoretical orientations

repeatedly discovered their favorite motivating
agents, but rarely found evidence for the

hypothesized causes emphasized by proponents of
competing views. The content of a particular
client’s insights and emergent “unconscious” could
therefore be better predicted from knowledge of

the therapist’s belief system than from the client's
actual social learning history.
It eventually became apparent that if progress
in the understanding of human behavior was to be

accelerated, more stringent requirements would
have to be applied in evaluating the adequacy of
explanatory systems. Theories must demonstrate
predictive power, and they must accurately
identify causal factors, as shown by the fact that
varying the postulated determinants produces
corresponding changes in behavior.
The attribution of behavior to inner forces can
perhaps be likened to early explanatory schemes

in other branches of science. At one time diverse
chemical reactions were supposedly caused by
movements of a material substance called
phlogiston, physical objects were internally
propelled by intangible essences, and

physiological functioning was ascribed to the
action of humors.
Developments in learning theory shifted the
focus of causal analysis from hypothesized inner

determinants to detailed examination of external
influences on responsiveness. Human behavior
was extensively analyzed in terms of the stimulus
events that evoke it and the reinforcing
consequences that alter it. Researchers repeatedly
demonstrated that response patterns generally
attributed to underlying forces could be induced,
eliminated, and reinstated simply by varying
external sources of influence. These impressive

findings led many psychologists to the view that
the causes of behavior are found not in the
organism, but in environmental forces.
The idea that man’s actions are under external

control, though amply documented, was not
enthusiastically received for a variety of reasons.

To most people it unfortunately implied a one-way
influence process that reduced man to a helpless
reactor to the vagaries of external rewards and
punishments. Popular descriptions of the potential
for social influence conjured up macabre

associations of 1984 and Brave New World, in
which people are manipulated at will by occult

technocrats. By associating the term behaviorism
with odious images of salivating dogs and animals

driven by carrots and sticks, critics of behavioral
approaches skillfully employ Pavlovian
conditioning procedures on their receptive
audiences to endow this point of view with

degrading properties. The fact that valuation of
places, persons, or things is affected by one's
emotional experiences, whether they be fearful,
humiliating, disgusting, mournful, or pleasurable,
does not mean that such learning outcomes reflect

a base animal process. To expect people to remain
unaffected by paired experiences is to require that
they be less than human. Moreover, to be sensitive
to the consequences of one’s actions indicates
intelligence rather than subhuman functioning.

Nevertheless, promoters of various causes often
lobby against social practices they do not like by

designating them as Pavlovian or as harrowing
precursors of 1984.
The view that behavior is environmentally
determined also appeared to contradict firm but
ill-founded beliefs that people possess generalized
personality traits leading them to behave in a
consistent manner, however variable the social

influences might be. Fortunately for survival
purposes, cultural practices are much too diverse
to produce undiscerning generalized traits
(Mischel, 1968). A person who behaved in a
church the same way he acted in a nightclub

would be a forced recipient of an extended rest in
a psychiatric facility. A high degree of behavioral
flexibility is required if a person is to deal
effectively with the complexities of ever-changing
environmental demands.
A more valid criticism of extreme situational
determinism is that, in a vigorous effort to avoid

spurious inner causes, it neglected determinants
of man’s behavior arising from his cognitive
functioning. Man is a thinking organism
possessing capabilities that provide him with
some power of self-direction. People can represent
external influences symbolically and later use such
representations to guide their actions; they can

solve problems mentally without having to enact
the various alternatives; and they can foresee the
probable consequences of different actions and
alter their behavior accordingly. These higher
mental processes permit both insightful and

foresighted behavior. By managing the stimulus
determinants of given activities and producing
consequences for their own actions, people are
able to control their own behavior to some degree.
As will be illustrated later, cognitive and self-

regulative influences often serve important
functions in causal sequences. To the extent that

traditional behavioral theories could be faulted, it
was for providing an incomplete rather than an
inaccurate account of human behavior.
In the social learning view, man is neither
driven by inner forces nor buffeted helplessly by
environmental influences. Rather, psychological
functioning is best understood in terms of

continuous reciprocal interaction between
behavior and its controlling conditions. Early
attempts to incorporate both individual and
environmental determinants in personality theory
simply depicted behavior as caused by these two

sets of influences. The problem with this type of
formulation is that it treated response dispositions
and the environment as independent entities.
Contrary to this assumption, the environment is
only a potentiality, not a fixed property that

inevitably impinges upon individuals and to which
their behavior eventually adapts. Behavior partly

creates the environment and the resultant
environment, in turn, influences the behavior. In
this two-way causal process the environment is

influenceable, just as the behavior it controls is.
In examining social interactions, Rausch
(1965) found that the immediately preceding act
of one person was the major determinant of the

other person’s response. In approximately 75
percent of the instances, hostile behavior elicited

unfriendly responses, whereas cordial acts seldom
did. Aggressive children thus created through
their actions a hostile environment, whereas
children who displayed friendly interpersonal

modes of response generated an amicable social
milieu. With little effort one could readily identify
individuals who predictably create negative social

climates wherever they go. People thus play an
active role in constructing their own

reinforcement contingencies through their
characteristic modes of response.
A complete theory of aggression, whatever its
orientation, must explain how aggressive patterns
of behavior are developed, what provokes people
to behave aggressively, and what maintains their
aggressive actions. These major issues are treated
at length in subsequent chapters. The remainder

of this chapter is devoted to a general discussion
of social learning principles of how patterns of
behavior are acquired and how their expression is
continuously regulated by the interplay of self-
generated and external sources of influence.
Development of new modes of behavior.
Patterns of behavior can be acquired through
direct experience or by observing the behavior of

others. The more rudimentary form of learning,
rooted in direct experience, is largely governed by
the rewarding and punishing consequences that
follow any given action. People are repeatedly
confronted with situations with which they must
deal in one way or another. Some of the responses
they try prove unsuccessful, while others produce
more favorable effects. Through this process of
differential reinforcement, successful modes of
behavior are eventually selected from exploratory

activities, while ineffectual ones are discarded.
Although behavior can be shaped into new
patterns to some extent by rewarding and

punishing consequences, learning would be
exceedingly laborious and hazardous if it
proceeded solely on this basis. Most of the

intricate responses people display are learned,
either deliberately or inadvertently, through the
influence of example. Indeed, virtually all learning
phenomena resulting from direct experiences can
occur on a vicarious basis through observation of

other people’s behavior and its consequences for
them. Man’s capacity to learn by observation

enables him to acquire complex patterns of
behavior by watching the performances of

exemplary models. Emotional responses toward
certain places, persons, or things can also be
developed by witnessing the affective reactions of
others punished for their actions. And, finally, the

expression of previously learned responses can be
socially regulated through the actions of
influential models.
The preceding remarks are not meant to imply
that new modes of behavior are fashioned solely

through experience, either of a direct or
observational sort. Biological structure obviously
sets limits on the types of aggressive responses
that can be successfully perfected, and genetic

endowment influences the rate at which learning
progresses.
Regulatory functions.
A comprehensive theory of behavior must
explain not only how response patterns are
acquired, but how their expression is regulated
and maintained. In social learning theory, human
functioning relies on three regulatory systems.
They include antecedent inducements, response
feedback influences, and cognitive processes that

guide and regulate action. Human aggression is a
learned conduct that, like other forms of social

behavior, is under stimulus, reinforcement, and
cognitive control. These control functions will be
discussed separately for explanatory purposes,
although in reality they are closely interrelated.
Stimulus control.
To function effectively a person must be able to

anticipate the probable consequences of different
events and courses of action and regulate his

behavior accordingly. Without a capacity for
anticipatory or foresighted behavior, man could

not profit much from experience. Information
about probable consequences is conveyed by
environmental stimuli, such as verbal
communications; pictorial cues; distinctive places,
persons, or things; or the actions of others.
In the earliest years of development,
environmental events, except those that are
inherently painful, exert little or no influence on

infants and young children. As a result of paired
experiences, direct, symbolic, or vicarious,
formerly neutral stimuli begin to acquire

motivating and response-directive properties.
Environmental stimuli gain the capacity to activate
physiological reactions and emotional behavior
through association with evocative events. Such
learning often occurs on the basis of direct

experience. People come to fear and to avoid
individuals who are commonly associated in their

experience with pain or distress. Through a
similar learning process they become easily

angered by the sight or thought of individuals with
whom they have had hostile encounters.
Social characteristics generally acquire
evocative power through processes that are more

subtle and complex than is commonly believed.
Emotional responses are frequently acquired on
the basis of vicarious rather than direct
experiences. The emotional responses exhibited
by others toward certain people tend to arouse in

observers strong emotional reactions that can
become conditioned to the same targets. It is not
uncommon for unpopular minority groups or
nationalities to become endowed with anger-

evoking potency in the absence of personal
contact through exposure to modeled animosity.
Emotion-arousing words and pictures that conjure

up hostile reactions likewise often function as the
vehicle for symbolic conditioning of hatreds.
The emotional responses that become
established to paired events can be evoked by not
only direct experience, observation of another’s
affective expression, and symbolic stimuli, but also

by provocative thoughts. People can easily make
themselves nauseous by imagining revolting
experiences. They can become sexually aroused by
generating erotic fantasies. They can frighten

themselves by fear-provoking thoughts. And they
can work themselves up into a state of anger by
ruminating about mistreatment from offensive
provocateurs. The cognitive capacities of humans
thus enable them to invest things with positive or

negative valence by pairing them repeatedly with
thought-produced emotions. This self-arousal

process is illustrated in a husband’s slaying of a
friend who kissed his wife at a New Year’s Eve

party (Portland Press Herald, 1963). The husband
had brooded almost incessantly about the kiss

over a period of two years. As Thanksgiving
approached he further intensified his anger by
imagining his foe enjoying a family Thanksgiving
dinner while his own family life was irreparably

ruined. Shortly after observing how easy it was to
kill a man from seeing Oswald shot on television,

the brooding husband sought and shot his former
friend.
The preceding discussion has shown how

things can become invested with response-
activating properties on the basis of paired
experiences. Environmental cues also acquire
response-directing functions when they are
associated with differential response

consequences. The same actions can produce
markedly different results depending upon the
time, the place, and the persons toward whom
they are expressed. Insulting an irascible
aggressor, for example, will have painfully
different effects from treating a submissive
individual in the same manner. People therefore
pay close attention to cues that signify probable
consequences, and they partly regulate their

behavior on the basis of such information. Stimuli
indicating that given actions will be punished or
unrewarded tend to inhibit their performance,
whereas those signifying that the actions are
permissible or rewardable facilitate their

occurrence. The following quotation provides a
telling example of an autistic boy who freely
expressed destructive behavior with his lenient
mother but rarely did so in the presence of his
father, who tolerated no aggression.
Whenever her husband was home, Billy was

a model youngster. He knew that his father
would punish him quickly and
dispassionately for misbehaving. But when
his father left the house, Billy would go to the
window and watch until the car pulled out.
As soon as it did. he was suddenly
transformed. . . . “He’d go into my closet and
tear up my evening dresses and urinate on
my clothes. He’d smash furniture and run
around biting the walls until the house was
destruction from one end to the other. He
knew that I liked to dress him in nice clothes,
so he used to rip the buttons off his shirts,
and used to go in his pants.” (Moser, 1965, p.
96)
Social behavior is extensively regulated by
verbal cues. We influence people’s actions in

innumerable situations by suggestions, requests,
commands, and written directives. Often these

operate in subtle ways. To take a common
example, parents are quick to issue commands to
their children, but they do not always see to it that

their requests are heeded. Children are therefore
inclined to ignore demands voiced in mild or

moderate tones. The parents’ mounting anger
usually serves as the cue that they will enforce

compliance, so that only shouts produce results.
Because of the differential signal value of parental
directives, many households are run at a fairly
high decibel level.
Of the numerous cues that influence how

people will behave at any given moment, none is
more ubiquitous or effective than the actions of
others. People applaud when others clap, they exit
from social functions when they see others
leaving, they wear their hair like others, they dress

alike, and on countless other occasions their
behavior is prompted and channeled by the power
of example. Modeling influences play an especially
important role in the rapid contagion of

aggression. The actions of others acquire
response-directing properties through selective
reinforcement in much the same way as do

physical and symbolic cues in nonsocial forms.
When behaving like others produces rewarding
outcomes, modeling cues become powerful
determinants of analogous behavior; conversely,
when imitative actions are treated negatively but
dissimilar behavior proves rewarding, models’
responses prompt divergent performances in

observers.
In everyday life the likely consequences of a
given course of action depend upon the presence
of several factors, including combinations of
temporal, social, and situational features. It may

be permissible to aggress toward a peer during a
physical contact sport, for example, but punishable
to do so at other times or in other settings. When
subtle variations in stimulus events carry diverse

outcomes for similar conduct, it is easy to
misjudge what effects aggressive actions will have.
Moreover, the force of activating circumstances

may at times prompt a person to behave
aggressively without paying much heed to
subsequent consequences.
Reinforcement control.
A second control system involves behavior

feedback influences, mainly in the form of
reinforcing consequences. An organism that
responded foresightedly on the basis of
informative environmental cues but remained

unaffected by the results of its actions would be
too obtuse to survive for long. In fact, behavior is

extensively controlled by its consequences.
Responses that cause unrewarding or punishing
effects tend to be discarded, whereas those that

produce rewarding outcomes are retained and
strengthened. Human behavior therefore cannot
be fully understood without examining the
regulatory influence of reinforcement.
Reinforcement control of behavior is most

convincingly demonstrated by intrasubject
replication. In these types of studies interpersonal

modes of response, many of long standing, are
successively eliminated and reinstated by altering
the effects they produce. The susceptibility of
behavior to reinforcement control is further

shown by the fact that even subtle variations in
the frequency and patterning of outcomes result in
different types of performances. Those who have
been rewarded each time they respond are likely

to become easily discouraged and to give up
quickly when their efforts fail. On the other hand,
individuals whose behavior has been reinforced
intermittently tend to persist for a considerable
time despite setbacks and only occasional success.
In the minds of most people, reinforcement is
usually equated with tangible rewards and

punishments. Actually, human behavior is largely
sustained and modified by symbolic reinforcers.

As a result of repeated association with primary
experiences, social reactions in the form of verbal

approval, reprimands, attention, affection, and
rejection acquire powerful reinforcing functions.
Such interpersonal reinforcers assume a
prominent role in regulating the interactions of

everyday life, including aggressive responsiveness.
Similarly, after accomplishments become a source
of personal satisfaction, knowledge that one has
done well can function as a reward that sustains
activities independently of social or material

incentives. Many forms of behavior (such as
communicative and motor skills) that enable an
individual to deal effectively with his environment
likewise persist with little external support
because they produce desired results. Finally, the

sensory experiences that are naturally produced
by the behavior itself can be effective in modifying

and maintaining it over a long period. One rarely
has to resort to extrinsic rewards to get people to
read interesting books, to watch entertaining

television programs, or to play pleasing musical
selections.
Vicarious reinforcement.
Human functioning would be exceedingly
inefficient, not to say dangerous, if behavior were

influenced only by directly experienced
consequences. Fortunately, one can profit greatly
from the experiences of others. People repeatedly

observe the actions of others and the occasions on
which they are rewarded, ignored, or punished.
Observed reinforcement influences behavior in
much the same way as outcomes that are directly
experienced (Bandura, 1971b; Kanfer, 1965).

Observed rewards generally enhance, and
observed punishments reduce, similar behavior in
observers.
Observed consequences also provide reference
standards that determine whether particular

outcomes will assume positive or negative value.
The same compliment, for instance, is likely to be
punishing for persons who have seen similar
performances by others more highly acclaimed,
but rewarding when others have been less

generously praised. Thus, through social
comparison processes, observation of other

people’s response outcomes can drastically alter
the effectiveness of direct reinforcements. Since
both direct and vicarious reinforcement inevitably
occur together in everyday life, the interactive

effects of these two sources of influence on human
behavior are of much greater significance than
their independent controlling power.
Self-reinforcement.
At the highest level of psychological
functioning, individuals regulate their own

behavior by self-evaluative and other self-
produced consequences (Bandura, 1971b). In this
process people set themselves certain standards of
conduct and respond to their own behavior in self-
satisfied or self-critical ways in accordance with

their self-imposed demands. Comparative studies
show that people can manage their own behavior
by self-reinforcement as well as or better than
through consequences arising from external

sources.
After a self-monitored reinforcement system

has been established, a given action produces two
sets of consequences—a self-evaluative reaction
as well as some external outcome. These two
sources of reinforcement can occur in several
different patterns. Sometimes people are

rewarded socially or materially for behavior that
they devalue. Anticipation of self-reproach for

personally repudiated actions provides an
important motivating influence to keep behavior

in line with adopted standards in the face of
opposing influences. There is no more devastating
punishment than self-contempt. Under conditions
in which self-devaluative consequences outweigh

the force of rewards for accommodating behavior,
external influences prove relatively ineffective. On
the other hand, when external inducements,
whether rewarding or coercive, prevail over self-
reinforcing influences, individuals exhibit

cheerless compliance. Humans, of course, possess
facile cognitive capacities for reconciling
distressing conflicts. Devalued actions can be, and
often are, justified so that losses in self-respect are
minimized as long as the self-deception remains

convincing.
An opposite type of conflict between external

and self-produced consequences arises when
people are punished for engaging in activities they
value highly. The rebels, the dissenters, and the
nonconformists often experience this type of
counterinfluence. Here, the relative strengths of
self-approval and external censure determine
whether the behavior will be discarded or

maintained. If the negative sanctions are relatively
weak, socially disapproved but personally valued
actions will be readily expressed. If the threatened
consequences are severe, however, transgressive
behavior is apt to be inhibited under high risk of

penalty and freely performed in the absence of
prohibitive agents. There are individuals whose
sense of self-worth is so strongly linked to certain
convictions that they will submit to prolonged
maltreatment rather than accede to social

practices they regard as unjust.
External reinforcement exerts its greatest

influence when it is consonant with self-produced
consequences—as when rewarded actions are a
source of self-pride, and punished ones are self-
censured. Indeed, people strive actively to achieve
and to maintain such conditions. They do this by
selectively associating with persons who share
similar standards of conduct, thus ensuring social

support for their own system of self-evaluation.
Some of the most drastic changes in behavior are
achieved in large part by modifying a person’s
basis for self-evaluation. By legitimizing
aggression and dehumanizing potential victims,

individuals who have been strictly socialized
against brutalizing and slaying people' can be led
to do so in military situations without
experiencing tormenting self-devaluative
consequences. A later chapter presents the various

antidotes for self-contempt. Among those who
fully adopt the new standards of conduct, skillfully

executed carnage may even serve as a basis for
self-commendation.
Cognitive control.
If human behavior could be fully explained in
terms of external inducements and response
consequences, there would be no need to

postulate any additional regulatory mechanisms.
However, actions are not always predictable from

these external sources of influence. Man’s
cognitive capacities tremendously increase the
information he can derive from his experiences,
and thus partly determine how he will be affected

by them. There are several ways in which
cognitive functioning enters into the regulation of
human behavior. These are discussed next.
Cognitive representation of reinforcement

contingencies.
Popular portrayals of behavioral approaches
would lead one to believe that people can be easily

conditioned and manipulated without their
awareness. This melodramatic sketch receives
little support in numerous studies of how
awareness of reinforcement contingencies affects
the process of behavior change. In fact, repeated
paired stimulation generally fails to produce

conditioned emotional responses as long as the
connection between stimulus events goes
unnoticed. The responses that get conditioned are
to a large extent cognitively induced rather than

directly elicited by external stimuli. Response
consequences similarly have weak effects on
behavior when the relationship between one’s
actions and outcomes is not recognized. On the
other hand, awareness of conditions of

reinforcement typically results in rapid changes in
behavior, which is indicative of insightful

functioning. People who are aware of what is
wanted and who value the contingent rewards

change their behavior in the reinforced direction;
those who are equally aware of the reinforcement

contingencies but who devalue the required
behavior or the reinforcers show little change;
those who remain unaware achieve, at best, small
increments in performance even though the

appropriate responses are reinforced whenever
they occur.
Human behavior is regulated to a large extent
by anticipated consequences of prospective
actions. Individuals may accurately assess the

customary effects of given activities but fail to act
in accordance with existing conditions or
reinforcement because of hope that their actions
may eventually bring favorable results. Many a
social reformer has sustained his efforts in the face

of repeated failures by the belief that the rightness
of his cause will ultimately produce desired
changes. Sometimes people lead themselves astray
by inaccurate expectations when they wrongly
assume that certain changes in their behavior will
alter future consequences. The deterrent value of
threatened punishment, for example, is weakened
when lawbreakers misjudge their chances of
escaping apprehension for antisocial acts.

Judgment of consequences can also be seriously
impaired by intoxicants. Felonies are typically
committed (72 percent) under the influence of
alcohol (Shupe, 1954). Under laboratory
conditions drunk subjects are more willing to risk

punitive attack than sober ones (Shuntich and
Taylor, 1972).
The notion that behavior is controlled by its
immediate consequences holds up better under

close scrutiny for anticipated consequences than
for those that actually impinge upon the organism.
In most instances customary outcomes are

reasonably good predictors of behavior because
the consequences that people anticipate for their
actions are accurately derived from, and therefore
correspond closely to, prevailing conditions of
reinforcement. However, belief and actuality do
not always coincide because anticipated
consequences are also partly inferred from

observed response consequences of others, from
what one reads or is told, and from a variety of
other cues that, on the basis of past experiences,
are considered reliable forecasters of likely
outcomes. When actions are guided by anticipated

consequences derived from predictors that do not
accurately reflect existing contingencies of
reinforcement, behavior is weakly controlled by
its actual consequences until cumulative
experiences produce more realistic expectations.
In some of the more severe behavior disorders,
psychotic actions are so powerfully controlled by

bizarre subjective contingencies that the behavior
remains essentially unaffected by its external
consequences. This process is vividly illustrated in
a patient’s account of his psychotic experiences in
an insane asylum during the early nineteenth
century (Bateson, 1961). The narrator had
received a scrupulously moralistic upbringing,

according to which actions ordinarily viewed as
fully acceptable were judged by him to be sinful
and likely to provoke the wrath of God.
Consequently, many innocuous acts elicited
dreadful apprehensions of hellish torment; these

in turn motivated and maintained self-torturing
rituals for hours on end, designed to forestall the
imagined disastrous consequences. Reduction in
acute distress accompanying the nonoccurrence of
subjectively feared, but objectively nonexistent,

threats powerfully reinforced the self-punitive
defensive behavior. Given fictional contingencies

with a powerful internal reinforcing system, the
patient persisted in his atonement rituals even in
the face of severe external punishment and blatant

disconfirming experiences. The attendants’
punitive measures were pale compared to the
feared Hadean torture imaginally represented in
visions of clanking iron and massive flames fanned

by huge forge bellows. When the prophecies of
divine inner voices failed to materialize, the

disconfirming experiences were discounted as
tests by the Almighty of the strength of his
religious convictions.
Instances of bizarre thought control of

grotesque murders occur from time to time. In one
such case reported by Reich and Hepps (1972), a
student became convinced that his commune
companions were, in fact, conspirators from

another planet plotting to destroy his mind with
LSD. Seeking refuge in Europe, he fled in terror
from city to city in the belief that he was escaping
a Nazi scheme to exterminate him. Upon arriving
in Tel Aviv, he concluded that the Nazis had
already taken over and were gassing the
population. While sitting in a sidewalk cafe
resignedly drinking what he thought was poisoned
wine, he interpreted an overheard conversation

between Israeli soldiers as Nazis boasting about
their kill rates. Enraged by imagined massacres,
the student drew a knife, killing the soldier and
wounding two other persons.
Cognitive guidance of behavior.
Cognitive processes play a prominent role in
the acquisition and retention of response patterns
as well as in their expression. The memory trace of

momentary influences is short-lived, but such
experiences often have lasting behavioral effects.
This is made possible by the fact that transitory
external events are coded and stored in symbolic
form for memory representation. Patterns of
behavior that have been observed and other
experiences long past can thus be reinstated by
visualizing them or by representing them verbally.
These internal models of the outside world can
serve as guides to overt action on later occasions.

It will be recalled from the earlier discussion of
learning processes that internal representations of
patterned behavior are constructed from observed
examples and from informative feedback to one’s

trial-and-error performances.
Cognitive functioning is especially important in

observational learning in which a person reads
about or observes a pattern of behavior, but does
not perform it overtly until appropriate
circumstances arise. Evidence will be cited later to
illustrate how modeled activities are acquired in

symbolic form without behavioral enactment, how
they can be strengthened by mental rehearsal, and

how they provide the basis for action on later
occasions given suitable inducements.
Thought control of action through mental problem-
solving.
Man’s efforts to understand and to manage his

environment would be exceedingly wearisome if
optimal solutions to problems could be arrived at
only by performing alternative actions and
suffering the consequences. Actually, most
problem-solving occurs in thought rather than in

action. Man’s higher mental capacities, for
example, enable him to design sturdy dwellings
and bridges without having to build them until he
hits upon a structure that does not collapse.
Alternative courses of action are generally tested

in symbolic exploration and either discarded or
retained on the basis of calculated consequences.

The best symbolic solution is then executed in
action.
The three major systems by which behavior is
regulated do not operate independently; most
actions are simultaneously controlled by two or
more of the component influences. Moreover, the
various systems are closely interdependent in

acquiring and retaining their power to determine
behavior. In order to establish and to maintain
effective stimulus control, for example, the same
actions must produce different consequences

depending on the cues that are present. Stimulus
and cognitive influences, in turn, can alter the
impact of prevailing conditions of reinforcement.
Certain stimuli can acquire such powerful control
over defensive behavior that people avoid

renewed encounters with feared or hated persons,
places, or things. In instances in which the original

threats no longer exist, their self-protective
behavior is insulated from realistic reinforcement

influences.
Even when the things one dislikes or fears are

not completely avoided, cues having strong
emotion-arousing potential provoke defensive
behaviors that predictably create adverse
reinforcement contingencies where they may not

ordinarily exist. To the extent that an individual’s
distrust of certain people leads him to behave in
ways that provoke hostile counterreactions from
them, their negative valence is further
strengthened and it, in turn, prompts actions that

produce reciprocal negative reinforcement. Both
processes thus support each other.
The way in which beliefs and conscious
recognition of environmental contingencies can

enhance, distort, or even negate the influence of
reinforcing consequences has already been amply
documented and needs no further illustration.

Cognitive events, however, do not function as
autonomous causes of behavior. Their nature,
their emotion-arousing properties, and their
occurrence are under stimulus and reinforcement
control. Analysis of cognitive control of behavior is
therefore incomplete without specifying what
influences a person’s thoughts. Detailed exposition

of social learning theory falls beyond the scope of
this book, but the principles are presented in other
sources (Bandura, 1969a, 1971c).
DIFFERENTIAL IMPLICATIONS OF DRIVE

AND SOCIAL LEARNING THEORIES
A number of predictions that follow from the

social learning formulation differ from the
traditional frustration-aggression hypothesis. It
will be recalled that drive theories of aggression
assume that frustration arouses an aggressive

drive that can be reduced only through some form
of aggressive behavior. Frustration, in this view, is
a necessary and sufficient condition for

aggression. The diverse events subsumed under
the omnibus term frustration have one feature in
common—they are all aversive in varying degrees.
In social learning theory, rather than frustration
generating an aggressive drive, aversive treatment
produces a general state of emotional arousal that

can facilitate a variety of behaviors, depending on
the types of responses the person has learned for
coping with stress and their relative effectiveness.
When distressed, some people seek help and
support; others increase achievement strivings;

others show withdrawal and resignation; some
aggress; others experience heightened somatic
activity; still others anesthetize themselves against
a miserable existence with drugs or alcohol; and
most intensify constructive efforts to overcome

their adversities. The major differences among the
instinctual, reactive drive, and social learning

theories in the ways they conceptualize the
motivational component of aggression are
depicted schematically in Figure 1.1.
There are several lines of evidence that lend
FIGURE 1.1 Diagrammatic representation of motivational
determinants of aggression in instinct, reactive drive, and
social learning theories.
support to the social learning formulation.
Psychophysiological studies have been conducted
in which people undergo fear- and anger-

provoking experiences while changes in their
physiological reactions are simultaneously
recorded. Interpretation of findings is complicated
because no efforts have been made to ensure that
the two emotional experiences are equally intense.

Should differences be obtained, they may arise
from dissimilarities in the nature of the affective
experiences or in their relative intensities.
Moreover, as Duffy (1962) has noted, unless it has
been shown that dissimilar stimuli within the

same emotional class produce identical patterns of
physiological arousal, the generality of findings

yielded by a single fear stimulus and a single anger
stimulus is open to question.
In a study in which adults were either
frightened or treated in a rude manner, Ax (1953)
found a few subtle differences in somatic
reactions, but the physiological manifestations of
these two emotions were much alike. Schachter

(1957), using analogous methods, obtained
essentially similar cardiovascular responsiveness
to fear and anger stimulation, although on the
basis of subjective categorization of the data, he
extracts greater physiological specificity than the

actual findings warrant. If the emotions of fear and
anger had different physiological determinants,
injections of drugs that selectively induce the
distinguishing visceral reactions should produce
qualitatively different emotions and social

behavior. Although such an experiment has not
been performed, it would most likely reveal that

people do not rely on visceral cues for judging
what emotion they are experiencing.
It would appear from the available evidence
that fear and anger have similar physiological
correlates. Looking at the physiological records
alone, one could not distinguish whether the
individuals had been frightened or angered. The

varied array of emotions experienced
phenomenologically apparently stem from a
common diffuse state of emotional arousal rather
than from distinct drive states. It seems unlikely
that small differences in an otherwise identical

pattern of physiological arousal are sufficiently
distinguishable, if at all, to serve as the cues for
differentiating among diverse emotional states.
Whether people experience their emotional
arousal as fear, anger, euphoria, or some other

state depends not on particular somatic cues, but
on a number of external defining influences.
People judge their emotions partly from the
nature of the instigating conditions. Visceral
arousal generated by threat is experienced as fear,
arousal produced by thwarting is experienced as
anger, and that resulting from irretrievable loss of
valued objects as sorrow (Hunt, Cole, and Reis,
1958). Even the same source of physiological

arousal may be experienced as different emotions
depending on the emotional reactions of others.
Affective modeling cues can give definition to
emotional states of uncertain origin, or in
ambiguous situations where people know the

source of their arousal but are unsure how they
are supposed to react to it. Schachter and Singer
(1962) provide suggestive evidence that people
under drug-induced arousal, who do not know
what to attribute it to, experience the arousal as

aggression when others respond hostilely, but
they experience the same state as euphoria when

they see others behaving in a jocular manner.
Emotional states produced by environmental
stress are likewise susceptible to relabeling to

some extent (Ross, Rodin, and Zimbardo, 1969;
Nisbett and Schachter, 1966).
Evidence that the same physiological arousal

can be experienced subjectively as different kinds
of emotion has several implications for a theory of
aggression. First, it calls into question the validity

of invoking an inborn aggressive drive from
observations that infants display diffuse emotional
upset to physical restraints and painful

stimulation (Munroe, 1955; Feshbach, 1970).
Second, the general arousal model predicts that
under conditions in which individuals are prone to
behave aggressively, any source of emotional
arousal may enhance aggressive behavior. Results

of several experiments have a bearing on this
prediction. Zillmann (1971) exposed adults to
physiologically arousing aggressive scenes, to
erotic displays, or to nonarousing contents, and
then provided them with an opportunity to behave
punitively by shocking their antagonist. Both
aggressive and sex arousal enhanced punitiveness,
compared to nonaroused subjects. Indeed,
aggressiveness was determined more by level of

arousal than by its source. A series of ingenious
experiments by Tannenbaum (1972) designed to
test the general instigational effects of arousal
further demonstrate that nonaggressive sources of
arousal enhance aggressive behavior under

conditions in which this is the only mode of
response available. Geen and O’Neal (1969)
similarly found that arousing though nonirritating
noise increased subjects’ punitiveness, especially
if they had prior exposure to aggressive models.
It might be argued that high sex arousal and
noise have some frustrating properties. An

experiment performed by Christy, Gelfand, and
Hartmann (1971) therefore provides a more
definitive test of predictions from the frustration-
aggression and the arousal-prepotent response
formulations. After observing aggressive or
nonaggressive models, groups of children engaged
in competitive activities in which one member

consistently won and the other was a consistent
loser. Other pairs played noncompetitively.
Following these experiences, the children’s
imitative aggressiveness was measured.
Competitors showed more imitative aggression

than noncompetitors; however, victors increased
their aggression just as much as those who
experienced failure-induced frustration. These
findings are especially interesting, since Lorenz
(1966), among others, prescribes rivalrous

activities as a means of reducing aggressive
behavior.
There are, of course, certain limiting conditions
to the emotional facilitation of aggressive
behavior. Some sources of arousal, such as joyful
or rewarding experiences, tend to be incompatible
with hostile actions and may thus serve as
inhibitors rather than as instigators of aggressive
responses. Even more important, in everyday life

informative cues and response consequences
serve as influential regulators of aggressive
actions. By arranging social learning determinants
such that arousal-linked aggression is negatively
sanctioned but aggression without arousal is well

received, one could undoubtedly reverse the
relationship between physiological state and
action. In other words, arousal not only has the
capacity to facilitate whatever responses are
dominant in a given setting; through selective

reinforcement, it also acquires informative value
in determining the type of behavior that will be

activated. Hence, if arousal-linked aggression
repeatedly produces negative results, experienced
arousal is likely to reduce rather than to increase

the likelihood of aggression. Evidence will be
presented later to show that activation of
aggressive behavior is, in fact, better explained in
terms of anticipated consequences than in terms

of drive.
The role of emotional arousal in aggression is

magnified when aggression is studied as a function
of arousal-producing conditions within a
permissive context only. It was previously noted

that theories generally give disproportionate
attention to aversively motivated aggression. In
the social learning analysis of motivation,
incentives also constitute important impellers of
action. A great deal of aggression is prompted by

its anticipated benefits. Here, the instigator is the
pull of expected success, rather than the push of
aversive treatment. As depicted in Figure 1-1,
anticipatory functions provide a mechanism of
response selection from among a variety of
alternatives.
Under circumstances in which arousal
facilitates aggression, the social learning
formulation predicts that arousal decreased

through non-injurious means will reduce
aggression as well as, or even better than, acting

aggressively. In drive theories the actuated
aggressive drive presumably endures until
discharged by some form of aggressive activity.
From the social learning perspective, anger

arousal dissipates, but it can be repeatedly
regenerated on later occasions by ruminating on
anger-provoking incidents. By thinking about past

insulting treatment, a person can work himself
into a rage long after the initial emotional

reactions have subsided. Given this cognitive
capacity, it is not necessary to invoke a drive to
explain arousal that outlasts its original

instigators. Persistence of elevated anger arousal,
in this view, reflects self-generated arousal rather
than the existence of an undischarged reservoir of
aggressive energy. To illustrate the different
views, let us consider the example of a person who

becomes angered by an apparent social slight only
to discover that the invitation to the social

function has arrived in the next mail. He is likely to
show an immediate drop in anger arousal and
aggressiveness without having to assault or
denounce someone so as to reduce a lingering

aggressive drive.
A number of investigators have compared the
relative effectiveness of different ways of reducing
aggressive behavior. Mallick and McCandless

(1966) found that explaining why a provocateur
behaved obnoxiously reduced children’s
aggression toward him, whereas free expression
of physical aggression did not lessen their punitive
behavior. This finding is in agreement with
Kaufmann and Feshbach (1963), who reduced
students’ inclination to behave aggressively by
cognitively restructuring the provocative
situation.
Research to be cited later reveals that
ventilative therapies aimed at draining aggressive
drives by having people behave aggressively may
be inadvertently reinforcing their aggressive
tendencies. By contrast, social learning

treatments, which have proved successful
(Bandura, 1969a), help people from the outset to
acquire better ways of dealing with social
problems so that they have less to get angry about.

Aggressive behavior of long standing is reliably
reduced by a variety of approaches, including
modeling of alternative coping responses;

selective reinforcement in which aggressive
actions are nonrewarded while better solutions
are actively supported; eliminating fantasied
instigators of violent outbursts; developing
competencies that provide new sources of reward;
and reducing aversive social conditions.
There is a third implication of social learning

theory that differs from traditional views.
Frustration or anger arousal is a facilitative but
not a necessary condition for aggression.
Frustration is most likely to provoke aggression in
people who have learned to respond to aversive

treatment with aggressive attitudes and actions. In
an early study bearing on this issue, Davitz (1952)
demonstrated that following arbitrarily insulting
treatment, aggressively trained children behaved

more aggressively, whereas cooperatively trained
children behaved more cooperatively. Responses
to aversive treatment are also frequently

patterned on the characteristic actions of salient
models (Bandura and Walters, 1963).
In writings on the causes of aggression, factors
such as broken homes, parental rejection, adverse
socioeconomic conditions, thwarted strivings, and
intrapsychic conflicts are usually identified as

important determinants. If one examines the
antecedents that have been invoked for other
types of behavior disorders, such as drug
addiction, alcoholism, and schizophrenia,

essentially the same adverse conditions appear in
roughly comparable rates. Nor are adversity and
familial discord necessarily rare in the
backgrounds of people who develop patterns of
behavior that are valued in the larger society.

Findings of these etiological studies suggest that
aversive conditions increase the probability that

deviant patterns will arise, but they do not
determine what form the behavior will take.

Ghetto conditions, for example, probably produce
more alcoholism, drug addiction, and resignation

and withdrawal than aggression. One cannot
account for the presence of elaborate modes of
behavior on the basis of deficits or frustrative
conditions alone. Response patterns that are

socially labeled as deviant, like emulative conduct,
require the presence of appropriate social learning

influences for their acquisition and maintenance.
One can distinguish among aggressive acts,
emotional arousal experienced as anger, and

aggressive attitudes exemplified in negative
evaluations. No evidence is available that
behavioral, affective, and cognitive responses are
inherently linked in one way or another. They can
be coupled through common consequences and

dissociated by differential consequences. When
people are free to voice aggressive attitudes but
injurious acts are strictly prohibited,
correspondence between word and deed will be
low. It was previously shown how emotional
arousal can either facilitate or inhibit aggressive
acts depending upon its contingent outcomes.
Thus, people often experience anger without
taking any aggressive action; conversely, they can

be induced to perform injurious acts without
accompanying anger. Even the physiological
manifestations that are subjectively experienced
as anger arousal can be dissociated through
selective reinforcement. If individuals are

rewarded only when they increase both their
heart rate and blood pressure, these two
autonomic functions become linked, but if rewards
are given only when one response goes up and the
other goes down, these physiological reactions

become differentiated (Schwartz, 1972). Miller
(1969) provides further impressive

demonstrations of how specific autonomic
responses can be selectively increased or reduced
by differential reinforcement. Consonance

between affect and attitudes will likewise be
reduced should emotionally charged attitudes
produce negative results while calmly expressed
ones are well received.
In short, people do not have to be angered or
emotionally aroused to behave aggressively. A

culture can produce highly aggressive people,
while keeping frustration at a low level, by valuing
aggressive accomplishments, furnishing successful

aggressive models, and ensuring that aggressive
actions secure rewarding effects. Since aggression
does not originate internally and its social
determinants are alterable, social learning theory
holds a more optimistic view of man’s capacity to

reduce the level of human destructiveness.
chapter two
Origins of Aggression
People are not born with preformed
repertoires of aggressive behavior; they must
learn them in one way or another. Some of the
elementary forms of physical aggression can be
perfected with minimal guidance, but most
aggressive activities—dueling with switchblade
knives, sparring with opponents, engaging in

military combat, or indulging in vengeful ridicule
—entail intricate skills that require extensive

social learning. Various types of aggression can, of
course, be attempted without skill learning at a
high risk of serious or fatal injury, just as people
sometimes hazard treacherous mountain slopes
without knowing how to ski or venture into
perilous traffic before they have developed

proficient driving skills.
Examination of the origins of aggression must

consider not only the behavior of free-lancing
aggressors, but also that of professionals who are
authorized to use aggression as a means of social

control or who are officially trained for mass
destruction in the service of national policies.
Societies rely on military training establishments
rather than on innate response repertoires to
produce good fighters. It requires a great deal of
complex learning to develop efficient weapons of

destruction as well as the technical skills to use
them. Nor were the infamous extermination
procedures in Nazi concentration camps

instinctively bestowed; they were perfected in
laboratories using human victims as test subjects
(Wechsberg, 1967). In this chapter we will
examine the processes whereby aggressive

patterns of behavior are learned. Before turning to
substantive issues, however, the methods used to
discover laws of human behavior require some
discussion to clarify certain misunderstandings
about experimental inquiry.
Strategies of Research
A major purpose of psychological research,
whatever the approach may be, is to identify the

determinants of human behavior. The search for
the causes of behavior is complicated by a number
of factors. In everyday life a wide variety of
conditions impinges on individuals, and it is

therefore exceedingly difficult to unravel which of
them exert some influence and which are
inconsequential. There is always the danger that a
person who relies solely on observation of
naturally occurring events will wrongly conclude
that one or another of the supposedly causative

factors is the critical determinant when, in fact,
unidentified influences are responsible for the
observed effects. In addition, it is often difficult to
ascertain the direction of causal relationships
derived from covariations of events in the natural

environment. It has been found, for example, that
hyperaggressive boys regularly view more
televised violence than those who are disinclined
to behave aggressively. What is cause and what is
effect? One cannot determine on the basis of

correlational evidence alone whether aggressive
boys are attracted to displays of violence, or

whether repeated observation of televised
violence fosters aggressive behavior in viewers. To
verify that a causal relationship does exist and, if

so, whether it involves a unidirectional or a
reciprocal influence process requires controlled
studies in which potential determinants are
systematically varied and their effects are

measured. Interpretation of evident covariations
is further complicated because correlated events

may not be causally related at all; rather, both can
be co-effects of a third factor. Crime and physical
health, for instance, may be positively related
because both are affected by socioeconomic status.
For the reasons given above, causal
relationships can be established most reliably
through experimental inquiry. Under laboratory
conditions it is possible to hold constant a variety

of influences while others are varied and their
effects assessed. If this method is skillfully applied,
it is possible to verify with a high degree of
confidence whether or not an alleged determinant
causes people to behave in a certain manner.
Causal relationships are most impressively
demonstrated through replicative control, in
which a selected behavior is successively
produced and eliminated by introducing and

removing its controlling conditions. The potential
benefits of experimental inquiry do not derive
from whether the phenomenon under
investigation occurs inside or outside a laboratory,
but from the degree of control exercised over

concurrent determinants. Poorly designed
experiments involving uncontrolled or
confounded variables can therefore yield
erroneous generalizations, whereas carefully
regulated variation of conditions in natural

settings can provide valuable information.
Unfortunately, the advantages of controlled field

studies often go unrealized because faulty
procedures are more readily excused than in
laboratory investigations.
Findings from experimental studies are often
discounted on the grounds that the artificiality of
laboratory conditions precludes extrapolation to
everyday situations. This attitude reflects a

misunderstanding of the manner in which
knowledge is advanced. Experiments are not
intended to duplicate events as they occur in real
life, and they would lose their value if they did.
Laws are formulated on the basis of simulated

conditions and then evaluated in terms of how
well they enable one to predict and to control
phenomena as they occur under natural
circumstances. In other words, it is predictive
power, not likeness, that should be the guiding

criterion of research. For this reason, investigators
study basic processes rather than exact

reproductions of natural events. This view of
experimentation is taken for granted in all other
branches of science. Airliners are built on

aerodynamic principles developed largely in
artificial wind tunnels; bridges and skyscrapers
are erected on structural principles derived from
experiments that bear little resemblance to the

actual constructions; and knowledge about
physiological functioning is principally gained

from artificially induced changes, often in animals.
Indeed, preoccupation with matching actual
conditions can retard advancement of knowledge
—witness the demise of venturesome fliers who

tried to remain airborne by flapping wings
strapped to their arms in the likeness of soaring
birds.
Experiments that precisely duplicated events

as they occur naturally would have limited
informative value. This point might be illustrated
by considering how a scientist could best
determine the effects of brain damage on
behavior. In everyday life, brain lesions can occur
as a result of a large variety of adverse
circumstances, such as automobile accidents,
falling objects, or as a by-product of disease. The
brain lesions that nature provides vary

considerably in extent and location, thereby
precluding unequivocal generalizations
concerning their effects on behavior. Knowledge
would be more effectively advanced if, instead of
waiting for injuries to occur, an experimenter

were to “artificially” produce lesions through
surgical operations in which the locus and amount
of damage were systematically varied. One would
not expect him to recreate the misfortunes that
occur in nature—for example, by striking or

dropping his subjects on their heads, or infecting
them with syphilis and waiting for paresis to

develop. Of course, ethical considerations preclude
experimental induction of brain injury in humans;
researchers study subhuman organisms, on which

they perform controlled operations.
Experiments should be judged not on the basis
of artificiality criteria, but in terms of the
explanatory and predictive value of the principles

they yield. Air travelers rarely discount research
utilizing artificial models in wind tunnels provided

it produces reliable airliners; rather, they would
question the wisdom of researchers who insisted
on using actual airliners for discovering basic
aeronautical principles. Nor do they require that

airplanes flap their wings because that is the
manner in which flying occurs in nature. If the
processes governing behavior under natural

conditions are not created in the laboratory
situations, then, of course, the experimental

analyses will furnish little useful data.
The benefits of experimental inquiry often
cannot be fully realized in social science because,
for ethical and humane reasons, some types of
causal conditions may not be applied or, even if
the determinants were unobjectionable, certain
outcomes may be prohibited. A psychologist who
wishes to study the effects of anxiety on

intellectual functioning is debarred from using the
whole gamut of threats people experience in their
everyday lives. Experimental conditions therefore
can cover only a limited range of those found in
society, and usually the weaker forms. Nor are

researchers at liberty to test, through controlled
experimentation, the necessary and sufficient
conditions for producing homosexuality,
schizophrenia, alcoholism, and other conditions
that may be socially or personally detrimental. The

relatively slow progress in understanding human
behavior may perhaps be as much attributed to

justified limitations on the use of powerful
experimental methods as to the complexity of
human functioning.
As in medical science, behavioral phenomena
that cannot be created in humans are investigated
to some extent through animal experimentation.
Such studies can provide helpful information in

instances in which the behaviors are analogous
and their determinants are similar in animals and
in man. On the other hand, when conditions
governing a behavioral phenomenon differ across
species, as previously noted in the case of sexual

behavior, then prediction from animal
experimentation to humans may be misleading.
As long as conditions that obtain in everyday
situations cannot be produced in laboratory

settings, controlled investigations alone cannot
provide a full understanding of aggression.
Theoretically based naturalistic and longitudinal

studies are thus indispensable adjuncts to
laboratory methods. Consequently, it is important
to seek conditions in nature in which certain social
practices are present and others in which they are
absent and to note corresponding variations in
aggressive behavior. Ideally, a concerted attack on
the problem of aggression should include both

field and laboratory studies, and the latter should
be designed in such a way as to reproduce the
essential processes operative in real-life
phenomena. Results obtained from laboratory and
field studies will not always be directly

comparable, for controlled experiments often
reveal relationships that are obscured by
unnoticed contravening influences in everyday
circumstances. Positive results yielded by
laboratory studies not only increase confidence in

conclusions drawn from field studies, but
frequently elucidate causal processes and bring to

light important interaction effects, the precise
nature of which cannot be assessed from field
data. However, it should be noted that negative

results obtained in laboratory studies may
sometimes simply reflect the weakness and the
limited character of the experimental influences or
deficiencies in the way in which the effects are

measured.
Learning-Performance Distinction
The social learning theory of aggression

distinguishes between acquisition of behaviors
that have destructive and injurious potential and
factors that determine whether a person will

perform what he has learned. This distinction,
which is generally overlooked in discussions of
aggression, is emphasized because not all that
people learn is exhibited in their actions.
Discrepancies between learning and performance

are most likely to arise under conditions in which
the acquired behaviors have limited functional

value or carry high risk of punishment. The
importance of this distinction is illustrated in a
study of how imitative aggression is affected by
seeing response consequences experienced by the
aggressor (Bandura, 1965a).
Children observed a filmed model who
behaved in a novel physically and verbally
aggressive manner. In one condition, the model
was punished for displaying the aggressive
behavior; in the second, the model was praised

and rewarded for his actions; in the third
condition, the model experienced no evident
consequences. Compared to children who
observed the modeled aggression punished, those

who saw aggression either rewarded or without
consequences spontaneously performed a greater
variety of imitative aggressive responses.

Moreover, boys reproduced substantially more of
the model’s behavior than girls, the differences
being particularly marked when the model’s
aggressiveness was punished (Figure 2.1).
FIGURE 2.1 Mean number of different aggressive responses
adapted by children as a function of response consequences to
the model and positive incentives for the children to display all
the aggressive acts they had learned. (Bandura, 1965a)
Following the performance test, children in all
three groups were offered rewards for each
modeled response they could accurately

reproduce, in order to activate into performance
all the behavior they had acquired through

observation. As shown graphically in Figure 2.1,
the rewards completely eliminated the previous
performance differences, revealing an equivalent
amount of learning among children regardless of

how the model’s aggression had been treated. The
initially large sex differential in performed
aggression was likewise virtually eliminated when

the girls were provided with positive incentives to
behave aggressively.
The results, considered together, thus show
that boys, who are generally encouraged to
emulate feats of physical prowess, spontaneously
performed all they had learned when they saw
aggression well received. When models were
punished for their aggressive actions, boys
performed less than they had learned, but they
later readily exhibited additional imitative

responses when they produced rewarding results.
By contrast, girls, for whom physical aggression is
traditionally regarded as sex-inappropriate and
hence negatively sanctioned, kept much of what
they had learned to themselves, regardless of how

the male model’s behavior was treated. Their
learning was not manifested in action until they
received direct assurance that it was acceptable. In
predicting the occurrence of aggression, one
should be more concerned with predisposing

conditions than with predisposed individuals.
Given that aggressive modes of conduct have been

learned, social circumstances largely determine
whether and when they will be performed.
It might be mentioned in passing that sex
differences in modeling of aggression are
commonly attributed to deficits in masculine role
identification in girls. Results of the above study
indicate, however, that in many instances the

apparent lack of response capabilities primarily
reflects performance rather than learning deficits.
This conclusion is supported by findings of other
studies showing that low aggressive modeling by
females reflects differential inhibition rather than

differential learning of aggression (Dubanoski and
Parton, 1971; Madsen, 1968). With few
exceptions, boys spontaneously perform
substantially more imitative aggression than girls,
but learning tests yield small or no sex differences.

With simultaneous exposure to multiple models,
sex differences in observational learning of

aggression may arise because observers
selectively attend to same-sex models who display
different styles of behavior. When female models

behave aggressively, sex differences disappear
(Maccoby and Wilson, 1957).
The discussion thus far has centered on how

discrepancies between learning and doing result
from the inhibiting effects of adverse
consequences accompanying the actions. Many of

the things that people have learned are not
revealed behaviorally either because the
appropriate situations do not occur or because the

equipment needed to execute the acquired pattern
is lacking. Television viewers, for example, may
learn from Western and crime series gun-fighting
skills that are never exhibited because the viewers
do not possess firearms—or, if they do, the

occasion to use them does not arise.
Because of concern over the social
consequences of aggression, most researchers
have explored influences that lead people to
aggress toward each other, but the mechanisms
through which aggressive behavior is acquired in
the first place have received comparatively little
attention. Since the conditions determining the

acquisition of aggressive patterns differ
substantially from those governing their
subsequent performance, failure to distinguish
between these two processes has created some

confusion (to be discussed later) about the
implications of research findings and the optimal
methodologies for the study of human aggression.
Aggressive behavior is learned through
essentially the same processes as those regulating
the acquisition of any other form of behavior.
People learn by observation and by direct

experience. These two modes of learning are
analyzed next in some detail as they apply

specifically to the development of aggressive
patterns of behavior.
LEARNING THROUGH MODELING
One of the fundamental means by which new
behaviors are acquired and existing patterns are
modified entails modeling and vicarious

processes. It is evident from informal observation
that human behavior is to a large extent socially

transmitted, either deliberately or inadvertently,
through the behavioral examples provided by
influential models. Indeed, as Reichard (1938)
noted some years ago, in many languages, “the

word for ‘teach’ is the same as the word for ‘show.’
There are several reasons why modeling
influences play a paramount role in learning in
everyday life. When mistakes are costly or

dangerous, skillful performances can be
established without needless errors by providing
competent models who demonstrate how the

required activities should be performed. If
learning proceeded solely through direct
experience, most people would never survive their
formative years because mistakes often result in

fatal consequences. Some complex behaviors, of
course, can be produced only through the
influence of models. If children had no opportunity
to hear speech, for example, it would be virtually

impossible to teach them the linguistic skills that
constitute a language. Where certain forms of
behavior can be conveyed only by social cues,
modeling is an indispensable aspect of learning.
Even in instances in which it is possible to

establish new skills through other means, the
process of acquisition can be considerably

shortened by providing appropriate models
(Bandura and McDonald, 1963; Luchins and

Luchins, 1961).
Three Major Effects of Modeling Influences
Modeling influences can produce three kinds of
effects in observers, each of which explains
different aspects of aggression. First, as noted

above, observers can acquire new patterns of
behavior through observation. A second major

function of modeling influences is to strengthen or
to weaken inhibitions of behavior that observers
have previously learned. Inhibitory and
disinhibitory effects are largely determined by

observation of rewarding and punishing
consequences accompanying models’ responses.
The actions of others also serve as social prompts

that facilitate similar behavior in observers.
Response facilitation effects can be distinguished

from observational learning and disinhibition by
the fact that no new responses are acquired, and
the appearance of analogous actions is not

attributable to weakening of inhibitions because
the behavior is socially acceptable and hence
unencumbered by restraints. Observational
learning of aggression is analyzed in the present
chapter; disinhibition and social facilitation are
discussed in a later chapter concerned with

instigators of aggressive actions.
Conceptualization of Observational Learning
Learning by example depends upon four
interrelated subprocesses. Before discussing the
specific role played by modeling influences in the

development of aggression, the four subsystems
governing observational learning will be reviewed
briefly because they specify the conditions that
must obtain if the examples set by others are to
have enduring behavioral effects.
Attentional processes.
A person cannot learn much by observation if
he does not attend to, or recognize, the important

features of the model’s behavior. One of the main
component functions in observational learning
therefore involves attentional processes. Exposure
to models does not in itself ensure that people will
attend closely to them, that they will necessarily

select from models’ numerous characteristics the
most relevant ones, or that they will even perceive

accurately the aspects they happen to notice.
Among the numerous factors that determine
observational experiences, a person’s
associational preferences are undoubtedly of

major importance. The people with whom a
person regularly associates delimit the types of
behavior that he will repeatedly observe and
hence learn most thoroughly. Opportunities for
learning aggressive modes of behavior obviously

differ markedly for affiliates of delinquent gangs
and of Quaker groups.
Within any social group, certain members are
more likely to be selected as sources of behavior
than others. The functional value of the behaviors
displayed by different models is highly influential

in determining which models will be closely
observed and which will be ignored. A model’s
efficacy is inferred partly from tangible evidence
of the results his actions typically produce and

partly from status-conferring symbols that signify
competence and past successes. The behavior of
models who possess high status in prestige,
power, and competence hierarchies is more likely
to be successful and therefore to command greater

attention from others than the behavior of models
who are socially, occupationally, and intellectually

inept.
Attention to models is channeled not only by

the utilitarian value of their behavior, but also by
their interpersonal attraction. Models who possess

interesting and winsome qualities are actively
sought, whereas those who lack rewarding
characteristics tend to be ignored or actively
rejected, even though they may excel in other

ways. Control of attention through rewarding
qualities is perhaps nowhere better illustrated
than in televised modeling. Most televised
stimulation commands the attention of people of
all ages for extended periods. Indeed, models

presented in televised form are so effective in
holding attention that viewers learn the depicted
behavior regardless of whether or not they are
given extra incentives to do so (Bandura, Grusec,
and Menlove, 1966). Psychological characteristics

of viewers also exert selective influence on what
they are most likely to attend to.
Observational learning is not a passive
receptive process in which people simply absorb
the vast array of models they encounter in their
daily interactions. When people are exposed to a
variety of models, as is invariably the case, they
rarely confine their learning to a single source, nor
do they reproduce all the characteristics even of

preferred models. In experiments in which
multiple models display different behaviors, the
imitative patterns of observers generally
represent amalgams of elements from the
different models (Bandura, Ross, and Ross,

1963b). Moreover, the particular admixture of
behavioral attributes varies from person to
person. Conflicting modeling influences arising
from adult and peer sources are likewise more apt
to produce composite patterns than exclusively

adult or peer identifiers (Bandura, Grusec, and
Menlove, 1966).
Traditional theories of personality generally
depict identification as a pervasive outcome that is
firmly established early in the child’s life in
relation to parental figures and that then
predetermines the direction of later development.
Research findings reveal that modeling influences
are considerably more complicated. Modeling is a

continuous process in which new behaviors are
acquired and existing patterns to some extent
modified by exposure to influences from diverse
actual and symbolic models at all periods of life
(Bandura, 1969b).
Retention processes.
A person cannot be much influenced by the
behavior of a model if he has no memory of it.

Another requisite function involved in
observational learning, therefore, concerns long-
term retention of activities that have been
modeled at one time or another. In order to
reproduce social behavior when the model is no
longer present to serve as a guide, the response
patterns must be represented in memory in
symbolic form. Past events can achieve some
permanence by being represented in images or in
descriptive verbal symbols for future use. After

modeled behaviors have been so transformed,
these memory codes function as internal guides
for imitative actions. Observers who preserve
modeled activities in words and vivid imagery

learn and retain the behavior better than those
who passively observe or are mentally
preoccupied with other matters while watching
the performances of others (Bandura, Grusec, and
Menlove, 1966; Bandura and Jeffery, 1972; Gerst,

1971).
In addition to symbolic coding, rehearsal also

serves as an important memory aid. People who
mentally rehearse or enact modeled patterns of

behavior are much less likely to forget them than
those who neither think about nor practice what

they have seen. Most violent forms of behavior
cannot be overtly practiced because of social
prohibitions. It is therefore of considerable import
that mental rehearsal, which can be readily

engaged in when behavioral enactment is either
impeded or impracticable, may increase retention
of modeled behavior.
Motor reproduction processes.
The third component of modeling is concerned

with behavioral enactment of what one has
learned. To achieve behavioral reproduction, a
learner must put together a given set of responses

according to the modeled patterns. The amount of
observational learning that a person can exhibit
behaviorally depends on whether or not he has
the required component skills. If he has the
subskills, modeled behavior can be more faithfully
enacted than if they are lacking. In the latter
instance, the symbolic learning remains in
abeyance until prerequisite skills are developed
through further observation or practice.
Even though symbolic representations of

modeled activities are developed and retained,
behavioral enactment may be impeded because
individuals do not have the physical capabilities

with which to carry out the necessary activities. A
young child can learn observationally the behavior
for driving an automobile and be adept at

executing the component responses, but if he is
too short to operate the controls, he will be unable
to coordinate the action sequences needed to
maneuver a vehicle successfully.
Reinforcement and motivational processes.
A person can acquire, retain, and possess the

capabilities for skillful execution of modeled
behavior, but the learning may rarely be activated
into overt performance if it is negatively
sanctioned or otherwise unfavorably received.
When positive incentives are introduced,
observational learning that previously remained
unexpressed is likely to emerge in action.

Reinforcement influences not only regulate the
overt expression of matching behavior, but also

affect observational learning by determining the
amount of attention people will pay to various
models they encounter in their everyday life.
Moreover, incentives aid selective retention of

what has been learned by motivating observers to
code and to rehearse modeled responses that have
high functional value for them.
For reasons given above, exposure to models,
even prestigious ones, does not automatically

produce matching performances. In any given
instance absence of imitative behavior may result
from faulty observation, retention losses due to
inadequate symbolic representation and
rehearsal, motor deficiencies, or simply
unwillingness to perform the exemplified behavior
because of its unfavorable consequences.
Experimental Analysis of Aggressive Modeling
Social transmission of aggression through the
power of example has been most clearly
demonstrated in controlled experimental
situations. Typically these studies make use of a

modeling paradigm originally employed by
Bandura, Ross, and Ross (1961) in which children
observe models behaving in a physically and

verbally aggressive manner toward a large inflated
plastic figure. Since the studies are mainly
concerned with issues of learning, the aggressive
acts modeled are those that are rarely, if ever,

displayed by children who have had no exposure
to the modeled performances. In the general
procedure, after addressing the figure
belligerently, the model pummels it on the head
with a mallet, hurls it down, sits on it and punches
it in the nose repeatedly, kicks it across the room,

flings it into the air, and bombards it with balls.
These physically assaultive behaviors are
accompanied by distinctive hostile remarks.
Following exposure to the modeling influence,
children are provided with opportunities to

display what they have learned in a situation
containing a variety of materials that can be used
either for aggressive or nonaggressive purposes.
Learning effects are measured either by recording
the children’s spontaneous behavior or by asking

them to reproduce all the modeled aggressive
actions and remarks they can recall (Bandura,

1965a). The latter measure provides the better
index of observational learning because people
generally learn more than they spontaneously

perform.
In technologically developed societies,
behavior is usually modeled in a variety of forms.
Much social learning is fostered through the

examples set by individuals one encounters in
everyday life. People also pattern their behavior

after symbolic models they read about or see in
audiovisual displays. As linguistic competencies
are acquired, written accounts describing the
actions and successes of others can serve as guides

for new modes of conduct. Another influential
source of social learning, at all age levels, is the
abundant and diverse modeling provided in

television. Both children and adults can acquire
attitudes, emotional responses, and complex

patterns of behavior through exposure to
pictorially presented models. Indeed, comparative
studies (Bandura and Mischel, 1965; Bandura,

Ross, and Ross, 1963a; Prentice, 1972) have
shown that people can learn equal amounts from
behavioral demonstration, pictorial
representation, and verbal description, provided
that they convey the same amount of response
information, that they are equally effective in

commanding attention, and that the learners are
sufficiently adept at processing information
transmitted by these alternative modes of

representation. When some of these conditions do
not obtain, aggression is learned more thoroughly
from observed actions than from verbal
descriptions (Grusec, 1972). Since large audiences

in widely dispersed areas can be reached
simultaneously through symbolic modeling, the
aggression contagion potential of media
presentations is greater than that of direct

behavioral modeling.
One of the early analyses of observational
learning of aggression (Bandura, 1962; Bandura,
Ross, and Ross, 1963a) examined the relative
potency of aggressive models presented in
different forms. Nursery school children were
matched individually in terms of their

interpersonal aggressiveness and assigned to one
of five conditions. One group observed adult
models behaving aggressively toward the plastic
figure. A second group saw a film of the same
models performing the same aggressive acts. Half

the children in each condition saw aggressive
models of the same sex, the remainder were
exposed to opposite sex models. A third group
observed the model costumed as a cartoon cat
enacting the same aggressive responses on the

screen of a television console. This condition was
included to test the notion that the more remote

the models are from reality, the weaker is the
tendency for children to imitate their behavior. In
addition to the three modeling treatments, two

control groups were included. The behavior of one
group of children was measured without any prior
exposure to the models to provide a baseline for
the amount and form of aggression that children

display in the same test situation when they have
not experienced the modeling influence. It is

conceivable that merely seeing the aggressive
materials in the modeling situation could later
increase children's use of them, quite apart from
the model’s actions. Therefore, in the fifth

condition, the filmed models behaved in a calm,
nonaggressive manner and did not handle the
aggressive materials that were visibly displayed.
After the exposure, children in all groups were

mildly frustrated and were then individually
brought to a different situation containing a
variety of play materials, where their behavior
was recorded. The major findings are summarized
graphically in Figure 2.2.
FIGURE 2.2 Mean imitative and total aggression performed by
children who were either exposed to aggressive models,
nonaggressive models, or who observed no models. (Plotted
from data by Bandura, Ross, and Ross, 1963a)
Results of this study show that exposure to
aggressive models had two important effects on
viewers. First, it taught them new ways of

aggressing. Most of the children who had observed
the aggressive models later emulated their novel

assaultive behavior and hostile remarks, whereas
these unusual aggressive acts were rarely
exhibited by children in the control groups.
Illustrations of how children patterned their

aggressive actions after the model’s example are
presented in Figure 2.3. The top frames show the
female model performing four uncommon

aggressive responses; the lower frames depict a
boy and a girl reproducing the behavior of the

female model whom they had observed in the film
presentation. Further analyses showed that a
person displaying aggression on film was as

influential in teaching distinctive forms of
aggression as one exhibiting it in real life. The
FIGURE 2.3 Photographs of children imitating the aggressive
behavior of the female model they had observed on film.
(Bandura, Ross, and Ross, 1963a)
children were less inclined, however, to imitate
the cartoon character than the real-life model.
Aggressive models not only furnished
examples for learning, they also reduced the
children’s inhibitions against performing

aggressive acts that they had previously learned
but that were never modeled in the experiment
itself. These nonimitative aggressive actions

included such things as spanking and shooting
dolls, killing animals, smashing automobiles, and
other assaultive behaviors, sometimes
accompanied by the children’s own vituperative
remarks. As shown in Figure 2.2, children who

were exposed to the aggressive models
subsequently exhibited substantially more total
aggression than children in the nonaggressive
model condition or the control group.
Interestingly, the cartoon model served as a
somewhat weaker teacher but an equally effective
disinhibitor of aggression compared with the live
and filmed counterparts. In addition, children who
observed the nonaggressive adult displayed the

restrained behavior characteristic of their model
and expressed significantly less aggression than
the no-model controls.
It was previously noted that cognitive
representation of modeled responses aids in their

acquisition and retention. A study by Bandura,
Grusec, and Menlove (1966) specifically
demonstrates that translating modeled actions
into words facilitates observational learning of

aggression. Children saw a filmed model perform a
series of novel actions, some of which were
physically aggressive. During the presentation,

different groups of children simply observed the
model’s behavior attentively, verbalized every
aggressive action as the model performed it, or
were mentally preoccupied with other matters
while watching the film to prevent them from
coding the modeled responses into symbolic
equivalents. After the exposure, children, were

asked to perform all the modeled aggressive
responses they could recall. Those who converted
the physical actions into words for memory
representation retained more aggressive
responses than the nonverbalizing viewers, who,

in turn, showed a higher level of aggression
learning than children who were mentally
preoccupied while observing the modeled
performances. Verbalizers reproduced a high
proportion (62 percent) of the modeled aggressive

responses that they had preserved in words,
whereas noncoded responses were much less

likely to be retrieved (26 percent). It was also
interesting to find that, although boys and girls
coded the same number of aggressive actions, girls

retrieved somewhat fewer responses. Further
research is needed to determine whether girls
have poorer memory than boys for aggressive
contents, whether they are more inclined to

inhibit symbolic codes of aggressive actions
because they are sex-role incongruent, or whether

they are simply less disposed to display such
behavior.
Whether aggressive responses acquired

through modeling are retained or lost with the
passage of time also depends on the extent to
which rehearsal memory aids are employed. If a
person casually observed a particular form of
aggression but never practiced it or thought about

it again, it is doubtful that lie would retain it in
memory for long. On the other hand, the contents
of observational learning can become firmly
established through either mental rehearsal or
overt enactment of modeled responses. It might be
noted here that assassins in some of the
sensational mass slayings originally got the idea
from reports of a mass killing. The incident
remained salient in their thinking long after it had

been forgotten by others, and it was repeatedly
revivified and elaborated until, under appropriate
instigating conditions, it served as the basis for an
analogous murderous action. Robert Smith, an
adolescent who shot four women and a young

child in a beauty salon, reported being affected in
this way by accounts of Speck’s murder of eight
nurses and Whitman’s mass killing of students
from atop a university tower {The New York Times,
1966d).
Relatively little research has been conducted
on retention of observationally learned aggression

over long periods, and even less on the factors that
aid memory for past experiences. Hicks (1965)
provides some evidence of enduring retention of
modeled aggression in a study comparing models
varying in age and sex as transmitters of
aggressive modes of behavior. Children observed
adult or peer aggressors presented on film and

their imitative behavior was measured
immediately after exposure and again six months
later. In the postexposure test, groups that
received the modeling influence performed more
imitative aggression than the controls, who did not

display a single imitative response. The different
models had a comparable effect except for the
male peer model, who produced the highest level
of imitative aggression. When the children were
retested six months later without any reference to

the previous modeling experience, they
spontaneously performed fewer imitative

responses and it was the male adult model who
had the most enduring effect. However, when the
children were explicitly asked to demonstrate all

they remembered, they revealed more of the
modeled aggression than they displayed in action.
In a second study, Hicks (1968a) reports that
children were able to reproduce more than 70

percent of modeled aggressive responses shortly
after exposure, and still retained about 40 percent

of them eight months later.
The surprising finding of these experiments is
not that some decrease occurred in

observationally learned responses over time, but
that children had any memory at all of a brief
modeling experience that one would not expect to
be especially salient or meaningful to them. There
is some evidence, however, that filmed violence,

particularly in realistic forms, is emotionally
arousing to young children and that they retain
substantially more aggressive than nonaggressive
content (Osborn and Endsley, 1971). Since one is
more apt to think about arousing than uninvolving
events, emotionality probably enhances retention
through its effects on symbolic rehearsal. In
everyday life, of course, because people repeatedly
observe the same models demonstrating their

preferred modes of coping with diverse
environmental demands, they have ample
opportunities to relearn whatever they may have
forgotten. If the model’s behavior appears to have
functional value, as it often does, observers have

strong incentives to practice the modeled patterns
and to overlearn them. Conditions of observational
learning combining repeated exposure with
opportunities for overt practice and symbolic
rehearsal ensure more or less permanent

retention of modeled activities.
Modeling of aggressive behavior has been

studied as a function of a number of other
variables. In some of the earlier experiments,
when it was commonly believed that frustration
was necessary for the occurrence of aggression,
children were mildly frustrated after exposure to
aggressive models. Kuhn, Madsen, and Becker
(1967) found that modeling alone produced

significant increases in aggression, whereas
frustration, which took the form of arbitrary
criticism and delay of reward, had no effect. If
anything, frustration appeared to disrupt rather
than to augment the effects of the prior modeling

influence. This pattern of results is consistent with
the view that frustration can have diverse effects
and that it may inhibit responsiveness in young
children, especially if it contains aversive features.
Experiments reported by Nelson, Gelfand, and

Hartmann (1969) illustrate how prior emotional
arousal may affect responsiveness to aggressive

modeling. Children either succeeded or failed in
competitive activities, or they engaged in
noncompetitive play. After these experiences they
observed a male model behaving either
aggressively or nonaggressively. Children who
were losers in competitive play were most prone
to adopt aggressive behavior modeled by others;

winners exhibited an intermediate level of
imitative aggressiveness; children who observed
the nonaggressive model showed virtually none of
the novel assaultive responses regardless of
whether they were winners, losers, or

noncompetitors (Figure 2.4).
Modeling and competitiveness had similar
effects on nonimitative aggression, though the
differences are less striking and more dependent

on sex characteristics than in the case of imitative
responsiveness. The nonimitative aggression of
boys seemed to be most affected by

competitiveness and that of girls by aggressive
modeling.
In a subsequent experiment designed to test
FIGURE 2.4 Mean imitative aggressive responses performed by
children exposed to aggressive and nonaggressive models
after they either succeeded or failed in competitive activities,
or engaged in noncompetitive play. (Plotted from data by
Nelson, Gelfand, and Hartmann, 1969)
predictions from the frustration-aggression theory
and the arousal-prepotent response theory,
Christy, Gelfand, and Hartmann (1971) had

children participate in competitive or
noncompetitive games, after which they were
exposed to models who were aggressive, active
but nonaggressive, or restrained in their behavior.

Competitive experiences enhanced imitation of the
types of behavior the children saw modeled.
Compared to noncompetitors, the competitors
behaved more aggressively when exposed to
aggressive models and more energetically when

exposed to active models, but they showed low
aggressive and animated responsiveness when

they were provided with restrained models. The
effects of failure-induced frustration were thus

largely determined by the behavioral examples set
by models. Moreover, repeated success in

rivalrous situations, in conjunction with modeling
influences, enhanced both aggressive and
nonaggressive behavior analogously to repeated
failure. These findings are in closer agreement

with derivations from social learning than from
frustration-aggression formulations.
Another set of studies employing the standard
modeling paradigm has explored the effects of
social sanctions and response consequences to the

model on children’s imitative aggression.
Observed punishment of assaultive actions
generally reduces imitative aggression, whereas
observed reward or the absence of adverse
consequences tends to facilitate imitative

responding (Bandura, 1965a; Bandura, Ross and
Ross, 1963b; Rosekrans and Hartup, 1967). The
influences of vicarious reinforcement on modeling
are discussed at length in Chapter 4, which is
primarily concerned with factors regulating the
expression of aggression.
The role of sanctions in the performance of
modeled aggression is elucidated by Hicks
(1968b). While children observed a televised male

model soundly pummeling his adversary, an adult,
who also watched the program, responded

approvingly, disapprovingly, or made no
evaluative comments regarding the model’s
conduct. The behavioral evaluations had a
substantial effect on children’s subsequent

imitative aggression provided the sanctioner was
present in the test situation; Those who previously
heard him praise the aggressor displayed a high

level of imitative aggression (mean = 35); those
who heard him condemn the aggressor seldom

behaved in a similar manner (mean = 7); and
when the adult had shown no reaction, children
performed a moderately high amount of imitative

aggression (mean = 25). By contrast, when the
sanctioning adult was absent from the test
situation, children paid no heed to his evaluative
judgments and exhibited a similar amount of
imitative aggression regardless of whether he had

approved, condemned, or been indifferent toward
the model’s actions.
Most of the studies reviewed in this section
probably underestimate the amount of aggression
that was actually learned from the examples

provided by models. With few exceptions,
observational learning was measured in terms of
the number of matching aggressive responses
children were willing to perform spontaneously.
As previously noted, people learn more than they

usually perform, unless given positive incentives
to do so. Therefore, the most precise measure of
observational learning is obtained by explicitly
instructing individuals, under optimal incentive
conditions, to demonstrate all the responses they
acquired from observing the modeled
performances.
Some of the variables that have been shown to
influence the incidence of imitative aggression are

most likely operating on performance of
observationally acquired behavior rather than on
the learning process itself. This is shown in the
experiment referred to earlier (Bandura, 1965a),
in which learning and performance of modeled

aggression were examined as a function of sex
characteristics of observers and response
consequences to the models. Results of a study by
Madsen (1968) also have a bearing on this issue.

Children were much more inclined to imitate a
familiar aggressive model than an unfamiliar one.
This was especially true of boys, who performed

approximately three times as many matching
responses as the girls. On the other hand, a prior
nurturant relationship with the model, extending
over a period of six weeks, had no appreciable
effect on degree of aggressive modeling. A test for
learning in which children were encouraged and
rewarded for enacting the modeled patterns they

had previously observed disclosed an equivalent
amount of observational learning irrespective of
model familiarity and nurturance; nor did boys
and girls differ in this regard.
Kniveton and Stephenson (1970) tested the

hypothesis that children would be more inclined
to display imitative aggression in an unfamiliar
environment than in a familiar one on the grounds
that previous experiences create competing

interests that “inoculate the observers against the
model’s example.” Children who watched
aggressive modeling, regardless of whether or not

they had previously played in the situation
portrayed in the film, performed more imitative
aggression than control children who had received
the familiarizing preexperience without exposure
to filmed aggression. However, aggressive
modeling without preexperience produced the
highest imitative aggression. The authors report

that the effect of the filmed influence was short-
lived, but their findings actually support the
opposite conclusion.
Control children in the foregoing experiment
displayed a surprisingly high initial level of

imitative aggression. This was probably because
punching responses, which are considered
common and therefore excluded from the
imitative category in other studies, were here

scored as imitative acts. Children assigned to the
control group, in fact, performed approximately
twice as much “imitative” aggression as the

preexperience modeling group to begin with, but
this sizable initial difference in aggressiveness was
ignored in the analysis. A different picture
emerges when children’s subsequent aggressive
responsiveness is compared against their initial
baseline aggression. Whereas controls showed a
marked drop in imitative aggression when

retested (—24) and virtually no change a week
later (2), children who received the familiarizing
experience with aggressive modeling exhibited an
increase in imitative aggression both immediately
after exposure to the filmed aggression (23) and in

the follow-up test (31). Similarly, when all forms
of imitative behavior are considered, controls
manifested a progressive loss in succeeding tests
(—18, —28), while children who were exposed to
the filmed modeling progressively increased in

imitative responding (33, 46).
The generality of the hypothesis that

familiarity attenuates modeling may also be
questioned. Whether familiarity enhances or
impedes imitation will largely depend on the
functional value of the modeled behavior. It seems
highly improbable that people who have
previously patronized boxing gymnasiums would
be less inclined to imitate a boxing instructor than

those who had never been in such places, or that
flight trainees who have had prior exposure to the
numerous instruments in cockpits would be
handicapped by that experience in learning how to
fly through modeling. On the contrary,

preexperience may enhance and focus
attentiveness to critical features of the models’
performances and thereby augment modeling
effects. On the other hand, people who have
developed successful ways of behaving in given

situations might be less willing to adopt new
modeled patterns of unknown value than would

observers who are uncertain about what actions
would be most appropriate or effective.
Reluctance to give up familiar responses for

modeled ones does not mean that exemplified
patterns have not been learned or that they will
not be adopted on future occasions if the originally
preferred modes should prove ineffective.

Modeling influences add to a person's repertoire a
new response that may vary in preferability in

different circumstances and at different times,
depending on its value relative to alternative
courses of action.
The generalizability of findings based on the

modeling paradigm is sometimes mistakenly
questioned on the assumption that tests of
modeling effects must include human targets
whom subjects actually hurt with injurious intent

(Hartley, 1964). These requirements fail to
recognize that aggression is rarely learned under
such conditions even in everyday life. Behavior
that has dangerous or costly consequences is
typically acquired and perfected in simulated
learning situations. Airline pilots, for example,
develop basic flying skills in simulators that
reproduce rolls, yaws, pitches, and vertical and
lateral motions of actual airplanes. Similarly,

aggressive behavior is largely learned under
nonfrustrating conditions in the absence of
injurious intent and often with inanimate objects.
Army recruits develop and perfect combat skills
through many hours of target practice and

simulated battle skirmishes. Military instructors
who angered enlistees and then gave them live
grenades to toss at humans would be judged
grossly deranged. Aggression simulators are by no
means confined to military training. Boxers

develop pugilistic skills by using punching bags
and sparring partners whom they do not

necessarily intend to injure; hunters acquire the
basic rudiments of hunting by shooting at
inanimate targets before they venture into the

woods in search of game; and parents rarely
follow their children around to teach them how to
fight in the context of actual battles. Indeed, if
aggressive skills were taught only while

individuals were angered and entertained
injurious designs, many of the tutors and learners

would probably be maimed during the acquisition
phase.
In studying aggression, researchers must

decide whether they are investigating learning or
performance effects and choose their
experimental procedures accordingly. Learning
effects of aggressive modeling are best assessed
under test conditions in which subjects are able

and willing to reveal all they have learned. To use
human targets for this purpose would be as
nonsensical as to require bombardiers to bomb
San Francisco, New York, or some other inhabited
area in testing whether they have acquired
bombing skills.
On the other hand, studies designed to identify
conditions that lead people to use behavior they
have learned for injurious purposes do require

human targets. In such investigations, which
receive detailed consideration in Chapters 3 and 4,

people are provided with opportunities to hurt
others. These experiments help explain what
causes people to behave punitively, but they throw
no light on learning determinants because the

aggressive actions are already part of the subjects’
repertoires. A complete understanding of
aggression therefore requires both types of

experimentation.
Although research using nonhuman targets is

principally designed to advance understanding of
learning processes, the question is sometimes
raised whether such influences do affect
interpersonal aggression. Considering that in
everyday life the actions used to hurt others are
usually developed under simulated conditions,
one might expect some transfer to occur. There is
evidence to indicate that this is indeed so. Walters

and Brown (1963) found that boys who had been
intermittently rewarded for punching an
automated Bobo doll—a training situation devoid
of pernicious intent and social injury—later
exhibited more physically aggressive behavior

toward other children in a competitive situation
than boys who received no prior training in
punching responses. Conversely, elimination of
aggressive styles of behavior toward nonhuman
targets can produce corresponding decreases in

interpersonal assaultiveness. Hyperaggressive
children who were taught through reinforced

modeling to favor cooperative over aggressive
solutions to problems in doll-play situations were
much less prone to respond aggressively to

interpersonal thwarting, both in structured
situations and in their regular social interactions
(Chittenden, 1942).
For those who attach special significance to
observational learning exhibited toward human
targets, the modeling paradigm has been applied

in this manner too. After children have seen filmed
models assault a person in novel ways, the
imitative and nonimitative aggression they direct

at a similar human target is recorded. Viewing
interpersonal assaults fosters imitative aggressive
acts toward people, whereas the unusual conduct
rarely occurs when the example is not provided
(Hanratty, O’Neal, and Sulzer, 1972; Savitsky,

Rogers, Izard, and Liebert, 1972). For reasons
given earlier, children reveal more observational
learning of aggression in relation to inanimate
than to human targets (Hanratty, Liebert, Morris,
and Fernandez, 1969). Learning is a necessary
though not a sufficient condition for performance.
In later chapters we shall specify the conditions
under which learned behavior is expressed in
action.
Another issue that requires some comment
concerns the criteria used for determining
whether observers have learned anything new
from modeled examples. Some writers (Aronfreed,
1969; Patterson, Littman, and Bricker, 1967) have

questioned whether behavior formed through
unique combinations of available actions
represents response learning, inasmuch as the
components already exist in subjects’ repertoires.

According to this line of reasoning, a pianist who
has mastered a piano concerto has learned
nothing new because all the finger movements

already existed in his repertoire. Most intricate
responses are composed of common behavioral
elements.
An observer can learn a variety of things, even
from relatively simple modeled performances.
First, he can learn through example how to

combine different actions, which themselves are
not novel, into new forms of patterned behavior.

Response novelty must be defined in terms of
empirical criteria rather than a priori estimations.
Any behavior that has an extremely low or zero
probability of occurrence given appropriate

stimulus conditions qualifies as a novel response.
Second, a person can learn to make common
responses to situations to which he previously did

not respond at all, or to which he reacted in a
substantially different manner. In the latter case,

learning involves bringing existing behavior under
new stimulus control rather than forming new
response configurations. And third, he can learn

the consequences that are likely to follow certain
actions.
Findings of modeling studies are sometimes
incorrectly interpreted on the grounds that
inflated objects invite physical aggression

(Aronfreed, 1969). This type of analysis not only
disregards the experimental controls instituted for

the target influence, but reflects a view of
aggression that is refuted empirically. Aggressive
acts are not internally impelled in the absence of
any external instigators. On the contrary,

aggressive responsiveness is generally under
appropriate stimulus control. For this reason,
people rarely go around assaulting innocuous

targets. In real life the targets, whether they be
boxing opponents, enemy troops, oppressors, or

more common provocateurs, all have aggression-
activating potential. Aggression does not occur in a
vacuum. Research in this field is principally

designed to explain variations in the form,
direction, intensity, and frequency of aggression
under certain eliciting conditions. Demonstrating
that observationally learned aggression is seldom
displayed when there are no suitable targets for it
would have no more practical or theoretical

significance than showing that hunters are
disinclined to fire their guns as long as they see no

game.
The effects of aggressive modeling are

evaluated by measuring how people behave in the
same situation as compared to their previous
conduct or to that of others who have never been
exposed to assaultive examples. Obviously, one
cannot attribute differences in aggression under

modeling and control conditions to the nature of
the target, because it is identical in both situations.
Moreover, the models purposely display
uncommon ways of aggressing toward the figure,
rather than the usual punching responses. Some of
the responses involve unusual patterns of actions
accompanied by distinctive hostile remarks. These
particular configurations are almost never
spontaneously performed by children who have

not seen them modeled. Other modeled activities
include common responses, but ones that children
rarely if ever direct toward inflated dolls without
the influence of example. Indeed, the behavior
displayed by control children in a variety of

modeling experiments consistently shows that
inflated dolls, by themselves, do not invoke
pummeling with mallets, bombardment with balls,
or irate punting.
Modeling General Tactics
The research reported thus far has been
concerned with the transmission through example

of specific aggressive actions. However, models
can teach more general lessons as well. From
observing the behavior of others one can learn
general strategies that provide guides for actions
which go well beyond the specific modeled
examples.
In studies of this higher form of modeling
(Bandura, 1971d), people observe models respond
successfully in a consistent manner to diverse
incidents. When observers are later confronted
with dissimilar situations, they are inclined to

respond in a style resembling the models’
dispositions, although their specific reactions
cannot be mimicked because of the altered
circumstances. In this modeling process, observers

abstract the common features exemplified in
specific modeled responses and formulate a
general principle for fashioning similar models of

behavior. Responses performed by subjects that
embody the observationally derived principle
resemble the behavior that models would be apt
to exhibit under similar circumstances, even
though subjects have never observed the models
responding to these particular events. Naturalistic
data will he cited later to illustrate how people

adopt from models both specific aggressive
actions and general tactics for dealing with diverse
interpersonal situations.
Modeling of Attitudes and Values
Exposure to modeled aggression can affect not

only observers’ actions, but also their attitudes
and values. If certain people are portrayed as
behaving in aggressive and violent ways, one

would expect such observational experiences to
produce, through modeled associations, certain
attitudes toward them. Siegel (1958), for example,
had children listen to radio presentations in which

taxi drivers resolved conflicts either by physical
violence or through nonaggressive means.
Children who had been exposed to the aggressive
modeling were more inclined to ascribe aggressive
characteristics to different taxi drivers
encountered in new situations than those

observing amiable examples. Many of the attitudes
that people hold toward members of different
occupations, races, and ethnic groups are
undoubtedly cultivated through the modeling of
stereotypes.
Results of a study by Bandura, Ross, and Ross

(1963c) indicate that the evaluative reactions
observers develop toward aggressors is partly
dependent upon the success of their behavior.

Children who had observed the aggressor’s
actions produce punishing consequences did not
imitate his behavior, and they rejected him as a

model for emulation when asked to select the
person they would prefer to be like. On the other
hand, when the aggressor’s behavior was effective
in amassing rewarding resources, he was readily
chosen by most of the children as a desirable
model for imitation. The surprising finding,
however, is that without exception these children

were critical of the tactics he used, which they
characterized as harsh, rough, mean, and wicked.
Evidently, the utility of aggression rather than its
moral value served as the primary basis for
emulation. These findings lend support to Toch’s

(1969) view that behavior is often discarded not
because it is bad or wrong, but because it is
ineffective. Moral justifications can always be
found for practices that are beneficial to the user.
Children resolved the conflict arising from

emulation of a person employing reprehensible
tactics by derogating the unfortunate victim,

thereby justifying the aggressor’s exploitive
assaultive actions. They criticized the victim for
his inability to control the aggressor, for his
unwillingness to cede some of his resources, and
they ascribed to him demeaning characteristics
that would excuse the aggressor’s behavior. This
study illustrates how rewarding consequences to

the model may outweigh the value systems
instilled in observers—children adopted the
behavior of a successful aggressor and held him in
high regard even though they considered his
specific actions objectionable and morally

reprehensible.
Anxiety-reduction Theory of Aggressive

Identification
Modeling of aggression has been widely

interpreted in psychoanalytically oriented
theories in terms of a process of defensive
identification originating in the Oedipus complex

(Freud, 1923; A Freud, 1946). According to this
formulation, boys experience high anxiety over
anticipated punishment by castration for their
incestuous wishes toward their mothers and
rivalrous feelings toward their fathers. Oedipal
boys both reduce their anxiety and derive
vicarious gratification of their sexual feelings by

emulating the characteristics of the threatening
father. Behaving like a feared aggressor
presumably alleviates anxiety by assimilating the
external threat. Alternatively, one might
hypothesize that, except for adoption of parental

prohibitions, acting like a threatening antagonist
would accent the rivalry and thus exacerbate
rather than diminish anticipatory fear of
punishment.
Evidence for an anxiety-reduction mechanism

in aggressive modeling is based almost entirely on
anecdotal reports furnished by Anna Freud (1946)

and Bettelheim (1943). Most of the examples of
aggressive behavior that these authors attribute to
defensive identification can be adequately
accounted for without invoking an identificatory
process at all. And in instances in which actions
have been patterned on an aggressor, more
parsimonious explanations can be offered for the

occurrence of modeled behavior.
In a number of Freud’s cases, for example, the
person who supposedly serves as the model in fact
displays no aggression, and it is simply assumed
that the aggressive child expects him to be

assaultive. Anticipatory aggression can be
interpreted as a defensive maneuver, but it could
hardly represent identification with the aggressor,
any more than swatting a menacing mosquito

represents insect identification by a threatened
adult. In others of Freud’s illustrations a child who
was accidentally hurt by a teacher in an outdoor

game wears military apparel the following day,
and a boy who had undergone dental treatment
later displays aggressively demanding and
destructive behavior during a therapeutic session.
It is, of course, entirely possible that the boys’
behavior and the previous hurtful experiences are
totally unrelated. Even if the events were causally

linked, the theory of defensive identification does
not explain why the forms of the boys’ aggression
were unlike that of the menacing models.
Numerous studies, to be reviewed later, disclose
that painful stimulation can elicit aggressive

behavior, especially in animals. Explanations of
this relationship never assume, however, that the
animals’ aggressive behavior represents defensive
identification with the punisher.
The clearest example Freud gives of actual

imitative behavior involves the case of a boy who
mimicked his teacher’s angry grimaces while the

latter was reprimanding him. Freud’s
interpretation that the boy “through his grimaces
was assimilating himself to or identifying himself
with the dreaded external object” (p. 118) is
complicated because the boy’s imitative grimaces
provoked bursts of laughter from his classmates,
thus providing him with social rewards. It is

therefore doubtful, even in the latter case, that the
imitative behavior was maintained by anxiety
reduction.
It is likewise apparent in Bettelheim’s (1943)
account of prisoners’ behavior in a Nazi

concentration camp, which is also frequently cited
as evidence of defensive identification, that most
of the behavioral outcomes may not, in fact, have
involved identificatory processes. Bettelheim

reports that many of the older prisoners were
verbally and physically aggressive toward
newcomers and potential troublemakers,

sometimes behaving more aggressively than their
guards when placed in charge of others. They
enforced nonsensical rules that the Gestapo had at
one time or another imposed on the group; some
of the older captives even modified their uniforms
to resemble those of the guards and resented
sympathetic foreign correspondents who

criticized the Germans.
It is true that the older prisoners often
imposed on their fellow captives aversive controls
similar to those they themselves had endured, but
it is by no means clear that their behavior

represented identification with the aggressor in
the sense that the concept is employed in
psychoanalytic theory. The Gestapo consistently
imposed group-oriented punishments in which the

transgressions of any individual resulted in brutal
torture of the entire group. When two prisoners
attempted to escape, for example, all the prisoners

were punished by being forced to stand at
attention for hours in a snowstorm without
overcoats, during which time some died from
exposure, and several hundred later had to
undergo amputations of their badly frozen
extremities. Since the group consequences were
often severely painful and the demands of the

guards highly capricious, it is not surprising that,
in order to avoid brutal and degrading treatment,
experienced prisoners often enforced their
captors’ demands. The prisoners’ punitive rule
enforcement may thus represent straightforward

avoidance behavior designed to minimize
transgressions that would endanger the whole
group, rather than emulative behavior. Indeed, the
explicit purpose of the hostage and group-
punishment system was to make every prisoner

feel responsible for the acts committed by others.
Similarly, antagonism toward foreign

correspondents and former fellow prisoners who
had publicly reported cruelties perpetrated in the
concentration camps, also interpreted by

Bettelheim as an example of identification with the
aggressor’s ideology, may reflect a protective
measure in that unfavorable newspaper accounts
written by these persons brought severe

punishment on the prisoners.
Bettelheim does provide evidence that, in some

cases, prisoners went to great lengths to emulate
the guards. Some of the older captives, for
example, collected pieces of Gestapo uniforms and

sewed their own uniforms so as to resemble those
of the guards. However, such imitative behavior
was punished by the guards and hence could
hardly have served anxiety-reducing functions.
Indeed, since this particular modeled behavior

persisted in the face of aversive consequences, it
requires an alternative explanation. The Gestapo
elite possessed potent rewarding and coercive
power. It has been shown in controlled
experiments (Bandura, Ross, and Ross, 1963b;
Mischel and Liebert, 1967) that people tend to
model their behavior after sources of power. Given
the additional evidence that authoritarian
personalities are prone to imitative behavior of

aggressive models (Epstein, 1966), one might
expect that prisoners who had developed
authoritarian attitudes prior to their
imprisonment (some were formerly prominent
politicians) would be disposed to admire and to

imitate these attributes in domineering guards.
Striving to emulate the elite who control desired
privileges and resources is also noted in upwardly
mobile persons who, like the prisoners in
question, persist in imitative behavior despite the

rebuffs they incur from their peers and high-status
models.
The case material discussed above and
empirical investigation of defensive identification
(Sarnoff, 1951) illustrate the loose criteria often
used in designating aggressive patterns as
outcomes of a defensive identificatory process.
There is considerable evidence, of course, that
both children and adults imitate aggressive

models presented on film, who obviously
constitute no personal threat. Moreover, studies in
which modeled aggressive acts incur either
rewarding or punishing consequences
demonstrate that the success of the model’s

behavior is an influential factor in determining the
degree to which an aggressive pattern of behavior
will be spontaneously reproduced by others. On
the basis of the response-consequences
interpretation of modeling effects, it would be

predicted that if aggressors are highly successful
in securing desired outcomes, observers will

identify with them even if they ordinarily regard
such attributes with disfavor (Bandura, Ross, and
Ross, 1963c; Baron and Kepner, 1970). If, on the

other hand, the aggressor’s behavior fails to gain
power and control over persons and their
resources, or in fact produces punishing outcomes,
identification with the aggressor will not occur.

There is a sense in which aggressive identification
may reduce anxiety. A person who adopts effective

aggressive patterns is less apt to be victimized by
others and is therefore less vulnerable to external
threats. In this interpretation aggressive
identification reduces anxiety not by assimilating

the external threat, but by appropriating forceful
tactics that ensure better control over the social
environment.
In naturalistic studies the possible influence of

the threat value of an aggressive model and the
success of his behavioral style are confounded.
That is, an aggressor may represent a threat to the
observer, but he also repeatedly demonstrates
that he can get what he wants by coercive means.
It has been shown that fear of an aggressive model
is not a necessary condition for aggressive
identification. Whether fear is a facilitative,
impeding, or irrelevant factor in the identification

process can be best answered through controlled
studies in which models’ threat and the functional
value of their aggressive behavior are
independently varied.
LEARNING THROUGH PRACTICE
People rarely teach social behaviors that are
never exemplified by anyone in their environment.

Although modeling influences are universally
present, it is possible to establish new response

patterns solely on the basis of trial-and-error
experiences. Moreover, observational learning
usually produces rough approximations of desired
activities that are then further refined through
reinforced practice. In learning by experience,
behavior is shaped into new patterns by its
consequences. During the course of trial-and-error
experimentation, unsuccessful responses tend to
be discarded, whereas rewarded alternatives are

progressively strengthened. New response
patterns are similarly created by making
reinforcement contingent upon novel
combinations of actions. Since in this mode of
learning the appropriate behavior must be

discerned from differential consequences that
accompany various actions, it is much less efficient
than having a good example to follow.
The literature contains ample evidence that

aggressive behavior is powerfully controlled by its
consequences, but no experimental attempts have
been made to fashion novel forms of aggression

through differential reinforcement alone. The
principal reason for this lack is that demonstration
rather than unguided experience is the best
teacher. It would be foolhardy to try to instruct
novices in how to handle firearms or hand
grenades by selectively reinforcing their trial-and-
error efforts. In the case of less hazardous

aggressive activities, however, they may be
formed to some extent by response feedback.
Learning through combat experience has been
explored to a limited extent in experiments with
subhuman species designed to train nonaggressive

animals to become ferocious fighters. This
outcome is achieved by arranging a series of bouts
with progressively more experienced fighters
under conditions in which the trainee can win

fights without being hurt. Initially, the animals
used to provoke defensive fighting are either
physically restrained or such weak combatants

that the trainees can easily emerge as victors. As
fighting skills are developed and reinforced
through repeated victories, tougher opponents
that can be beaten only by vigorous, hard fighting
are introduced. Results show that as training
progresses, formerly noncombative animals
become more and more vicious in their aggressive

behavior (Ginsburg and Allee, 1942; Kahn, 1951;
Scott and Marston, 1953).
Some indications of the importance of social
learning in animal aggressiveness are provided in
studies of the long-term effects of infantile fighting

experiences and in comparisons of physiological
and training determinants of combativeness.
Fredericson (1951) found that a brief experience
in competitive aggression over food during infancy

produced mice that fought over food in adulthood
even when not motivated by hunger. By contrast,
animals that did not receive the infantile fighting

experience never battled over food when they
were not hungry. Results of an experiment that
varied castration and androgen injections before
and after aggression training showed victorious
fighting experiences to be the most influential
determinant of subsequent aggressiveness,
regardless of the point at which the physiological

modifications occurred (Bevan, Daves, and Levy,
1960).
Although successful fighting produces brutal
aggressors, severe defeats create enduring
submissiveness. The power of injury to inhibit

aggression is determined by a number of factors.
Kahn (1951) found that the younger the animals
are when they first suffer defeat, the more
passively they behave later in the face of attack.

Animals that have been subjected to many painful
defeats likewise do not fight even when
confronted with harmless opponents, whereas less

frequent defeats result in only partial inhibition of
aggressive behavior (Scott and Marston, 1953).
Indeed, fighting success largely determines
animals’ social organization. In the course of a
series of tournament bouts a stable social order
eventually emerges that corresponds closely to the
rate of triumphs and defeats of individual

members (Ginsburg and Allee, 1942; Scott, 1944).
Consistent victors assume a dominant social
position; repeated losers become subordinates
that submit passively; animals with a history of
both wins and losses occupy intermediate ranks.

Naturally evoked dominance relationships can be
modified by systematically altering the reinforcing
consequences of fighting. It is somewhat easier,
however, to lower the aggressiveness and social
status of superior fighters by a series of defeats

than to transform meek appeasers into
swaggering aggressors. Their extreme reluctance

to perform any aggressive responses limits
opportunities to influence them through positive
reinforcement.
People are often confronted with provocative
situations calling for some type of aggressive
action. To the extent that forcible responsiveness
produces good results, the rewarding

consequences can shape aggressive styles of
behavior. Patterson, Littman, and Bricker (1967)
report a field study of the development of
aggression in passive children through a process
of victimization and successful counteraggression.

Nursery school children were observed over a
nine-month period for the frequency with which
they aggressed toward each other and the
immediate consequences that such behavior
produced for them. Passive children who seldom

resorted to defensive aggression because they
were rarely picked on and those who were often

victimized but whose counterattacks generally
failed to terminate assaults by aggressors
remained submissive in their behavior. By

contrast, equally passive children who were
frequently victimized by their peers but who
counteraggressed successfully showed a marked
increase in aggressive tendencies. Considering

that the children had innumerable opportunities
to learn observationally how to aggress from

examples provided by their peers and assailants, it
is debatable whether the particular forms of
aggression they displayed were created solely by
reinforcement influences. The findings

nevertheless clearly show that if aggression,
however learned, is positively reinforced, it will
become a preferred mode of response. The

development of aggressors through a process of
tyrannization and reinforcement for

counteraggression bears some similarity to how
the more active members of oppressed segments
of society are shaped into aggressors by people

who themselves model, but decry, such behavior.
LEARNING OF AGGRESSION UNDER
NATURALLY OCCURRING CONDITIONS
There are many naturalistic studies
demonstrating that people behave in aggressive
ways that resemble patterns exemplified by

salient models. Naturally occurring resemblances
are difficult to interpret, however, for a number of

reasons. Unlike laboratory studies in which the
nature •of the modeling influences is carefully
delineated, in everyday life people are
concurrently exposed to a number of models,

including adults, peers, and all sorts of television
personalities. There is no reliable way either of
identifying which of these examples are selected
as sources of behavior or of gauging their relative
influence.
It was previously shown that when the

examples set by others differ in significant ways,
observers often pattern their behavior after a

composite model embodying features taken from
the different sources. The net result may be a
novel pattern of aggression that does not bear
exact likeness to any of the individual models,

even though the composite pattern is created
entirely from the examples provided.
Resemblances arising from modeling influences
are also less discernible when observers adopt the

general tactics but not the specific actions
modeled by others. It is not uncommon for people
whose efforts to produce desired changes have
been repeatedly thwarted to emulate coercive
strategies, but in their own way, after seeing other

groups successfully attain what they want through
aggressive means.
Behavioral similarity does not invariably
indicate a modeling outcome. Some performances,

especially relatively simple ones, can be acquired
on the basis of common direct experiences. To

complicate matters further, the effects of modeling
are often counteracted, attenuated, or augmented
by the effects that imitative aggression produces
when it is tried. The changes created by modeling,

in turn, obscure the effects of reinforcement. In
instances in which aggressive behavior is learned
observationally but inhibited due to unfavorable
consequences, positive reinforcement later may
increase aggressive actions simply by conveying

permissiveness for the previously learned
behaviors rather than acting through its
reinforcing and shaping functions, as is commonly
assumed.
Within a modern society, three major sources

of aggressive behavior are drawn upon to varying
degrees. One is the aggression modeled and

reinforced by family members. Though familial
influences play a major role in setting the
direction of social development, the family is
embedded in a network of other social systems.
The subculture in which a person resides and with
which he has repeated contact provides a second
source of aggression. The types of behaviors that

are exemplified and valued in community
subsystems may support or counteract familial
influences. The third source of aggressive behavior
is the symbolic modeling provided by the mass
media, especially television. In heterogeneous and

rapidly changing societies the specific patterns of
behavior exemplified by parental models may
have limited value for their offspring. Television
can teach viewers how to behave in a variety of
situations that extend well beyond the learning

experiences available within the family or the
immediate community.
Familial Transmission of Aggression
There are several published studies specifically
comparing the quality of parental models of
children who develop hyperaggressive
personalities with those of nonaggressive models.
In an analysis of information on familial conditions

collected over an extended period, McCord,
McCord, and Zola (1959) found that sons of
criminals tended to become criminals themselves,
especially if their fathers were cruel and
neglecting (85 percent), whereas similar adverse

treatment was associated with a lower rate of
criminality (40 percent) in families where the
father did not provide a grossly deviant model of
behavior. Other investigators (Glueck and Glueck,
1950) have reported a much higher incidence of

aggressive paternal modeling for delinquent than
for nondelinquent boys. That violence breeds

violence is further suggested by Silver, Dublin, and
Lourie (1969). Their longitudinal study of child
abuse cases over three generations shows that

children who suffer brutal treatment at the hands
of assaultive parents are themselves inclined to
use abusive behavior in the future.
Most assaultive youngsters do not have
criminally violent parents. In slum districts, where
children generally spend more time in the streets

than in their homes, peers and various
extrafamilial adults serve as the primary
transmitters of aggressive life styles. And in

middle-class families that produce violence-prone
offspring, parental aggressive modeling usually
takes less blatant forms. This is shown in two
studies (Bandura, 1960; Bandura and Walters,
1959) that examined the social learning

determinants of aggressive disorders in boys who
lived in privileged neighborhoods that are not
conducive to antisocial behavior. Rather than
examine why people who grow up under adverse
familial and community conditions develop
delinquent tendencies, we asked instead why
intelligent boys who came from intact middle-
class homes where the neighborhood standards of
conduct support law-abiding behavior become

antisocially aggressive (Bandura and Walters,
1959). Families of adolescents who exhibited
repetitive antisocial aggression were compared
with those of boys who were neither markedly
aggressive nor passive.
The families differed most strikingly in the

extent to which they trained their sons, through
precept and example, to be aggressors. Parents of
nonaggressive boys encouraged their sons to be

firm in defending their principles, but they did not
condone physical aggression as a means of settling
disputes. In familial interactions the parents

modeled considerate behavior and relied mainly
on reasoning in handling social problems. By
contrast, parents of the aggressive boys displayed
little antisocial aggression, but they repeatedly
modeled and reinforced combative attitudes and
behavior. While intolerant of aggression toward
themselves, one or the other of the parents almost

invariably encouraged their sons to aggress
toward peers, teachers, and other adults outside
the family. During the early childhood years, the
instigation often took the form of demands that
the boy use his fists in dealing with antagonists, as

illustrated in the following interview excerpts.
Interviewer (I): Have you ever encouraged Glen to

use his fists to defend himself?
Father (F): . . . Well, Glen has always been a kid where

kids have pushed him over at school. We’ve
had to encourage him, “Now, son, you just
stand right up there.” That comes right back
to the point that he’s never played with
children until five years ago. In fact, he was
the type of kid that, when he was small, he’d
rather sit down and play paper dolls than get
in the yard and play with the boys, kicking a
football around…. But lots of times he’d come
home and kids would throw a left or right at
him and he didn’t know what to do about it. I
used to want to teach him to protect himself
but my wife didn’t go for it. She said, “You’re
putting things in his head,” that I was training
him to fight. So we dropped it. (Bandura and
Walters, 1959, pp. 106-7)
This father’s method of training was depicted in
greater detail by his wife, and it is easy to see why
she felt that she had to intervene.
Mother (M): When he was about six or seven years

old, all the kids were fussing and fighting and
he would never fight. His sister would always
have to take up his battle for him. Always
fighting. So one day my husband took off his
belt and said, “Listen, you’re coming home
and crying all the time, saying ‘Somebody hit
me.’ ” So my husband was watching through
the bedroom window one day and he saw two
little boys. They were really fighting him. So
he went up, took off his belt, and he said,
“Glen, I’m going to tell you something. You’re
going to whip these boys or else I’m going to
whip you.” So he made him stand up and fight
both of them, (p. 107)
Although because of his wife’s objections the
father apparently toned down the fighting
instructions, he continued to demand that his son
accept and enjoy aggressive activities.
F: One day we went to see Bobo Olson fight. We tried

to get him to go. “No, I don't want to go.” Well,
he hated it. I said, “No, you’re going.” Of
course, if he didn’t want to go, that’s fine. But
he’d rather be home calling his friends over
the telephone. That’s what just burns me up:
it makes me so mad. I never turned down
anything like that when I was a kid. You get a
chance maybe one out of five years to see
Olson fight, the champion of the world, and he
turns it down flat. (pp. 106-8)
Some of the most aggressive boys came from
families in which both parents demanded,
instigated, and condoned assaultive behavior.
I: Have you ever encouraged Earl to stand up for

himself?
M: Yes. I’ve taught young Earl, and his Dad has. I feel
he should stand up for his rights, so you can
get along in this world.
I: How have you encouraged him?
M: I 've told him to look after himself and don’t let

anybody shove him around or anything like
that, but not to look for trouble. I don’t want
him to be a sissy.
I: Have you ever encouraged Earl to use his fists to

defend himself?
M: Oh yes. Oh yes. He knows how to fight.
I: What have you done to encourage him?
M: When he was a little boy, he had a little pair of

boxing gloves. His dad has been an athlete all
his life, so his dad taught him.
I: Has he ever come to you and complained that

another fellow was giving him a rough time?
M: Oh yes, when he was younger. I told him, “Go on

out and fight it out yourself.”
I: If Earl got into a fight with one of the neighbor’s

boys, how would you handle it?
M: Oh, he should fight it out himself. When he was a

little fellow he used to fight his own battle . . .
I: What would you do if you found Earl teasing

another fellow or calling him bad names?
M: That would be up to Earl. If the other boy wants to
lick him, that would be up to Earl. He
deserves it. (pp. 115-16)
In disputes with adults, the parents would insist
that their sons not allow themselves to be “pushed
around”; they would side with the boys against

their teachers; and they would criticize
protagonists in ways that excused their sons’

assaultive actions.
The generality of these findings is confirmed
by a second study (Bandura, 1960) that compared
the social learning antecedents of hyperaggression

and withdrawal in preadolescent boys. Parents of
the aggressive boys consistently punished
aggression directed toward themselves, but they
encouraged and intermittently reinforced
aggression outside the home. The withdrawn boys,

on the other hand, had undergone pervasive
inhibitory training in relation to aggression,
dependency, and sex behavior. In addition, parents
of the aggressive boys, with their relatively
frequent use of physical and verbal punishment,
provided more aggressive models for their sons

than parents of the withdrawn group, who were
emotionally restrained in their behavior.
During early childhood years, perhaps the
most salient parental modeling of interpersonal
behavior occurs in the context of disciplinary
activities. Here children are furnished with vivid

examples of how one might attempt to influence
and control the behavior of others. And, indeed,
children often do draw on parental practices in
coping with the interpersonal problems they
encounter in their daily interactions. Hoffman

(1960) found that mothers who forced compliance
with their demands through domineering
methods, which included verbal and physical
aggression, had children who employed similar
aggressive tactics in controlling the behavior of
their peers.
In many instances the modeling provided in

disciplinary actions is inconsistent with, and
therefore contravenes, the effects of direct
training. If, for example, a parent beats his child
for having struck a playmate, the intended
objective of the punishment is that the child

should refrain from hitting others. Concomitantly
with the intended training, however, the parent is
unwittingly exemplifying the very behavior he is
trying to eliminate. From fear of retaliation,

children are unlikely to counteraggress toward
their parents, but the influence of parental
example may outweigh parental dictates in

guiding behavior in extrafamilial situations.
Findings of laboratory studies in which modeling
and direct tuition are systematically varied
(McMains and Liebert, 1968; Mischel and Liebert,
1966; Rosenhan, Frederick, and Burrowes, 1968)
reveal that people are most influential when they
are consistent in what they practice and what they

preach, whereas the impact of their behavioral
prescriptions is considerably weakened by
discrepant modeling.
Johnson and Szurek (1952) have identified
another set of subtle familial influences that may

create and maintain antisocial aggression in
children even though extrafamilial conditions
provide weak support for such behavior. It is not
uncommon for parents to foster in their children

patterns of behavior that they approve but cannot
themselves engage in for various reasons, and to
derive vicarious gratification from their children’s

successes. Such processes are clearly operative in
the case of achievement behavior. However,
socialized parents cannot openly instigate and
delight in the aggressive exploits of their children
without incurring censure. For this reason, the
process, according to the observations of Johnson
and Szurek, operates in a covert manner.
Antisocial behavior is often initiated by
repeated warning of possible transgressions
presented more in the form of suggestive
prescriptions than prohibitions. Provocative
negative modeling thus informs the child of

courses of action that otherwise might never have
occurred to him. Transgressive behavior is less
likely to recur when desired behavior is clearly
defined and actions that exceed acceptable limits

are dealt with firmly without apology, guilt, or
vacillation. The unwitting condoner, however,
tends to respond to antisocial behavior in ways

that serve to maintain it. Transgressions are
overlooked, dismissed as unimportant, or excused;
detailed accounts of the exploits are relished with
feigned distress; and antisocial actions are met
with covert permissiveness and approval. These
detrimental parental practices are presumably
reinforced by gratifications derived from the

child’s actions.
Unwitting condoning of socially disallowed
forms of aggression is, of course, not confined to
familial situations. Disputes often arise over
whether societal agents entrusted with coercive

power, such as the police, employ unwarranted
forms of aggression in apprehending suspected
lawbreakers and in controlling protesters.
Proposed solutions to this problem generally

center around offering police recruits better
training, especially along psychiatric lines,
elevating the status of police work, and

establishing grievance review boards to monitor
police departments. Although such programs can
improve law enforcement practices, it should be
recognized that excessive use of aggression is
primarily determined by sanctions rather than by
education. Police administrators who condone
brutal tactics will find a great deal of such

behavior exhibited by their officers, whereas those
who censure such behavior will encounter little of
it. The cruel third degree, for example, was
eliminated by firmly imposed prohibitions, not by
psychiatric training of interrogators.
Subcultural Transmission of Aggression
The highest rates of aggressive behavior are
found in environments where aggressive models

abound and where aggressiveness is regarded as a
highly valued attribute. In these delinquent
subcultures (Short, 1968; Wolfgang and Ferracuti,
1967) one gains status primarily through fighting

prowess. Consequently, good aggressors are the
prestigious models upon whom members pattern
their behavior. The combination of prestigious
aggressive modeling with positive reinforcement
of fighting and other manifestations of toughness
creates the most effective condition for cultivating

aggressiveness.
The challenging question requiring
explanation is not why people who inhabit an
aggression-breeding milieu should develop
aggressive modes of response, but why anyone

residing in such an environment should adopt a
markedly different style of life. How does a child
develop in a divergent direction when models of
alternative modes of behavior are lacking in his

immediate environment and prosocial patterns
are not only weakly rewarded but may even be
devalued, whereas adherence to norms of the

corner gang provides comradeship and the valued
recognition of peers? Sociological studies of lower-
class neighborhoods (Whyte, 1955) show that
although most of the young residents conform to
streetcorner codes of behavior, some do come to
value scholastic performance and strive hard for
self-advancement through educational pursuits.
The processes whereby differential
development is achieved within the same milieu
are clarified in studies investigating the sources of
high educational aspirations among lower-class
children (Ellis and Lane, 1963, Krauss, 1964). In

these families the parents themselves cannot
serve as models for class-typed habits of speech,
interests, and social skills that are required for
successful upward mobility. Parents usually

initiate development in prosocial directions by
valuing and encouraging educational aspirations.
By direct guidance and example, admired teachers

and other prestigious adults further reinforce
college ambitions in lower-class youths. Finally,
selective association with college-oriented peers
provides the role models and incentives for
selectively acquiring the values and complex skills
necessary to attain the desired status. When one's
background is assigned an inferior status, the

more successful the upward social transformation
is, the greater the familial alienation will be. With
increased legitimization of subcultural patterns
and willingness of successful models to return to
their own communities, members of minority

groups may more easily attain educational
aspirations without having to renounce their
cultural heritage.
The preceding discussion was concerned with

subcultural training in aggressive styles of conduct
that are disavowed or condemned by the larger
society. Most societies maintain elaborate social

agencies to which they officially assign aggression-
training functions. These include military
enterprises with their many supporting
subsystems. Military training practices, however,
have received little psychological analysis. This is
all the more remarkable considering that military
establishments can, within a relatively short

period, transform people who have been taught to
deplore killing as morally reprehensible into
skilled combatants who feel little compunction or
even a sense of pride in taking human life. Such
radical changes have more profound implications

for accounts of aggression than the actions of
assaultive individuals or youthful groups residing
in dismal neighborhoods.
The task of converting socialized men into

proficient combatants is achieved not by altering
personality structures, aggressive drives, or traits.
Rather, willingness and ability to kill in combat are

attained by direct training combining several
important features. In the first place, the moral
value of killing is changed so that people can do it
free from the restraints of self-condemning
consequences. This is accomplished through
indoctrination that assigns a high moral purpose
to warfare. One must fight for his country to

defend its survival, to preserve world peace, to
save humanity from enslavement by cruel
oppressors, and for national honor. The force of
the moral appeals is strengthened by portraying
the enemy as servile fanatics or subhumans driven

by ruthless leaders. The moral redefinition of
killing is nowhere more dramatically illustrated
than in the case of Sergeant York, one of the most
phenomenal fighters in the history of modern
warfare (Skeyhill, 1928). Because of his deep

religious convictions he registered as a
conscientious objector, but his numerous appeals

were rejected. At camp his battalion commander
quoted chapter and verse from the Bible to
persuade him that under certain conditions it was

Christian to fight and kill. A marathon
mountainside prayer finally convinced him that he
could serve both God and country by becoming a
dedicated fighter.
Self-censuring reactions to brutal acts are
repeatedly neutralized by a variety of self-

absolving devices that will be examined later, as
well as by ideological justifications. Powerful
social sanctions are also effectively used to

promote acceptance of warfare. Combat heroism is
glorified, while opposition to military mission is
treated as an unpatriotic social disgrace. A nation
that fails to convince its members of the justice of
its cause encounters considerable difficulty in

mobilizing them to fight.
The social influence process is greatly
facilitated by immersing recruits in a totally new
reality. Arrivals are promptly dispossessed of
most civilian accouterments and outfitted with
military gear. During the period of intensive
training, rookies are isolated from family, friends,
and normal community life, thus removing the

customary social supports for their behavior and
beliefs. Instead, almost every aspect of their daily
life is closely regulated in accordance with the new
reinforcement structure. Throughout this process

they are subjected to obedience tests and firmly
disciplined for noncompliance.
In addition to legitimization of military killing,
recruits receive intensive training in the intricate
techniques of warfare. A host of coordinated skills
must be mastered, such as hand-to-hand fighting,
tactical maneuvers, reconnaissance patrols, field

fortifications, and artillery and aerial
bombardment. Training proceeds by

demonstrating how combat activities are executed
and by having recruits practice attacks against

simulated targets until proficiency is attained.
The third feature of military training is
concerned with reducing fear of battle. Various
methods are used for this purpose. Action

extinguishes fear. Inductees are therefore drilled
repeatedly in combat performances until they

reach the level of automatic action. In some
training programs (Rees, 1945) combat
desensitization was aided by exposure to recorded
sounds of actual warfare during mock battles.

Frightening aspects of heavy bombardment were
similarly neutralized by beginning with small fire
and then gradually increasing its intensity until

terrifying explosives could be discharged with
equanimity.
The anxiety-neutralizing power of
desensitization procedures is further revealed in
successful extinction of anxiety disorders arising
from terrifying battle experiences (Saul, Rome,
and Leuser, 1946). In a secure setting, soldiers
were shown a graded series of movies of battle
scenes beginning with exposures they could
tolerate. At first the films depicted preparatory

combat activities, followed by scenes of surface
and aerial bombardment from which displays of
injuries and destruction were deleted. In later
sessions the soldiers were gradually presented
with more frightening combat scenes. Reexposure

to the sounds of battle was regulated in a similar
manner. At first the warfare was shown silently,
and only gradually was sound introduced. Day by
day the noise of gunfire and aerial bombardment
was increased until full intensity was reached. As a

further safeguard against excessive emotional
arousal, each soldier was provided his own

volume control with which he could regulate the
amount of aversive stimulation. After about a
dozen showings the soldiers, who earlier exhibited

acute anxiety, hypersensitivity, and recurrent
nightmares, were unperturbed by battle scenes
that had previously terrified them. Moreover, the
emotional extinction effect generalized to

everyday situations. The soldiers were able to
attend commercial movies, which most of them

had avoided because of newsreels, and they no
longer responded agitatedly to loud sounds or
music. Schwartz (1945) also extinguished battle
anxieties by repeated exposure to combat films

graded in intensity as part of a larger treatment
program.
Soldiers are returned to civilian life without
putting them through a resocialization process

designed to reinstate aggression restraints or to
restore commitment to the dignity of human life.
Nevertheless, the vast majority promptly revert to
their civilian self-reinforcement systems so that
they are deterred from behaving cruelly by self-
condemning consequences. Achievement of such
marked shifts in destructive behavior through
moral sanctions without greatly changing the
person provides striking testimony for the social

control of aggression.
Symbolic Transmission of Aggression
The field studies cited earlier identified some
of the familial conditions that breed antisocial
aggression when the prevailing subculture is not

conducive to such behavior. People can learn
varied forms of aggression even when examples
and direct tuition of assaultive actions are lacking

in both the home and the immediate subculture.
The advent of television has greatly expanded the
range of models available to the growing child.
Whereas his predecessors, especially those in

middle-class homes, had limited opportunity to
observe brutal aggression, the modern child has
witnessed innumerable stabbings, shootings,
stompings, stranglings, muggings, and less blatant
but equally destructive forms of cruelty before he
has reached kindergarten age. If parents could

select from diverse alternatives the kinds of
symbolic models they would wish to influence the
development of their children, it is exceedingly
improbable that they would choose Western
gunslingers, remorseless psychopaths, deranged

sadists, and unprincipled racketeers, unless they
harbored odd ambitions for their growing
progeny. The types of symbolic models regularly
furnished to households are evidently determined
more by economic than by moral values.

Therefore, both children and adults, regardless of
their social backgrounds, have unlimited

opportunities to learn from televised modeling
aggressive coping styles and the whole gamut of
felonious behavior within the comfort of their

homes.
Evidence that people can learn as much from
symbolic as from actual models indicates that
television is an important source of social

behavior. Indeed, underdeveloped countries, in
which educational resources are limited, rely

heavily upon televised tutors for expanding
children’s cognitive functioning. Consistent with
findings of studies on aggression, television-
mediated learning is usually as effective as that

achieved by actual tutors (Chu and Schramm,
1967).
Several field studies (Stein, Friedrich and
Vondracek, 1972; Steuer, Applefield, and Smith,

1971) have shown that exposure to televised
violence increases interpersonal aggressiveness,
but the authors do not report the extent to which
the aggressive acts are explicitly imitative. From
time to time, however, numerous incidents in
which people have patterned their aggressive
behavior directly on that of televised models are
reported. Children have been apprehended for
writing bad checks to obtain money for candy, for

sniping at strangers with BB guns, for sending
threatening letters to teachers, and for injurious
switchblade fights after witnessing similar
performances on television (San Francisco
Chronicle, 1961; Schramm, Lyle, and Parker,

1961). Cousins (1949) cites examples in which
children were prepared to carry out televised
homicidal solutions to their own problems. After
receiving an uncomplimentary report card, a
youngster suggested to his father that they send

the teacher a box of poisoned chocolates for
Christmas as he had recently seen on television a

man successfully kill his wife with poisoned candy,
and “she didn’t know who did it.”
That repeated portrayal of slayings may
devalue human life is suggested in the case of a
policeman’s son who asked his father for actual
bullets because his sister “doesn’t die for real
when I shoot her like they do when Hopalong

Cassidy kills ’em.” In still another case, “a
housemaid caught a seven-year-old boy in the act
of sprinkling ground glass into the family’s lamb
stew” to determine whether it would work as
effectively as it did on television. Children have

been maimed and sometimes killed while
attempting to reenact the heroics of televised
superstars, and many pets owe their shortened
lives to the influence of televised Western
hangings.
The modeling effects of televised aggression
apparently are not restricted to the games of

youngsters. It is not uncommon for criminal
activities to be patterned after ingenious styles
portrayed on television. Members of an assaultive
gang styled themselves after the gangsters on the
television series, “The Untouchables,” under the
same name (Cleveland Press, 1961). Another ring
of youths successfully employed burglary tactics

learned from the program “It Takes a Thief”
(Washington Post, 1971). The specific entry
procedure, adopted from a televised model,
involved taping a window to avoid the noise of
shattering glass and then breaking it with a

blowtorch. Susan Atkins, a member of the
infamous Manson family, later confessed that the
Sharon Tate killings were committed partly out of
love, in the belief that new murders might cause
police to free another member of their family who

had been jailed for an earlier slaying. She reported
that the idea for the “copycat” murders was

suggested to them by a television movie in which
police concluded they were holding the wrong
suspect after additional murders were committed

in the same manner (San Francisco Chronicle,
1971a).
Television often introduces highly novel forms

of aggression, the effects of which could be easily
measured given advance knowledge of the content
and time of showing. When the depicted activities

pose serious social dangers, the occurrence of
analogous actions must be recorded by
requirement of law. One such experiment was

inadvertently conducted with the showing of the
program “Doomsday Flight,” over vigorous
protests by the airline pilots’ association. The
televised drama depicted an extortionist
threatening airline officials that a pressure-

sensitive bomb would explode on a
transcontinental airliner as the plane descended to
a certain altitude while landing. In the end the
pilot outwits the bomber by selecting an airport
located at an elevation above the critical altitude
setting. A substantial increase in telephoned bomb
threats occurred in the week following nationwide
presentation of the program (The New York Times,
1966e). Shortly after reruns of “Doomsday Flight”

were shown in Anchorage and in Sydney,
Australia, two airlines in these cities received
extortion threats that a barometric bomb was
concealed on planes in flight and set to explode at
a certain altitude. The extortion bomb hoax cost

Western Airlines $25,000 and Qantas $560,000 for
information on where the bombs were planted in
the aircraft and how they could be defused. In the
Australian incident the anonymous caller added
force to his demand by also placing a bomb with a

barometrically controlled trigger system in an
airport locker and directing aviation officials to it.
National Airlines also received telephoned
threats in San Francisco and Miami patterned after
the television plot when it was rerun in these
cities (San Francisco Chronicle, 1971c). Following a
showing of the “Doomsday Flight” on Montreal
television, an extortionist used the bomb plot in an
effort to extract a quarter of a million dollars from

the British Overseas Airways by warning that a
barometric bomb was set to explode on a jet
bound from Montreal to London when it
descended below 5,000 feet. The hoaxer was
unsuccessful, however, because airline officials,

having gained thorough familiarity with the oft-
repeated scenario, diverted the plane to a high-
altitude landing at the 5,339-foot elevation airport
in Denver (San Francisco Chronicle, 1971d). A
California extortionist came closer to succeeding

by not disclosing the altitude at which the
pressure-triggered bomb would detonate (Palo

Alto Times, 1971). During this period there was a
dramatic rise in airline extortion attempts from a
normal baseline of approximately 2 per month to

an average of 16 per month for the next two
months (Murphy, 1971). Frequency data could
similarly be collected for baseline and post-
showing rates of other dramatic forms of violence,

such as stabbings in subways and the like.
Instigation of felonious acts in foreign

countries by exported American programs
illustrates the potential influence of cross-cultural
modeling through television. The medium of

television enlarges the range of models available
to members of other societies for better or for
worse, depending upon the modes of behavior that
are portrayed. The impact of foreign television on
the habits of a society can be verified by

measuring the appearance of styles of behavior
alien to the native culture until the advent of
transcultural televised models. As is true of
intrasocietal modeling, one would expect the
power of foreign televised models to vary with the
international prestige of the exporting nation and
the utilitarian value of the modeled practices in
the new cultural setting. Susceptibility to foreign
modeling is also likely to be increased when the

indigenous culture is assigned an inferior status,
thus prompting aspiring members to accept
external media portrayals as exemplifications of
modernity.
People who ordinarily dismiss the

psychological impact of televised influences
nevertheless protest vigorously when its contents
pose personal threats. Airline pilots and travelers
object to portrayals of bomb plantings on

airplanes; subway riders denounce programs
depicting subway stabbings; and members of
different ethnic and occupational groups condemn

televised characterizations that repeatedly
portray representatives of their groups in an
unfavorable light.
Modeling Effects in Collective Aggression
Social responsiveness is extensively guided by
modeling influences, a process that is particularly
noticeable in collective aggression. The present

chapter is concerned only with the forms that
collective aggression takes. Its various causes and
instigators will be analyzed later. Social contagion
of new styles of aggression conforms to a pattern
that characterizes the transitory changes of most

other types of activities: New behavior is initiated
by a salient example; it spreads rapidly in a
contagious fashion; and after it has been widely
adopted it is promptly discarded, often to be

replaced by a new form that follows a similar
course.
Modeled solutions to problems that achieve
some measure of success are not only widely
adopted by people facing similar difficulties, but
tend to spread as well to other troublesome areas.
The civil rights movement, which itself was
modeled after Gandhi’s crusades of nonviolent
resistance, in turn provided the example for other

protest campaigns aimed at eliminating injustices
and undesired social practices. Through collective
protest, blacks eventually gained lawful rights
they had been dispossessed of in voting, in the use
of public facilities, and in education. In addition,

opportunities were opened up for which they
were qualified but from which they had been
barred by discrimination. Admittedly, the changes
were slow in coming and fewer than desired, but
much greater than if no coercive action had been

taken. The model of collective protest as a means
of forcing social reforms spread to the antiwar

movement and to disadvantaged groups including
Chicanos, Indians, homosexuals, and women.
The turbulent sixties provide numerous
additional illustrations of the contagion of
collective aggression. The urban disorder in Watts,
for instance, grew rapidly into a national
phenomenon following a common pattern of

property burning, looting, and scattered sniper
fire. Urban rioting ceased rather abruptly even
though the desired social and political reforms
were not secured and societal counterresponses, if
anything, provided greater instigation and

justification for violence than had existed
originally. The campus protest movement at
Berkeley served as the model for the sit-in method
of protest in universities throughout the country.
The peaceful sit-in was supplanted by

progressively more aggressive forms of protest,
graduating to combative disruptions of university

functions, and eventually to a rapidly spreading
“trashing.” In this form of aggression, which was
equally transitory, buildings were bombarded

with rocks and firebombed by protesters
employing mobile guerrilla tactics. Following a
New York rally in which construction workers
beat up antiwar demonstrators, assaults on

dissident students by hard-hatters spread
nationwide.
Airline hijacking provides another striking
example of the rapid rise and decline of collective
modeling of aggression. Air piracy was unheard of

in the United States until a commercial airliner
was hijacked to Havana in 1961. Prior to this
incident there had been a rash of hijackings of
Cuban airliners to Miami. These incidents were
followed by a wave of hijackings both in the

United States and abroad, a wave that reached its
height in 1969, when a total of 87 airplanes was
pirated (Figure 2.5). Thereafter, hijackings
declined in the United States but continued to
spread to other countries, so that international air
piracy became relatively common. Foreign
FIGURE 2.5 Incidence of hijacking over a span of 25 years. The
rise in the foreign hijackings during the 1949-50 period
occurred in Slavic countries at the time of the Hungarian
uprising and the second flare-up between 1958-61 comprised
almost entirely Cuban hijackings to Miami. A sudden
widespread diffusion of hijackings occurred between 1969-71
eventually involving airliners from a total of 55 different
countries. (Plotted from data furnished by the Federal Aviation
Administration)
skyjackings then dropped precipitously. News of
an inventive hijacker who successfully parachuted
from an airliner with a large bundle of extorted

money temporarily revived a declining
phenomenon in the United States, as others
became inspired by his successful example (San
Francisco Chronicle, 1971e).
In Brazil a new form of political bargaining

tactic was devised when a United States
ambassador was abducted and later freed in
exchange for political prisoners. This practice
quickly spread across Latin America as other

consular and ambassadorial envoys were
kidnapped in Argentina, Brazil, Guatemala,
Uruguay, and the Dominican Republic, and held

hostage for the release of political prisoners.
Canada and Spain, both containing separatist
political factions, soon joined the ranks of South
American countries in consular abductions.
There are several alternative explanations for
the occurrence of analogous forms of aggression in
close temporal proximity but in widely dispersed

places. It might be argued that the likeness in
events is sheer coincidence and that, in fact, they

are in no way causally connected. While this is not
inconceivable, the high stylistic similarity in tactics
and simultaneity of episodes make the validity of
such an interpretation highly improbable. A

second explanation, most often advanced in
political circles, is that the rash of aggressive
incidents are related through a conspiratorial plan

under the direction of revolutionary advocates.
The spread of some types of collective aggression,

as in campus protests and urban disorders, is
sometimes aided by visits from militant
spokesmen without any highly organized strategic

collusion. Moreover, it is not uncommon for
modeled violence to be adopted in settings where
there has been no interchange of advocates.
Although conclusive evidence for the
determinants of mass social phenomena is difficult

to obtain, congruence in the style of collective
aggression can be best accounted for in terms of

modeling influences.
New aggressive tactics are usually
disseminated to members of a given community

on the basis of direct contact with actual models.
However, symbolic modeling serves to spread the
modeled patterns to widely dispersed groups.
Newspaper accounts, radio reports, and vivid
televised portrayals furnish detailed examples of

how the new modes of aggressive behavior are
performed, as well as the consequences they are
likely to produce. Hijackers, for instance, learn
piracy behavior from the mass media rather than
from direct observation of a hijacker in action on
an airliner.
News of a sensational crime is often followed
by a sharp rise in criminal violence that continues
to grow at an accelerating rate for some time and

then tapers off (Berkowitz and Macaulay, 1972). It
is noteworthy, however, that the jump in violent

offenses generally occurs a short time later rather
than immediately after the horrifying incident.
Assuming that the two sets of events are not
coincidental, the delayed contagion may be

interpreted in several ways. One possibility is that
a heinous crime, by its shocking nature,
temporarily increases restraints. A second

interpretation is that it may require the
cumulative impact of several modeling influences

to initiate a significant rise in violent actions;
thereafter, spreading violence breeds violence.
Delayed effects may also occur because a variety

of situational influences determines when violent
acts will be performed. In the following example, a
jealous husband lost his fear of killing by watching
Oswald murdered on television, but he did not slay
his wife’s admirer until they met a short time later.

The husband spontaneously commented on the
disinhibitory effects of observing an actual

homicide: “I saw how easy it was to kill a man
when Oswald was shot on TV” (Portland Press-
Herald, 1963). Reports of a masterly extortion
parachute skyjack did not send viewers running

instantly to their local airports with demands for
cash and parachutes. Within the next few months,
however, a number of hijackers, emboldened by

the successful example, copied the specific tactics,
including threats of bombing unless passengers

were exchanged for ransom money and
parachutes (Palo Alto Times, 1972). Any of the
preceding explanations, either singly or in

combination, may account for the temporal
covariation between the inciting examples and
subsequent imitations.
Cross-cultural Studies of Aggression
Variations in aggression training practices

within a given society may cover a limited range of
those found in different cultures. Causal

relationships can be further verified by comparing
the incidence and form of aggression in cultures
providing social learning conditions conducive to
aggression with those in which they are much less

prevalent.
The effects that aggressive modeling and
reinforcement patterns have on cultural
expressions of aggression are strikingly revealed
in ethnographic reports of societies that pursue a

warlike way of life and those that display a pacific
style of behavior. The Dugum Dani, who live in the
New Guinea highlands, exemplify a warrior society
(Gardner and Heider, 1969). In the Dani culture
men from villages separated by agricultural
gardens regularly engage in intertribal warfare
that is one of the most valued activities in Dani life.

The Dani do not fight for land, food resources, or
conquest of opponents; rather, fighting serves
social and spiritual purposes.
Dani warfare, which is highly stylized, is

performed in designated battlefields adjacent to
the villages. Sentries maintain continuous

surveillance from high watchtowers to safeguard
against enemy ambushes. Men are armed
wherever they go so they can be readily
summoned to combat surprise raids. Indeed, much
of Dani life is organized around warfare, including

such activities as extended guard duty, fashioning
weapons, cutting grass to prevent ambushes, and

performing magical practices to secure defense
systems. Formal battles are initiated by shouting a

challenge across the no man’s land. After ritualistic
confrontations between advance bands of
warriors, the combatants, armed with spears and
bows and arrows, engage in repeated brief clashes

of deadly fighting throughout the day.
Though the origin of the institutionalized
warfare remains unknown, fighting is instigated
and perpetuated largely by feared consequences
of unavenged spirits. It is believed that when a

dead warrior is cremated, a ghost is released that
has the power to cause accidents, sickness, crop
damage, and other misfortunes for living relatives
until avenged by the taking of an enemy life. As a

further method of placating malicious ghosts, a
finger of a little girl is amputated and burned, so
that females lose several fingers before they reach

adulthood. In addition to threats of unavenged
ghosts, fighting is prompted by the goading of
women who seek retaliation for family deaths, and
by the status rewards bestowed on able fighters.
Children of the Dani undergo a graduated
training program in learning how to be warriors
that bears on several issues discussed earlier. This

step-wise training is illustrated photographically
in Figure 2.6. Fighting skills are learned and

perfected under simulated conditions long before
the boys engage in actual combat. Young boys are
initiated into fighting techniques by participating
with older peers in a variety of war games that

duplicate conditions of adult combat. In one of the
battle games (spear the seed), berry seeds serve as
opposing warriors who are “killed” by skewering

them with sharpened sticks. The game even
includes a miniature watchtower manned by a

seed warrior. A second game (kill the hoop) is
used to develop skill with spears. The object of this
game is to impale with a lance a bounding vine

hoop tossed by an opposing group. In another
game that duplicates even more closely adult
FIGURE 2.6 A-E Graduated training program used by the Dani
in teaching their children warrior fighting skills. The tutelage
proceeds through war games with seed militia (A); practice of
spear assaults on inanimate moving targets (B); simulated
battles with actual combatants using grass-stem spears to
safeguard against physical injury (C); and observational
learning of fighting tactics by watching from a distance
warriors fighting on the battlefield
warfare patterns, groups of boys battle each other
with grass-stem spears which are not firm enough
to inflict injury. These simulated battles teach the

boys, through example and practice, throwing and
dodging skills, as well as defensive and raiding

strategies of war. The final stage of aggression
training involves observational learning on the
battlefield itself. While warriors of a fighting unit
are engaged in combat, young men watch the

battle from a distance to learn tactics for later use
when they join the ranks of professional warriors.
The Polynesians of the Society Islands present
a marked contrast to the Dani in the socialization

of aggression (Levy, 1969). In the Tahitian society,
physical aggression is rarely modeled, aggressive
behavior is devalued, and unlike the belief system
of the Dani, ancestral spirits serve as punishers
rather than inciters of aggression. Tahitians are
characterized as affable people who are slow to
anger, who quickly get over any ill feelings, and
who lack vengefulness and hostile aggressiveness.
They are disinclined to create anger-provoking

situations, and when they show aggression, it is
generally expressed in words rather than physical
fights. On the infrequent occasions when they
aggress physically, they do so in relatively
harmless and designedly inept ways.
From an early age, Tahitians are made to fear

the consequences of anger arousal and aggressive
actions. Anger is presumed to have toxic effects on
the body and is therefore best avoided. Moreover,

if others are provoked, their spoken threats arouse
ancestral spirits who can hurt the provocateur.
Such anticipated consequences foster conciliatory

behavior, even in highly irritating situations.
Although parents are initially permissive of
mock aggressiveness in young children, they
strongly discourage hurtful actions and prolonged
anger. Control of aggression is largely achieved
with threats of punishment, especially through the

action of spirits who punish aggressors by causing
illnesses and various injuries. Ordinarily, most
verbal threats are ignored without any
consequences, and hence they either lose their

controlling power or retain their effectiveness
only in the presence of the threatening agents.
Tahitian parents, however, periodically reinforce
their threats by capitalizing on false contingencies.
When people get sick or suffer an accident, the

misfortunes are depicted as caused by their earlier
reprehensible behavior. Since the punishing

spiritual agents are omnipresent, they serve as a
more generalized source of control over

aggression than parents or peers. Thus, by
providing few opportunities for learning

aggressive behavior and by removing its
functional value, aggression assumes a minor role
in the Tahitian way of life.
Numerous other examples can be cited in
which different societies foster different
temperaments. In cultural settings where

interpersonal aggression is discouraged and
devalued, people live peaceably (Alland, 1972;
Mead, 1935; Lantis, 1959; Turnbull, 1961). In

other societies that provide extensive training in
aggression and make it an index of manliness or
personal worth, people spend a great deal of time
threatening, fighting, maiming, and killing each
other (Bateson, 1936; Chagnon, 1968; Whiting,

1941). That divergent cultural practices produce
different human natures is even more
convincingly demonstrated by the coexistence for
generations of dissimilar aggressive patterns in
people residing within the same segment or
neighboring segments of a society. Contrasting
socialization practices produced markedly
different characteristics in the American Indian.
The Comanche and Apache raised their children to

be warriors (Goodwin, 1942; Linton, 1945),
whereas the Hopi and Zuni, who embraced
peaceful orientations, reared children with gentle
dispositions (Goldfrank, 1945). To take a more
contemporary example, among the American

Hutterites, who stress pacifism as a style of life,
aggressive conduct goes consistently unrewarded
(Eaton and Weil, 1955). Even though children in
this subculture are subjected to severe and
presumably frustrating socialization pressures,

they show virtually no interpersonal aggression.
On the other hand, within the larger American

society, where aggressive masculinity is widely
promoted as a valued quality, combativeness is
not uncommon. Social class differences in

aggressiveness similarly reflect variations in the
extent to which such conduct is modeled and
reinforced.
One can only speculate on the origins of
diverse cultural practices. Various explanations
have been advanced in terms of geographic,

biological, technological, and social determinism.
To the extent that geographic conditions impose
different survival demands, they may favor certain

social and economic systems over others.
Proponents of biological determinism ascribe
cultural variations in behavior to inherent racial
qualities of constituent members. As previously
shown, man’s biological endowment creates

response potentialities, not preformed patterns of
social behavior. Technological innovations can
have profound consequences that extinguish old
ways of life and force new social arrangements
and practices. Finally, social events and the efforts
of influential men can set in motion influences that
alter the aims, the social organization, and the
reinforcement practices of a society.
Whatever their origins, once social practices

become institutionalized with authorized
sanctions, they tend to be self-perpetuating long
after the historical determinants cease to exist.
The power of social determinism is clearly
demonstrated by groups migrating to new

societies where they transplant their traditional
ways of life and, given sufficient collective
isolation, preserve the patterns over generations.
Thus, the Hutterites adhere to nonaggressive

codes of conduct within a larger society in which
rivalrous and aggressive behaviors have relatively
high functional value. When a subgroup is unable

to maintain its social isolation, as is commonly the
case, the impact of powerful foreign influences
produces fundamental changes in what is valued
and in prevailing modes of behavior. Warlike
tribes eventually adopted peaceful ways of life
under government edict that removed all social
supports for violent activities (Goldfrank, 1943).

Cultures in which aggression is less firmly
entrenched in prestige systems discard fighting
even more rapidly under outside influence. The
Fore, an equalitarian society in the New Guinea
highlands, is a perfect case in point. They promptly

gave up warfare on the arrival of Australian
officials (Sorenson, 1971).
Individuals raised in aggressive societies are
prone to attribute fighting to man's biological

makeup and have difficulty conceiving of people
living peaceably. Researchers coming from these
settings who subscribe to the belief that man

possesses an aggressive drive requiring periodic
discharge selectively search for evidence of
psychological disorders when they study the
people of pacific societies. Considering the
omnipresence of problems of living, the dubious
validity of personality tests, and the elasticity of
referents for psychiatric conditions, one who sets

out to demonstrate that noncombativeness is
hazardous to mental health should have no
difficulty in finding confirmatory evidence,
regardless of the merits of the belief. The
reinforcement customs and habits of aggressive

societies are rarely, if ever, studied by observers
from gentle cultures. Were they to conduct
anthropological field research revealing that in
societies in which aggressiveness is idealized and
cultivated people recurrently humiliate, injure,

and kill each other, they would undoubtedly be
struck with how aggression is generated by man’s

social customs. From the social learning
perspective, human nature is characterized as a
vast potentiality that can be fashioned by social

influences into a variety of forms.
chapter three
Instigators of Aggression
People rarely aggress in blind, indiscriminate
ways. Rather, aggressive actions tend to occur at
certain times, in certain settings, toward certain
objects or individuals, and in response to certain
forms of provocation. As Toch (1969) has noted, if
one wished to observe a stabbing, he could best do
so by visiting a bar or a street corner frequented

by a teenage gang in a slum neighborhood on a
weekend night. Contrary to popular accounts of

aggression as explosive emotional reactions, like
other forms of social conduct, performance of
injurious actions is extensively regulated by
environmental cues (Bandura and Walters, 1959).
A theory of aggression must therefore explain not
only how aggressive patterns are developed, but

also how it is that some stimuli become elicitors of
aggression while others do not. This chapter

examines the processes whereby originally
neutral events acquire aggression-evoking
potential. In addition, it reviews in some detail
research on the types of incidents that function as

predictable elicitors of aggressive actions as well
as on the predisposing conditions.
ESTABLISHMENT OF STIMULUS CONTROL

OF AGGRESSION
Most of the events that evoke aggressive
behavior in humans gain this capacity through

learning experiences rather than from genetic
endowment. Potent aggression elicitors, such as
personal insults, verbal challenges, status threats,
unjust treatment, and provocative aggressive
displays, obviously do not instigate two-year-olds
to fight. As development progresses, however,

people no longer remain indifferent to either
direct or symbolic assaults. A wealth of research
findings shows how aggression can be provoked in
others, but relatively few studies have been
addressed to the issue of how elicitors that are not

innately valenced acquire aggression-arousing
properties in the first place. As is true of other
types of behavior, aggression can be brought
under stimulus control through either paired
experiences or response consequences.
Stimulus-contingent Experiences
The discussion of learning principles in the

introductory chapter explained how formerly
ineffective stimuli acquire the capacity to control
behavior through paired experiences. The
operation of this learning process in aggression is
illustrated by studies of conditioned fighting. Since
researchers cannot record all the learning

experiences of everyday life that serve to endow
stimuli with controlling functions, the basic
mechanism is best examined under laboratory
conditions. In these experiments a previously
ineffective stimulus, such as a tone, is repeatedly

associated with pain-elicited aggression in pairs of
animals (Creer, Hitzing, and Schaeffer, 1966;
Farris, Gideon, and Ulrich, 1970; Lyon and Ozolins,
1970; Vernon and Ulrich, 1966). After a number of
such paired experiences, in which the tone

foreshadows provoked assault, the appearance of
the tone alone tends to produce fighting. Delgado

(1963) similarly found that aggressiveness in
monkeys could be classically conditioned to a
formerly neutral stimulus by associating it with

aggressive behavior originally elicited by brain
stimulation.
In natural situations physical assaults are

ordinarily preceded by a sequence of provocative
threats. Studies reported by Thompson (1969)
show how anticipatory aggressive displays can be

brought under discriminative stimulus control.
Animals learned to exhibit aggressive threats in
the presence of a red light that previously signaled

aggressive activities, but they rarely behaved in a
provocative manner to a green light that had never
been associated with aggression-provoking
experiences. The same learning mechanism is
involved in neutralizing conditioned signals.

Aggression elicitors lose their provocative value as
a result of new experiences in which they are no
longer associated with aggression (Farris, Fullmer,
and Ulrich, 1970).
The learning histories of different individuals
often vary substantially, and hence the incidents
that set off aggressive actions in one person may
have relatively little effect on another. Toch’s

(1969) studies of chronic assaulters give some
indication of how aggression elicitors are
conditioned in the course of natural social
interactions. One of the cases reported suffered a

humiliating beating as a youngster at the hands of
an imposing opponent, a painful incident that
determined his selection of future victims.
Thereafter, he would become violent at any slight
provocation by a large-sized person. These

characteristics acquired such powerful control
over his aggressive behavior that they often

overrode the hazardous consequences of attacking
powerfully built opponents.
Although people learn to dislike and to attack
certain types of individuals on the basis of direct

aggressive encounters with them, aggression
instigators are perhaps more often established
through symbolic and vicarious experiences. In the
former case, the names and attributes of given

individuals are paired with words or pictures
likely to evoke in observers strong negative
reactions on the basis of prior learning. Thus, for
example, formerly neutral events take on negative
valence through repetitive pairings with adjectives

having repugnant connotations (for example, ugly,
dirty), whereas the same things are evaluated
favorably after they have been associated with
words that conjure up pleasant reactions (Insko
and Oakes, 1966; Staats and Staats, 1957). Likes

and dislikes toward familiar names of persons and
nations have also been created using emotionally

charged words (Berkowitz and Knurek, 1969;
Staats and Staats, 1958).
A study by Das and Nanda (1963) further
reveals that evaluative responses established
through symbolic means tend to generalize along
previously established associative networks, thus
producing effects that extend beyond the specific

experience. After associating neutral syllables with
the names of two aboriginal tribes, positive and
negative reactions were developed toward the
syllables by pairing them with affective words.
The tribes took on positive or negative values in

accordance with the evaluative responses attached
to the syllables with which they had been
associated. Similarly, by observing the hatreds
expressed by others, people acquire strong
emotional reactions toward certain things or

classes of people on the basis of little or no
personal contact with them.
Response-contingent Experiences
A second way in which aggressive actions are
brought under the control of environmental cues
is through association with differential response
consequences. When the same behavior is treated
differently depending on the times, places, or

persons toward whom it is directed, these
informative cues come to signify the probable
consequences accompanying certain actions, and
people regulate their behavior accordingly. They
tend to aggress toward persons and in contexts

where it is relatively safe and rewarding to do so,
but they are disinclined to display such behavior
when it carries high risk of punishing
consequences.
The precision of stimulus control of fighting is

again most clearly revealed in animal
experimentation, where the requisite learning

conditions can easily be created. Reynolds,
Catania, and Skinner (1963) reinforced hungry
animals with food whenever they attacked
another animal in blue light, whereas fighting in
green light was never rewarded. Under these
conditions of selective reinforcement, animals
fought predominantly in blue illumination, but

rarely did so in green light. Later, opponents
learned to fight in the presence of the green light
as well when it signified that attack would bring
rewarding results. Thereafter, the animals’
aggressive behavior could be effectively regulated

simply by changing color cues that signaled
probable outcomes—one group attacked in blue
light, the other attacked in green light. Dominance
hierarchies created by the reinforcement
contingencies could be repeatedly reversed simply

by altering the discriminative fighting cues. In blue
situations the blue-rewarded animals behaved in

an aggressively domineering fashion, while the
green-rewarded animals were submissive; under
green circumstances, the social ranking was

reversed. The latter findings would suggest that
the relative stability of dominance relations noted
among animals in their natural habitat partly
results because the social circumstances under

which domineering behavior is originally
established do not change much over time. In a

further extension of the aforementioned
experiment, paired animals were reinforced for
retiring to opposite corners of their cage when
white light was present. By alternating the color

fighting cues and white illumination, the animals
could be induced to fight and then return to their
corners during white light periods between

rounds.
It has been shown that animals will aggress

toward a companion when they must work hard in
order to secure food reinforcement. After a time,
stimuli that have been correlated with aversive
work demands for meager rewards provoke attack
reactions, whereas stimuli that signal a more
beneficent rate of reinforcement virtually never
produce combative behavior (Cole and Litchfield,
1969; Flory, 1969a).
The discussion thus far has focused on the
development of aggression elicitors. Stimulus
control also serves to regulate and reduce
aggression. In social systems based on force,
aggression largely determines social organization.

This relationship is most strikingly illustrated in
field studies of primate behavior. Past successes
and defeats in combat establish a well-defined
dominance hierarchy that inhibits intragroup

fighting. As long as the social order is maintained,
ranking members effectively control aggression of
those subordinate to them by threat displays

alone. Any overt aggression, when it does occur, is
directed downward, usually to the immediately
lower member who represents the most serious
status threat. If the social order is disrupted by
introducing strangers, intragroup fighting erupts
and persists until a new dominance hierarchy is
established (Southwick, 1967). Interestingly,

status threats produced by social changes in
membership have a much greater impact on levels
of intragroup fighting than do environmental
changes such as food deprivation and crowding.
It is easy to find parallels between aggression

in the primate social order and in human
functioning. Like primates, man often fights
competitively for leadership when dominant
figures depart or are overthrown; coercive threats

are used extensively to constrain aggressive
encounters between contending factions; and
forceful challenges of those in positions of

authority bring prompt counterattacks. It would
be a mistake, however, to conclude that aggression
is an inevitable determinant of social organization.
Societies do exist in which the prevailing
reinforcement systems actively promote
nonaggressive qualities as the basis for social
status (Alland, 1972). Perhaps even fighting

among animals could be drastically reduced, if not
eliminated, by removing the rewarding payoffs for
aggression. If aggression is not inevitable, then
social ranking by threat is adaptive for superiors
in maintaining a privilege system to their liking,

but detrimental to the hapless subordinates who
must content themselves with meager offerings.
Writers who extol the adaptive benefits of ranking
created and perpetuated by force are interpreting
the situation from the perspective of victors rather

than victims. Evidence for reinforcement control
of animal aggression suggests that by making

access to food, mates, and the like contingent upon
sharing behavior, a primate society could
conceivably be organized as well around

cooperative as around fighting skills. Kuo’s (1960)
reported success in establishing stable
equalitarian relationships among animals merits
further investigation. Research on aggression

must not only record what exists, but determine
what is possible under different social conditions.
In everyday life the potential consequences of
aggressive actions are most often conveyed by
distinctive features of potential targets—the

positions they occupy in dominance hierarchies,
their social roles, their sex status, their familiarity,
their combative reputation, their size, and so on.
Aggressive behavior is therefore extensively
regulated by anticipated consequences derived

from informative social cues. Without this type of
stimulus control, people would repeatedly aggress
foolhardily without concern for the probable
effects of their actions.
MODELING INFLUENCES
The preceding chapter contained an analysis of
how new patterns of aggression are learned
through the examples provided by aggressive

models. The expression of preexisting aggressive
responses, however learned, is also extensively

regulated by the actions of others. Numerous
experiments have investigated the effects of
exposure to live and symbolic models on
subsequent interpersonal aggressiveness.
A major difficulty in studying the conditions
that lead people to behave punitively arises
because, for obvious humane reasons, one cannot
incite individuals to inflict injury on each other.

Buss (1961) has devised a procedure that
provides individuals with opportunities to engage
in physically injurious behavior without the victim
experiencing any pain. The aggression apparatus
used for this purpose consists of a panel
containing ten shock buttons arranged in
increasing intensity, together with lights that
signal right and wrong responses ostensibly made
by the victim in an adjacent room. The task is

typically presented as a study of the effects of
punishment on learning, in which participants
shock the “learner” every time the apparatus
signals that he has made a mistake. They are free
to vary the length and the intensity of the shocks

administered. In some studies, participants simply
judge the quality of another person’s performance
by administering more shocks of their own
choosing the more they devalue his work.
Aggressors usually sample a few shocks at the

outset to familiarize themselves with the severity
of pain corresponding to the different shock

buttons. Unknown to the aggressor, however,
there either is no target person in the adjacent
room or the shock electrodes are inoperative and

the intended victim suffers no pain.
During the task a given number of wrong
responses is signaled on the aggressor’s panel, and
the duration and intensity of the shocks he inflicts

on the victim are recorded. Since aggressors are
clearly aware of what shock intensities they are

administering, whereas they may be relatively
inaccurate in judging the duration of each shock,
the intensity measure generally yields stronger
relationships to variables that are known to

determine aggressive responsiveness. Participants
who do not wish to behave cruelly can easily meet
the task requirements by choosing low-intensity

buttons that would cause no discomfort and by
curtailing their duration. On the other hand, those

who select painful shocks and keep their thumbs
on the shock button for relatively long periods are
acting in an unnecessarily punitive manner.
In a somewhat different procedure (Taylor,
1967; Shuntich and Taylor, 1972), participants
compete in a task ostensibly with another person
in an adjoining room, though in fact there is no
opponent. Participants select at the beginning of

each trial any of five intensities of shock they wish
their opponent to receive when he is defeated. On
trials in which the opponent presumably wins, the
participants receive the shock their opponent set
for them. With this type of procedure one can

study initiation and retaliation of aggression in
sequential interchanges under varying degrees of
opponent provocation.
A number of studies using some of these

approaches have been conducted to elucidate how
modeling influences, either singly or in
conjunction with other determinants, affect

people’s interpersonal aggressiveness. Epstein
(1966) investigated the interaction effects of
observers’ authoritarian attitudes and the models’
ethnic and socioeconomic characteristics on
aggression directed toward a black victim. White
college students varying in authoritarianism
observed either a white or a black model

aggressing toward a black person. For half the
students the models represented, in dress and
interview replies, a lower-class rank; for the
remaining students the models conveyed a
middle-class status. On the assumption that high

authoritarians tend to be conforming and sensitive
to power relations, they were expected to be more
imitative of punitive models, especially middle-
class ones, than low authoritarians. It was further
reasoned that seeing a minority model attack a

member of his own group would legitimize such
behavior; a black aggressor would elicit from

white observers more punitive treatment of a
black victim than would a white aggressor.
Compared to students who had no exposure to
aggressive modeling, those who observed punitive
models subsequently administered two to three
times as many shocks that were labeled “very
strong’’ and that evoked vocal expressions of pain

from the victim. Although the models’ social status
did not exert a differential effect, the observers’
authoritarianism and the models’ ethnic
characteristics were important determinants of
imitative aggression. High authoritarians were

more imitatively aggressive than lows, and the
black model evoked greater punitive behavior
than the white aggressor. Some data also
suggested evidence that the imitative aggression
of people who have strong authoritarian attitudes

is relatively indiscriminative. Whereas low
authoritarians refrained from imitating the white

model but were responsive to the influence of the
black one, high authoritarians readily adopted the
punitive actions of the models regardless of their

ethnic characteristics.
The present study cannot establish whether
the black model induced more imitative
aggression because the victim of his attack was

ethnically similar, was of minority status, or
because of other factors. To clarify this

relationship would require an experimental design
in which the ethnic characteristics of observers
are varied and the target persons include whites
as well as members of minority groups that differ

from the model’s ethnic affiliation.
Modeled punitiveness can disinhibit
aggression in children as well as in adults. Hoelle
(1969) measured the willingness of young boys to

shock a peer for incorrect answers after they
observed a model using either high or low shock
intensities exclusively, or when no model was
observed. Punitive modeling produced more
intense aggressive responsiveness than displays of
subdued aggressiveness or no modeling, which did
not differ from each other.
Most studies of the behavioral effects of
modeling present the aggressive models on film

either to ensure standard portrayal of the action
or because of intrinsic interest in the influence
potential of symbolic modeling. In an early
'experiment of this type, Walters and Llewellyn-
Thomas (1963) found that both adults and

adolescents who watched the knife-fight scene
from the motion picture Rebel Without a Cause
behaved more punitively toward another person,
whereas a control film on art work had no

influence on their aggressiveness. The aggression-
enhancing effects of both real-life and symbolic
aggressive modeling have since been abundantly

documented by other investigators studying
diverse populations.
It has been assumed by advocates of the
catharsis hypothesis (Buss, 1961; Feshbach, 1964,
1970) that the cathartic or drive-reducing effects
of aggressive modeling occur only under certain

specified conditions. Observation of aggression
supposedly reduces subsequent injurious
behavior only if observers are angered at the time
of exposure. If, on the other hand, the aggressive

drive has not been aroused, vicarious participation
in aggressive activities will heighten ensuing
aggressiveness. The catharsis theory thus
presupposes that the functional properties of
aggressive modeling can be radically altered by

transitory emotional states of the observer.
An aggressive response sequence generally

contains two important elements—the aggressive
actions of the attacker, and the pain expressions of

the victim. Although a number of researchers have
reported changes in punitiveness following

exposure to fight sequences, no attempt had been
made to determine whether the pain expressions
or the instrumental aggression depicted were
primarily responsible for the obtained effects. It is

apparent from results of experiments employing
inanimate targets that observation of aggressive
acts in the absence of pain cues can produce
substantial increases in aggressive behavior.
There are conflicting views, however, on whether

expressive pain reactions augment or counteract
the effects of aggressive displays. According to the
theory of aggression proposed by Sears, Maccoby,
and Levin (1957), signs of pain and injury
resulting from a child’s aggressive behavior occur

sufficiently often in conjunction with the removal
of his frustrations that seeing others in pain can

acquire rewarding value and thus enhance
aggressiveness. In contrast, it would be predicted
from principles of generalization and vicarious

arousal that expressions of pain would elicit
empathetic and anxiety reactions that serve as
aggression inhibitors. Functional relationships
could be even more complicated if the rewarding

and inhibiting properties of witnessed pain were
significantly altered by the observers’ level of

anger arousal.
Hartmann (1969) examined the independent
and interactive effects of anger instigation,

aggressive actions, and pain expressions on
subsequent interpersonal punitiveness.
Adolescent delinquent boys who had or had not
been angered were shown one of three films, each
of which portrayed two adolescent boys tossing

balls on a basketball court. In the control film the
boys engage in active but noncompetitive play,
whereas in the other two films the boys get into an
argument that develops into a fistfight. The pain-
cues film focuses almost exclusively on the
victim’s pain reactions as he is vigorously
pummeled and kicked by his opponent. The
instrumental-aggression film, on the other hand,
centers on the assailant’s actions, including angry

facial expressions, foot thrusts, flying fists, and
hostile remarks. After exposure to their respective
films, the boys were provided with an opportunity
to shock the peer who earlier had either angered
them or treated them in a considerate manner.
Results of this experiment, based on several

indexes of aggressiveness, are contrary to the
catharsis hypothesis, but consistent with social
learning theory. Boys who had observed either the

aggressive acts or the pain expressions
administered more severe shocks, both under
aroused and nonaroused conditions, than boys

who watched the nonaggressive film (Figure 3.1).
Moreover, angered viewers behaved more
aggressively than nonangered ones after
observing aggressive activities, a result that is
FIGURE 3.1 Mean shock intensities administered by angered
and nonangered adolescents after observing either
instrumental aggression, pain expressions of the victim as he is
being assaulted, or a nonaggressive control film. (Hartmann,
1969)
directly counter to the catharsis hypothesis. Prior
anger arousal also heightened aggression in
control subjects but not significantly. The finding

that anger instigation combined with aggressive
modeling increased punitiveness, but anger alone
did not, provides additional evidence for the role
of modeling influences in determining aggressive
responsiveness to emotional arousal.
Although boys who watched either aggressive

actions or their painful consequences showed
similar increases in the intensity of shocks
administered, these two groups differed when

punitiveness was measured in terms of both
intensity and duration of shock stimulation
(Figure 3.2). Nonangered boys were more punitive

when they had witnessed another person severely
beaten. The relationship between punitiveness
and suffering feedback will depend upon, among
other factors, who the assailant is, the strength of
FIGURE 3.2 Mean punitiveness (shock level x duration)
displayed by angered and nonangered adolescents after
observing either instrumental aggression, pain expressions of
the victim as he is being assaulted, or a nonaggressive control
film. (Hartmann, 1969)
his aggression restraints, and the severity of the
injuries he causes. Evidence will be cited later
indicating that a victim’s suffering is more likely to

lessen attacks when the injurious consequences
are produced by the aggressor’s own actions than
by observed assaults. Moreover, seeing horrid
results of aggression by others may reduce
punitiveness in observers by exposing the

potential dangers of such conduct, whereas
expressions of suffering alone may have few or no
inhibitory effects (Goranson, 1970).
It will be recalled that in the motivational

analysis of aggression from a social learning
perspective, the expression of aggressive behavior
is partly controlled by emotional arousal and by

reinforcement influences. A number of modeling
studies have been reported that are pertinent to
these assumptions. One would expect people to
aggress more freely toward victims who cannot
retaliate than toward those who are apt to
counteraggress. Baron (1971b) examined the
extent to which responsiveness to modeling

influences is affected by threat of retaliation.
Seeing another person inflict punishment on a
provocateur raised the level of punitiveness in
angered observers when they were also given an
opportunity to get back at him, but threat of

retaliation significantly reduced aggression only in
subjects who had no exposure to aggressive
modeling.
It is surprising that the almost certain

likelihood of reprisal did not significantly
attenuate the instigative effects of modeling.
Possibly this finding reflects the operation of

arousal processes. Under high anger arousal,
aggressive behavior may be under relatively poor
outcome control. That is, an enraged person is
likely to attack and then to worry about possible
consequences of his actions if he sees others
aggress unconcernedly toward a common
antagonist. Under low instigation, however,

observers are likely to be more heedful of how
their actions may affect both themselves and their
victims.
The research discussed thus far demonstrates
that aggressive modeling raises the level of

aggression shown by others. Conversely,
aggression can be reduced in observers by
exposure to models who behave in a restrained,
nonaggressive fashion in the face of provocation.

This is revealed in a study by Baron and Kepner
(1970), who measured the punitiveness of
angered college students after they had seen a

model punish the insulter severely, respond to
him in a nonaggressive way, or had observed no
models. As shown in Figure 3.3, students who
watched the punitive model exhibited the highest
FIGURE 3.3 Mean intensity of shocks administered by adults
depending on whether they had previously observed
aggressive models, nonaggressive models, or no models, and
whether they had high or low attraction toward the model.
Within each of the three modeling conditions, the solid lines
represent the high attraction subgroups and the dashed lines
represent the low attraction subgroups. (Baron and Kepner,
1970)
amount of aggression, whereas those exposed to a
nonaggressive model performed relatively little
aggression and were significantly lower in this

respect than students who did not see the
reactions of others. Whether observers liked or
disliked the model had no appreciable effect on
the degree to which they were affected by his
behavior. The latter finding is consistent with

previous results showing that aggression can be
readily disinhibited by models regardless of their
attractive qualities (Bandura and Huston, 1961;
Madsen, 1968), and even by successful but
disliked models (Bandura, Ross, and Ross, 1963b).
Baron (1971a) presents additional suggestive
evidence that under multiple modeling influences,

the presence of a restrained model can counteract
the instigative effects of an aggressive one. Thus,
angered students exposed successively to a model
who aggressed against their annoyer and one who
behaved temperately were not only less punitive
than those observing only an aggressive example,
but did not differ in this respect from subjects

provided with no social guides for their actions.
Although modeled restraint can often serve to
avert violence, it is apt to be less effective when
observers are highly angered (Waldman and
Baron, 1971) or when aggression produces good

results.
Explanation of Modeling Effects
Social learning theory postulates several

processes whereby modeling influences enhance
expression of previously learned aggressive
responses. As noted earlier, the actions of others
function as social cues in facilitating similar

behavior in observers. People often do what they
see others doing. In this response facilitation
process, a model’s behavior simply provides an
external inducement for similar actions that can
be performed with ease.
Modeling influences acquire response-

activating power through prior differential
reinforcement. The behavior patterns widely
adopted within a given culture are usually those of
proven functional value, so that new members are
spared the hazards and travail of unguided trial-

and-error groping. In most instances, therefore,
behaving like others is rewarding, whereas
dissimilar actions are often less effective or may
even bring disapproval. As a result of such

differential consequences, modeled actions come
to serve as informative cues that similar
performances are likely to be positively

reinforced. Conversely, when contingencies are
reversed, so that analogous responses are never
reinforced but dissimilar behavior is consistently
rewarded, imitativeness is rarely exhibited (Miller
and Dollard, 1941).
Observers obviously differ in the extent to
which their behavior is guided by the actions of

others, and not all models are equally influential in
evoking the types of behavior that they themselves
exemplify. Responsiveness to modeling influences
is largely governed by three sets of variables.
These include, in decreasing order of importance,

the reinforcement contingencies associated with
matching behavior, the attributes of models, and
observer characteristics.
The principles that explain how modeling

stimuli acquire the capacity to evoke similar
responses account as well for the differential
influence of various model and observer
characteristics. Models differ in the extent to
which their behavior is likely to be successful in
producing valued outcomes. People are most
frequently rewarded for following the behavior of
models who are intelligent, who possess certain
social and technical competencies, who command

social power, and who, by virtue of their
adroitness, occupy high positions in various status
hierarchies. In contrast, the behaviors of models
who are ineffectual, uninformed, and who have

attained low vocational, intellectual, and social
status are apt to have considerably less functional
value. Identifying attributes and status-conferring
symbols thus assume informative value in
signifying the probable consequences resulting

from behavior modeled by different people.
Through a process of generalization, the effect of a

model’s prestige carries over from one area of
behavior to another and to unfamiliar persons, to

the extent that they share similar characteristics
with models whose behavior has been effective in

the past. Similarly, individuals who achieve better
outcomes with minimal effort costs from following
the examples of others than from their own
independent endeavors are apt to be most

responsive to modeling influences.
Aggressive behavior, especially when cruel,

invidious, and lacking adequate justification, is
often socially censured if not self-condemned.
Consequently, anticipated punishment exerts a

restraining influence on aggressive actions. A
second major function of modeling influences is to
strengthen or to weaken inhibitions of responses
that already exist in observers’ repertoires
(Bandura, 1971c). The effects that modeling exerts

on behavioral restraints are largely determined by
observing the consequences engendered by the
models’ responses. Seeing the behavior of models
punished generally suppresses similar
responsiveness in observers. On the other hand,
observing models engage in threatening or
disapproved activities without experiencing any
adverse consequences reduces observers’
restraints toward behaving in a like manner. In

social groups where aggressive conduct is
regarded as emulative behavior, aggressive
modeling influences are primarily instigational,
whereas in those groups that censure such actions,
modeling may also serve a disinhibitory function.

Usually both instigational and disinhibitory
processes are involved.
Modeling may produce disinhibitory effects in
several different ways. When people respond

approvingly or even indifferently to the actions of
assailants, they convey the impression that
aggression is not only acceptable but expected in

similar situations. By thus legitimizing aggressive
conduct, observers anticipate less risk of
reprimand or loss of self-respect for such actions.
In addition, observing models perform anxiety-
provoking responses without experiencing any
untoward consequences results in vicarious
extinction of fears and behavioral inhibitions.

These sorts of changes are most strikingly
illustrated in the treatment of severe phobic
conditions through modeling procedures
(Bandura, 1971a). People who are terrified of
snakes, for instance, and cannot approach them

are generally able to handle them without
experiencing fear arousal after seeing bold
performers engage in threatening interactions
with snakes. Cline, Croft, and Courrier (1972)
report a similar extinction process in response to

aggressive modeling. Children initially showed
high emotional arousal to filmed aggression, but

they exhibited progressively less emotional
responsiveness during successive viewings of the
same violence.
Reduction of inhibitions does not mean that
observers will attack people indiscriminately, as
some authors mistakenly contend (Berkowitz,
1965b; Berkowitz and Geen, 1966). Nor does the

appearance of aggressive behavior as a result of
restraint reduction imply that a formerly bridled
urge to attack is set free or that individuals are
internally impelled to aggress by liberated “habits”
(Berkowitz, 1970). These misunderstandings can

best be clarified by considering cases in which
rigid restraints over other forms of behavior are
eliminated. Obviously, snake phobics who have
had their inhibitions removed do not go around
randomly stroking reptiles, whether venomous or

not. On the occasions when they do handle snakes,
they are externally prompted to do so rather than

internally propelled by aroused serpent-caressing
habits seeking expression. Indeed, how can
“habits,” whatever they may be, activate

themselves?
When inhibitions are reduced, behavior is
more subject to cognitively mediated control.
Actions are now self-regulated by anticipated

social and other external consequences, by self-
evaluative responses to one’s own behavior, and

by other considerations. Behavior is not only
discriminatively performed in accordance with the
effects it is likely to produce, but must still be
activated and guided by various stimulus

conditions. What a disinhibitory interpretation
does imply is that, given appropriate instigation,
responses will be performed more rapidly, more

strongly, or more often when behavioral
inhibitions are weak than when they are strong.
Aggressive modeling can have additional
psychological effects, though these may be of
lesser importance. The behavior of models directs
observers’ attention to the particular objects used
by performers. As a result, observers may
subsequently use the same objects to a greater
extent, though not necessarily in an imitative way.
In one experiment (Bandura, 1962), for example,

children who had observed a model pummel a
plastic figure with a mallet spent more time
pounding other things with a mallet than those
who did not see a person handle this particular
weapon. Stimulus enhancing effects are

distinguished from response facilitation in that the
observers’ behavior in the former case, although
the same implements are used, may bear little or
no resemblance to the modeled activities.
It has been further shown that observation of

others fighting produces emotional arousal, as
measured physiologically, in children (Cline, Croft,

and Courrier, 1972; Osborn and Endsley, 1971), in
adults (Ziilmann, 1971), and in animals (Welch
and Welch, 1968). The resultant emotional arousal
may augment the strength of aggressive
responding. Much of the effects of exposure to
violent displays is, in fact, attributed by
Tannenbaum (1972) to their arousal rather than

their content properties. This interpretation,
though partially valid, requires certain
qualifications. Social conduct is learned from
modeled contents, not from physiological
activation per se. Combat strategies, for example,

are better acquired by observing the actions of
skillful fighters than from seeing equally arousing
amorous behavior enacted in a boudoir.
Differential sanctions for aggressive behavior,
conveyed through content, can substantially affect

the degree of punitiveness elicited by assaultive
models under the same level of excitation. And

finally, depicted catastrophic consequences tend
to reduce aggressiveness even though they
generate higher emotional arousal than less

frightening outcomes resulting from the same
assaultive actions. The empirical evidence, taken
together, reveals that models can serve as
teachers, as elicitors, as disinhibitors, as stimulus

enhancers, and as emotion arousers. A complete
account of how models affect observers must

encompass the diverse functions of modeling
influences.
Researchers who have advanced views on

aggression differ in the relative emphasis they
place on the eliciting and disinhibitory functions of
modeling influences. In explaining contagion of
aggression, Wheeler (1966), for example, assigns a
prominent role to the reduction of internal

restraints against aggressive actions. People who
have been provoked presumably experience a
conflict between an urge to aggress and their fear
of displaying such behavior. Observation of others
responding aggressively reduces observers’ fears,
thereby enabling them to perform similar actions.
Wheeler and his associates have explored
several implications of an explanation of
aggressive modeling in terms of disinhibitory

processes. In the procedure employed, which is
concerned solely with modeling of verbal

aggression, adults (typically naval recruits)
express opinions on a number of topics via an
intercom system and then supposedly evaluate
each other’s judgments. Unknown to the listener,

however, the verbal interchanges he hears are
prerecorded to create different social influences.
One of the recorded participants, who assumes the

instigator role, arrogantly voices offensive views,
while the second recorded participant, the

aggressive model, derides him with varying
outcomes. Following exposure to aggressive
modeling, listeners' verbal aggression toward the

obnoxious annoyer is measured.
As previously noted, behavioral restraints are
largely determined by the effects that modeled
aggression produces. Results of these studies
provide only partial support for a restraint

reduction process. Listeners were equally verbally
abusive toward the obnoxious character
regardless of whether they heard him
counterattack, temper his original remarks, or
show no response after the aggressive model

berated him (Wheeler and Caggiula, 1966).
Although listeners were more restrained after
hearing an authority figure reprimand the model
for attacking the provocateur, neither model self-
censure nor peer disapproval of the model

affected hostile responsiveness in the listener
(Wheeler and Smith, 1967). Wheeler and Levine

(1967) hypothesized that in circumstances where
restraints arise from fear of punishment by an
external agent, similar and dissimilar models will

disinhibit aggression with equal ease. As it turned
out, subjects who observed a dissimilar aggressor
later expressed more hostility than those who
heard an aggressor resembling them in many

characteristics. The authors reasoned that seeing a
person who is quite different get angry at an

irritator provides greater justification for one’s
own annoyance than observing someone with a
similar outlook.
Some of the discrepancies between Wheeler’s

findings and those of others regarding the
inhibitory effects of model punishment are
probably attributable to the nature of the
treatment conditions. The annoyer behaved so

obnoxiously that the modeled reprimands would
appear justified regardless of the sentiments of
others. Model punishment, of course, does not
automatically produce response inhibition. The
degree to which observed outcomes affect
aggression restraints will be largely determined
by the comparability, assumed or real, of response
consequences customarily applied to models and
observers. Social learning theory therefore assigns

greater weight to similarity of probable
consequences than to similarity of personal
attributes in predicting modeling effects. Those
who censured the model for his hostile attack
differed in potential control over subjects’ future

well-being. The authority figure, presented as a
commanding officer, could produce repercussions
for naval enlistees beyond the experimental
situation should they misbehave, whereas
unfamiliar peers could in no way get back at them.

It would be of interest to study the effects of
observing response consequences to aggressors

under different levels of provocation, and where
the characteristics of censuring agents and their
controlling power are varied. It might be noted

parenthetically that aggression can be most
effectively disinhibited by praising and otherwise
rewarding the aggressor for his daring actions.
Berkowitz (1965a; 1970) conceives of
modeling stimuli principally as activators of
implicit aggressive responses, although sometimes

he ascribes changes in aggression to disinhibitory
processes as well. According to his motivational
analysis, aggressive responsiveness requires both

a state of anger arousal and the presence of
environmental stimuli that possess aggression-
activating properties. Anger arousal “primes” the
organism for aggressive actions; the stimuli
trigger them off. Under low anger arousal a

powerful releaser is considered necessary to elicit
aggressive responses, but a relatively weak
external stimulus will suffice under high
instigation. Berkowitz’s formulation thus provides
a revised frustration-aggression theory that differs
from the traditional hypothesis mainly in the role
assigned to external cues. Whereas the traditional
version depicts aggression as largely impelled by
internal excitation, Berkowitz views frustration

produced arousal as merely a predisposing state
requiring an appropriate external releaser for
aggression to occur.
The assumptions of Berkowitz’s formulation
have been tested in a series of experiments in

which punitiveness toward an antagonist is
measured as a function of anger arousal and
exposure to stimuli varying in aggressive cue
value. In the standard design (Berkowitz, 1970),

college students are treated insultingly or
neutrally and then shown a prize fight scene from
the movie Champion or a nonaggressive film. In

addition, the provocateur’s cue value for
aggression is increased either by assigning him the
same name as the victim of the filmed violence or
an aggressive role such as a college boxer. The
combined influence of anger arousal, aggressive
modeling, and aggressive labeling of the
antagonist generally produces the highest level of

aggression. However, evidence that punitiveness
can be enhanced by anger arousal and aggressive
modeling in the absence of aggression-activating
cues and by aggressive cues in the absence of
anger arousal (Berkowitz, 1965b; Berkowitz and

Geen, 1967) indicates that these factors should be
regarded as facilitative rather than necessary
conditions for the occurrence of aggression.
The justifiability of modeled aggression was

varied in some of the above studies in order to
determine whether modeling influences may
produce their effects partly through reduction of

inhibitions. For half the observers the victim in the
filmed fight was portrayed as an unprincipled
scoundrel who presumably deserved the
thrashing; for others, the victim’s behavior was
depicted as an outgrowth of adverse
circumstances that made the aggressor look
somewhat malevolent. Observation of socially

justified aggression evoked more punitive
reactions in observers than unwarranted assaults.
In a related study, Hoyt (1970) found that
exposure to filmed aggression in which the
victorious aggressor was portrayed as avenging a

past unfair beating generated more aggression in
viewers than exposure to a film in which the
aggressor was portrayed as defending himself
against a dangerous attack, or to the same film

shown without references to any justifying
circumstances. Evaluation of these results with
respect to the disinhibitory effects of justifiability

must await independent evidence that aggression
for vengeance is more excusable than for self-
defense. The finding that self-defense justification
did not produce more aggression than portrayal of
brutal assault without extenuation presents
difficulties for an interpretation of the data solely
in terms of justification processes. A study by Geen

and Stonner (1972) reveals further complexities of
the problem. Modeled assaultiveness depicted as
vengeful was more effective in disinhibiting
retaliatory aggression in angered individuals than
if it was presented as a professional activity

devoid of animosity. Unangered people, on the
other hand, rated themselves as less restrained
and behaved more aggressively after seeing
professional rather than vengeful violence. The
direction of the causal relationship is uncertain,

however, because self-reported restraints were
measured after the people had aggressed, and it is

conceivable that they judged their restraints from
their punitive actions.
Taken as a whole, the data indicate that
modeling of defensible violence adds legitimacy
and thereby weakens restraints over the use of
aggressive solutions for social problems. Findings
of this line of research suggest that broadcasting

codes requiring eventual punishment of criminal
acts can inadvertently reduce aggression
restraints in viewers when respected enforcement
agents inflict violence on culprits with justified
impunity. There is always the uncertainty as to

whether righteous violence in the moral lesson
will legitimize aggressive conduct or depicted
punishment will inhibit transgressive behavior.
At this point, it might be instructive to compare

the different views advanced on the conditions
under which aggressive modeling will enhance
punitive responsiveness. Both Wheeler and

Berkowitz postulate anger provocation as
necessary for aggressive modeling to have an
effect. Social learning theory, on the other hand,
regards anger provocation as one form of
emotional arousal that serves as a facilitator
rather than as a prerequisite for aggressive
modeling. There is substantial empirical evidence

that differs from the former line of theorizing, but
lends validity to the social learning interpretation.
Although both Wheeler and Berkowitz generally
find that modeling without provocation is
relatively ineffective, other investigators (Epstein,

1966; Hartmann, 1969; Hoelle, 1969; Walters and
Llewellyn-Thomas, 1963) working with diverse
populations have clearly demonstrated that
unangered observers behave more aggressively
after seeing aggressive models than those who

have had no such exposure.
The divergent results may arise in part from

different procedures in measuring aggressive
responses. In Berkowitz’s studies subjects are
asked to evaluate the victim’s solution of a
problem by delivering shocks ranging from one
shock for an excellent solution to ten for a poor
solution. The task therefore requires subjects to
administer multiple shocks when they judge the

solutions insufficient regardless of whether or not
they may want to hurt the target person. The
measure thus incorporates both performance
evaluations and punitiveness. Some people may
administer many shocks because they judge the

performer’s solution in fact inadequate, while
others may do so because they want to hurt him.
This ambiguity about what the shock response
represents does not exist in studies where
subjects are free to respond to the victim’s

performances by administering very weak shocks
that produce no discomfort or highly painful ones

that clearly represent needlessly punitive
treatment. The possible confounding effects of
performance evaluation could be easily reduced in

the Berkowitz procedure by allowing subjects free
choice in the intensity of shocks they may deliver.
Of subjects who considered a given solution
sufficiently inadequate to warrant seven shocks,

for example, those who selected the weakest
intensities would be behaving nonaggressively,

whereas those who inflicted painful shocks would
be acting punitively. Studies employing
unequivocal measures of aggression probably
yield the more conclusive findings.
Interpretive problems also arise when anger
arousal is induced in subjects by shocking them,
because it introduces modeling influences.
Receiving many or few shocks for one’s own

solutions in an ambiguous situation can
significantly alter the standards for judging the
work of others, independently of whether one is
angered or not. Provocateurs administering many
shocks are conveying a much higher achievement
requirement than those delivering few shocks.
Observers tend to match the shock intensities
administered by others (Baron and Kepner, 1970;
Epstein, 1966; Hoelle, 1969). There is every

reason to expect that they would match the
frequency of shocks as well. A person who
receives many shocks may therefore deliver many
because he has been angered, because the
behavior of the other person sets a high standard

for him, or because of both factors. Confounding of
stringent modeling by the antagonist with the
anger aroused by his treatment makes it difficult
to evaluate findings of studies in which aggressors
are presumably angered by differential shocks

from the person who later serves as their target.
Differences obtained in such experiments are

attributed to anger arousal, whereas they may
reflect, in varying degrees, the influence of
differential aggressive modeling. This

methodological problem does not exist when
arousal is induced by insult, by arbitrary
thwarting, or by other means that differ from the
form in which aggression is later measured. When

the provocateur angers verbally and the assailant
attacks him physically, specific modeling is

essentially precluded. In any event, substantial
evidence exists that contagion of aggression does
not require anger arousal.
Another issue that deserves comment concerns

the conceptualization and functional role of
external stimuli in the regulation of aggressive
behavior, especially as they relate to symbolic
modeling. Wheeler conceives of modeling

influences principally as disinhibitors of
aggression. Berkowitz considers aggressive
modeling stimuli to be elicitors, but incapable of
producing aggressive actions. In this approach,
even the combined effect of anger arousal and
aggressive modeling is considered insufficient to
increase aggressive responding. As Berkowitz
summarizes his own view: “Observed violence
serves mainly to arouse the observer’s previously

acquired aggressive habits—to set them in
operation, but only in ‘low gear,’ so to speak.
Before these habits can erupt into a manifestly
aggressive act, certain elicitory cues must also be
present in the environment” (Geen and Berkowitz,

1966, p. 456). The implied mechanisms that limit
the function of modeling stimuli are not explained,
however. If aggressive habits are activated but not
in action, where are they manifested and in what
forms? In more recent elaborations of his position

(Berkowitz, 1972), the implicitly evoked events
include aggressive thoughts. The question

remains, however, if aggressive modeling stimuli
can activate implicit aggressive responses, why is
it that they cannot evoke aggressive actions as

well?
Unlike Berkowitz’s theory, which assigns
response-priming functions to anger provocation
and modeling stimuli, and response-eliciting

functions to other environmental stimuli, social
learning theory regards all three sources of

influence as having response-activating
properties. Aversive stimulation, in the form of
personal insult or physical assault, is in fact a
much more potent elicitor of aggression than

some environmental cue that has merely been
previously associated with anger-arousing
experiences. Thus, people often counteraggress

when insulted or struck without requiring
additional stimulus prompts; they aggress on

slight provocation when they see that such actions
are effective in gaining desired outcomes. And
individuals who possess certain characteristics

that are disliked by others as a result of past
experiences are quick to elicit aggressive
responses. In the social learning view, the more of
these instigators that are present, and the greater
their potency, the higher the likelihood that
aggression will occur. The tendency to view anger

provocation as a predisposing drive state rather
than as a response activator probably stems from
the fact that its effects on behavior may persist

after the inciting incident has ceased. The point
will be developed later that the effects of
instigation endure not because of an undischarged
drive, but because the provocation is re-created

symbolically and, analogously to their
environmental counterparts, infuriating thoughts
also serve as symbolic stimuli for aggressive

actions.
The equation of aggression with aggressive
drives and injurious aims has restricted the range
of determinants typically investigated in
laboratory studies. Since, in social learning theory,
aggression has multiple determinants and
multiple functions, comparative tests of the
predictive power of alternative theories require a

greater array of experimental treatments than is
ordinarily employed. Aggression ascribed
specifically to anger may be more generally
determined by emotional arousal. Data on this
issue are sparse, but the evidence that does exist is

consistent with the latter position. Aggression has
been enhanced by sexual arousal (Zillmann, 1971),
by nonirritating noise (Geen and O’Neal, 1969), by
epinephrine (Schachter and Singer, 1962), and by
success experiences (Christy, Gelfand, and

Hartmann, 1971) in much the same way as by
anger provocation. Empirical tests of catharsis and

frustration-aggression hypotheses should
therefore include conditions containing
nonfrustrative sources of arousal as well.
In social learning theory, response
consequences constitute a powerful determinant
of aggressive responsiveness. It would be of
theoretical interest to compare the relative power

of anger arousal and positive reinforcement
without anger provocation in producing
aggressive behavior that is socially injurious.
Considerable evidence showing that hurtful
aggression can be enhanced, sustained, and

eliminated simply by altering its consequences
calls into question theories that postulate anger
arousal as an essential and primary determinant
of aggression.
In Berkowitz’s formulation disliked persons

serve as releasers of aggressive responses as a
result of verbal and other associations with

previous instigators. It will be recalled from an
earlier section that there are two basic operations
by which formerly neutral stimuli can acquire
aggression-controlling functions. They can become
informative or discriminative stimuli through
association with differential response
consequences. People experience different

outcomes for aggressing toward individuals who
differ in status, power, age, sex, and many other
attributes. A variety of personal characteristics
come to signify probable consequences for
aggressive behavior. Through a similar process of

selective reinforcement, situational, temporal, and
other cues acquire informative value that aid in
regulating one’s actions. People therefore attack
not only those whom they have learned to dislike,
but also those whom it is relatively safe to attack

and those whom it is advantageous to attack. The
effects of anticipated consequences, depending on

their nature, may enhance, attenuate, or even
override the eliciting power of social
characteristics. A less disliked person who cannot

easily counteraggress is more likely to be selected
as a target than a highly provoking person who
has the power to retaliate.
Cues can become eliciting stimuli for
aggression through contiguous association with
aggression-provoking experiences. Classically

conditioned stimuli presumably elicit aggressive
behavior regardless of its consequences.
Berkowitz (1970, 1972) favors a classical

conditioning explanation, though his speculations
about how the control process works carry
additional implications. He advances the view that
aggressive responses are automatically and
involuntarily elicited by aggressively valenced

stimuli. On the other hand, he notes further that
such stimuli are not in fact directly connected to
overt behavior; rather, they evoke thoughts and
feelings according to the meaning attributed to
them, and it is these intervening events that serve
as the prompts for aggressive responding. In a
two-stage causal process, external stimuli may
very well function as reliable elicitors of hostile
thoughts, but there is abundant evidence that

thoughts do not automatically produce
corresponding actions. A major reason for the
weak relationship is that irritating treatment
arouses not only retaliatory ideas, but also
thoughts of likely response consequences which,

depending on their nature, can either facilitate or
inhibit hurtful actions. The two stimulus functions
are not mutually exclusive. Indeed, as the
preceding analysis shows, both eliciting and
response-regulating influences usually operate

jointly. Hence, a complete theory of aggression
must consider both stimulus functions in the

activation of aggressive responses.
In the social learning formulation, stimulus
control is extensively cognitively mediated by
judgments of how particular behavior is likely to
be received under different circumstances.
Therefore, aggressive actions are not
automatically and rigidly controlled by external

cues. This view of the regulatory process has a
bearing on the controversy of whether the mere
presence of aggressively valenced cues, such as a
gun, enhances expression of other forms of
aggression (Berkowitz, 1971; Buss, Booker, and

Buss, 1972; Page and Scheidt, 1971). Stimulus
prompts of this sort will facilitate aggression if
presented in ways that convey permissive or
expectant reactions toward such behavior, but not
if they are introduced in a manner that makes the

behavior either personally or socially
unacceptable.
Before turning to other substantive issues,
some of the reservations that have been raised
concerning the generality of results from the
shock paradigm deserve comment. Critics of this
approach (Feshbach and Singer, 1971) question
the import of injurious behavior when it is
performed in a permissive setting as part of a

sanctioned social task. What this argument
overlooks is that these are precisely the conditions
under which aggression is usually elicited and
encouraged in everyday life. Aggressors are not
recruited by being told to perform acts designated

as ruthless, despicable, criminal, or bloodthirsty.
Punitive treatment is typically carried out in the
name of discipline rather than personal
vengefulness; military slaughter is structured as a
patriotic endeavor rather than as unconscionable

murder; political activists win supporters by
presenting their coercive tactics as principled

protest against social injustice rather than as
willful destructiveness; and people are prompted
to assaultiveness by designating it as a sign of

manliness rather than psychopathy. It therefore
comes as no surprise that aggressive modeling
increases aggressiveness in everyday situations in
much the same way as it does in laboratory

settings (Harris, 1972). While it is of interest to
study aggression when such conduct is explicitly

defined as human cruelty, it would be a mistake to
regard dissuading conditions as prerequisites for
identifying determinants of aggression as they
operate in natural circumstances.
Another questionable practice is to invoke
“demand characteristics” as a cause of behavior in
experimental settings. This type of explanation
recognizes that people infer from cues the kind of

behavior that may be expected in particular
situations and act compliantly or oppositionally to
the perceived “demand.” It is true that cues guide
conduct in the laboratory as they do in everyday
life. People often behave on conjectures of what
others want them to do, but when this happens
outside a laboratory it is rarely dismissed as a
“demand effect.”
The term “demand characteristics” has been

used in two ways. Sometimes it is applied to
possible determinants of behavior and other times
to the processes by which designated determinants
produce their effects. In poorly designed studies it
is difficult to evaluate whether behavioral changes

result from the conditions deliberately altered or
from some other uncontrolled factors. It makes
little difference whether unsuspected
determinants are called demand characteristics or

by some other name. Confusion arises when the
effects of social influences that are explicitly
varied are ascribed to demand characteristics. To

designate changes as demand effects does not
explain them. Moreover, all forms of social
influence (for example, comments, instructions,
persuasive appeals, conditioning, reinforcement,
modeling) represent demands in the sense that
they function as prompts for particular actions.
Social influences might be better analyzed in

terms of their explicitness and coerciveness and in
terms of whether they change behavior directly or
through cognitive processing.
The significance of controlled investigations of
violence portrayed in symbolic form is more

vigorously debated because of its social
implications for television programming. Media
spokesmen are especially quick to minimize such
findings, on the grounds that the inputs do not

involve standard television programs and
injurious behavior is not assessed in spontaneous
social interactions. It is doubtful that medical

researchers, or those in other branches of science,
would have made much headway in discovering
general laws if their experiments adhered to the
naturalness criteria. The determinants of a
phenomenon isolated by controlled inquiry must,
of course, also be examined under naturally
occurring conditions. Unfortunately, much time is

expended in unproductive controversies over the
virtues of laboratory and field studies when, in
fact, knowledge is best advanced by using both
approaches in a complementary fashion.
Aggressive Modeling and Vicarious Catharsis
Numerous field studies have been conducted

to determine whether the causal relationships
established under controlled conditions are

replicated under more natural circumstances.
According to Feshbach, a leading proponent of the
catharsis hypothesis, televised violence operates
through fantasy functions to decrease the

likelihood of aggressive behavior (Feshbach and
Singer, 1971). Fantasied injury, it is argued,
partially substitutes for actual infliction of damage
as a means of discharging the aggressive drive.
The resultant drive reduction reinforces violent
thoughts, which are considered to be incompatible

with overt aggression. By delivering violent
material, television thus furnishes rewarding
substitutes for aggressive action. This service is
presumably most beneficial to the less intelligent
and unimaginative, who, lacking symbolic

capacities, need external aids for generating
violent fantasies.
In a large though loosely controlled field study,
for six weeks Feshbach and Singer (1971) showed

either aggressive or predominantly nonaggressive
television programs to boys enrolled in residential
private schools and in homes for the

underprivileged. During this period supervisory
personnel in these settings rated the boys’ verbal
and physical aggression; they were also
administered a variety of personality tests. The
authors report that their findings provide
evidence that exposure to televised violence
reduces aggressive behavior. Actually, when boys

in the two viewing conditions are statistically
equated for initial differences in aggressiveness,
those in the private schools did not differ on any of
the measures of overt aggression. Institutionalized
boys shown the violent offerings expressed less

verbal aggression toward peers than those in the
control condition, but they did not differ in
physical aggression toward peers, nor did high
aggressors, who should be experiencing the
greatest cathartic discharge from violence

viewing, express less physical or verbal aggression
toward authorities. Even the few significant

findings are difficult to interpret, however,
because the boys assigned to the control condition
were more aggressive toward each other to begin

with. As Ball-Rokeach (1971) points out, a higher
proportion of hyperaggressive boys in an ongoing
group is likely to generate an elevated level of
hostile interaction among the members that

cannot be corrected statistically.
Relatively few differences were obtained on

the personality variables and, in the instances in
which some television effects are reported, the
relationships vary depending upon subjects’ age,

institutional affiliation, and personality
subgrouping. Given a diverse pattern of results in
which many of the comparisons fail to support
hypothesized relationships and the few obtained
effects hold only for small subsamples, such data

permit any variety of interpretations. However,
the fact that the televised influence had a
consistent impact on fantasy behavior deserves
brief comment because of its direct relevance to
the catharsis hypothesis. Feshbach, among other
personality theorists, has generally regarded
fantasy aggression as an index of aggressive drive
that is presumably reduced by viewing televised
violence. Contrary to expectation, Feshbach and

Singer found that exposure to televised violence
produced a higher level of fantasy aggression than
the control offerings. They reconcile the
contradictory evidence by reinterpreting fantasy
activities as serving to delay and to control

impulse expression. The absence of any television
effects on overt aggression of boys in private
schools is likewise attributed to their greater
cognitive capacities that enable them to self-
generate violent fantasies without the aid of

television. Since the authors do not report any
significant differences between the two samples of

boys in their production of aggressive fantasies on
thematic tests administered at the beginning of
the study, the assumption that underprivileged

boys are deficient in this cognitive capacity does
not seem borne out by the data.
Research evidence indicates that fantasying

goal responses serves neither drive-reduction nor
cognitive-control functions. The notion that
fantasy provides substitute gratifications that

reduce corresponding behavior has been
extensively tested and found wanting. Fantasying
desired goal activities tends to increase rather

than to reduce actions designed to produce them
(Mischel and Baker, 1972; Mischel, Ebbesen, and
Zeiss, 1972). For this reason, athletes in combative
sports deliberately conjure up angry fantasies
toward their opponents just prior to a contest in

order to heighten aggressive responsiveness.
Based on evidence that angry thoughts can
instigate aggressive behavior, therapists have
successfully reduced intractable violent actions by
eliminating their fantasy determinants (Agras,
1967). Drive-reducing functions are further
disconfirmed by evidence cited later that not only
does fantasy fail to reduce anger arousal, but
injurious actions, which presumably should be

even more powerful drive reducers, rarely lower
aggressiveness.
A number of methodological deficiencies,
which are discussed in detail elsewhere (Ball-
Rokeach, 1971; Liebert, Sobol, and Davidson,

1972) and will be only briefly noted here, most
likely account for the variable results of this
project. One of the more serious problems arises
from the failure to control, or at least to measure,

the aggressive contents the two groups of children
were in fact viewing. Programs were classified as
aggressive or nonaggressive on the basis of

general format rather than in terms of their actual
content. Thus, cartoons, which contain the highest
violence rates, were assigned to the nonaggressive
viewing condition. And in some of the serialized
programs control boys were watching, the
characters do resort from time to time to threats,
verbal abuse, and aggressive actions (Gerbner,

1972a). Although the action-adventure type
programs viewed by the aggressive group
undoubtedly contain more expressions of lethal
violence, they may present fewer examples of
moderate aggression that would be immediately

serviceable to boys in coping with their own
interpersonal conflicts. Seeing people handle
familiar problems through coercive means may
have greater effects on aggressive behavior in
institutional situations than viewing killers shoot

victims in exotic settings.
Televised influences were additionally

confounded by inadequate control of what shows
were watched, with the result that control boys
admitted viewing some of the aggressive
programs. Without detailed recording of program
selections, the amount of unacknowledged
violence viewing by the controls cannot be
determined. Results of an experiment are

uninterpretable when the amount and type of
aggressive portrayals remain unknown. This is
analogous to studying the effects of protein intake
on physical health by serving two sets of
smorgasbord, both containing varying amounts of

protein food, under loosely supervised conditions
permitting some cross-over eating, without
measuring how much protein the two groups
actually consumed. To confuse further the
televised input, in three of the institutions the

experimenters yielded to demands from the
controls to reinstate a then popular serial

saturated with violent episodes. Not only was their
so-called “nonaggressive diet” further enriched
with violent offerings, but the rewarded protest

demonstrated that aggression does pay. These
factors, along with the initial higher
aggressiveness of the controls, might partly
account for some of the aggression they displayed.
This study is plagued by numerous other
methodological problems, including nonrandom

assignment of institutionalized boys to treatment
conditions; many subjects withdrew from the
experiment at different periods; aggressive

behavior was recorded retrospectively by
relatively untrained personnel whose ratings are
of unknown reliability; there were inadequate
controls for possible rating biases arising from
knowledge of treatment conditions; and wide

variations occurred in the number and types of
measures administered to subjects both within
and across institutions. Considering the
deficiencies in the regulation of the television
input, in the implementation of procedures, and in
the assessment of effects, it is hoped that the
publisher’s claim that the findings “will
undoubtedly have a serious bearing on future
programming” are not seriously accepted as

guides either for therapeutic practice or for social
policy. Indeed, the findings of Feshbach and Singer
are disconfirmed by Wells (1971), who replicated
their study with better controls for some of the
major methodological defects. In two field studies

conducted by Parke and his associates (Parke et
al., 1972) delinquent adolescents who had been
repeatedly exposed to filmed violence became
more physically assaultive than those who
received a nonviolent fare.
Feshbach (1972) reports a laboratory study
designed to test the hypothesis that exposure to

realistic violence facilitates aggression through
modeling and disinhibition, but fictional violence
reduces aggressiveness by delaying and
substituting for action. Children watched a filmed
sequence of a campus riot presented as either a
fictional or a realistic event. Compared to a no-film
control, the realistic set stimulated aggression

while the fantasy set reduced aggressive behavior,
as measured by children’s administration of
louder sounds to an adult for incorrect guesses on
a game. If these differences are reproducible,
further research would be needed to determine

how unrecognized factors in what the children
were told created the differences, because the
view that fictional violence reduces aggression is
disputed by a large body of evidence. In previously
cited experiments innumerable college students

have been stimulated to punitive action by
viewing familiar actors behaving violently in

presentations they know are fictional. None, for
example would mistake Kirk Douglas for a real
boxer fighting an actual championship bout in the

often used segment from the movie Champion.
Furthermore, children have been shown to
become more interpersonally aggressive as a
result of exposure to violent cartoons that bear

little resemblance to reality. Although realistic
violence may function as a more effective

aggression activator, exposure to fictional violence
does not reduce such conduct (Ellis and Sekyra,
1972; Hapkiewicz and Stone, 1972). Aggression is
enhanced through observation of filmed violence

mainly by its justification, not by whether it is real
or fictional (Meyer, 1972).
There have been several well-designed
experiments in which children are exposed daily

over an extended time either to commercially
televised aggression or to nonviolent offerings.
The amount of aggression that children display in
their everyday interactions during this period is
systematically observed and recorded. Results of
these field studies essentially corroborate those of
laboratory investigations. Steuer, Appleheld, and
Smith (1971) found that repeated exposure to
aggressive cartoons increased children’s physical

assaultiveness, whereas nonviolent contents
produced no change in interpersonal aggression.
In an experiment by Stein, Friedrich, and
Vondracek (1972) employing a similar design,
nonaggressive children showed no behavioral

effects of viewing violent cartoons over a four-
week period, but those above average in
aggressiveness were more physically and verbally
aggressive when exposed to televised violence
than to neutral or socially positive programs. A

follow-up assessment disclosed that television-
heightened aggressiveness persisted after the

violent programs were discontinued. The authors
speculate that the passive children may have
undergone severe inhibitory training which, if

true, would require extended exposure to
successful violent conduct to weaken restraints
over behaving aggressively. Other studies (Ellis
and Sekyra, 1972; Liebert and Baron, 1972)

showing that even short exposure to violent
scenes can increase children’s interpersonal

aggression lend further weight to the instigative
potential of televised aggression.
Correlational studies relating television

viewing to social behavior as it occurs under
natural conditions provide supplementary
evidence concerning television effects. Because of
the gross indexes of violence viewing and
aggressive behavior used in these studies, it is

difficult to estimate how closely the obtained
correlations approximate the true relationship.
Their consistency, however, is noteworthy. Several
projects based on large samples representing
different communities, age levels, and sexes have
found modest positive correlations between
amount of violence viewing and interpersonal
aggressiveness (Chaffee and McLeod, 1971;
McIntyre and Teevan, 1972; Robinson and

Bachman, 1972). The higher the exposure to
televised violence, the more children are willing to
use aggressive behavior, to suggest it as a solution
to interpersonal conflict, and to view it as effective
(Dominick and Greenberg, 1972).
Correlational data do not, of course, identify

causal links between covarying events. Whether
aggressive children are attracted to violent
contents, whether repeated observation of people

resorting to violence cultivates an aggressive
disposition, or whether both activities are
products of a common influence cannot be

conclusively answered by a correlational study.
Experimental evidence is sparse for the former
relationship, though it is well founded
experimentally that exposure to aggressive
modeling increases aggressiveness. As part of a
large field study exploring the social determinants
of aggression, Eron (1963) correlated children’s

television viewing habits with their tendency to
behave aggressively, as rated by their peers. Boys
who preferred television programs containing a
high level of violence displayed significantly more
interpersonal aggression than those who regularly

viewed programs low in violence. Not surprisingly,
the overall findings reveal that the critical factor is
the content of what children are observing, not the
sheer amount of exposure to televised stimulation.
No consistent relationships were noted between

the viewing habits of girls and their aggressive
conduct. Since aggressiveness in adolescent girls

has been shown to correlate positively with
amount of violence viewing but not with high
preference for violent programs (Chaffee and

McLeod, 1971), differential results may be partly
attributable to the measure of violence viewing
used as well as to other factors, such as more
severe socialization of aggression in girls and

preponderance of televised male aggressors.
In a follow-up study conducted ten years later,

Eron and his associates (Eron, Huesmann,
Lefkowitz, and Walder, 1972) report longitudinal
correlations that help clarify the direction of

causal relationships. High exposure to televised
violence at age eight was positively related with
interpersonal aggressiveness in boys at age
nineteen (Figure 3.4). The positive correlation
remains significant when level of childhood

aggression is partialled out, thus removing the
possibility that initial aggressiveness determines
both child viewing preferences and adult conduct.
Other partial correlations, controlling for a variety
of familial causal contributors, further indicate
that habitual exposure to televised violence
FIGURE 3.4 Relationship between boys’ violence viewing at
age eight and their interpersonal aggressiveness as measured
ten years later. (Plotted from data by Eron, Huesmann,
Lefkowitz, and Walder, 1972)
promotes aggressive habits and eliminates
alternative hypotheses that aggressiveness causes
attraction to violent portrayals or that the

obtained association is due to some other common
determinant. Chaffee and McLeod (1971) arrive at
a similar conclusion after demonstrating that
significant correlations between amount of
violence viewing and aggressiveness persist after

ruling out the potential influence of socioeconomic
variables, level of intellectual functioning, and a
host of familial relationship factors. Model
preferences and behavioral dispositions are more
likely reciprocally interrelated for real life than for

televised models. That is, aggressors may
selectively gravitate toward aggressive models

who, in turn, further shape and disinhibit
aggressive modes of conduct.
In the controlled field studies interpersonal
aggression is measured solely as a function of
televised influences without regard to the social
consequences that aggressive actions produce.
The course that aggression takes over a period of

time will be determined by the combined effects of
televised inducements and the reinforcing
consequences of a reactive environment. The same
televised input can therefore result in different
outcomes under diverse reinforcement conditions.

Accurate prediction of television effects on either
group or individual performance depends on how
the concurrent influences interact. If television-
prompted aggression proves effective for users in
their daily interactions, aggressive modes of

behavior are likely to increase and persist long
after the televised influence has ceased. That

television can give impetus to a long-term process
is suggested by Eron’s finding that exposure to
televised violence in early formative years

predicts aggressiveness in boys ten years later.
These correlational data are corroborated by a
short-term experimental demonstration by Stein,
Friedrich, and Vondracek (1972) that televised

influences can contribute to the formation of
enduring aggressive styles of behavior.
Under conditions where television-stimulated
aggression is firmly disapproved or creates new
sanctions against injurious conduct, the instigative

effects of television are counteracted by
environmental punishment. The net effect of
aggressive modeling in such circumstances may be
little noticeable change or even a decline in the
incidence of aggressive actions. Since

reinforcement influences are mediated by
stimulus control, prediction of outcomes must be
situation specific. Televised aggression that is
suppressed in prohibitive settings may be
displayed when there is little risk of punishment
(Hicks, 1968a).
Considering the multiple control of aggression,
further progress in understanding television
effects is unlikely to be achieved by additional

studies in which media inputs are systematically
varied but concurrent reinforcement determinants

go unrecorded. Finer analysis of interacting
influences is required to bring orderliness to
variable results. For example, individual data
reported by Steuer, Applefield, and Smith (1971)

show that four of the five children exposed to
televised violence became more assaultive toward
peers than did their matched partners who

watched nonviolent programs (Figure 3.5).
However, they varied considerably in rate and

level of aggression increments. Some showed
immediate increases, whereas others required
multiple exposures before the effects upon

behavior were evident. It might be noted
parenthetically that experiments in which
FIGURE 3.5 Cumulative frequency of physical interpersonal
aggression by children who were exposed either to violent
cartoons or to nonviolent offerings. (Steuer, Applefield, and
Smith, 1971)
observation of a single violent cartoon increased
interpersonal aggression, but short of
conventional significance levels (Siegel, 1956),

would probably have produced stronger effects
with longer exposure. The two weak responders in

Steuer’s study exhibited no aggression during
baseline assessments, which suggests that limited
impact of televised violence on their overt
behavior may be attributable to low aggression

proneness arising from deficient skills, strong
aggression restraints, or both factors. Some
displayed continuing marked increases in

assaultive behavior with repeated exposure, and
others manifested comparable heightening of

aggressiveness, but eventually stabilized at an
intermediate level.
When rate of aggression rises with longer
exposure to violent conduct, the elevation may
reflect the increasing cumulative impact of
television, or viewers may be adopting aggressive
solutions they saw modeled and using them more
often because they produce good results. To

elucidate the causal processes would require
behavior observations of the frequency with
which aggressive actions bring rewarding or
punishing consequences. In sum, the impact of
televised violence would be best predicted by

taking into account viewers’ preexisting
aggression skills, their aggression restraints, and
the proportion of positive and negative
reinforcements provided for aggressive conduct in
daily interactions. Multiple causation can be even

better unraveled by experimentally varying
exposure to televised violence in conjunction with

environmental reward, extinction, or punishment
of aggression for subjects differing in aggression
skill and inhibition.
Exposure to modeled violence presented in
ways that legitimize its use not only raises the
probability of aggressive conduct, but also tends to
increase preference for aggressive toys (Larder,

1962; Lovaas, 1961), stressful mood states
(Tannenbaum and Gaer, 1965), and selection of
aggressive solutions to depicted conflicts (Leifer
and Roberts, 1972).
The combined findings of laboratory

experiments, controlled field studies, and
correlational investigations provide substantial
testimony that violence viewing tends to foster
aggressiveness. Research that clarifies causal

relationships increases rather than reduces the
level of controversy when profitable practices of a
major industry are shown to have some adverse

effects. Vested interest groups become
understandably apprehensive about the
implications of research findings for social policy.
Although the issues continue to be vigorously
debated in terms of empirical verification, at this
stage the dispute is more a political matter than a
scientific one.
Performance Catharsis
The catharsis hypothesis assumes that the
drive to behave aggressively is reduced not only
by vicarious experience, but also by direct
expression of aggression. Although performance

catharsis does not necessarily involve modeling
processes, it is discussed in the present context
because results of this line of research have an

important bearing on the mechanism presumed to
operate in vicarious catharsis through modeling.
The hydraulic model of personality and belief

in the benefits of cathartic drainage of aggressive
impulses were widely accepted in both

professional and lay circles. Guided by this view,
many parents, educators, and mental health
workers urge hyperaggressive children to engage
in aggressive activities and to aggress in
psychotherapeutic playrooms, in the hope of

reducing aggressive impulses that are presumably
maintaining the troublesome behavior. Failure to
release pent-up drives, it is contended, risks
explosive aggressive discharges.
Evidence from research studies of children
indicates that, far from producing a cathartic

reduction of aggression, participation in
aggressive activities within a permissive setting
maintains the behavior at its original level or
actually increases it (Feshbach, 1956; Freeman,
1962; Kenny, 1952, Mallick and McCandless, 1966;

Nelsen, 1969). Early studies with adults in which
evidence for cathartic processes was claimed

relied on indirect indexes of aggression, such as
replies to word association tests that are subject to

response set biases (Feshbach, 1955; 1961) or
quality of intellectual performance, on the
assumption that anger arousal diminishes
functioning (Worchel, 1957). Since emotional

arousal can facilitate as well as disrupt
performance, depending on its level, the
complexity of the task, and the stress coping
habits of subjects, the indirect indicators of
aggression reduction have dubious value. In some

studies presented as demonstrating a catharsis
effect the data do not, in fact, support the
conclusions. Angered subjects exhibit the same
level of dislike of their tormentor after seeing him
undergo either painful, neutral, or euphoric

experiences (Bramel, Taub, and Blum, 1968).
Reported substantiation of catharsis effects in

other instances depends more on selective
analysis of data than on what they actually show.
Results of experiments with adults using direct
measures of aggression are essentially similar to
those obtained with children. When adults are
given repeated opportunities to shock a person
under nonretaliative conditions, the more they

aggress, the more punitive they become (Buss,
1966a). Angered aggressors, who should be
achieving drive reduction if catharsis were
operative, exhibit greater increases in
punitiveness on successive attacks than

nonangered subjects (Loew, 1967). Further
evidence that encouraged or permissively viewed
aggressive displays reduce restraints over such
conduct is provided by Zimbardo (1969) in an
experiment of a direct interpersonal nature.

Intense verbal and physical assaults directed at
passive resisters by disguised assailants tended to

increase their subsequent hurtful actions.
Nor has the catharsis hypothesis fared well
when the effects of aggressive displays are
measured in terms of changes in hostile attitudes
rather than injurious actions. De Charms and
Wilkins (1963) report evidence that angered
people who vent their hostility become

increasingly hostile. The higher the volume of
unpleasant remarks they direct at their tormentor,
the more they dislike him. Hearing others berate
the offender for his inconsiderate behavior
increased hostility in maltreated observers even

more powerfully. Kahn’s (1966) results are
especially relevant to the catharsis issue because
subjects voiced their resentment to a sympathetic
person, an interaction similar to those that occur
in everyday experience. Those who expressed

hostility with support and encouragement disliked
their annoyer significantly more and were slower

to calm down physiologically than equally angered
subjects who were left to themselves for an
equivalent period.
Under certain conditions expression of
aggression can, of course, decrease its incidence,
although the reductive effects do not occur
through drainage of aggressive drive forces. Angry

displays, especially if accompanied by threats, may
intimidate antagonists so they cease behaving in
provocative ways. Aggressive action may not only
turn off abuse but promote beneficial
improvements in social treatment, thereby

removing chronic instigators of the behavior.
Interpersonal interactions are somewhat
ambiguous as to exactly who did what to whom
and why, with the result that events can be easily
misjudged or distorted. When malign intent is

misattributed to the actions of others, anger
arousal can be ruminatively generated to the point

where the aggrieved person acts with
inappropriate hostility—as if he had been
deliberately maltreated. To the extent that

verbalized resentment clarifies misconceptions or
creates new understanding of why others behaved
as they did, it can diminish autistically produced
aggression. Moreover, behaving aggressively can

arouse fear and self-censure over the injurious
consequences of one’s actions, and thus have

reductive effects through inhibitory processes.
After a provocateur has been duly punished for his
insulting actions, a counteraggressor may temper
further assaults on him without reducing his anger

or dislike because the culprit has already paid
sufficiently for his wrongdoings. Finally, repeated
unrewarded expressions of a behavior produce a

decrease in responsiveness due to aversive fatigue
and declining attraction of the activity. Extinction

is an operative factor in all nonreinforced
performances.
The belief that human behavior is impelled by
inner aggressive forces is nonetheless so strongly
ingrained in psychological thinking that
aggression decrements, when they do occur, are
automatically attributed to cathartic discharges
without measuring more plausible processes

(Doob and Wood, 1972; Konecni and Doob, 1972).
Progress in the understanding of how observed or
performed aggression affects subsequent
punitiveness would be better advanced if a
moratorium were declared on catharsis

explanations until proponents of the catharsis
hypothesis devise a measure of aggressive drive
independent of the behavior it supposedly causes.
Attributing response decrements to drive
reductions provides a ready explanation; one need

therefore search no further for the determinants
of aggressive behavior.
The discussion thus far has examined the
behavioral and attitudinal effects of aggressive
displays. Another aspect of the catharsis
hypothesis is concerned with the physiological
changes accompanying expressions of aggression.
A series of experiments reported by Hokanson
(1970) and others (Baker and Schaie, 1969;

Gambaro and Rabin, 1969) demonstrate that
angered adults often achieved faster decrease in
blood pressure to baseline levels if they
counteraggressed toward their antagonist than if
they were given no opportunities to respond.

However, venting aggression did not facilitate
decline in vascular arousal if the aggression was
expressed in fantasy, if the attack was displaced to
substitute targets varying in similarity to the
tormentor, if the aggressor was female, or if the

attacked frustrator was a person of high status.
The latter finding illustrates the difficulties of

measuring aggressive drive strength in terms of
physiological arousal because it does not
distinguish between anger arousal and aggression-

produced fear arousal—or other sources of
activation, for that matter. Further studies,
reported by Kahn (1966) and Holmes (1966),
disclosed that direct aggression, even toward

nonthreatening targets, does not always bring
physiological relief. On the contrary,

counteraggression sustained elevated arousal,
whereas nonaggressive activities reduced it.
The catharsis hypothesis is more seriously

challenged by the arousal reactions of angered
nonaggressors than by the diverse physiological
effects of counteraggression. After an aggressive
drive is aroused, presumably it should remain at
an elevated state until discharged through some

form of aggressive activity. In reality, all angered
subjects, regardless of whether or not they vent
their hostility, display rapid decline in
physiological arousal. Given that arousal reduction
is the general rule, theories should perhaps be
more concerned with explaining what prolongs
arousal rather than what reduces it. Social
learning theory explains both phenomena in terms
of self-arousal mechanisms. It also specifies the

additional experimental controls required to
demonstrate a catharsis effect.
As previously noted, the anger arousal created
by maltreatment dissipates, but it can be sustained
and reactivated symbolically on later occasions by

dwelling on past angering experiences.
Persistence of elevated anger arousal, in the social
learning view, arises from self-generated
stimulation rather than from an undischarged

reservoir of aggressive energy. If a person should
become immersed in activities that take his mind
off past provocations, the arousing ruminations

are thereby eliminated and he experiences a
noticeable reduction in tension. Similarly, on the
assumption that diminution of emotional arousal
accompanying observed performances results
from changes in symbolic engagement, not
vicarious energy releases, one would expect
aroused individuals to experience equally salutary

effects from getting involved in absorbing books,
movies, or television programs lacking aggressive
displays.
Subjects who have been treated insultingly and
then left with nothing to occupy their attention are

more likely to dwell on affronts they have just
experienced than those assigned a task that
provides opportunities to express aggression. To
confirm that arousal decrements attributed to

energy discharges are not due to differences in
opportunities for resentful self-stimulation
requires an experimental design in which changes

in physiological arousal are compared in angered
subjects who counteraggress and in those who
engage in equally absorbing but nonaggressive
activities. An idle group left either to foment their
resentment or to distract themselves with
irrelevant trains of thought does not furnish a
neutral baseline against which to evaluate the

effects of counteraggression.
Results of numerous experiments lend support
to the self-arousal interpretation. When people are
given an opportunity to respond to provocation
they reduce their hostility, but the decrease cannot

be attributed to catharsis because most voice little
or no hostility (Rosenbaum and de Charms, 1960;
Thibaut and Coules, 1952). Elevated blood
pressure levels are maintained more effectively by

exposure to an antagonist, whose presence is
likely to reactivate anger-provoking thoughts, than
by a lack of opportunity to express aggression in

response to indignant treatment (Vantress and
Williams, 1972). A study by Zillmann and Johnson
(1973) explicitly designed to test the self-arousal
hypothesis provides the most pertinent
substantiating evidence. Aggressive portrayals
sustained elevated physiological arousal and
punitive behavior in angered adults, whereas

exposure to nonaggressive contents resulted in
substantial reduction in both arousal and
retaliatory aggression.
In experiments permitting different
counterresponses, angered males experience

faster arousal reduction by aggressive than by
friendly reactions to a punitive opponent, whereas
females achieve greater tension relief by friendly
reciprocal actions (Hokanson and Edelman, 1966).

The differential physiological concomitants of
defensive reactions most likely reflect the residual
of past reinforcement experiences. Having

developed stronger aggressive restraints, females
might be expected to generate fear-arousing
thoughts after behaving in a physically aggressive
manner. Sosa’s (1968) finding that passive male
prisoners are more apt to experience tension
reduction by friendly than by aggressive reactions
to provocation confirms the impact of socialization

experiences.
Thus, from a social learning perspective,
physiological tension level is determined not only
by what one does, but also by what one thinks.
Diverting or tranquilizing thoughts will reduce

arousal, whereas those that revivify past
provocations or dwell on threatening future
consequences of contemplated or accomplished
actions will generate high states of arousal. The

consequences people anticipate for their actions
are derived largely from observed or directly
experienced reinforcements. Studies in which the

outcomes of reciprocal interchanges are
systematically varied demonstrate that the same
counterresponse, regardless of its content, can
acquire either arousing or tranquilizing properties
depending upon its consequences (Hokanson,
Willers, and Koropsak, 1968). Both kindly and
aggressive responses to provocation produce

cathartic-like physiological relief after they have
consistently elicited positive reactions in others;
conversely, they both become physiologically
arousing when they consistently draw punitive
responses. By reversing reinforcing outcomes, the

tension relief value of the same mode of response
is radically altered.
Self-directed aggression can also acquire relief
value through its functional utility. When adults

could avoid painful shocks from others by
administering to themselves shocks of lesser
intensity, self-punitive responses not only

increased but were physiologically relaxing (Stone
and Hokanson, 1969). The way in which self-
punishment can be maintained by averting
anticipated threats is graphically demonstrated
with subhuman subjects by Sandler and Quagliano
(1964). After monkeys learned to press a lever to
avoid being shocked, a second contingency

involving self-administered punishment was
introduced. A lever press prevented the
occurrence of the original shock, but it also
produced an electric shock of lesser magnitude. As
the experiment progressed, the self-administered

shock was gradually increased in intensity until it
equaled the one being avoided. The animals,
however, showed no reduction in self-punishment
even though this behavior no longer served as a
“lesser of two evils.” Even more interesting, after

the avoided shock was permanently discontinued
but lever-pressing responses (which had now

become objectively functionless) still produced
painful consequences, the animals continued to
punish themselves needlessly with shock

intensities that they had previously worked hard
to avoid. This experiment reveals how self-
punishment can become divorced from actual
conditions of reinforcement and be maintained

through its capacity to forestall imagined threats
that no longer exist. A puzzled observer, who had

no knowledge of the prior learning conditions for
this apparently senseless behavior, would be
tempted to attribute it to a misdirected aggressive
drive.
In psychotic disorders, self-punishment is
often powerfully maintained by delusional
contingencies that have little relationship to
reality. The case cited earlier (Bateson, 1961)

illustrated how a man who judged trivial acts as
heinous sins could relieve his fright of hellish
torment and feelings of self-contempt only by
performing exceedingly self-punitive behaviors for
long hours. Responses that are reinforced by
reduction of thought-generated distress tend to be
self-perpetuating.
These findings indicate that the physiological
accompaniments of aggression are explainable in

terms of the same principles governing aggressive
actions. When aggression succeeds in curbing
attack, it becomes a preferred tension-reducing
style of behavior. On the other hand, when
aggressive actions beget punitive treatment, they

become physiologically perturbing and are
generally not favored. The physiological correlates
of counteraggression further illustrate the
differences between drive and social learning

theories. The former view assumes that affective
discharges diminish aggressiveness. In the social
learning analysis, tension reduction can serve as a

reinforcer to increase the likelihood that
aggression will be used on subsequent occasions.
Relief of aversive arousal may thus supplement, or
even override, the effects of other consequences in
maintaining aggressive responsiveness.
Discomforting arousal, however, would tend to
exert inhibitory effects on hostile actions.
Controversies over modeling influences are
not confined to aggressive behavior. Although
experimental-data are scarce, there is every
reason to expect, for example, that sexual
modeling, which has begun to be studied, would

affect sexual behavior in similar ways. By
observing the sexual activities and accompanying
reactions of others, people can learn specific
amorous techniques; their sexual anxieties can be

diminished (Walters, Bowen, and Parke, 1964);
and the portrayals can shape the types of sexual
expression that will be practiced in a culture by

teaching which forms are permissible and which
exceed socially acceptable bounds. Viewing erotic
films produces transitory increases in sexual
arousal and sexual activities such as masturbation,
petting, and marital intercourse (Mann, Sidman,
and Starr, 1971; Schmidt and Sigusch, 1970), but
habitual exposure can reduce the sex-arousing

potential of nudity or erotic displays (Reifler,
Howard, Lipton, Liptzin, and Widmann, 1971). In
addition, sexual viewing fosters more permissive
attitudes toward erotic depictions (Lockhart,
1970).
Although the processes and outcomes of sexual

and aggressive modeling are undoubtedly similar,
the social implications may differ markedly
depending upon the moral codes applied to these

two areas of functioning. Because of the hazardous
consequences of physical aggression, a society
must continuously control injurious actions in

children and adults alike. Harmonious functioning
therefore requires reduction of social influences
that promote cruelty and destructiveness. By
contrast, most cultures, including our own,
present discontinuities in the socialization of
sexual behavior. Overt sexual expressions are
firmly prohibited during childhood and

adolescence, but are not only expected but
considered essential to a satisfying heterosexual
relationship in adulthood. Therefore, the more
successfully inhibitory training is achieved in early
formative years, the more likely is heterosexual

behavior to serve as a source of guilt and anxiety
in adulthood. Given prolonged negative
conditioning of sexual attitudes and behavior,
many people feel that exposure to sexual material
under proper conditions may help to counteract

ignorance and dysfunctional inhibitions. On the
other hand, proponents of stringent moral codes,

fearing that erotic viewing will propagate
premarital intercourse and antisocial sexual
transgressions, vigorously oppose any lessening of

prohibitions against sexual modeling.
Taken together, available research findings
indicate that when erotic stimuli are effective, they
tend to activate established modes of sexuality

(Lockhart, 1970). Thus, exposure to erotic
material temporarily increases masturbation

principally in unmarried persons who habitually
engage in such activities and in the sexually
conversant lacking amorous mates. Among
married couples and sexually experienced persons

with available partners, viewing erotic displays
produces transitory increases in sexual
intercourse. There is no evidence, however, that

erotic viewing stimulates deviant sexual practices
or sex offenses.
AVERSIVE TREATMENT
The concept of frustration, which is frequently
invoked as a principal cause of aggression, is so
broad as to have little value. As previously noted,
the diverse events subsumed under this all-
inclusive term do have one feature in common—
they are all aversive in varying degrees. Since,

however, they differ in other important ways that
can produce dissimilar behavioral effects, the
determinants of injurious behavior can be better
ascertained by relating aggression to specific
aversive antecedents rather than to the omnibus

term. This section examines the extent to which
aggressive behavior is aversively controlled by
unpleasant, thwarting, offensive, and physically
painful treatment.
Physical Assaults
If one wished to provoke aggression, the most
dependable way to do so would be simply to

physically assault another person, who would then
be likely to oblige with a vigorous counterattack.
To the extent that counteraggression discourages
further assaults, it is powerfully reinforced by pain
reduction and thereby assumes high functional
value in social interactions. Although the pain-

aggression relationship is well established, there is
some dispute over whether it is innate or
acquired.
Azrin and Ulrich, exponents of the nativistic
view, have extensively investigated the

relationship between shock-produced pain and
aggression (Azrin, 1967; Ulrich, 1966; Ulrich,
Hutchinson, and Azrin, 1965). In studies
employing a social paradigm, pairs of animals are

placed in a chamber containing a grid floor
through which foot shocks can be delivered. At
periodic intervals brief shocks are presented and

any fighting responses displayed by the animals
are recorded and compared with their behavior at
times when shocks are not administered. During
the attacks, which are typically brief and cease
shortly after shock termination, animals stand
erect on their hind legs, lunge at each other with
bared teeth, and strike with their forepaws.
Elicitors of aggression are often difficult to
assess in social situations because target animals
can influence the course of fighting in uncontrolled
ways by their own actions, especially if they
manifest threatening gestures or strike back, as

they understandably are inclined to do. Social
paradigms also cannot be used with highly
destructive animals that can fatally injure their
victims. In some studies, therefore, aggression is

measured in terms of biting attacks directed
against inanimate objects. Results based on the
latter measure, however, may have limited

generality because it appears to be relatively
insensitive to known determinants of fighting
behavior.
A considerable amount of information has
been accumulated over the years on the nature
and generality of the shock-attack relationship.

The more intense the shocks and the longer their
duration, the greater is the likelihood that they

will provoke stereotyped fighting. Attack
responses tend to be performed somewhat
indiscriminately toward targets of varying
physical characteristics. Animals will assault their

own species, members of other species, and even
inanimate objects. Additionally, shock-attack
reactions have been found to be present in a

variety of species. Other studies disclose that
attack can be provoked by different types of

aversive stimulation including, in addition to
painful shock, intense heat and physical blows.
Unlike fighting under natural conditions, which is

partly determined by sex of combatants, their
inbred aggressiveness, and previous familiarity
with each other, the incidence of shock-induced
fighting is essentially unaffected by such factors.

Azrin and Ulrich reasoned that since combative
behavior is elicited by aversive stimuli with some
regularity, emerges without any prior training,

and persists without any evident reinforcement,
pain-induced aggression must be an unlearned
reflexive behavior.
Although the pain-aggression relationship

appears to have some degree of generality,
negative findings—often minimized or dismissed
as unexplainable—are also very much in evidence.
The latter data, which are reviewed next, dispute

the view that animals are genetically programmed
to respond aggressively to painful stimulation
(Ulrich, 1967). Close examination of research
findings reveals that aversive stimuli do not exert
the “push-button” control over aggression
sometimes claimed. In some species noted for
their aggressiveness, such as fighting cocks and
Siamese fighting fish, pain not only fails to
provoke aggression, but suppresses such

responses (Azrin, 1967). Shock also fails to
produce fighting in certain docile species (Ulrich
and Azrin, 1962).
Shock-attack reactions are much too variable
to be attributable to genetic determination. In fact,

numerous environmental influences determine
whether or not aversive experiences will provoke
combativeness. Young animals rarely, if ever, fight
when shocked unless they have had some fighting

experience (Hutchinson, Ulrich, and Azrin, 1965;
Powell and Creer, 1969), and in some studies
repeated shocks produced little or no fighting in

20 to 30 percent of mature animals (Azrin,
Hutchinson, and Hake, 1963; Azrin, Hake, and
Hutchinson, 1965). The likelihood that physically
painful events will provoke attack, especially at
lower intensities, is partly determined by a
number of accidental factors, such as whether
animals happen to be facing each other at the

moment of shock delivery, whether they are
moving, standing upright, or are close to each
other. Thus, rats that are physically restrained and
forced to face an inanimate target placed
immediately in front of their mouths consistently

bite it when shocked (Azrin, Rubin, and
Hutchinson, 1968), whereas they never do so
when repeatedly shocked in a confined but free-
responding situation (Ulrich and Azrin, 1962). The
former procedure, which is increasingly used for

the convenience of automatic recording,
apparently produces behaviors quite

unrepresentative of those occurring under
circumstances the experiments were originally
designed to explain.
If aggression is an unlearned dominant
response to painful stimulation, then shock would
be expected to elicit attack during initial trials.
Unfortunately, the numerous studies that have

been published contain no information on how
animals initially respond to painful treatment,
despite the relevance of such data to instinctual
interpretations of aggression. But some findings in
addition to those cited above cast doubt on the

view that aversive stimuli are innate elicitors of
aggression. In one report in which responding
during initial trials is briefly described (Azrin,
Hutchinson, and Hake, 1963), one pair of monkeys
jumped and squealed but did not fight when shock

was first presented; another pair never fought
during the first ten trials; a third pair never

attacked each other even though they were
repeatedly shocked. They displayed shock-attack
reactions only later, after they had been paired

with trained sparring partners.
The most decisive evidence that pain-
aggression reactions may be more situationally
determined than innately programmed is

provided by investigations of shock-induced
fighting as a function of size of the chamber

(Ulrich and Azrin, 1962). In a small enclosure (6
in. x 6 in.) approximately 90 percent of the shocks
provoked fighting, whereas in a larger chamber
(24 in. x 24 in.) animals ignored each other and

only 2 percent of the shocks elicited fighting.
Powell and Creer (1969) likewise found pain-
attack reactions to be a rarity in young animals

shocked in large cages. If animals, in fact,
possessed an unconditioned fighting reflex, they

should be equally combative in varying, though
still relatively small, enclosures when subjected to
painful stimulation.
Premature attribution of pain-induced fighting
to reflexive mechanisms has retarded
investigation of other possible determinants and
the functional utility of fighting in the
experimental settings. Pain-elicited attack appears

to differ qualitatively from naturally instigated
fighting. In the former case, fighting episodes are
usually very brief; fighting ceases almost
immediately after the shock is stopped; and it is
conspicuously absent during periods between

shocks. If pain consistently provokes aggression,
one would expect the biting and mauling initiated
by shock to be sufficiently painful to perpetuate
reciprocal fighting for some time after shock
stimulation has ceased. Given evidence that

animals can acquire aggression-eliciting
properties through association with combat

experiences (Ulrich, Hutchinson, and Azrin, 1965),
it is also surprising that fighting does not come
under social stimulus control so that animals

eventually attack each other when shock is absent.
In many respects, pain-induced postures appear to
be more closely related to escape and defensive
responses in which the combative features are

secondary (Reynierse, 1971).
Ulrich and Azrin (1962) have considered—and

discounted—possible undetected sources of
reinforcement of pain-induced fighting. By
assuming upright postures and jumping on each

other, animals may periodically reduce contact
with the electrified grid and thus have their
sparring intermittently reinforced by partial
escape from painful stimulation. As the authors
point out, this explanation would not account for

fighting induced by shock delivered through
electrodes attached to the animals’ bodies.
Findings of the latter studies are less persuasive,
however, because they are based on response-
constraining procedures that force biting attacks
on inanimate targets, attacks which are not often
duplicated under unrestrained conditions.
A second interpretation is that fighting is
partly maintained by accidental reinforcement

through occasional coincidences of attack
reactions with shock termination. Protective
behavior is most readily perpetuated by
superstitious contingencies when threats cannot
be controlled by one’s actions; coincidences

between actions and outcomes tend to be
interpreted as genuine contingencies that regulate
behavior as though a given action will, in fact,
forestall a feared outcome. The authors doubt

whether such a process is operative on the
grounds that the shocks are too brief to provide
much superstitious reinforcement. Evidence of

sequential changes in fighting responses as a
function of shock duration suggests, however, that
the temporal relation between attack behavior
and shock termination may be a critical factor in
sustaining pain-provoked fighting (Azrin, Ulrich,
Hutchinson, and Norman, 1964). Initially, brief
shocks rarely elicited fighting responses, but long

shocks almost always provoked attack. As the
experiment progressed, long-lasting shocks
rapidly lost their effectiveness, whereas short ones
increased and maintained their power to instigate
fighting responses. On the occasions when shock

elicits attacks, they tend to appear immediately
and cease quickly. If the assumption is valid that
fighting sometimes coincides with pain reduction
given short shocks, but with continued painful
stimulation when shocks last a long time, the

obtained changes in the evocative power of shocks
are consistent with an adventitious reinforcement

interpretation. This speculation is tentatively
supported by research showing that the
contingency, not pain per se, is the critical

determinant of aggression. Noncontingent painful
stimulation provokes aggression, but—contrary to
the reflexive elicitation hypothesis—shocks
presented after each attack reduce and eliminate

rather than increase aggressive responding (Azrin,
1970; Baenninger and Grossman, 1969; Ulrich,

Wolfe, and Dulaney, 1969).
Aversive stimulation ordinarily activates
behavior designed to eliminate or at least to

reduce the source of discomfort. Thus, if a room is
unbearably hot, people will typically turn on air
conditioning devices in preference to assaulting
someone nearby. The experiments reported thus
far permitted attack but not escape. In order to

determine whether aggression predominates in
aversive situations, researchers examined the
relationship between combative and flight
reactions to painful treatment. The results show
that, given a choice, animals would rather escape
than fight. Under conditions in which the escape
response requires social confrontations, shock can
provoke some fighting, but successful escape
quickly displaces attack reactions and fighting

reappears only when escape is not possible (Azrin,
Hutchinson, and Hake, 1967; Wolfe, Ulrich, and
Dulaney, 1971). If paired animals can flee without
the necessity of confrontation, they do so when
shocked without ever aggressing toward each

other (Logan and Boice, 1969). Knutson (1971)
has further shown that, even under close social
contact, when environmental inducements to fight
are minimized, avoidance and escape responses to
shock stimulation take priority over attack.
In everyday situations the target person is
usually the source of the distress rather than

simply an innocent bystander. Under these
circumstances successful counterattacks provide
escape from injurious assaults. As might be
expected, aggressive responses that serve a
protective function are highly resistant to change
(Azrin, Hutchinson, and Hake, 1967).
There are some further indications that as the

determinants of pain-attack reactions are more
critically examined, they will most likely be
dispossessed of their reflexive status.
Developmental studies (Hutchinson, Ulrich, and
Azrin, 1965) reveal that painful events rarely

produce fighting in young animals, and they are
relatively weak elicitors of aggression in animals
of all ages if reared alone after weaning (Figure
3.6). The authors attribute the increasing capacity

of shock to elicit fighting with advancing age to
higher levels of testosterone, the male sexual
hormone. Evidence bearing on this issue would

seem to favor experiential rather than hormonal
factors. Daily injections of testosterone create
sexual precocity but do not enhance the attack-
eliciting potential of shock in young animals
FIGURE 3.6 The percent of shocks that evoked fighting in
groups of animals that were reared socially or isolated at 22
days of age and lived alone until tested for aggression to 100
shocks, presented at 30, 60, and 90 days of age, respectively.
(Plotted from data of Hutchinson, Ulrich, and Azrin, 1965)
(Powell and Creer, 1969).
The role of fighting experiences in determining
whether pain will provoke attack or nonaggressive

reactions was examined by Powell and Creer
(1969), who assessed the interaction of
developmental and environmental variables in
shock-elicited aggression. As illustrated in Figure
3.7, young animals in standard enclosures almost

never fought when shocked, but those in small
cages, who also at first responded
nonaggressively, eventually became fighters as the
sessions progressed. This differential

combativeness indicates that young animals are
physically capable of aggressing but that painful
treatment alone is insufficient to elicit such

behavior. Somewhat older animals yielded a
similar pattern of results, whereas the oldest
members, those most likely to have had previous
fighting experience, increased their combativeness
FIGURE 3.7 Frequency with which shock induced fighting in
paired animals of varying ages in small and standard
enclosures. One hundred shocks were administered during
each daily session in a series of 10 sessions. (Powell and Creer,
1969)
over the sessions regardless of enclosure size. A
related experiment demonstrated that aversive
stimuli essentially lose their capacity to arouse

aggression if they have been previously
experienced under social circumstances in which
aggressive responses could not be performed.
These findings are consistent with those of Maier,
Seligman, and Solomon (1969); they show that

after animals have been repeatedly shocked no
matter what they do, they later respond passively
to hurtful treatment even though successful
coping responses are made available.
The overall results support the view that

physically painful experiences are, at best,
facilitative but not sufficient to provoke aggression

in animals. To produce regular fighting through
painful treatment, animals must be directly
confronted with the target in a constricted area,
the aversive stimuli must be very brief and not
follow attack reactions, the animals must have
some fighting experience, and all flight reactions
must be blocked. Fighting thus appears to be a

forced rather than a primal unlearned response to
pain. The restricted circumstances under which
aversive stimuli apparently exercise control over
aggressive responses cast some doubt on whether
experimentation confined to such conditions will

substantially advance understanding of what
causes animals to aggress in their natural habitat,
much less explain human aggressiveness. Under
natural conditions aggression may be aversively
controlled after animals have begun to attack each

other and thus cause pain, but the factors that
begin fights in the absence of any immediate

aversive experiences still require explanation.
Aggression is typically provoked by threats
intruders pose to food supplies, to nesting sites, to

offspring, to mates, and to dominance status,
rather than by actual physical pain. In these
instances, events that forebode painful
consequences provoke combat. The latter

phenomena can perhaps best be understood in
terms of the response-controlling functions of

informative stimuli and reinforcing consequences.
From field studies of social interactions
resulting in violent conduct it is well documented

that physical assault serves as a powerful
instigator of aggression. For obvious humane
reasons, virtually no experiments have
investigated the conditions determining whether
physical attack will elicit fight, flight, submission,

or some other type of reaction in humans. The
strength of anticipatory retaliative aggression is
undoubtedly a critical controlling factor. The only
studies that have included some physical
provocation are ones in which adults could
counteraggress toward either a punitive
competitor or a person who had previously
devalued their intellectual performance by
administering painful shocks to them. People

aggress more strongly toward their assailants
when they are physically punished than when they
are treated considerately (Berkowitz and Le Page,
1967; Pisano and Taylor, 1971). When escape
from the situation is blocked, the more frequently

they are attacked, the more physical aggression
they inflict on their adversaries (Helm, Bonoma,
and Tedeschi, 1972). It remains unclear, however,
how much of the retaliation is provoked by the
physical maltreatment, how much by the implied

belittling of their intelligence, and how much by
the aggressive modeling of the provocateur.
Verbal Threats and Insults
Social interchanges are typically escalated into
physical aggression by verbal threats and insults.
Toch (1969) retrospectively analyzed sequences
of actions and reactions of chronic assaulters to
determine what sets them off on a course

productive of violence. Among those studied were
assault-prone police officers, as well as parolees
and prison inmates who recurrently get embroiled
in acts of violence. Humiliating affronts and
threats to reputation and manly status emerged as

major precipitants of violence.
In the case of assaultive policemen, the
interaction generally begins with the officer
issuing orders coercively in some petty incident.

Most people acquiesce to forceful show of
authority, but individuals who are prone to
violence respond to arbitrary commands as an

unwarranted infringement of their rights. They
regard their actions as perhaps mildly provocative
but essentially within legal boundaries, and the
policeman's coercive behavior as arbitrary and
unjustified. Negative definition of police actions is
especially prevalent among underprivileged youth
who, due to experienced, observed, or affirmed

harassment, are inclined to view police as
guardians of an oppressive system. Consequently,
they question the legitimacy of police authority.
Policemen with a penchant for provoking fights
also tend to assert their authority in inept ways

that cause needless humiliation and defiance. As
Toch puts it, assault-prone officers generally use
last resorts as initial moves. Through his opening
frontal approach the policeman thus places
himself in a hazardous position where, given

noncompliance, he has no alternative but to
invoke additional force that eventually culminates

in violence. Verbal orders are escalated to threats
and physical struggles, in the course of which
instigated antagonism is transformed into law

violations. The policeman finally extricates himself
from a dangerous situation largely of his own
creation by overpowering his challenger and
placing him under arrest.
Toch reports that in other types of social
interactions people who are prone to act

aggressively are similarly incited to violence when
their reputation, status, and manliness are
challenged or they are otherwise affronted and

slighted. High sensitivity to embarrassing
treatment is usually combined with deficient
verbal and social skills for dealing with
altercations that can easily erupt into violence.
Where verbally skilled persons often resolve

disputes and restore their self-esteem by verbal
means without having to dispose of their
antagonists physically, those who have
impoverished coping techniques quickly resort to
violent solutions.
Insult provokes counterattack as well in
individuals who ordinarily eschew aggression
(Berkowitz, 1965b; Geen, 1968). In the latter

cases, however, insult alone is sometimes
ineffective in increasing punitiveness, but it does

heighten aggressive responding when supported
by hostile modeling and other disinhibitory
influences (Hartmann, 1969; Wheeler and
Caggiula, 1966).
No one has traced, either developmentally or
experimentally, the process whereby insults
acquire aggression-provoking potential. The most
plausible explanation is in terms of differential

reinforcement. Affronts that are not successfully
counteracted can have far-reaching consequences
for victims. Among other things, ineffectual
recipients become easy targets for further
victimization; they are apt to forfeit their standing
in social hierarchies along with the recognition
and power to control resources that go with it; and
they suffer a loss in self-respect because of their
inability to curb humiliating treatment. To the

extent that punishment of insults by
counteraggression reduces the likelihood of
further mistreatment, the insult-aggression
reaction becomes well established. Conversely,
when retaliative responses are consistently and

severely punished, insult produces acquiescence
with its attendant social and self-devaluative
consequences.
Unfavorable Changes in Positive Reinforcement
After individuals have come to expect their
actions to be rewarded at a certain level,
withdrawal or reduction of positive reinforcement

serves as an aversive event. Like other types of
unpleasant treatments, reinforcement withdrawal
functions as both a punisher for responses and a
conditioner of negative properties to formerly
neutral events with which it is regularly
associated (Azrin and Holz, 1966; Leitenberg,

1965).
Since aggression is to some extent aversively
controlled, it is not surprising to find that
unfavorable changes in reinforcement can
provoke assaultive behavior. Hutchinson, Azrin,

and Hunt (1968) report that hungry animals were
disinclined to attack a member of their species
located in the same chamber as long as their
instrumental responses never produced food

rewards. On the other hand, when food
reinforcement was given and later withdrawn,
they frequently attacked their companion upon

transition to extinction. In successive phases of the
experiment, attacks rapidly declined during times
when food reinforcement was consistently
withheld but quickly reappeared when food
rewards were alternately given and revoked. The
more prior reinforcement they received, the
stronger was the assaultive behavior when

rewards were terminated. Extinction-induced
aggression was evident even in animals that were
raised in social isolation and thus had no prior
history of competitive battles over food. Close
examination of temporal relations between

different modes of response discloses that when
customary rewards are first withdrawn, animals
intensify instrumental responding that previously
secured reinforcement; but as their energetic
efforts fail, they resort to combative actions

(Thompson and Bloom, 1966).
In the preceding studies attack reactions were

precipitated by removing all positive
reinforcement for performances that had been
consistently rewarded. Less drastic reductions in
conditions of reinforcement may likewise provoke
aggression. Schedules of positive reinforcement
requiring high expenditure of effort for minimal
rewards are generally experienced as aversive.

Hutchinson, Azrin, and Hunt (1968) found that, as
work requirements for reinforcement were
progressively increased, primates engaged in
biting attacks even though their instrumental
behavior was occasionally rewarded. Return to a

more favorable payoff schedule reduced attack
reactions. Other studies of schedule-induced
aggression (Flory, 1969a; Gentry, 1968; Gentry
and Schaeffer, 1969) verify that sparing
reinforcements for effortful performances can act

as elicitors of aggressive behavior. Physical attack
is provoked by increasing the time between

rewards as well as by raising work demands
(Flory, 1969b). Although meager payoffs provoke
aggression, terminating them altogether is also

aggression arousing (Knutson, 1970). With
continued nonreward, however, both work
performances and assaultive behavior diminish
and after a time virtually disappear. Thompson

and Bloom (1966) have further shown that when
efforts to secure reinforcement are repeatedly

extinguished, behavioral responsiveness,
instrumental as well as combative, is substantially
depressed and not easily reinstated by occasional
reward. These phenomena are not unlike

naturalistic observations that people who are
chronically subjected to adverse conditions of
reinforcement develop a feeling of helplessness

and become passively resigned to their dismal
living conditions.
Extinction-induced aggression in humans has
received little systematic study. A number of
behavior therapists (Bucher and Lovaas, 1968;
Williams, 1959; Wolf, Risky, and Mees, 1964) have
noted that withdrawal of social reinforcement for
deviant behavior sometimes produces a
temporary intensification of aggressive reactions.
An especially hazardous disorder, prevalent

among autistic and grossly retarded children,
involves self-injurious behavior. The conditions
eliciting and maintaining self-mutilation are not
fully understood, but case studies by Lovaas and
his co-workers (Lovaas and Simmons, 1969) have

established reinforcement withdrawal as one
possible instigator of self-mutilation. Withholding
positive attention for other behaviors was
accompanied by substantial increases in self-
hurtful actions. Lovaas speculates that through

past experiences autistic children learned to
behave self-destructively as a desperate way of

reinstating attentiveness from unresponsive
caretakers.
Overcrowding is an environmental condition
often invoked as an aversive instigator of
aggression. There are certain features of density
groupings that could facilitate combativeness. At
the familial level, children who live in cramped

quarters are likely to spend more time on
neighborhood streets than those residing in
spacious households. To the extent that
neighborhood influences foster antisocial styles of
life, the probability of developing aggressive

conduct is thereby increased. Moreover, parents
who have to cope with many children in congested
quarters are often forced to punitive measures to
halt squabbles between family members who
repeatedly get in one another’s way. At the

broader social level, the immense logistic
problems of servicing hordes of people inevitably

create annoyances. Crowding heightens
competitiveness for services and desired
resources, which can also be exasperating. Densely

crowded conditions not only increase the
likelihood of interpersonal friction, but pollute and
otherwise intensify the aversive qualities of the
physical environment. In addition to the irritants

of crowded living, it fosters a sense of anonymity
and impersonality conducive to inconsiderate

behavior.
Despite widespread concern over the
ecological dangers posed by overpopulation, the

possible effects of population density on human
aggressiveness have received little experimental
attention. The psychological consequences of
overcrowding are difficult to isolate because it is
usually associated with, among other factors,

poverty, ethnicity, unemployment, and inadequate
law enforcement, which confound the causes. The
problem of untangling the multiple causation is
further complicated by the fact that populousness
may help to create some of the correlated
conditions. Comparison of homicide rates for
different nations and for different regions within
the same country indicate that crowding in itself
does not necessarily increase combativeness.

Highly populated locales that traditionally eschew
aggressive conduct have relatively low incidences
of violence, whereas inhabitants of less crowded
places who view aggressive conduct permissively
and make lethal weapons easily available

experience higher rates of interpersonal assaults.
Such data indicate that the level of combativeness
within a society can be better predicted from
prevailing social sanctions than from population
density. For reasons given earlier, crowding would

be expected to enhance aggressiveness in
conjunction with other aggression-promoting

influences.
The problem is more complicated than it might
at first appear, however. Crowded living is not
invariably stressful. Man is immensely adaptable.
People who become accustomed to a city life style
may at first dislike isolated country living and vice
versa. Those who discover sufficient rewarding

features in the new environment eventually
change their preferences. Interpersonal clashes
are curtailed even under tremendous crowding
through social organization and systems of
control. Various segments of the population are

spatially separated in different areas of a
community on the basis of ethnic, religious, or
socioeconomic characteristics, thus reducing the
opportunity for conflict. The threat of personal
retaliation and legal restraints additionally

discourage people from taking what they want by
forceful means. It is therefore not uncommon for

extreme poverty to exist amid affluence without
serious challenges to inequities. Similar patterns
have been noted in the behavior of animals living

in crowded colonies (Calhoun, 1950, 1962).
Dominant animals gain control over resources,
and after a few painful skirmishes, subordinate
members resign themselves to a meager existence

on the fringes of the colony. The probability of
widespread aggression increases under conditions

in which massive population growth overwhelms
organizational constraints.
Thwarting of Goal-directed Behavior
Proponents of the frustration-aggression

theory (Berkowitz, 1965a; Dollard et al., 1939;
Feshbach, 1970) define frustration in terms of
interference or blocking of goal-seeking activities.
In this view, people are provoked to aggression
when obstructed, delayed, or otherwise thwarted

from getting what they want. Research bearing on
this issue shows that thwarting can have a variety

of behavioral effects, though under certain
conditions it provokes forceful actions.
Frustrative conditions essentially prevent or
delay desired reinforcement. The obstacles may
take a variety of forms. Sometimes they are

created by environmental barriers, physical or
social, such as impoverished living conditions,

obstructions and privations resulting from
overcrowding, and restrictive laws and
discriminatory social practices. Personal
limitations, both physical and psychological, serve

as other potential barriers. Positions in various
status and power hierarchies are to a large extent
determined by social and vocational

competencies. The degree of control that a person
can exercise over his own activities, the power to

modify his environment, and the accessibility to,
and control over, valued resources increases with
higher status positions. People who lack

sociovocational proficiencies are relegated to a
subordinate status, in which not only is their
welfare subject to arbitrary external controls, but
they are irreversibly channeled into an economic
and social mode of living that further restricts
their opportunities to improve their lot in life. Self-
restraints, arising from conditioned inhibitions

and self-censuring reactions, may also curtail the
types of gratifications that people can secure from
among those that are socially permissible.
A number of experiments, conducted with both

children and adults, in which blocking of goal-
directed behavior and social learning factors are
independently varied disclose that thwarting is a
relatively weak and unpredictable determinant of

aggression. Although it sometimes heightens
aggression (Buss, 1963; Geen and Berkowitz,
1967; Mallick and McCandless, 1966; Ulrich and
Flavell, 1970), thwarting often exerts no
significant influence, whereas variables within the
same studies such as modeling, prior aggressive
training, and the functional value of aggressive
actions emerge as major determinants of
aggression (Buss, 1966a; Gentry, 1970; Jegard and

Walters, 1960; Kuhn, Madsen, and Becker, 1967;
Walters and Brown, 1963).
People are more easily provoked by insult than
by blocking their goal-directed behavior. As was
previously noted, they are also more inclined to

respond aggressively to unwarranted than to
justifiable hindrances, even though their goal-
seeking behavior is equally thwarted. The latter
findings suggest that in instances where thwarting

arouses aggression, the implied insult rather than
response interference is perhaps the more critical
instigating factor.
The role of thwarting can be best understood
by considering its effects on both aggressive and
nonaggressive responses. A number of
investigators have studied the intensity of
responding after expected rewards have been
withheld or delayed at different points from the

goal. Children blocked near to the goal respond
more vigorously than if they are thwarted at some
distance from it (Haner and Brown, 1955;
Longstreth, 1966). In a related study, Penney

(1960) showed that blocking may reside in the
intensification of a response which is not part of,
but immediately succeeds, the thwarted
instrumental activity, provided a strong
expectation of reward was created. Holton (1961)

similarly demonstrated that the more often
children had been reinforced in the past, which

presumably created reward expectation, the more
forcefully they responded to nonreward,

especially if blocked close to goal attainment.
The vigor of instrumental responses may be

increased by delay as well as by omission of
anticipated reward. The forcefulness with which
children perform a behavior that brings them
rewards is heightened with increasing delay of

reward and reduced with decreasing delay. The
effects of frustrative deprivation are not confined
to performers. Observers also respond more
intensely after seeing another person barred from
achieving a goal that is within easy reach

(Kobasigawa, 1965).
It is evident from these findings that thwarting
can temporarily intensify performance. Since
responses of high magnitude tend to be labeled as

aggressive, when people react more vigorously
after their efforts fail or are obstructed or their
successful performances are simply nonrewarded,

the strengthened responding is typically
construed as aggression. The major questions of
interest are twofold: Is intensified responsiveness,
when it does occur, attributable to motivational or
directive properties of frustrative events, and is
the thwarting-vigorous-reaction relationship
inherent or established through prior learning?

There is some evidence (Amsel, 1958; Blixt and
Ley, 1969) that thwarting can function as an
aversive arouser that energizes behavior, but
learning experiences appear to be the principal
determinants of the kinds of responses that are

activated (Bolles, 1967; Davitz, 1952). Forceful
modes of behavior generally prove more
successful in surmounting obstacles than
dispirited efforts; consequently, people are
differentially reinforced for responding vigorously

to frustrative conditions. Various obstacles and
thwarting agents may eventually become cues for

forceful behavior independently of changes in
motivational level. Natural contingencies
ordinarily favor vigorous reactions to thwarting,

and it is perhaps for this reason that the
relationship is frequently observed. It can be
reversed, however, by altering the contingencies
so that weak rather than forceful responding to

frustration pays off (Marx, 1956). Frustrative
nonreward can thus come to elicit diverse

reactions depending upon what has been found
through repeated experience to work best.
In the preceding laboratory studies, thwarting

failed to provoke forceful actions in people who
had not experienced sufficient positive
reinforcement to develop reward expectations and
in those who were blocked far enough from the
goal that it appeared unattainable. Indeed,

prolonged thwarting and nonreward of efforts to
secure valued outcomes eventually leads to
apathy. These findings are in accord with field
studies examining the influence of relative
deprivation and behavioral efficacy on militant
actions.
Explanations of civil disorders frequently
invoke as principal causal factors the pervasive
impoverishment, exploitation, and feelings of

powerlessness to control conditions that affect
one’s life. Ransford (1968), for example, reports

that, among black residents interviewed in Watts,
those who felt most isolated, powerless, and
dissatisfied were most willing to endorse violent
methods. The view that discontent breeds

disorder requires qualifications, however. It
follows from this type of causal explanation that
collective violence would be more likely to occur

in cities where such conditions are prevalent than
in those in which they are present to a lesser

degree. Lieberson and Silverman (1965) provide
some evidence, though admittedly based on gross
measures, that adverse conditions of life are poor

predictors of the location and timing of riots. Cities
experiencing violent civil disturbances did not
differ significantly in social and economic
conditions commonly assumed to be sources of
collective aggression from cities similar in size and

region that had no riots. If anything, civil
disruptions were more likely to occur where

black-white differences in income and
occupational pursuits were smaller, but
discrimination practices remained in the system.
Comparisons of rioters and nonrioters within the

same communities yield essentially similar results.
The groups did not differ on factors such as
employment and income, but those who

participated in urban violence were generally
better educated, had greater racial pride, and

favored coercive action to secure civil rights
(Caplan, 1970). In considering why some strained
situations erupt into riots and others do not, it

should be noted that almost every urban
disturbance was started by a police encounter
with a ghetto resident that provoked onlookers to
retaliatory violence. The fact that fortuitous events
serve as precipitants of social disturbances makes

prediction difficult from preconditions alone.
The major limitation of the privation theory of

social unrest is clearly apparent when one
considers that, despite condemnation of their
degrading and exploited conditions of life,

comparatively few of the sufferers take active
measures to force warranted changes. A vast
majority of the disadvantaged do not engage in
disruptive public protest, and even in cities
experiencing civil disturbances, only about 15 to

20 percent of ghetto residents actively participate
in the aggressive activities (Bowen, Bowen,
Gawser, and Masotti, 1968; Sears and McConahay,
1969; McCord and Howard, 1968). The critical
question for social scientists to answer is not why
some people who are subjected to gross
maltreatment aggress, but rather why a sizable
majority of them acquiesce to dismal living
conditions in the midst of affluent styles of life. To

facilely invoke the frustration-aggression
hypothesis, as is commonly done, is to disregard
the more striking evidence that severe frustration
is generally accompanied by feelings of
hopelessness and massive servility. Pervasive

discontent may be a necessary but obviously not a
sufficient cause of collective aggression.
Informal observation would indicate that
discontent produces aggression not in those who

have lost hope, but in those whose assertive
efforts at social and economic betterment have
been periodically reinforced; consequently, they

have some reason to expect that coercive action
will force additional social change. Psychological
research on reinforcement processes has
established that the value of given outcomes is
determined relationally rather than by their
absolute properties (Bandura, 1971b; Premack,
1965). The same level of compensation may thus

be experienced as rewarding or punishing
depending on the nature, frequency, and
magnitude of past reinforcement. People not only
compare their present gains with those they
secured in the past, but also continually observe

the rewards and punishments accruing to others.
Unfavorable discrepancies between observed and
experienced outcomes tend to create discontent,
whereas individuals may be satisfied with limited
rewards as long as they are equivalent to, or

exceed, what others are receiving.
Selective processes operate in the standards

used to evaluate one’s outcomes. According to
social comparison theory (Festinger, 1954;
Pettigrew, 1967) people tend to select for
comparison those who are somewhat similar to
themselves. Disadvantaged people are therefore
more likely to be dissatisfied with their life
situations when members of their own group

achieve differential advancements than when all
remain subordinated. Upward mobility may also
breed discontent, especially when self-betterment
made possible by the development of requisite
competencies raises aspirations, the attainment of

which is then thwarted by prejudicial barriers.
While the disadvantaged slowly improve their lot
in life, the more affluent members of society
usually make more rapid progress, so that the
disparity between the groups widens. Hard-won

gains become relative losses (Pettigrew, 1963).
Betterment of social conditions can thus give rise

to discontent. When coercive pressures are
applied to secure further reforms, protesters are
criticized as ungrateful for the benefits they have

been granted, and warned that their militant
actions will not only jeopardize additional
advances but may even cost them what they have
already gained.
Current explanations of violent protest
emphasize relative deprivation rather than the

actual level of aversive conditions as the instigator
of collective aggression. In a survey study,
Crawford and Naditch (1970) report that blacks

who judged their present status as relatively close
to their aspirations were inclined to disapprove of
militant action and to believe that social changes
could be better achieved through persuasion than
through force. On the other hand, blacks who felt

that their present life situation fell short of their
aspirations displayed more militant attitudes. A
majority of them regarded riotous actions as a
legitimate and effective means of achieving
reforms.
In an analysis of conditions preceding major
revolutions, Davies (1962, 1969) shows that
enduring severe privations is more likely to create

resignation and despair than revolution. He
attributes people’s submissiveness to their

preoccupation with physical survival. An
alternative explanation is that their efforts to
effect social change are extinguished to the point
of apathy. Davies builds a convincing case that

revolutions are most likely to occur when a period
of social and economic improvement that instills
rising expectations is followed by a sharp reversal.

This pattern of changes would both create
aversive social conditions that breed the

widespread discontent required for collective
action and reinforce a sense of behavioral efficacy
that intensifies efforts to regain previously

experienced benefits. Collective discontent,
however, appears to play a more influential role in
revolutionary violence than in protest activities,
which are determined more by the coercive

potential of challenging groups and the social
sanction of forceful tactics (Gurr, 1970a, b).
Although aggression is more likely to be

provoked by relative than by absolute deprivation,
the predictive value of this determinant is limited
by its conceptual and operational ambiguity. There

is little consensus among theorists on the
comparative standards used to measure level of
dissatisfaction. Consequently, variable results may
be obtained depending upon whether people
evaluate their life circumstances in relation to

their aspirations, to their past gratifications, or to
the life situations of others whom they select for
social comparison. Discontent created by raised
aspirations, by reduction of rewards and
privileges from accustomed levels, and by
deceleration in the rate of anticipated
improvement undoubtedly has variant effects.
Moreover, different sources of inequity (for
example, social, economic, political) may have

differential aggression-activating power. A
comprehensive theory would also have to specify
the form of the relationship between size of
negative disparity and likelihood of aggressive
response. Informal evidence suggests a threshold

rather than a linear function. That is, people
lacking coercive power may acquiesce to
considerable inequity only to respond violently
when it exceeds a certain intolerable level.
The behavioral effects of aversive

circumstances are determined not only
relationally, but by their justifiability as well.

Privations judged to be warranted are generally
accepted without indignation, whereas less severe
hardships regarded as unjustified are resented. It
is inequitable deprivation of things people believe
they rightfully deserve that is most conducive to
aggressive action. Even unjustified inequities,
however, may be accepted without undue protest

given expectation either that the maltreatment
will be corrected within the foreseeable future, or
that opportunities will be created which promise
eventual improvement in life circumstances. Hope
thus reduces the aggression-arousing potential of

aversive privation. To complicate matters further,
response to inequitable deprivation is highly
influenced by other controlling variables. A
relatively low level of discontent combined with
coercive power is more apt to produce vigorous

protest than high, though powerless,
dissatisfaction. Finally, variations in the scope and

severity of aversive conditions partly determine
whether people will attempt to force changes in
specific practices within a system or more radical

shifts in institutional principles.
Obviously, most people who feel relatively
deprived do not resort to violent action.
Unfortunately, researchers rarely examine the

reinforcements that are provided for alternative
means of securing desired goals. It would be

predicted from social learning theory that people
who are consistently nonrewarded or punished
regardless of what they do would eventually lower
their aspirations and acquiesce in their

impoverished conditions of life; those who achieve
some measure of success through the
development of their potentialities would eschew

violent tactics; and those who had found
aggression to be the only effective means available

for getting what they want or for expressing their
grievances would favor violent actions.
Several lines of evidence indicate that some
reinforcement of personal efficacy is an important
factor in aggressive response to aversive social
conditions. In many nations it is university
students rather than the grossly underprivileged
segments of the society who initiate protest

movements that force social reforms and topple
disliked governments (Lipset, 1966). Student
activists in the United States came
disproportionately from advantaged backgrounds
and prestigious universities that attract the ablest

students (Keniston, 1967). Admittedly, factors
other than personal efficacy foster student
activism. The lack of constraining responsibilities
and availability of free time make it easier to act
on idealistic convictions when risks are involved.

By tradition, students are generally granted some
degree of immunity from community control of

their protest behavior. The broader awareness
gained through critical study can create in highly
principled students estrangement, if not outrage,

over social inequities, even though they may
occupy relatively privileged positions within the
system. Since the vast majority of students,
however, do not engage in protest activities, an

adequate explanation of student activism must
include personal as well as common

organizational determinants. These issues are
discussed at some length in the next chapter.
Studies of urban violence provide further

evidence bearing on reinforced efficacy and
aggressive responsiveness to maltreatment.
Contrary to the popular belief that riotous
behavior is performed by disreputable elements of
ghetto society, those who participate in civil

disturbances have achieved some economic and
educational gains to reinforce a sense of racial
pride, but they feel unjustly excluded from
opportunities within their capabilities by a
widespread system of discrimination (Caplan and
Paige, 1968; Tomlinson, 1968). After collective
aggression is begun, the disinhibiting effects of
widespread modeling and the immediate rewards
resulting from aggressive activities may instigate

aggressive actions in ordinarily passive
individuals. The personal determinants of social
violence are therefore obscured when studies of
the characteristics of participants in collective
aggression fail to distinguish between initiators

and followers.
People are usually discontented with many
aspects of their life situation, but only a few of
these, and not necessarily the most pressing ones,

ever reach the level of public protest. Explanations
of collective aggression must therefore address
themselves to the timing and selection of specific

problem areas for social action. Blumer (1971)
distinguishes several stages in the selective
process. The initial phase concerns problem
recognition. A variety of factors may determine
why a particular adversity attains visibility at a
certain time, while other concerns receive little
public attention. Adventitious events, coupled

with the crusading efforts of reformers, often help
to dramatize objectionable social conditions. Thus,
prison revolts, mine disasters, shocking crimes,
gross injustices, and incidents involving blatant
misuse of power expose reprehensible practices.

They gain further recognition through agitation of
concerned groups. People generally learn about
the plight of others through the press and
television. Consequently, the mass media, with its
immense dissemination capacity, serves as a

powerful force in creating, mollifying, and
reinforcing public concerns.
After a problem has gained public recognition,
it must then survive the process of social
validation. Unless the discontented can convince
others of the justifiability of their cause, they will
fail to gain the widespread sympathy and support
needed to mobilize dissatisfaction into corrective
action. Groups whose vested interests are

threatened, of course, work actively to minimize
the grievances and to discredit the protesters as
self-serving agitators. Battles over the legitimation
of social problems are fought in public assemblies,
in the press, and on television by various

advocates, including community organizers,
students, lawyers, clergy, politicians, and other
influential backers of contending factions. Since
tactics that are good attention-getters tend to be
self-discrediting by their immoderate nature, it is

not uncommon for protesters to undermine the
justification of their grievances in their effort to

publicize them.
Problems can be well publicized and
characterized in ways that win the support of
socially concerned individuals as well as the
aggrieved, but little is done about them. In fact,
more often than not, drawing attention to
warranted grievances generates a temporary

outpouring of moral indignation but little social
action. Discomfort over appalling disclosures is
thereby relieved without incurring the costs of
challenging individuals or institutions
contributing to the problems. Personal influences

must be brought to bear to get sympathizers to
participate in forceful action. Since quick victories
are difficult to achieve in confrontations with
those who command authorized power,
considerable positive reinforcement is also

required to sustain reformative efforts in the face
of repeated failure. When adequate supportive

rewards are lacking, discouragement and
weariness take a heavy toll of the ranks of
activists.
Considering the complex interplay of opposing
influences at each stage of the change process, it is
hardly surprising that deprivation, whether
defined in absolute or relative terms, is a weak

predictor of collective aggression (McPhail, 1971).
The substantial evidence reviewed in this section
and elsewhere in this book supports the
conclusion that aversive events commonly
subsumed under the term frustration—be they

physical assaults, insults, reinforcement
withdrawal, or thwarting of goal-directed
activities—are, at best, facilitative rather than
necessary or sufficient conditions for aggression.
“Frustration” is most likely to provoke aggression

in individuals who have learned to behave
aggressively and for whom aggression has

functional value.
INSTRUCTIONAL CONTROL OF AGGRESSION
During the process of socialization people are
extensively trained to obey orders. In the early
formative years, when children cannot foresee the
consequences of their actions, parents impose

strict obedience demands, particularly in
hazardous situations. Behavior is brought under
verbal control through differential consequences
for obedient and noncompliant reactions. By
rewarding obedience to directives and punishing

noncompliance, orders eventually acquire
powerful response-controlling properties (Ayllon
and Azrin, 1964). After this form of social
influence is well established, people use it
extensively to get others to do what they want

them to do in a wide range of situations. Since
verbal orders quickly lose their effectiveness if

they go unheeded without penalty, defiance is
seldom treated lightly. In order to preserve
authoritative control, social agents are careful to

avoid situations that risk open defiance, and they
apply strong enforcement measures to ensure that
the demands they do impose are obeyed. This is
especially true in coercive societies where strict

obedience is of considerable importance in
maintaining the existing social system.
Given that people will obey orders, legitimate
authorities can successfully command aggression
from others, especially if the actions are presented

as justified and necessary and the enforcing agents
possess strong coercive power. Indeed, as Snow
(1961) has perceptively observed, “When you
think of the long and gloomy history of man, you
will find more hideous crimes have been

committed in the name of obedience than have
been committed in the name of rebellion” (p. 24).
He cites as an example the atrocities obediently
perpetrated by agents of the Third Reich. The
documentation of this thesis might also include
the atomic holocausts at Hiroshima and Nagasaki,
as well as the more recent wholesale slaughter at
My Lai.
A series of controversial experiments

conducted by Milgram provide graphic
demonstrations of how demands of legitimate
authorities can serve as powerful instigators of
inhumane aggressive actions. In the obedience
paradigm designed to study this phenomenon,

well-meaning adults are told to administer
increasingly severe shocks to a learner in an
adjoining room, ostensibly to examine how
memory is affected by punishment of mistakes.

The apparatus contains thirty graded switches
labeled with voltage ratings and verbal
designations varying from slight shocks to

dangerously severe ones. Each time the learner
errs, the aggressor is ordered to increase the
shock level one step. When the more painful levels
are reached, the learner—actually a confederate of
the experimenter, who never receives any shocks
—vehemently protests the punitive treatment and
emits increasingly agonized screams as the

sessions progress. The experimenter, unperturbed
by these reactions, instructs the aggressor to
disregard the protestations and to step up the
shock level after each error or refusal to respond.
Whenever aggressors voice reluctance to continue

the cruel actions, as they often do, they are simply
told to go on, until they absolutely refuse to
comply with the commands.
In the first experiment (Milgram, 1963),

conducted with paid volunteers of different
occupations, ages, and education, the
experimenter provided the authoritative sanction.

Much to everyone’s surprise, 65 percent of the
adults, though highly distressed over the suffering
they caused by their brutal actions, nevertheless
obeyed the orders to the end, and delivered
shocks designated as extremely dangerous. Those
who refused to go all the way did so only after the
victim began pounding on the wall in desperation.

Ordinarily sensible men will behave in a cruel and
callous manner when told to do so by a seemingly
legitimate authority in an organization of
unknown character as well as in a university
situation of high integrity (Milgram, 1965b). The

following transcript from an obedient subject
illustrates destructive compliance with verbal
disclaimers for one’s injurious actions:
150 volts delivered. You want me to keep

going?
165 volts delivered. That guy is hollering in

there. There’s a lot of them here. He’s liable
to have a heart condition. You want me to
go?
180 volts delivered. He can’t stand it: I’m not

going to kill that man in there: You hear him
hollering? He’s hollering. He can’t stand it.
What if something happens to him? …I’m not
going to get that man sick in there. He’s
hollering in there. You know what I mean? I
mean I refuse to take responsibility. He’s
getting hurt in there. He’s in there hollering.
Too many left here. Geez, if he gets them
wrong. There’s too many of them left. I mean
who is going to take responsibility if
anything happens to that gentleman? (The
experimenter accepts responsibility.) All
right.
195 volts delivered. You see he’s hollering.

Hear that. Gee, I don’t know. (The
experimenter says: “The experiment
requires that you go on.”) I know it does, sir,
but I mean—uh—he don’t know what he’s in
for. He’s up to 195 volts.
210 volts delivered.
225 volts delivered.
240 volts delivered. Aw, no. You mean I’ve got

to keep going up with the scale? No sir. I’m
not going to kill that man: I’m not going to
give him 450 volts. (The experimenter says:
“The experiment requires that you go on.”) I
know it does, but that man is hollering in
there, sir . . . (p. 67)
This unusually high level of destructive
obedience is especially astonishing considering
that the experimenter had no coercive power to

enforce his demands as is characteristic of
authoritative control in everyday life. A number of

factors undoubtedly contributed to the ruthless
obedience. Included among them is anonymity,
protection against retaliative aggression, the force
of a contractual obligation, verbal dissociation of

oneself from one’s actions by assigning
responsibility for them to the compelling
authority, and the assurance that a qualified

researcher would not demand perilous actions
without necessary safeguards.
People are often induced to perform cruel acts
by the collective pressures exerted by peers. A
second experiment demonstrated that people find
it difficult to behave counter to what the group
wants just as they are averse to defying
legitimatized authority (Milgram, 1964). In the
absence of group influences, adults delivered only
weak shocks; whereas under peer pressure calling

for increasingly hurtful actions, men punished the
victim with increasing severity despite his
desperate cries. Seeing others carrying out
punitive orders calmly likewise increases obedient
aggression (Powers and Geen, 1972). Peer

influences, depending on their nature, can also
undermine authoritative control. Thus, under
conditions in which others defy the
experimenter’s brutal demands, only 10 percent of
the aggressors followed the issued commands

unfailingly in the face of peer rebellion and
victims’ pleas (Milgram, 1965a). Modeled defiance,

however, is less effective in reducing obedient
aggression if resisters become emotionally upset
than if they are unperturbed by their disobedience

(Powers and Geen, 1972). Nervous defiance
apparently conveys the impression that either
opposition is risky or the challenger is faint-
hearted and lacks conviction in the stand he has

taken. The degree of compliance with demands for
aggression shown by men under various social

conditions is depicted graphically in Figure 3.8.
It is relatively easy to hurt a person when his
suffering is not visible and when causal actions

seem physically or temporally remote from their
deleterious effects. Mechanized forms of warfare,
in which masses of people can be put to death by
destructive forces released remotely, illustrate
such depersonalized aggression. When the

injurious consequences of one’s actions are fully
evident, vicariously aroused distress and loss of
self-respect serve as restraining influences over
aggressive conduct that is otherwise
authoritatively sanctioned. Milgram (1965b)
obtained diminishing obedience as the harmful
FIGURE 3.8 Percent of subjects who aggressed compliantly at
different shock levels. The arrows on the graph for the defiant
peers indicate the points at which each of the peers refused
the authority’s command to shock the victim more severely.
(Plotted from data by Milgram, 1963, 1964, 1965a)
consequences of punitive actions became
increasingly more salient and personalized. The
percent of aggressors who eventually disobeyed

brutal orders was 34 percent when the victim
pounded the wall in an adjacent room; 38 percent
when they heard agonized cries through a door
left ajar; 60 percent when the victim expressed the
same pleas and protests but in the aggressor’s

presence; and 70 percent when, in addition to
seeing the victim’s suffering, aggressors also had
to place the victim’s hand forcibly on the shock-
plate in order to punish him as ordered. And when
the commands to increase the severity of

punishment were issued by telephone without
social surveillance, subjects disregarded the

authoritative orders by selecting the weakest
shock levels.
When violence toward people designated as
foes is not only legitimized but is a source of
personal pride, as in warfare or sanctioned
oppression of minorities, aggressors share the
same values as their peers and superiors. Under

such circumstances, there are no strong self-
imposed restraints to counteract or circumvent
cruel dictates. Malevolent authorities as well as
those of good will who extenuate morally
reprehensible practices, can, given proper

conditions, get otherwise considerate people to
behave inhumanely. In other words, it requires
appropriate social influences rather than
monstrous people to produce heinous deeds.
NONRECURRENT AGGRESSORS
Interpersonal violence is commonly viewed as
an impulsive, irrational, and pathological type of
behavior. Social learning theory, on the other
hand, conceives of aggression as a learned conduct
that, like other forms of social behavior, is under
situational, reinforcement, and cognitive control.
The view of aggression as a reckless explosive
reaction is undoubtedly reinforced by periodic

dramatic episodes in which otherwise law-abiding
persons perform heinous crimes. Bizarre
mutilation of victims, frenzied overkill of
strangers, and sensational “crimes of passion” are
striking examples that figure prominently in the

mass media.
It has been repeatedly shown that, barring
significant changes in social conditions, the best
predictor of future behavior is past behavior

(Mischel, 1968). This finding is consistent with the
social learning view that a person’s behavior is
highly influential in creating social conditions that,

in turn, tend to maintain the behavior. Previous
performance has high predictive value for
recurrent aggressors, but it is of no aid in
identifying individuals who are likely to commit a
monstrous act of violence without any prior
history of assaultive behavior. Here the predictive
and analytic problems are complicated by the fact

that unusual forms of violence generally have
unusual idiosyncratic determinants.
An earlier analysis of grossly deviant behavior
illustrated how behavior can come under bizarre
symbolic control to the point where it remains

refractory to external reality influences. Every so
often tragic episodes occur in which individuals
are led by delusional beliefs to commit acts of
violence. Some follow divine inner voices

commanding them to murder. Others are
instigated to self-protective attacks by paranoid
suspicions that the victim is conspiring to harm

them. And still others are prompted by grandiose
convictions that it is their heroic responsibility to
eliminate evil individuals in positions of power.
A study of Presidential assassins (Weisz and
Taylor, 1970) shows that, with one exception, the
murderous assaults were partly under delusional

control. The assassins acted either under divine
mandate, through alarm that the President was in

a conspiracy with treacherous foreign agents to
overthrow the government, or on the conviction
that their own adversities resulted from
Presidential persecution. Ordinarily, erroneous

beliefs are modified through disconfirming
evidence presented by others. The assassins,
however, tend to be loners. Being unusually

seclusive in their behavior, they effectively
shielded their erroneous beliefs from corrective

influences. Serious personal failures, also a
prominent feature of this group, further
strengthened the tendency to assign their

misfortune to pernicious external causes. In the
case of autistically determined aggression, both
the instigators of the violent acts and the dreaded
consequences they are intended to forestall are to
a large extent internally generated.
A number of studies have been conducted on
whether chronic aggressors can be distinguished

from occasional extreme assaulters by their
personality characteristics (Blackburn, 1968;
Fisher, 1970; Megargee, 1966; Megargee and

Mendelsohn, 1962). The individuals included in
the latter group tend to overinhibit aggressive
responses, but when they are finally provoked by
cumulative or extreme instigation to commit an
aggressive act, they do so in a violent way. These

two types of aggressors can be easily
distinguished behaviorally, but they do not differ
much in their responses to conventional
psychological tests. The behavioral discontinuity
of the obsequious murderer may be somewhat
overemphasized because his past aggressive
actions, unless they take blatant or antisocial
forms, are seldom recorded. Incongruity is more
characteristic of journalistic accounts than of the

behavioral history of retiring assassins. It is not
uncommon for mild-mannered assailants to be
unusually well stocked with guns and to be highly
proficient in their use. This requires some prior
engrossment in activities productive of violence.

Charles Whitman, the student sniper who went on
a mass murder spree, is a good case in point.
Newspaper reports described Whitman as a
former altar boy and Eagle Scout who exemplified

the virtues of an All-American youth (The New
York Times, 1966a). In fact, he was much less
righteous than the summary characterization

would imply (The New York Times, 1966b, 1966c).
Whitman was raised in an atmosphere of familial
violence. Like his father, who repeatedly beat his
wife and displayed a “fanatic” interest in guns,
Whitman was known to assault his wife and
others with a minimum of provocation. From early
childhood he also shared his father’s fondness for

firearms. As a marine recruit he was court-
martialed for insubordination and fighting. Upon
discharge he had occasional brushes with the law
in connection with aggressive activities. Whitman
was once apprehended with two other students

for poaching deer when they were caught dressing
the carcass in a dormitory bathroom. On another
occasion he threw an exchange student out of a
classroom for sitting on his chair. At about the
time of the murder spree, Whitman was drifting

deeper into illegal activities, including, among
other things, passing bad checks for heavy

gambling debts and unsuccessful efforts to
become a pornography supplier to local
merchants. Contrary to journalistic accounts,

Whitman’s behavioral record was not that of a
meek altar boy suddenly gone berserk.
Newsman similarly depicted Howard Unruh,

who fatally shot thirteen people during a
murderous rampage on a Camden street, as a
“mild, soft-spoken veteran” with a reputation of

being a “model boy” (The New York Times, 1949).
Whatever other attributes he may have displayed,
his intense engrossment with instruments of

violence and neighborhood disputes belied the
pacific journalistic characterization. He managed
through various ruses to bring home from Europe
an arsenal of weapons that he collected while in
the Army. The walls of his room were adorned

with crossed pistols and bayonets and pictures of
armored artillery in action. Scattered about his
room were pistols, machetes, and ashtrays made
of cartridges. In the months preceding the
murderous episode, Unruh spent his time
seclusively reading Scriptures, practicing pistol
shooting in his back yard, and building up
resentment against neighborhood shopkeepers
who he believed were persecuting him by making

derogatory remarks about his character. His
persecutory beliefs were periodically reinforced
by arguments with his neighbors over his
disturbing behavior. After sitting through several
showings of a double feature, I Cheated the Law

and The Lady Gambles, into the small hours of the
morning, Unruh returned home, whereupon he
made up his mind to slay his tormentors. The next
morning he walked down the street of the
neighborhood shopping center methodically

shooting merchants and their customers with the
deadly accuracy that won him marksmanship

medals in the Army.
Unruh displayed most of the distinguishing
features of psychotic assassins. There is the
deterioration in socioeconomic functioning;
idleness and marked seclusiveness foster the
development of persecutory and autistic
resentments that become magnified in the absence

of disconfirming judgments of trusted friends;
eventually delusional constructions assume
greater reality than the actual social events and,
given appropriate instigating conditions, they take
violent action. This conceptualization of

homicidally extreme aggression by otherwise
withdrawn individuals differs in many important
respects from explanations in terms of
undischarged aggressive drives exceeding rigid
controls. Although they may have little prior

history of criminally assaultive behavior, the
allegedly quiet assassins such as Oswald,

Whitman, and Unruh were all skilled riflemen who
regularly engaged in far more aggressive activities
that could “discharge pent-up aggression” than

nonassaultive people who are also subjected to
aggression instigation.
One can undoubtedly identify some

distinguishing attributes among persons differing
in aggressive patterns, but they may have limited
value in predicting homicidal behavior in

individual cases. For every psychotic or
overinhibited person who behaves violently, there
are countless others with similar characteristics

who rarely display such conduct. People
diagnosed as mentally ill, for example, are neither
more nor less prone to acts of violence before,
during, or after hospitalization than the general
public (Gulevich and Bourne, 1970). Study of

personality characteristics can yield valid
predictors of homicidal behavior only if the causes
reside primarily in the individual. If social
conditions are major instigators of aggression, as
is generally the case, then prediction of when,
where, and by whom acts of violence will be
committed cannot be achieved without
assessment of instigating social influences. Bizarre
homicides are usually produced by the coexistence

of several unusual factors, such that if any one of
them were absent the event would not have
occurred. Unless there exist some common
determinants of peculiar aggression across
individual cases, the study of rare acts provoked

by an idiosyncratic combination of influences will
not yield much generalizable information.
chapter four
Maintaining Conditions
The analysis of aggression has centered thus
far on the development of injurious patterns of
behavior and their various antecedent
determinants. The third major feature of the social
learning formulation is concerned with the
conditions that sustain aggressive responding.
Maintenance of behavior depends in large part on

prevailing conditions of reinforcement. The
principle that behavior is strongly controlled by its

consequences applies equally to aggression.
Aggressive actions that are rewarded tend to be
repeated, whereas those that are unrewarded or
punished are generally discarded. Thus,
aggressive modes of response, like other forms of
social behavior, can be produced, eliminated, and

reinstated by altering the effects they produce.
This chapter discusses the wide range of

rewards and punishments that determine whether
or not people will behave aggressively under given
circumstances. In traditional theories of learning,

reinforcement influences are largely confined to
the effects of external outcomes impinging directly
upon the performer. Social learning theory
distinguishes three forms of reinforcement control
of aggression: the influence of direct external
reinforcement, vicarious or observed

reinforcement, and self-reinforcement.
EXTERNAL REINFORCEMENT
People aggress for a variety of reasons. Some
resort to force to appropriate tangible resources
they desire. Some behave aggressively because it
wins them approval and status rewards. Still

others may rely on aggressive conquests to bolster
their self-esteem and sense of manliness. And
some may derive satisfaction from seeing the
expressions of suffering they inflict on their

victims. Essentially the same aggressive actions
thus may have markedly different functional value
for different individuals and for the same
individual on different occasions.
There are no conceptual or empirical grounds
for regarding aggression maintained by certain
reinforcers as more genuine or important than

others. A complete theory must account for all
aggressive actions whatever purposes they serve.

The research findings reviewed in succeeding
sections reveal that, like other forms of social

behavior, aggression is maintained by diverse
conditions of reinforcement. It appears, however,
that aggressive actions are primarily reinforced by
their functional value, whereas accompanying

expressions of suffering by the victim tend, if
anything, to have inhibiting effects. Extrinsic
rewards assume special significance in
interpersonal aggression because such behavior,
by its very nature, generally creates a punishment

contingency. A person who gets into fights will
suffer some injury even though he eventually
triumphs over his opponents. Under noncoercive
conditions, positive incentives are therefore
needed to overcome the inhibitions aroused by the

inherent aversive effects of aggression. Rewards
become increasingly important as the costs of

aggression become greater.
Tangible Rewards for Aggression
Aggressive behaviors are often repeated
because, if skillfully performed, they serve as
effective means of securing desired tangible
rewards. By threats of physical violence or

aggressive actions children can forcibly
appropriate possessions of vulnerable peers;
delinquents and adult transgressors can support
themselves or costly drug habits on income
derived from aggressive pursuits; through forceful

means protesters can secure social reforms that
materially affect their lives; and nations are
sometimes able to gain control over prized
territories through warfare.
The degree to which individual or collective

aggression is affected by its material
consequences is difficult to evaluate under

naturally occurring conditions. Unless one can
observe the outcomes associated with all acts of
aggression—an immense task—sample
observations in a selected setting may provide an
incomplete picture of the actual schedule on which
aggression is reinforced. In the case of antisocial
pursuits, delinquents are understandably averse

to having the profits of their muggings recorded.
Establishing reliable functional relationships is
further complicated because aggressive actions
ordinarily produce diverse effects that are not
easily separable. Some of the concomitant

consequences might augment the aggressive
behavior, others might attenuate it, and still others
might exert little influence on it. For these reasons,
the reinforcement causes of aggression can be best
understood by controlled experimental

demonstrations.
Control of aggression by its material

consequences has rarely been studied with
humans, although several experiments with
animals have been addressed to this issue. Ulrich,
Johnston, Richardson, and Wolff (1963) report
that initially docile rats who had been deprived of
water attacked their partners with increased
intensity when fighting responses won them

drinks, but ceased aggressing when water rewards
for such behavior were withdrawn.
The schedule and frequency of reinforcement
apparently control aggressive behavior in
essentially the same predictable way as they do

nonaggressive responses (Azrin and Hutchinson,
1967). As shown in Figure 4.1, food-deprived
pigeons displayed no aggression when food was
made available on a noncontingent basis, whereas

they readily fought when aggressive attacks
produced food. The more often their fighting
responses were rewarded, the more aggressive

they became. When food was again given
regardless of how they behaved, their attacks
declined to a near zero level. Once fighting
responses are established, they can be effectively
FIGURE 4.1 Percent of time spent by two hungry pigeons
attacking a target pigeon during periods when attacks
produced food and when food was given independently of
how they behaved. (Azrin and Hutchinson, 1967)
maintained by occasional reinforcement (Stachnik,
Ulrich, and Mabry, 1966). In the study cited
earlier, Reynolds, Catania, and Skinner (1963) not

only trained pigeons to fight for food rewards, but
by reinforcing their aggressive responses only
under designated circumstances, brought the
fighting under precise situational control.
The power over fighting behavior wielded by

positive reinforcement may be somewhat
magnified in these studies because the opponents
selected were either relatively docile or physically
restrained. Pain resulting from more vigorous

defensive combat by the victim would most likely
attenuate the influence of rewards. However,
repeated victories with progressively tougher

opponents can produce vicious fighters not easily
dissuaded by counterattacks.
In everyday life, people are rarely, if ever,
rewarded every time they perform a given action.
Forceful tactics do not always succeed. Aggressors,
however skilled they may be, do not always

emerge unscathed from their frays. Because of the
variable nature of the social environment,
behavior is characteristically reinforced on an
intermittent basis. People respond quite

differently depending on the pattern and
frequency with which their actions are reinforced.
Those who have been consistently rewarded
become easily discouraged and give up quickly
when their efforts fail. Individuals whose behavior

has been reinforced only periodically tend to
persist for a considerable time despite setbacks

and only occasional success. In the absence of
better alternatives, behavior that has been

reinforced on a thin unpredictable schedule is
exceedingly hard to extinguish because one’s

efforts are sustained by the belief that the actions
will eventually prove successful.
The influential role of the reinforcement

schedule in the maintenance of aggressive
responses is revealed in laboratory studies with
children. Cowan and Walters (1963) rewarded

boys with marbles for punching an automated
clown. They were reinforced after every punch,
every third punch, or every sixth punch. Material

rewards were then discontinued to see how long
children would persist in their aggressive
behavior. During this extinction phase, boys who
were rewarded continuously were the least
aggressive, whereas those who had to perform six

aggressive responses to gain a reward displayed
the highest number of punching attacks. Boys
institutionalized for behavior disorders did not
differ from noninstitutionalized youngsters in
their initial aggressiveness; however, after
rewards for aggression were withdrawn,
institutionalized children continued to behave
more aggressively and took longer to cease their
punching assaults.
Intermittently reinforced aggressive responses
not only are more persistent, but also tend to
generalize to new situations. Walters and Brown
(1963) reinforced boys with marbles for hitting
the automated clown on a continuous schedule or

after every sixth punching response, or they did
not reward hitting behavior. Following aggression
training, children in each condition underwent
frustrating or enjoyable experiences and then

their physical aggressiveness toward another child
was measured in both competitive games and free
responding situations. Boys who were only

occasionally rewarded for aggression during
training were about twice as interpersonally
assaultive as boys who were continuously
reinforced, never reinforced, or received no
aggression training. Surprisingly, the latter three
groups did not differ from each other. Nor did
frustration, in which the experimenter interrupted

an absorbing movie, confiscated the children’s
candy, and threw it away, influence the incidence
of interpersonal aggression. Evidence that
occasional reinforcement of physical aggression in
nonfrustrating impersonal situations facilitates

interpersonal aggression has important
theoretical and social implications.
Social and Status Rewards
In every society some patterns of behavior are
admired, others arouse no special interest, and
still others are held up to ridicule. By behaving in
ways that are highly valued, people can gain status

in the circles in which they travel. There are
certain advantages in holding a high position in a
status hierarchy. Those who possess a superior
rank enjoy greater power to control the activities
of the group than do subordinate members; they
are granted many privileges; they get better

treatment; they have greater access to rewarding
resources; and if their stature is secured through
deserving accomplishments, they receive a great
deal of rewarding attention and adoration. People
therefore tend to aspire to higher standing by

pursuing activities that are status conferring and
by shunning those that lower their personal
stature or bring disfavor. Once they gain
advantageous positions, they will continue to do
whatever is needed to maintain their status in the

privilege system. For similar reasons, people will
engage in costly retaliation to restore their social

prestige following public humiliation.
Social rewards can be distinguished from
status rewards, although they have some features
in common. In the latter case, performance of
valued behavior gains one a social rank that
carries with it diverse benefits as long as that
position is occupied. In the case of social

reinforcement, a person wins approval for specific
actions without necessarily changing his rank in a
status hierarchy. In other words, one can be
praised without being promoted in a status
hierarchy.
Data are lacking on how strongly behavior is

affected by these two types of reinforcement
systems. Martin and her associates (1968),
however, report a study that has some bearing on

this issue. They compared the effectiveness of
treatment programs for delinquents in which their
academic and social performances were either

rewarded individually or won them promotions to
increasingly more privileged ranks. In the latter
system, five levels that required progressively
higher achievements were devised. As the
adolescents advanced through these ranks by
mastering the requisite patterns of behavior, they
were granted greater freedom and more generous

rewards. The rank-contingent reinforcement
produced greater changes in behavior than
previous practices in which specific responses
were individually reinforced.
The relative superiority of status rewards may

be attributed to several factors. If failure to
perform a given action results in loss of a specific
reward, the negative consequence is limited and of
no great import. Maintenance of a status position

typically depends on successful performance of a
broad range of role behaviors. When a few
foolhardy or faulty actions can produce a

demotion in rank, thus creating pervasive negative
consequences, the continuous threat of loss of
status and its attendant rewards effectively
maintains a wide repertoire of behaviors. Pressure
for exemplary performance is especially great
when status positions are limited and there are
many eager competitors for them. Martin et al.

(1968) ascribe some of the power of status reward
systems to their inherent modeling influences. By
designating the hierarchical role behaviors
required for different status positions, members
who aspire to higher status are provided with

explicit symbolic and actual models of what they
must do to gain more privileged ranks.
Some aggressive behaviors are controlled by
the positive social reactions they elicit from

others. Support for this interpretation comes from
several sources. Children who are praised for
hitting increase their assaultive attacks more than

those who receive no approval for such behavior
(Patterson, Ludwig, and Sonoda, 1961). The effects
of social reinforcement on aggression are by no
means confined to children. Adult males praised
for administering high-intensity shocks to another
person became progressively more punitive,
whereas nonreinforced subjects displayed a

relatively low level of aggression (Geen and
Stonner, 1971). These differences are depicted
graphically in Figure 4.2. Individuals rewarded for
aggressive behavior continued to respond more
punitively and generated more hostile thoughts to

verbal aggressive cues after praise for injurious
conduct was discontinued (Geen and Pigg, 1970).
Social reinforcement not only makes further
aggression more likely, but can enhance other

forms of aggression that were not explicitly
rewarded. Both children (Lovaas, 1961) and
adults (Loew, 1967; Parke, Ewall, and Slaby, 1972)

who are socially rewarded for hostile remarks
later exhibit more physical aggression than those
reinforced for positive or neutral verbalizations.
Even seeing the aggressive remarks of others
FIGURE 4.2 Mean intensities of shocks administered by adults
when they were praised for behaving punitively or were never
reinforced for escalating punitive responses. (Geen and
Stonner, 1971)
reinforced can increase physical aggression in
observers (Parke, Wiederholt, and Slaby, 1972).
The generalization effects of reinforcement

may be interpreted in several ways. It may be that
different types of aggression represent a general
response class so that rewarding any one form
automatically strengthens other forms as well. An
alternative, and more tenable, explanation can be

cast in terms of the informative and incentive
functions of reinforcement. Praising hostile
expressions conveys permissiveness or even
invitation to aggressive modes of response. The

likelihood of aggression is increased on future
occasions by expectations that such behavior,
though ordinarily discouraged, is not only

acceptable but brings commendation from others.
Geen and his associates advance the view that
approval of reinforcement automatically
strengthens aggressive habits rather than
reducing inhibitions, on the assumption that
disinhibition should increase aggressiveness to

hostile and neutral situations alike. In point of fact,
responses to markedly different stimuli are never
uniformly rewarded in everyday life; disinhibition
therefore tends to be selective rather than

indiscriminate. Thus, rewarding attacks on peers
will weaken aggression inhibitions toward them,
but will not necessarily eliminate restraints
against assaulting policemen. Since responding
aggressively to hostile cues would be considered

more appropriate and approvable than the same
response to neutral or positive stimuli, the

findings of Geen and Stonner (1971) are
explainable in terms of differential restraint

reduction. Staples and Walters (1964), indeed,
demonstrated that inhibitory processes partly

determine the effects of social reinforcement.
Although females intensified shocks administered
to a victim for praise, those exhibiting strong
aggression inhibition were less responsive to the

social reinforcement. Modeling studies cited
earlier (Bandura, 1965b; Madsen, 1968)

furthermore show that variations in aggression
under different reinforcement conditions
primarily reflect the effect of restraints and
anticipated consequences rather than response

availability. Whether generalization of aggression
accompanying social reinforcement is attributable
to its sanctioning or its habit-strengthening

function can easily be tested by structuring
aggression as commendable behavior without

praising specific actions. The disinhibition
hypothesis would predict comparable initial
increase in aggressive responding under praise

and sanctioning conditions. Since praise is not
only a form of sanction but also conveys a sense of
personal worth, it serves as a source of self-esteem
rewards that can promote aggression over time
more effectively than can a condoning attitude

alone.
It has been amply documented in studies of

verbal conditioning that the frequency with which
adults engage in verbal hostility is partly
determined by the positive interest shown in it

(Binder, McConnell, and Sjoholm, 1957; Buss and
Durkee, 1958; Simkins, 1961). Moreover,
sequential analysis of interactions in
psychotherapeutic interviews reveals that hostile
remarks increase when they are positively

received by therapists, and are either discarded or
directed toward different targets when aggressive
displays evoke disinterest or disapproval
(Bandura, Lipsher, and Miller, 1960; Goldman,
1961). It is interesting to speculate whether, in
interview treatments based on the catharsis
model, clients are venting anger arousal or trying
to please their therapists by telling them what
they want to hear. Given evidence that positive

reinforcement of verbal aggression increases the
likelihood of aggressive action, ventilative
therapies aimed at draining aggressive drives may
inadvertently reinforce aggressive tendencies.
Analyses of social reinforcement of aggressive

behavior in natural settings are in general
agreement with results of laboratory studies.
Bandura and Walters (1959) compared the
reinforcement practices of parents of violence-

prone adolescents and of nonaggressive boys.
Parents of nonaggressive adolescents rarely
reinforced their sons for resorting to physical

aggression in response to provocation. Parents of
assaultive delinquents, on the other hand,
tolerated no aggressive displays whatsoever in the
home, but condoned, actively encouraged, and
reinforced provocative and aggressive actions
toward others in the community. Similar findings
emerged from a study of the reinforcement

conditions prevailing in the families of young
aggressive and inhibited boys (Bandura, 1960).
Parents of inhibited boys consistently disapproved
any type of aggressiveness, thus affording little
opportunity for the development of such forms of

conduct. By contrast, parents of the aggressive
boys, while nonpermissive and punitive for
aggression toward themselves, condoned a great
deal of sibling aggression and encouraged and
rewarded their sons’ aggressive behavior when

this was directed toward other children.
The effects of differential parental

reinforcement were reflected in the boys’
aggressive behavior. They expressed considerably
more physical and verbal aggression toward their
peers, displayed more oppositional behavior
toward their teachers, and showed weaker
restraints over aggression in the face of
provocation than inhibited boys. However, they

displayed no more aggression toward their
parents, which is understandable, in view of the
parents’ punitiveness for aggression toward
themselves.
Ordinarily, activities such as scholastic, artistic,

and athletic accomplishments are highly valued,
and it is noteworthy performances in such
enterprises that win praise. For reasons given
earlier, people willingly forego many personal

gratifications in order to excel in status-conferring
tasks. The countless hours of practice by aspiring
musicians and stage performers, for example, are

partly sustained by visions of future applause.
Football players who have no professional
aspirations nevertheless risk crippling injuries
and endure grimy, arduous drills for social
recognition and some brief moments of glory on
the gridiron.
Cultures differ to some degree in what they

value. Even the same society may, under unusual
circumstances, radically alter its system of social
rewards. During wartime, otherwise
compassionate societies offer medals, promotions,
laudatory press reports, and other social

commendations for skill in killing. Those who
refuse to participate in the sanctioned
manslaughter, whether through fear or because
they value the dignity of human life, are subjected

to humiliating societal punishments. In the Nazi
structure of reinforcement, where enslavement
and execution of designated racial groups were

viewed as meritorious acts of patriotism,
promotions in concentration camps were made
partly on skill in performing mass murders. Camp
commandants proudly compared execution rates
like industrial production figures (Andrus, 1969).
Most societies contain subgroups in which the
members achieve status and recognition through

their skill in fighting. That interpersonal
aggression is often an approved form of conduct,
rather than an impulsive frustration reaction, is
nowhere better illustrated than in observational
studies of delinquent subcultures (Miller, 1958;

Short, 1968; Wolfgang and Ferracuti, 1967;
Yablonsky, 1962). Among the personal qualities
most highly prized in such groups are fighting
prowess, toughness, ability to outsmart others,

and a quest for excitement. Members are
rewarded for fighting exploits and lose stature for
faint-heartedness in the face of insults and combat

challenges. Because aggression assumes such a
central place in the system of values, active
participants show continued concern about their
fighting reputations. Status threats, from either
challengers within the group itself or rival gangs,
are quick to provoke aggressive actions. One need
not be a member of a fighting gang to derive

satisfaction from assaultive conquests. In circles in
which physical prowess commands respect,
beating up a person who has behaved irritatingly
reinforces a sense of manliness and personal
worth (Toch, 1969). Assaultiveness is most often

equated with manhood among males in lower
socioeconomic classes.
Many people have come to believe, on the basis
of popular accounts of sensational episodes, that

gang delinquents spend much of their time in
perilous battles defending their “turf” against rival
territorial intruders. Researchers who have

observed daily gang activities over an extended
period paint quite a different picture. Like the
aggressive responsiveness of most other people,
the aggression displayed by gang boys tends to be
highly discriminative, is restricted to a narrow
range of targets, and is largely limited to
attenuated forms. Miller, Geertz, and Cutter

(1961), who studied a representative example of a
juvenile gang, found that most of the gang’s
aggressive actions (70 percent) are directed at one
another, and only a small percentage of their
attacks are aimed at either adults (16 percent) or

other adolescents (12 percent). Not only does
most of the aggression occur within the gang, but
it is predominantly verbal. Of the total aggressive
episodes recorded, 93 percent involved verbal
attacks; physical assaults against persons or

property were relatively rare (7 percent).
Intragang aggression is primarily used to

maintain effective group functioning. Active
members are typically chastised for such things as
deviating from the group’s standards of conduct,
for unruly behavior that disrupts group activities,
for ineptitude, for shirking obligations, for boastful
self-aggrandizement, and for dealing dishonestly
with compatriots. Miller notes parenthetically that

these performance codes are not unlike the Boy
Scout guide, though the types of behavior by
which delinquents gain status and recognition
obviously differ. Based on his painstaking
observations, Miller concludes that aggressive

gang behavior is inadequately explained in terms
of either venting aggressive impulses or a lack of
norms and identity. Rather, it serves a positive
function to enforce adherence to the group’s
behavioral standards.
Any generalizations about gang aggression
must, of course, be accepted with reservation,

because gangs vary in a number of ways. Some of
the more important distinguishing features
include the size and stability of membership, the
degree of consensus on standards of conduct, and
the types of behaviors rewarded. Delinquent gangs
that glory in fighting exploits will bear little
resemblance to those valuing drug experiences.

Nevertheless, the findings of Miller and others
(Short, 1968) show the popular portrayal of
members of fighting gangs as combative predators
to be grossly exaggerated.
Getting smashed in the face or stabbed is

doubtless as frightening and painful to delinquent
boys as to persons leading more pacific styles of
life. Therefore, to induce others to perform actions
that can have self-injurious consequences requires

some coercive or rewarding power over them. In
the absence of strong reinforcement control,
individuals would be reluctant to affiliate closely

with a gang that confers status on the basis of
fighting prowess if they were, in fact, often
required to embark on perilous violence toward
rival groups. As might be expected, close
examination of intergang encounters reveals much
verbal bravado but little actual combat, even by
gangs whose status is built mainly around skill in

fighting (Miller, 1958; Short and Strodtbeck,
1964).
Core members have to engage in some fighting
from time to time when their status is publicly
challenged in order to preserve their fighting

reputation. However, it is the functional leaders
who do most of the fighting. The main reason for
this is that, as a rule, gang affiliation is relatively
weak and unstable (Klein and Crawford, 1967).

Many of the boys remain fringe members who
hang around more for protection from rival gangs
than for the rewards provided within the group.

With weak rewarding control, the leadership does
not have a sufficient hold on marginal members to
enlist them easily in battle. Nor are they willing,
for one reason or another, to risk beatings to
enhance their standing in the gang hierarchy.
On occasions when gangs are provoked into
physical violence by status threats, the combatants

are usually relieved when police interventions or
brandishing of guns provide them with legitimate
alibis for cutting short the fray. Both forces can
thus gracefully retreat with claims of victory that
uphold or even boost their respective reputations.

If intergang conflicts are skillfully orchestrated,
gang boys can maintain their prestige as warriors
on the basis of minimum actual combat. When one
plays the violence game, however, miscalculations

and unexpected circumstances sometimes result
in serious injury to the participants. On such
occasions, gang loyalties and preservation of

reputation as a group not to be trifled with
demand retaliatory violence despite high risk of
further injury.
Expressions of Injury
Aggressive behavior usually produces, among
other effects, expressions of pain and injury in
victims. Several views have been advanced on how

seeing the suffering caused by one’s aggressive
actions affects subsequent aggressiveness. Drive
theories (Feshbach, 1970) consider infliction of
injury the goal response that reduces the
aggressive drive. Just as eating relieves hunger,

hurting others presumably discharges the
aggressive impulse and hence becomes satisfying
in its own right. Sears, Maccoby, and Levin (1957)
suggest two ways in which signs of pain and injury

acquire rewarding value. In the first place,
aggressive actions can reduce uncomfortable

tensions resulting from conflict over desire to
strike out and fear of punishing consequences.
Second, aggressive behavior that removes

frustrations is often accompanied by signs of pain
and discomfort in those toward whom the attack
is directed. Pain cues thus become rewarding
through their repeated association with tension

relief and removal of frustrations. As a result,
people often behave aggressively for no other

apparent reason than to produce signs of pain and
distress in others.
At this point, it is important to distinguish

between acquired reinforcers and acquired drives.
If an aggressor achieves positive effects whenever
he hurts others, there is every reason to expect
that signs of suffering will develop reinforcing
properties. What remains unexplained in drive

theories is how people acquire a drive to inflict
injury. If drives are invoked for everything people
find satisfying, the result is an immense
proliferation of drives that have no greater
explanatory value than the conditioned
reinforcers from which they are inferred.
Moreover, as was shown earlier, frustration
(which is presumably the prime inducer of the
aggressive drive) is, at best, only a weak

determinant of aggression. Nor is there any
convincing evidence that people who refrain from
aggressive activities necessarily experience a
mounting urge to hurt others, as would be
expected from the characteristics usually assigned

to drives.
Feshbach (1970) interprets the rewarding
value of pain expressions in terms of self-esteem
processes. Through example and precept children

learn a retaliation norm: Infliction of injury
requires that the initial aggressor must be hurt. It
is further assumed that perception of pain in one’s

tormentors is experienced as satisfying because
retaliation restores self-esteem. Humiliating
defeat of a foe can, under certain conditions,
bolster aggressors’ self-evaluations. Undoubtedly,
there are also many instances in which reprisal is
designed more to reestablish one’s public image
than to enhance one’s self-regard. It is an open

question, however, whether assailants derive
much pride specifically from the agonizing cries
and gory wounds caused by their actions.
Retaliatory aggression can gain considerable value
from its success in reducing or terminating further

victimization. In clarifying the functional
properties of retaliation it would be important to
assess its reinforcing value depending upon its
successfulness and its injuriousness. These
separable effects should be distinguished because

they do not always occur together. That is,
vengeful injury can be administered in defeat, and

conversely, retributive triumphs can be achieved
with minimal infliction of injury. A comparative
analysis would reveal which combination of

outcomes makes further aggression more likely.
A contrasting view is that signs of suffering
ordinarily function as inhibitors rather than as
positive reinforcers of aggressive behavior.

Because of the potential dangers of violent and
destructive actions, most societies prohibit

physical forms of aggression except under
specified conditions. As the process of social
training takes effect, people learn to substitute the
more readily tolerated verbal means of aggression

for physical ones (Bandura and Walters, 1959;
Sears, Maccoby, and Levin, 1957). Even direct
verbal assaults meet with disapproval, so that

people eventually learn to express aggression in
subtler ways. Although defensive fighting may be

permissible under certain circumstances, cruel
and destructive acts of aggression are likely to
bring swift punishment. Most people also adopt

for self-evaluation societal standards in which
ruthless aggression is morally reprehensible.
Consequently, aggression that produces suffering
in others elicits fear of retaliation and self-
condemnation, both of which tend to inhibit
injurious attacks. Examples can be cited of societal

practices in which cruel, monstrous acts are
considered praiseworthy by those in positions of
power. Such bizarre reinforcement contingencies

can breed people who take pleasure in inflicting
pain and humiliation. Some of the most horrid
illustrations of this phenomenon are documented
in the proceedings of the Nuremberg trials.

Rudolph Hoess, a former commandant of
Auschwitz, for example, had a window installed in
a gas chamber so he could watch the gruesome

massacres (Andrus, 1969).
Lest the Nazi atrocities be dismissed as an
anomalous product of a deranged social system, it
should be noted here that otherwise socialized
people can be led to behave brutally and to take
pride in such actions under circumstances where
reinforcement practices are instituted that favor
sadistic forms of behavior. A United States infantry

battalion in South Viet Nam adopted its own
reward system for eliminating Viet Cong (San
Francisco Chronicle, 1970). Infantrymen who could
prove they had killed an enemy were not only
rewarded with rest and recreation, but were also

decorated with specially designed “Kill Cong”
emblems formally presented at company
formations. The barbarity of the reinforcement
system adopted by this particular battalion is
revealed by what had to be submitted as evidence

of a kill. To lay claim to a badge one had to bring
back an enemy ear. A string of ears was proudly

displayed at the headquarters like huntsmen’s
trophies. Within any society certain individuals, or
even subgroups, may function in a social

environment subscribing to pernicious
contingencies of reinforcement that make
infliction of suffering a source of satisfaction.
Some experimental attempts have been made
to assess whether signs of suffering by victims
function as reinforcers. Feshbach, Stiles, and Bitter

(1967) tested the notion that painful expressions
assume rewarding value under conditions of anger
arousal. Female students participated in a verbal

conditioning task after being treated in either a
hostile or a friendly manner by a confederate of
the experimenter. For half the subjects in each
condition the correct verbal response was
reinforced by showing the confederate being

shocked, while for the remaining students the
correct words merely produced a light flash.
Angered subjects increased verbal responses that
brought displays of the provocateur undergoing
pain, but the noninsulted subjects did not.
The above findings would seem to indicate that
pain cues serve as positive reinforcers for angered
people, but additional information is needed

before the conditioning differences can be
unequivocally attributed to the influence of anger

arousal. Seeing others suffer pain evokes
emotional reactions in observers (Berger, 1962).
Angered subjects in the light flash condition were
subjected to only one source of arousal, whereas

those who also observed the suffering provocateur
experienced two sources of emotion arousal. It has
been shown that people become conditioned more

easily under moderately high, as compared to low,
emotional arousal (Bandura and Rosenthal, 1966;

Spence, 1964). The obtained differences in
conditioning could conceivably reflect differences
in arousal level. Experiments examining the

influence of anger should include other forms of
arousal (for example, fear-aroused or sex-aroused
conditions) to establish whether the results are
due specifically to anger or to the general effects of
emotional arousal.
Researchers who have studied how signs of
injury influence further attacks on suffering

victims do not find much evidence that pain
expressions reinforce aggressive actions. Buss
(1966b) analyzed whether knowledge that one’s

actions were physically injurious to an initial
victim affects aggressors’ tendency to behave
punitively toward another victim. Aggressors
were equally punitive toward both victims when
they had no feedback on the results of their hurtful

behavior. Neither male nor female aggressors
reduced their punitiveness toward a male
knowing that their aggressiveness had previously
harmed a person. However, they behaved less
aggressively toward a female after learning about
the injurious consequences of their actions.
Apparently, the cultural stereotype that men
should be able to withstand painful treatment
mitigates compassion toward them.
In the preceding study aggression was
performed mainly in the service of social request.
Pain feedback may have different effects on
persons who are prompted to retaliatory
aggression by insults or physical attacks. As

previously noted, drive theories assume that
expressions of suffering are likely to function as
inhibitors of aggression in nonangered persons
but to reinforce punitiveness in those harboring

hostile feelings toward their adversaries. Several
studies provide evidence on this point. Geen
(1970) found that males who were treated

considerately, as well as those who had been
angered, behaved less aggressively toward their
antagonist when he expressed anguished cries
than when he suffered in silence. The aggression-
inhibiting effects of hurtful reactions are shown
even more convincingly by Baron (1971c), who
varied anger arousal and the intensity of pain

suffered by the victim. Both angered and
nonangered adults lessened the strength of their
attacks as the pain manifested by the victim
increased.
In a closely related experiment, Baron (197Id)

tested the notion that similarity between the
aggressor and the victim would arouse empathy
for the victim's suffering and thus reduce
aggression directed toward him. Students were

led to believe that the target person was either a
fellow student with similar attitudes and opinions,
or an apprentice electrician who advocated

opposing views on such matters as environmental
pollution and student power. The intensity of
anger toward the victim and the intensity of his
expression of pain were also varied. Signs of
suffering on the part of the victim diminished
interpersonal punitiveness regardless of whether
the victim resembled or differed from the

aggressor.
Although the scope of the experimental
treatments and the populations studied are too
limited to warrant sweeping generalizations, the
evidence indicates that expressions of suffering

serve as aggression inhibitors. These findings
assume added significance considering that the
pain reactions are conveyed only vocally or by
readings on a pain meter without any visual

contact with the victim. In actual life aggressors
not only hear the tormented cries of their victims,
but see the gruesome consequences of their

actions. People are even less inclined to behave
cruelly when they see their suffering victims than
when they merely hear the distress they have
caused them (Milgram, 1965b).
The conditions under which pain cues assume
rewarding value still remain to be demonstrated.
A gratuitous insult from a stranger in a laboratory

may not create sufficient animosity for the victim
to derive satisfaction from injurious retaliation. It

is a quite different matter when an antagonist
repeatedly tyrannizes others or wields his power
in ways that make life miserable for them. In such
instances, news of the misfortune, serious illness,

or death of an oppressor is joyfully received by
people who ordinarily respond more
compassionately to the adversities befalling

others.
Suffering of one’s enemy is most apt to be

rewarding when hurting him relieves discomfort
or benefits aggressors in other ways. On the other
hand, when aggressors suffer reprisals or self-
contempt for harming others, signs of suffering
function as negative reinforcers that deter
injurious attacks. In other words, the alleviation of
aversive treatment from an injured oppressor
rather than his suffering may be the primary

source of satisfaction. In studies showing that
injury to the victim reduces aggression, pain
expressions occur without the other extraneous
rewards accompanying victory over antagonists.
The best evidence that pain cues themselves

can become positively reinforcing comes from
societies in which sexual partners bite, scratch,
and lacerate each other during coital activities
(Ford and Beach, 1951). Expressions of pain thus

acquire rewarding value through repeated
association with sexual gratification. As a result,
erotic pleasure is derived from inflicting pain.

Within our own society it is not unusual to find
clinical cases who experience high sexual
excitement from inflicting bodily pain on
themselves or on others. From the standpoint of
social learning theory, the victim’s oppressiveness,
the social sanctions for punitive behavior, and the
consequences correlated with pain cues are more

pertinent than transitory anger arousal in
understanding the pleasurable value of injurious
actions.
Experiments have been conducted with
animals to determine whether opportunity to

aggress can act as a reinforcer for instrumental
behavior. Highly aggressive strains of animals will
perform responses that produce an attackable
target without much extraneous provocation

(Hogan, 1967; Myer and White, 1965). In less
aggressive species, however, opportunity to attack
does not serve as a reinforcing event unless

animals are subjected to aversive stimulation and
have been trained for aggressive behavior (Azrin,
Hutchinson, and McLaughlin, 1965; Cole and
Parker, 1971; Dreyer and Church, 1970). It is
under the combined influence of aversive arousal
and high aggressiveness that fighting is most apt
to be reinforcing (Legrand, 1970).
Although the preceding studies suggest that
under certain circumstances fighting can be
rewarding, this conclusion must be accepted with
reservation for several reasons. Some of the
experiments do not control for whether it is the

escape provided by the activity in general or
combat per se that serves as the reinforcer for
disconcerted animals. Cole and Parker (1971)
found that animals on whom unfavorable

schedules of reward were imposed pecked a key
to gain access to an attackable target but they also
did so to get to an empty goalbox. If it were shown

that aversively aroused animals do in fact work
harder to secure an attackable target, it would be
essential to include a condition in which they can
see but not assault the target to assess whether
combat or social contact is the source of reward.
Should the phenomenon survive the necessary
experimental controls, its significance can be best

assessed with paradigms that not only remove
marked situational constraints, but allow
alternative goal activities. Under natural
conditions animals rarely go out of their way to
secure victims; on the occasions when they do

encounter opponents, they generally try to avoid
fighting by resort to threat gestures. Given
freedom of movement and a choice of flight,
drinking, eating, courting, or fighting in response

to aversive stimulation, fighting may very well
turn out to be a weak reinforcer. When animals
are treated aversively in a confined enclosure

allowing only access to a restrained target, some
blows will likely be exchanged. Aggression
exhibited under these circumscribed conditions
may, however, lack generalizability. Moreover,
there are fundamental differences in the
complexity of reinforcement control of human and
animal behavior that make cross-species

generalizations hazardous. Animal aggression is
almost exclusively governed by direct physical
consequences. If a human inflicts suffering on
others, his actions not only produce immediate
physical effects, but can arouse self-devaluative

consequences and social reprobation, which serve
as formidable negative reinforcers. Animals are
not known to worry about what their cagemates
might think of them, or to berate themselves for
their cruelty. To the extent that inflicting suffering

boosts self-pride and public acclaim, the self-
rewarding effects of aggression will not be

revealed by animal experimentation.
Alleviation of Aversive Treatment
People are frequently subjected to distressing
treatment from which they seek relief. Forceful
action that is not unduly hazardous is the most
direct and quickest means of alleviating adverse

conditions, if only temporarily. Pain reduction has
been shown to be a powerful reinforcer of
behavior. Indeed, many of the activities in which
people engage are sustained by their functional
value in eliminating various irritants and

discomforts. Some aggressive behavior is
undoubtedly reinforced by its effectiveness in
removing the source of painful experiences.
Because of its aversive properties, aggression,

if skillfully executed, can both secure rewards and
remove painful stimuli. Evidence for this dual
reinforcement process is provided by Patterson

and his associates (Patterson, Cobb, and Ray,
1972; Patterson, Littman, and Bricker, 1967). In
these field studies children’s interactions are
observed for extended periods in natural settings.
The incidence of aggressive acts is recorded and
their immediate consequences are classified as
punishing, rewarding, or having no appreciable

effect.
Patterson reports that aggressors are most
likely to assault compliant peers. When attacked,
these children cry, give up desired objects, or
withdraw. If the victim does not immediately

acquiesce, the aggressor intensifies his attack until
the victim submits. Approximately 80 percent of
the aggressors’ assaultive actions produce
rewarding consequences for them. The victims’

submissiveness thus both invites and reinforces
the aggressors’ actions. This surprisingly high
level of positive reinforcement of physical

aggression by nursery school children remained
stable over the nine-month period sampled by the
observations.
Older children and adults are not always
accommodating to the desires of aggressors.
Assaults often elicit counterattacks. Observational

studies of parent-child relationships (Patterson,
Cobb, and Ray, 1972; Patterson and Reid, 1970), in

fact, disclose a pattern of mutual coercion rather
than quick acquiescence. These findings have led
Patterson to conceptualize aggression as a form of
coercive behavior that is primarily reinforced by

reduction of aversive stimulation. In reciprocal
fighting, for example, the victim attempts to
terminate painful blows by defensive attacks of his

own; the recipient, in turn, strikes back even more
forcefully to escape the pain caused by his

opponent's blows. Efforts at coercive control
escalate in intensity until one of the combatants
yields or is injured.
After people have begun to exchange blows,
aggressive actions may be reinforced to some
extent by their capacity to cushion or to reduce
attacks. Immediate pain reduction, however, does
not explain what initiates an aggressive course of

behavior, or why combatants continue to suffer
increasingly more painful beatings when any
number of nonaggressive responses provide more
effective means of escape. Explanation of
aggression in terms of instant pain reduction is

somewhat reminiscent of the view that people
inflict injury on themselves because it feels so
good when they stop.
In their reinforcement analysis of aggression

Patterson, Cobb, and Ray (1972) discount the
value of laboratory studies demonstrating that
aggressive responses are increased by social and

material rewards on the grounds that, in their field
studies, aggression was seldom socially reinforced
by peers or by adults. If this is the case, then the
generality of the observational studies rather than
the value of controlled experiments should be
questioned. It is probably true that three- and
four-year-olds rarely confer status rewards or

commend their playmates for grabbing a prized
toy and shoving other youngsters. Nor are
onlooking wives likely to shower their husbands
with praise for beating their children into
submission. Other field studies, sampling a

broader range of situations, show that parents
often do praise their children for aggressing
toward others outside the home; adolescents and
adults win admiration and respect by
demonstrated toughness; and in certain

subcultures members achieve status through
aggressive exploits.
In the social learning analysis, defensive
aggression is sustained to a greater extent by
anticipated consequences than by its
instantaneous effects. People will endure painful
treatment on expectation that their aggressive
efforts will eventually succeed. In many instances,
of course, performances are based on faulty

assessment of probable consequences. Aggressive
actions may also be partly sustained in the face of
hurtful counterattack by anticipated costs of
timidity. In aggression-oriented circles, failure to
fight back can arouse fear of future victimization

and humiliation. Moreover, cowardice, unless
convincingly self-justified, can arouse anticipatory
loss of self-esteem, especially when attacks are
unjust and degrading. A physical pummeling may
therefore be far less distressing than social

derision or self-contempt. In other words, humans
do not behave like unthinking servomechanisms

directed solely by immediate response feedback.
Rather, their cognitive capacity to represent future
consequences enables them to regulate their

actions by outcomes extended forward in time.
Under aversive conditions of life, people will
persist, at least for a time, in aggressive behavior
that produces immediate pain but prospective

relief from misery.
The manner in which aggressive behavior is

reinforced by its success in terminating aversive
treatment from aggressors is shown by Patterson,
Littman, and Bricker (1967). Passive children who

were repeatedly victimized and whose
counteraggression often proved effective in
halting attacks not only increased defensive
fighting, but eventually began to initiate attacks of
their own. On the other hand, passive children

who were seldom maltreated because they
avoided others and those whose
counteraggressive responses to victimization were
unsuccessful remained submissive in their
behavior.
VICARIOUS REINFORCEMENT
People repeatedly observe the actions of
others and the occasions on which they are
rewarded, ignored, or punished. Observed

outcomes influence behavior in much the same
way as directly experienced consequences

(Bandura, 1971b; Kanfer, 1965). In general,
observed rewards increase, and observed
punishment decreases, the tendency to behave in
similar or related ways. Vicarious reinforcement

not only influences similar performances in
observers, but can markedly alter how they
respond to the actual consequences accompanying
their actions. Seeing other people’s outcomes
determines the nature and effectiveness of direct

reinforcement by providing a standard for judging
whether the treatment one customarily receives is
equitable, beneficent, or unfair. The same

consequence may therefore be experienced as a
reward or as a punishment depending upon
whether comparable actions by others are treated
more or less favorably.
In experiments studying the influence of
vicarious reinforcement on aggression, individuals
observe aggressive models whose actions are
rewarded, punished, or unaccompanied by any
evident consequences (Bandura, 1965a; Bandura,

Ross, and Ross, 1963c). The amount of imitative
and nonimitative aggression that observers later
spontaneously perform in a test situation is
recorded. Witnessing aggression punished usually

produces less imitative aggression than seeing it
obtain social and material success or go unnoticed.
However, performance of aggressive responses

that differ from those displayed by the model is
not greatly affected by observed negative
consequences. Like the effect of directly
experienced punishment, the inhibitory impact of
vicarious punishment is thus largely confined to
the modeled activities prompting negative
sanctions. As we shall see later, the legal system of

deterrence rests heavily on the restraining
function of exemplary punishment (Packer, 1968).
Observed rewards generally produce a greater
increase in similar responding than if the
exemplified actions have no evident effects. It is

noteworthy that in the case of reprehensible
behavior—such as unprovoked physical assaults
or violation of prohibitions—seeing
transgressions go without consequences heightens

analogous actions in observers to the same degree
as witnessing the modeled behavior rewarded
(Bandura, 1965a; Thelen and Soltz, 1969; Walters,

Parke, and Cane, 1965). To the extent that absence
of anticipated punishment conveys
permissiveness and allays fears, behavioral
restraints are thereby reduced and aggressive
actions are performed more readily.
Because of the variety and complexity of social
influences, people are not always consistent in

how they respond to aggressive behavior.
Rosekrans and Hartup (1967) examined the
effects of discrepant observed consequences on
imitative aggression. Children who saw assaultive
behavior consistently rewarded were most

aggressive; those who saw it consistently
punished displayed virtually no imitative
behavior; while those who saw aggression
sometimes rewarded and sometimes punished

exhibited a moderate level of aggressiveness. It is
the latter mixed pattern of reinforcement that
predominates in everyday life.
In the preceding studies the aggressive model
was punished either verbally or physically by
someone else. There are times when people
respond with self-approving or self-devaluative
reactions to their own behavior. Numerous
experiments, discussed at length elsewhere

(Bandura, 1971b), demonstrate that witnessing
punishment self-administered by models has
inhibitory effects on observers, whereas seeing
others respond self-approvingly to their

unmerited achievements and transgressions
facilitates similar behavior in observers. Porro
(1968), for example, reports that, of children who
saw a model respond with self-praise to her
transgressions, 80 percent subsequently violated a

similar prohibition. By contrast, only 20 percent of
the children deviated when they had observed the

same model react self-critically to her
transgressive behavior. Although relevant data are

lacking, there is every reason to expect that
aggressors’ self-evaluative reactions toward their

own conduct may have similar effects on
aggressive dispositions of observers, especially if
they hold the models in high esteem.
Most people, especially those residing in
middle-class neighborhoods, rarely observe
physical violence in their daily experiences. The

vast majority have never directly witnessed any
murders or muggings, and few encounter physical
fights or violent confrontations between

antagonistic groups. Rather, they learn about
violence principally from televised portrayals and,
to a lesser extent, from newspaper accounts.
These symbolic sources provide abundant
opportunities to observe aggressive conduct and

its likely consequences.
A number of content analyses have been
conducted to determine how fictional violence is
portrayed on television. Larsen, Gray, and Fortis
(1968) found that socially approved goals were
achieved by illegal or violent methods 47 percent
of the time in programs designed for adult
audiences and 56 percent of the time in children’s

programs. Halloran and Croll (1972) similarly
report that violent means are presented as both
acceptable and successful in the short term,
though villainous instigators fare less well in final

outcomes. Witnesses to violence are more likely to
approve of such behavior or to join in the fray
rather than to seek alternative solutions. Violence
is not only shown to pay off, but is repeatedly used
by good characters who often dispose of

adversaries in a quick perfunctory way, as though
killing a human being were a routine act devoid of

any moral implications or tragic consequences.
Aggression is most readily legitimized as

acceptable conduct by having good triumph over
evil through aggressive means. When violent

virtue kills, which is often, the killers seldom
suffer negative consequences for their acts. On the
contrary, good aggressors emerge victorious 66
percent of the time when they initiate assaults and

84 percent of the time when they reciprocate
them. The physical effects of beating others are

underplayed as less bloody or messy than in real
life. Aggressive life styles are thus portrayed as
exceedingly prevalent, highly functional, and
relatively clean.
There has been surprisingly little analysis of
how the functions and consequences of actual
aggression are reported in news and documentary
programs. The handling of criminal violence poses

few dilemmas because it rarely arouses conflicts in
how such behavior should be judged. On the other
hand, it is difficult to remain impartial in the
coverage of social disturbances arising from
power conflicts within the community or society
at large. Depending on what is selected for
presentation and the manner in which the chosen
events are evaluated, the media can convey
support for coercive force either in the service of

social control or social change. These issues, and
other aspects of the legitimation of forcible means,
are discussed at length in the concluding chapter
of this book.
Explanation of Vicarious Reinforcement
Social learning theory posits several different

mechanisms by which witnessed rewards and
punishments alter the actions, feelings, and

thoughts of others (Bandura, 1971c). A vicarious
reinforcement event may vary in a number of
aspects, including the characteristics of the
aggressors and those who respond to them, the

type and intensity of consequences, their
justification, the situations in which the outcomes
occur, and the counteractions of aggressors. The
mechanisms that are operative in any given
instance will therefore depend upon the nature
and combination of these various factors.
One explanation of vicarious reinforcement is
in terms of the informative function of observed
outcomes. Response consequences accruing to
others convey information to observers about the
types of actions that are likely to be approved or

disapproved. Given knowledge about probable
response consequences, people will generally do
the things they have seen well received and avoid
those that they have seen punished.
Assessment of probable consequences from

observed experiences is affected by comparability
of sanctions customarily applied to models and to
observers. In any social group aggressive behavior
may be permitted for some members but
prohibited for others depending upon, among
other things, the status and legitimized roles they
occupy. When models and observers differ in
characteristics that foster different reactions to

their behavior, observed reinforcement may not
produce the usual effects. Knowledge gained from
witnessing the outcomes experienced by models
would be particularly influential in instances

where the observer believes that the model’s
contingencies apply to himself as well. It is highly
unlikely, for example, that seeing physical
aggression approved in a person occupying a
socially sanctioned role, such as a policeman,

would prompt much imitative aggressiveness in
citizens, because actions that are commended in

law enforcers may be prohibited for others. For
similar reasons, people are reluctant to adopt the

behavior of models who occupy divergent
occupational or social positions.
The same behavior can produce markedly
different consequences depending on where,
when, and toward whom it is performed. What is
permissible on a street corner may be censurable

in a church. Hence, adaptive functioning requires
not only response information, but also knowledge
about what actions are appropriate in what
setting. When individuals are rewarded for
aggressing in a given situation but ignored or

punished for exhibiting the same behavior in a
different setting, observers learn to differentiate
environments in which it is relatively safe to
behave aggressively from those in which such
actions are hazardous. In this way, seeing the

outcomes experienced by others facilitates
aggression in settings correlated with observed

rewards, and decreases aggression in situations in
which the likelihood of punishment is high.
Aggressive actions are accordingly regulated and

modified on the basis of the successes and
mistakes of others in designated settings.
Observed reinforcement is not only

informative, but can also have incentive
motivational effects. Seeing others positively
reinforced can serve as a motivator by arousing in

observers expectations that they will be similarly
rewarded for analogous performances. Given
appropriate reinforcement systems, people can be

as much inspired by observed social acclaim and
tangible benefits to aggress as to excel in athletic,
artistic, or academic pursuits. The more highly
observers value the rewards accruing to
successful aggressors, the more likely they are to

be prompted to aggressive actions when
opportunities arise. The frequency with which
others achieve desired outcomes can affect how
long observers will persist in the face of
discouraging results. Observational incentives
play an especially important role in social
activism, for here the chances of quick success are
poor, but protest behavior is partly sustained by
the long-range attainments of groups that have

persevered in their efforts. Historical as well as
contemporary successes in social change can
provide motivational support for reform activities.
Models generally exhibit emotional reactions
while undergoing rewarding or punishing

experiences. Observers are easily aroused by the
emotional expressions of others. Vicariously
elicited emotions can eventually become
conditioned either to the modeled behavior itself

or to situations that are regularly associated with
the performers’ distress. As a result of vicarious
emotional learning, situations in which aggressors

were punished, or similar aggressive
performances, are likely to frighten and to inhibit
observers. Fears and inhibitions can also be
extinguished by having fearful observers watch
others engaging in threatening activities without
experiencing any adverse consequences (Bandura,
1971a). Vicarious conditioning and extinction of

fear may therefore partially account for increases
and decreases in aggressiveness that result from
observing affective consequences accruing to
aggressive models.
People see not only the consequences of

another's behavior, but also how he responds to
his treatment. There is some evidence, presented
by Ditrichs, Simon, and Greene (1967), that
modeled responsiveness to social influence affects

how observers later react when they themselves
are rewarded for displaying similar behavior.
Children who observed models express

progressively more hostility for social approval
later increased their own output of hostile
responses under positive reinforcement. However,
when models appeared oppositional by reducing
hostile responses that brought them praise, or
reacted in random fashion as though they were
uninfluenced, observers did not increase their

expression of hostility even though they were
praised whenever they did so. Thus, susceptibility
to direct reinforcement influences can be increased
by observed responsiveness and reduced by
observed resistance.
In addition to the foregoing effects of vicarious

reinforcement, social status can be conferred on
aggressors by the manner in which their behavior
is reinforced. Punishment tends to devalue the

model and his behavior, whereas the same model
assumes emulative qualities when his actions are
praised and otherwise rewarded (Bandura, Ross,

and Ross, 1963c; Hastorf, 1965). Modification of
model status, in turn, influences the degree to
which observers pattern their own actions after
behavior exemplified by different models. There
are conditions, of course, in which observed
punitive treatment enhances rather than lowers
the recipient’s social status. Members of protest

groups who risk punishment for upholding their
basic rights and beliefs gain the admiration not
only of their peers but of others as well, especially
when attacks are directed at social practices that
violate the professed values of the society. It is for

this reason that authoritative agencies are usually
careful not to discipline challengers in ways that
might martyr them.
Observed reinforcements can alter the

valuation of reinforcing agents as well as
recipients. When societal agents misuse their
power to reward and punish, they undermine the

legitimacy of their authority and generate strong
resentment. Under these conditions, seeing
inequitable punishment, rather than prompting
compliance, may free incensed observers from
self-censure of their own actions and thus increase
aggressive behavior. This is most likely to occur
when retaliative aggression is self-justified as

rectifying past grievances or preventing further
maltreatment. Otherwise considerate people can
thus be readily provoked by observed injustice to
behave cruelly without remorse.
SELF-REINFORCEMENT
The discussion thus far has illustrated how
people regulate their aggressive behavior on the

basis of response consequences that they either
observe or experience at first hand. A theory that

attributed performance of aggression solely to the
influence of external rewards and punishments
would be incomplete because humans can, and do,
regulate their own actions to some extent by self-
produced consequences. People do things that
give them self-satisfaction and a feeling of self-
worth; conversely, they refrain from behaving in
ways that result in self-criticism and other self-
devaluative consequences.
Self-reinforcement functions are acquired in
several different ways (Bandura, 1971c). People
learn to respond evaluatively to their own
behavior partly on the basis of how others have
reacted to it. Parents and other socialization

agents subscribe to certain norms of what
constitute worthy or reprehensible performances.
They are inclined to respond approvingly when a
child meets moral standards, whereas they are

quick to reprimand him whenever he deviates
from the way he is expected to behave. As a result
of such differential treatment, children eventually

come to respond to their own actions with self-
approval and self-criticism in accordance with the
standards originally set by others.
Individuals not only prescribe self-evaluative
standards for others, they also exemplify them in
response to their own behavior. Modeling has

been shown to be another influential means of
transmitting systems of self-reinforcement

(Bandura, 1971b). People tend to adopt standards
of self-reinforcement displayed by exemplary
models, they evaluate their behavior relative to
those standards, and they serve as their own

reinforcing agents. Because of self-reactive
tendencies, aggressors must contend with
themselves as well as with others when they

behave in injurious ways.
Self-reward for Aggression
One can distinguish several ways in which self-
imposed sanctions enter into the self-regulation of

aggressive behavior. First, there are individuals
who have adopted a self-reinforcement system in
which aggressive actions are a source of personal
pride. Such people readily engage in aggressive
activities because they derive self-satisfaction

from aggressing successfully. Lacking self-
reprimands for hurtful conduct, they are deterred
from cruel actions mainly by fear of
counteraggression and its other punishing costs.
In the comparative study discussed earlier

(Bandura and Walters, 1959), assaultive
delinquents often showed little or no self-censure
for their injurious behavior, even when it took
severe forms.
I: Have there ever been times when you felt you’ve

got to do something just for the hell of it?
B: Yeah.
I: What kind of things?
B: Well, I feel like going out and getting drunk, and

going and beating the shit out of some guy.
I: How often have you done this?
B: About five or six times.
I: How do you feel about it afterwards?
B: I feel good. I always pick a person bigger than me,

so I can’t think I was picking up some guy
smaller. I don’t care if he’s that much bigger.
He’s always bigger than me. (Pp. 288-89)
In the above case, it is the size of the unwary
victim, not the suffering inflicted on him, that
serves as the major deterrent and determinant of

self-respecting reactions to physical conquest. The
absence of self-reproof for cruel treatment of
others in these boys stemmed not so much from
failure to internalize parental standards of

conduct, but rather from the adoption of their
parents’ positive valuations of aggressiveness. The

following interview excerpt illustrates this
process.
I: Are there things about yourself that you’re proud

of, and wouldn’t want to change?
B: Motorcycle riding.
I: Is there anything else?
B: Say, something like you’re proud of? You probably

won’t understand, but stomping. I’m proud of
it because, I don’t know, all the guys I hang
around with do that. Do you know what
stomping is?
I: No, I don’t.
B: Fighting with two feet without using your hands,

see. I’m not trying to be conceited or
anything, but I know I can use my feet better
than all the guys I hang around with, so I
wouldn’t want to change that. Like my Dad,
he said, “If you know how to fight with your
feet, then it’s in your hands, you’ve got it
made,” or something like that. “You never
need be afraid of anybody.” (Pp. 121-22)
Studies of subgroups in which aggression is

regarded as emulative behavior similarly report a
high incidence of self-reward for inhumane
actions. In the Nazi regime, under which horrific
massacres were politically sanctioned,
commandants of extermination camps felt
tremendous pride when they surpassed the kill
rates of other crematoria (Andrus, 1969). Indeed,
Eichmann confessed that death lists were his

favorite reading before he went to sleep
(Wechsberg, 1967). Gang delinquents have been
known to compile scrapbooks of newspaper

articles about their aggressive exploits, which they
regard with pride (Short, 1968). Toch (1969)
vividly documents how violent men often derive

self-satisfaction and enhanced feelings of worth
from physical conquest. The chronic assaulter,
quoted below, describes the tremendous boost in

self-esteem he experienced from administering a
brutal beating to a fellow prisoner over a trivial

incident during a card game.
And he turned around again, and he said, you

know, “I told you not to stand behind me.”
And he said, you know, “Fuck you man.” And
the dude got up and we were both on him,
man. And we beat him to a pulp. Fixed him
up bad, man. ... So we just, once we got going
we just wasted the dude. And that was that.
Sent him down to the hospital. And after that
I felt like a king, man. I felt like you know,
“I’m the man. You’re not going to mess with
me.” ... I felt like everybody looking up to me.
(Pp. 91-92)
The above illustrations depict idiosyncratic
self-systems of morality in individuals and deviant
subgroups. Inhumane self-reinforcement codes

are sometimes fostered on a society-wide basis. In
aggressive cultures where prestige is closely tied

to fighting prowess, members take pride in cruel
treatment of fellow human beings.
Self punishment for Aggression
In the course of socialization most individuals
adopt, through example and precept, negative

sanctions against cruel actions. As a result, they
are restrained from injurious aggression by

anticipated self-criticism. There is no more
devastating punishment than self-contempt.
Results of the study by Bandura and Walters
(1959) reveal how anticipatory self-reproach for
repudiated aggression serves as a motivating

influence to keep behavior in line with adopted
standards. Adolescents who were compassionate

in their dealings with others responded with self-
disapproval, remorse, and attempts at reparation
even when their aggressive actions were minor in
nature. Assaultive boys, on the other hand,

experienced relatively few negative self-reactions
over rather serious aggressive activities. Hicks
(1971) provides a laboratory demonstration of a

similar self-regulatory process in young children.
The more reprehensible girls judged aggressive

actions to be, the less likely they were to adopt
them when they were later exemplified by a peer
model.
Neutralization of Self-condemnation for Aggression
Rarely is aggression uniformly self-rewarded
or self-punished without regard to the victim or
the circumstances under which it is performed.

Humane, moral people sometimes behave cruelly
without undue upset; and callous, brutal assassins

may experience guilt over hurting pets and
persons of whom they are fond. Like other
reactions, self-reinforcing responses come under
discriminative control. This is especially evident

when social pressures impel people to act counter
to standards of conduct that would ordinarily
make them feel ashamed and self-contemptuous.

By engaging in self-deceptive machinations, they
can behave inhumanely without self-

condemnation. The different forms that self-
absolving practices take are discussed in the
sections that follow.
Slighting aggression by advantageous comparison.
A practice that is widely employed is to slight
one’s aggressive actions by comparison with more
hideous deeds. American executors of Asian

warfare and their ardent supporters, for example,
minimized the slaying of countless Indochinese as

checking massive Communist enslavement. Given
such a benevolent definition of destructive
practices, aggressors remained unperturbed by
the fact that the intended beneficiaries were

exterminated at an alarming rate. Militant
protesters, on the other hand, characterized their
domestic violence as trifling, or even laudable by

comparing it with the widespread carnage
perpetrated by military forces in foreign lands.

Justification of aggression by comparison to
outrageous acts is one of the more effective
methods for making violence acceptable because it

redefines the behavior in positive terms rather
than relying on disguises of purpose or personal
responsibility.
Since warfare requires collective contribution,

however remote it may appear from military
operations, almost anyone can be singled out as a
party to the national crime. Unfortunately, the

targets, whether they be national, institutional, or
individual, are often selected for attack more on
the basis of vulnerability and minimum risk than

on degree of culpability. In such conflicts,
reciprocal aggression usually escalates, with each
side lauding its own violence but condemning the
violence performed by its adversaries.
Justification of aggression in terms of higher

principles.
A closely related form of self-vindication is to
construe one’s aggression in terms of higher

values. Given sufficiently noble aims, almost any
form of aggression can be justified as righteous. To
take a historical example, many massacres were
devotedly perpetrated by crusading Christians in
the service of religious principles. Groups that
resisted the control of the clergy and the

ambitious nobles who surrounded them were
stigmatized as witches and heretics by the Pope,

who encouraged all true believers to assist in their
extermination (Mackay, 1932). Similarly, in
contemporary times violence is frequently
espoused in the name of freedom, social justice,

equality, and civil order.
That high moral principles do not alone curb
injurious behavior is shown more specifically in
studies relating level of moral reasoning to
aggressive conduct. Individuals exhibiting the

highest moral viewpoints are inclined to resist
arbitrary demands to behave punitively, but they
will aggress against people who violate their
personal principles (Keniston, 1970). Throughout
history many have suffered at the hands of self-
righteous crusaders bent on stamping out what
they considered evil. By the same token, social

reformers are able to use coercive, and even
violent, tactics to change inhumane practices by
reducing self-deterrent reactions through moral
conviction in their cause.
The self-devaluative consequences of cruel

deeds are sometimes eased by posing social
conflict as involving issues of property values
against human values. Little self-appeasement is
required to destroy property belonging to

adversaries if the destructive actions are justified
as aimed at stopping maltreatment of humans. If
this form of redefinition is to remain effective, the

aggressor must disregard the threat to human
rights created by the destruction of a man’s
belongings.
Displacement of responsibility.
People can be led to behave in injurious ways

provided that a legitimate authority is willing to
assume responsibility for their actions.
Participants in Milgram’s (1963) experiments,
who were deterred from punitive assaults by

distress over the suffering they had inflicted,
continued to escalate the shocks to hazardous

levels despite their victims’ agonizing cries after
the experimenter had assured them he would be
fully accountable for the consequences of their

behavior. Responsibility for cruel deeds is not
always assumed so explicitly because no one
wants to be answerable for such acts. To minimize
risks to themselves, superiors more often invite
and condone reprehensible conduct by their

subordinates in insidious ways that allow them to
claim ignorance for what was happening in the

event that disclosures arouse public
condemnation. The intended purpose of
sanctioned destructiveness is also disguised so
that neither issuers nor perpetrators regard their

actions as censurable. When outrageous practices
are publicized, they are officially dismissed as only
isolated events.
Exemption from self-devaluation for heinous
deeds is most gruesomely revealed in socially

sanctioned mass executions. Nazi prison
commandants and their staffs felt little personal
responsibility for their unprecedented
inhumanities. They were only carrying out lawful

orders. Guiltless obedience to horrific orders is
most evident in military atrocities. When
questioned about the massacre of the entire

population of the South Vietnamese hamlet of My
Lai, the American lieutenant in charge replied,
without any remorse, “That was my order, sir—
that was the order of the day, sir. They were all
enemy, they were all to be destroyed” (San
Francisco Chronicle, 1971b). In an effort to deter
institutionally sanctioned atrocities, the

Nuremberg Accords were established declaring
that obedience to inhumane orders even from the
highest authorities does not relieve subordinates
of the responsibility for their actions. However,
since victors are disinclined to try themselves as

criminals, such decrees have limited deterrent
value without an international judiciary system
empowered to impose penalties on victors and
losers alike.
Diffusion of responsibility.
Exemption from self-criticism can be achieved
to some extent by obscuring and diffusing
responsibility for aggressive practices. Collective
aggression requires numerous task functions that
must be supported by an organizational
apparatus. Departmentalization of destructive
activities works in several ways to reduce
participants’ sense of personal responsibility for

their complicitous behavior. Development,
production, and use of instruments of destruction
require the services of a considerable number of
people, each performing a fragmentary job that in

itself may appear harmless. The fractional
contribution can be easily isolated from the end
product, especially when one exercises little
personal judgment in carrying out a sub function
that is related to the terminal aggression through

remote, complex links. When the violence inflicted
on victims is horrifying, guiltless functionaries in

the destructive enterprise are quick to condemn
and to take vengeance on the individuals assigned

the gruesome task of performing the aggressive
actions.
Another bureaucratic practice for relieving
self-condemnation for aggression is to rely on
group decision-making, so that no single
individual feels responsible for what is eventually

done. Indeed, social organizations go to great
lengths to devise sophisticated mechanisms for
obscuring responsibility for decisions that affect
others adversely. Nor are people likely to
experience much self-punishment for injurious

outcomes resulting from direct participation in
collective violence. The destruction of life and
property can always be attributed to the other
members of the group. Through division of labor,
division of decision-making, and collective action,

people can be contributors to cruel practices and
bloodshed without feeling personally responsible

or self-contemptuous for their part in it.
Dehumanization of victims.
A further means of protection against self-
devaluation is to dehumanize the victim. People
selected as targets are often divested of human
qualities by being viewed not as individuals with

sensitivities, feelings, and hopes, but as
stereotyped objects bearing demeaning labels
such as “gooks” or “niggers.” If dispossessing
victims of humanness does not fully eliminate self-

reproof, it can be further reduced by attributing
subhuman or degrading characteristics to them.
Foes become "degenerates,” “pigs,” and other
bestial creatures. After victims have been so
devalued, they can be cruelly attacked without

much risk of self-punishment.
Bernard, Ottenberg, and Redl (1965)

enumerate many conditions of contemporary life
that are conducive to dehumanizing behavior.

Bureaucratization, technology, automation,
urbanization, and high social mobility all lead

people to relate to each other in anonymous,
impersonal ways. Under conditions of
overcrowding, people inevitably become aversive,
obstructive objects. It is difficult to remain

sensitive to human qualities in individuals who
are shoving you, blocking your progress, grabbing
things you want, and in other ways reducing your
comfort and freedom of movement. In addition,
social practices that divide people into in-group

and out-group members produce human
estrangement conducive to dehumanization.
Strangers can be more easily cast as subhuman
villains than can personal acquaintances.
Attribution of blame to victims.
Attribution of blame to victims is still another
expedient that can be used for self-assuaging
purposes. In this process, aggressors see
themselves as essentially persons of goodwill who
are forced into punitive actions by villainous
opponents. Victims are condemned for bringing
the suffering on themselves either by their
character defects or by their witless and

provocative behavior. Violent encounters involve a
series of reciprocally escalative assaults in which
victims are rarely faultless. Self-vindication can
therefore be achieved by selecting from the chain

of causes an instance of defensive aggression by
opponents and interpreting it as the original
incitement.
Observers of victimization can be affected in
much the same way as the aggressors. Seeing
victims suffer punitive treatment for which they
are held partially responsible leads observers to

devalue them (Lerner, 1971; Piliavin, Hardyck,
and Vadum, 1967). The indignation aroused by

ascribed culpability, in turn, provides moral
support for even more brutal acts by aggressors.

When bad practices are well entrenched, efforts
on the part of concerned individuals to halt them
by publicizing their destructive effects are more
likely to arouse derogation than sympathy for the

victims. To acknowledge the inhumanities arouses
self- critical reactions if one does nothing about
the situation. It is easier to reduce the discomfort
by designating the victim as a bad person than to
challenge bad practices that are an accepted part

of the social order.
Justified maltreatment can have more
devastating human consequences than
acknowledged cruelty. When blame is

convincingly ascribed, victims may eventually
come to believe the degrading characterizations
they hear about themselves (Hallie, 1971).

Negative attribution by itself may not be too
persuasive, but it is usually accompanied by
maltreatment that produces self-confirming
evidence of the victim’s defects or badness.
Vindicated inhumanity is thus more likely to instill
self-contempt in victims than if it does not justify
itself.
Graduated desensitization.
The practices cited above will not
instantaneously transform a compassionate

individual into a brutal aggressor. Rather, the
change is usually achieved through a gradual

desensitization process in which the participants
may not fully recognize the marked changes they
have undergone. Initially, individuals are

prompted to perform aggressive acts they can
tolerate without excessive self-censure. After their
discomfort and self-reproof are extinguished
through repeated performance, the level of
aggression is progressively increased until
eventually gruesome deeds, originally regarded as
abhorrent, can be performed without much
distress. Andrus (1969) describes this extinction
process succinctly in his study of mass
executioners: “Tentatively at first, then callously,

and finally quite easily, they performed these
terrible massacres” (p. 204).
The repetitive drills of military training, as in

bayonet practice and shooting simulated targets,
are intended not only to routinize fighting skills,
but to extinguish deterring self-reactions to such

behavior. Previously reported studies showed
how traumatized soldiers could be desensitized
within a brief period by reexposing them to
progressively more terrifying battle scenes. It is
perhaps increasing desensitization to injury more

than frustration that accounts for the rise over
time in military atrocities and the growing public

indifference to them.
There exist a number of factors—many of them
social and symbolic —that serve as disinhibitory
devices over and above those already cited.
Targets are usually presented as distilled military

abstractions. In the hygienic verbal labels used to
describe military slaughter, people are “wasted”
rather than put to death. The inhumanity of the
enemy, which makes it easier to kill him, is

bolstered by accounts of atrocities and orientation
training on how to deal with torture techniques as
a prisoner. Even considerate soldiers can
eventually be driven to brutal acts to avenge the
deaths of fighting companions. And the prayers of

clergy convey the symbolic impression that the
vested agents of institutional morality are backing

the combat enterprise. Finally, medicolegal
procedures are available for controlling

information sources and removing “undesirables”
who might act as contravening influences.
The research of Davitz (1952) and Wright
(1942) has shown that the presence of
cooperative and friendly bonds between people
serves to inhibit aggression toward each other in

the face of provocation from external sources. An
individual cannot easily mistreat a friend without
holding himself in low regard, given a sense of
implicit trust and heightened awareness of the
personal suffering caused by such actions.

However, just as high moral principles can be used
to support ruthless behavior, positive social bonds
can heighten aggressiveness under certain
conditions. Wright (1942) found that whereas
thwarting increased cooperativeness and reduced

discord in close friends, it had little effect on the
social behavior of those with weaker bonds of

friendship. But the mutual support of close
fellowship also counteracted restraints against
aggressing toward the person who thwarted them.

Close friends were more abusive toward their
antagonist than the less close friends, the
difference being particularly marked in the
expression of physical aggression. The latter group

often resorted to physical assaults, whereas the
aggression displayed by those with a weaker

friendship never went beyond hostile remarks.
Thus, while close acquaintance with potential
targets may serve as an inhibitor of aggression, the
mutual social reinforcement inherent in

comradeship can be effectively used, as in times of
national warfare or intergroup strife, to
strengthen aggression against a common foe.
Minimization and selective forgetting of

consequences.
The foregoing discussion reviewed the various
self-absolving devices that permit otherwise

socialized people to behave violently toward
others without suffering self-condemning
consequences. Once they have aggressed,
additional self-placating measures are available
that operate principally through distortion of the
effects produced by one’s actions. When people

are prompted to self-disapproved conduct under
conditions in which they have some choice on
whether or not to behave that way, they tend to
minimize the painfulness of its consequences; on

the other hand, when they are forced to act
punitively contrary to their values, they magnify
the severity of the effects (Brock and Buss, 1962).
Moreover, after being compelled to perform
aggression of which they disapprove, individuals

recall prior information given them concerning
potential benefits of the punishment, but they are

less able to recall its harmful effects (Brock and
Buss, 1964). Both minimization and selective

forgetting of consequences would be expected to
reduce distress over conduct that is incompatible

with one’s self-evaluative system.
Examples taken either from military atrocities
or political violence convey the impression that
sanctioning of human cruelty occurs only under

extraordinary circumstances. Quite the contrary. It
is a common practice supported by financial profit

as well as by governmental coercion or political
disaffection. Exploitation of human brutality by
the television industry illustrates the
authorization of aggression for commercial

purposes. Testimony of media practitioners at
congressional hearings and in private interviews
document how familiar self-absolving devices are

used in conjunction with organizational sanctions
to produce a heavy volume of violent material.
In the earlier days of television, producers and
writers were issued explicit instructions by
network officials to inject violence into their
programs in ways that contributed neither to
characterization nor to plot. In one such exchange
a producer with humanitarian concerns was
reminded that his program was scheduled against
“Wanted Dead or Alive,” a series in which the

“hotrod cowboy kills a man a minute” (U.S. Senate,
1963, p. 1964). Scenes of violence, he was told,
were therefore required for competitive purposes.
Scenes were rearranged to begin with violent
teasers repeating the fatal actions contained in the

body of the programs. Other producers reported
how plot development is often sacrificed for
redundant brutality. Concerned writers expressed
their indignation at being instructed to devise
more interesting ways to kill people and at finding

violence added to their scripts without their
knowledge.
Eventually the production of violence became
an occupational norm that no longer required
special orders. Like other salable commodities, its
manufacture required efficient routinization.
Writers and producers who find the artistic
constraints and exploitation of violence
objectionable are selectively eliminated; those

who remain as regular contributors become
socialized into the practices of the system to use
violence in amounts and forms that will not
provoke troublesome criticism.
When one must produce material containing

dehumanizing ingredients for money, losses in
self-respect are reduced in many ways, as revealed
in the remarks of media practitioners (Baldwin
and Lewis, 1972). Diffusion of the production

process reduces a sense of personal responsibility
for the final product. Rewrite men alter writers’
scripts; directors fill in the details of the scenarios;

and editors take a part in how filmed events are
depicted by what they select from the lengthy
footage. If the original writers are offended by
alterations of their scripts, they can use their
registered pseudonyms and thereby avoid any
personal embarrassment.
Producers often excuse commercialization of

violence by contrasting it with outrageous
inhumanities, as though one form of human
cruelty absolves another. Why pick on television,
the scapegoat disclaimer goes, when the
government is slaughtering foreigners and the

tobacco industry is disabling the natives (“To
examine violence where the end result is a dead
body on television glosses over the point. This
evades the culpability of a whole society which

permits wars” p. 343). Another variant in the
justification process is to sanctify brutalizing
excesses on television by pointing to revered

masterpieces containing some violent episodes.
The fact that people occasionally aggress in folk
tales and in the Bible presumably exonerates the
saturated overkill delivered daily to every
household in the country year after year.
Sometimes a high moral purpose is assigned to the
taking of human life in the likeness of a national

character-building service: “The government
wants kids to think that there are values worth
fighting for, and that's basically what the leads on
our show are doing. ... If people who break the
society’s code resist the law, we have to use

violence to suppress them. In doing so we are in
the mainstream of American morality” (p. 300).
Modeling violent solutions to problems not only
builds character and establishes the measure of
man; it is also purported to serve a public

therapeutic function by draining accumulated
aggressive drives in viewers.
The term violence rarely appears in the
vocabularies of media personnel. Rather, sanitized
labels such as “action” and “adventure” are used in
its place. Virtually all producers disclaim using
gratuitous violence by attributing evident excesses
to the characters they create. Ruthless individuals
or even peaceful folks confronted with mortal

jeopardy demand acts of violence. One of the more
candid scriptwriters discounted the asserted
dramatic requirement for violence as analogous to
saying, “I never put cotton in a wagon that’s not
prepared for cotton—but I never use anything but

a cotton wagon” (p. 311). In addition to sanitizing
the product, media personnel tend to divest
viewers of human sensitivities or invest them with
base qualities that justify serving them gory
offerings: “Man’s mind is connected to his

stomach, his groin, and his fists. It doesn’t float five
feet above his body. Violence, therefore, cannot be

eradicated” (p. 300)….“Not as much action as
some, but sufficient to keep the average
bloodthirsty viewer fairly happy” (U.S. Senate,

1963, p. 2329).
Producers of products that may have some
adverse consequences can avoid discomforting
self-reactions by ignoring evidence about effects. It

is therefore not surprising that most are unaware
of the results of existing research, that they do not

seem especially eager to find them, and that they
can think of many reasons why the social effects of
television can never be clearly demonstrated. As is
true of any profit-oriented enterprise, some of the

program manufacturers are not exactly plagued by
bothersome standards of social morality:
“Filmmaking for television is a business of

merchandising and profit making. We are
manufacturing a product and we want to attract

the largest possible audience, short of
prostitution” (p. 313).
The distributors and financial patrons of
televised violence attribute the prevalence of such
contents to the aggressive nature of man.
Television is simply responding to the desires of
its viewers, according to network officials. The fact
that action programs attract mass audiences

proves that people find violence inherently
rewarding. This questionable diagnosis is usually
accompanied by the democracy argument. Since
there is such a substantial consumer demand for
violence, the television industry has almost a

mandate to serve the public what it wants. Social
critics, accepting this causal analysis as valid,
nevertheless contend that a public resource
should be used for human betterment, not to
reinforce destructive tendencies in the pursuit of

profit.
These debates overlook that fact that the

prevalence of action formats demonstrates neither
that violence is the source of fascination, nor that
violent offerings are necessarily the most popular.
Suspense, excitement, conflict, and fast pace of
action, all prominent features of such programs,
attract and hold attention. The success of
Westerns and crime series that rarely show people

assaulting each other suggest that the nonviolent
components of the characterizations are the more
important sources of attraction. For example, in
the dramatization of Wyatt Earp, which attracted a
mass audience, this agent of the law shot only one

man during the entire six years the program was
on the air, and this single episode represented an
actual event from history (U.S. Senate, 1963).
It is probably more accurate to say that people

are fascinated with conflict and suspense rather
than with brutality per se. For this reason,
detective stories would hold little interest for

readers or viewers if the outcomes of the plots
were revealed to them in advance, even though
the violent episodes remained unchanged. Nor are
many sport fans willing to sit through complete
reruns of football games knowing the outcomes of
the contests. For years professional football, which
is claimed to attract spectators by appealing to

their inherent aggressiveness, in fact aroused little
public interest until it was commercially
promoted. If artificial sources of excitement
created by league rankings, national ratings, and
intersectional bowl games were discontinued

attendance would drop, however much
contestants shove each other around. On the other
hand, spectators frequent competitive games such
as baseball that involve little physical injury but
sufficient uncertainty of outcome to hold their

interest. Whatever the athletic activities might be,
teams that rank low in the publicized standings

typically draw small crowds regardless of their
level of aggressivity.
The myth that program content is determined
by the preferences of the viewing majority is
widely accepted. Television preferences reflect
what viewers are given to choose from. The
offerings that are broadcast are determined by

commercial considerations as well as popularity
ratings, which do not always match. Programs that
capture large audiences are discontinued if the
people they attract do not fit the age (18-49) or
income categories that advertisers wish to

influence. Thus, popular programs that deliver the
wrong kinds of viewers from a commercial
standpoint are retired for unfavorable
“demographics” despite their wide acceptance
because they fetch low advertising rates (Brown,

1971). Since people who lack purchasing power
hold little interest for advertisers, disadvantaged

minorities tend to be underrepresented in survey
samples of television preferences.
Within each broadcast hour approximately ten
minutes are sold for commercial breaks. Popular
shows that are expensive to produce will not make
much of a profit even at high advertising rates.
Program production cost per thousand viewers

therefore strongly influences what gets televised.
Action formats are prevalent because they are
relatively cheap to produce. With a couple of
horses, a makeshift saloon, a superhero, a
transient evil-doer, and the open range a producer

can grind out endless episodes at a price that
yields higher profits than more popular programs
incurring greater production costs. Many shows
high in violent content are retained because they
are financially attractive to the industry even

though they draw smaller audiences than
programs that do not feature violent action. We

might note here in passing that television
offerings, though generally accepted, fail to evoke
much involvement or enthusiasm (LoSciuto,

1972). A good predictor of what viewers will
watch is whatever appears on the same channel
next, hardly the picture of a motivated search for
violence.
Those who create the programs acknowledge
quite candidly that conflict is essential for

engrossing drama and physical violence is the
easiest and intellectually nondemanding way of
depicting it. As one producer put it: In dramatizing

conflict on television, man against nature is too
expensive, man against God is too intellectual, man
against himself is too psychological and leaves too
little opportunity for action, so man against man is
what one usually ends up with (Baldwin and

Lewis, 1972). Given severe time constraints,
physical violence is a handy device for creating a
conflict and a quick and decisive way of resolving
it. Gifted writers and producers who could portray
conflict more imaginatively are understandably
reluctant to devote their talents to grinding our
formula shows at costs that will yield high
financial returns to stockholders.
When violence is routinely used as a substitute

for creative plot development, it tends to increase
if left unchecked. Producers resort to incremental
violence to capture the attention of viewers who
have grown accustomed to deadly combat: “Last
week you killed three men; what are you going to

do this week? So producers began to lean more
heavily on the violence” (Baldwin and Lewis,
1972, pp. 321-22). It also forces them to create
new ways of killing people to inject variety into

overused formulas: “I wish we could come up with
a different device than running the man down
with the car as we have done this now in three

different shows. I like the idea of the sadism, but I
hope we can come up with another approach for
it” (U.S. Senate, 1963, pp. 2302-32).
The answer to the abundance of televised
violence perhaps lies more in production costs and
in constraints imposed by time schedules,

voluminous output, and limited talent pool than in
viewers’ cravings. The justifications of media

violence in terms of social demand also
misrepresents a two-way influence process as a
one-way cause and effect relationship. Television
creates as well as caters to preferences. In a major

program of research Zajonc (1968), among others,
presents several lines of evidence that mere
repeated exposure to material increases liking for

it. The abundant evidence that familiarity
produces liking is confirmed by Himmelweit,

Oppenheim, and Vince (1958) with regard to
television offerings. Programs that initially held
little interest for children and that they would not

choose to watch if an alternative were available
became more attractive with repeated viewing.
There exist many societal practices that have
widespread injurious consequences, but because
they are financially profitable or politically

expedient, are institutionally sanctioned and
routinely performed by decent people under self-

exonerating disguises. Institutionalized
discrimination takes a heavy toll of victims. A
variety of industries, each with its public-spirited
justifications, contribute to environmental

destruction that causes harmful effects on a large
scale.
Given the existence of so many self-absolving
devices, a society cannot rely entirely on

individuals, however noble their convictions, to
safeguard against destructiveness, just as
aggression is not rooted in the individual, neither
does its control reside solely there. Humaneness
requires, in addition to benevolent codes of self-
reinforcement, social reinforcement systems that
continuously uphold compassionate behavior and
discourage cruelty. Societies must create more
effective organizational safeguards that function

as restraining influences in the use of institutional
power for exploitive or destructive ends.
The earlier discussion has shown that
aggressive actions persist because they are
functional in producing rewarding effects,
however lean or peculiar these outcomes might be

in any given instance. Injurious forms of
aggression can therefore Ire most effectively

eliminated by removing the social conditions that
instigate them and the positive reinforcements
that maintain them. Unfortunately, most of the
people who suffer the results of violence rarely
exercise much direct control over the
circumstances that continue to generate it. Ghetto
victims, for example, lack the means with which to
remove influences that breed violence. Similarly,
university faculty, even if they could achieve high

consensus on foreign policy, cannot by themselves
promptly turn off military activities that arouse
violent protests on campuses. People most often
resort to punishment solutions when aggressive
behavior is disturbing to them and the conditions

fostering it are beyond their power to modify.
When aggression is either highly functional or
personally valued, threatened consequences serve
as its main deterrents. Behavior that produces

aversive effects is generally performed less
frequently, if not discarded altogether. There are
two principal ways in which negative sanctions

inhibit forbidden actions. Repeated punishment
for aggressing toward certain persons, places, or
things endows them with fear-arousing value. As a
result, inclinations to aggress toward these threats
evokes fear, which motivates inhibitory controls.
Man’s cognitive capacities enable him to
regulate his behavior to some extent without

requiring fear arousal. Punishments that are either
observed or experienced directly convey
information about when certain forms of
aggression are safe and when they are hazardous.
Aggressive actions are thus partly regulated on the

basis of anticipated consequences. Being under
cognitive and situational control, aggression
inhibition arising from external threats varies in
its durability and in how widely it generalizes

beyond the prohibitive situation.
The effectiveness of punishment in controlling
behavior is determined by a number of factors
(Bandura, 1969a; Campbell and Church, 1969). Of
special importance is the level of reward achieved
through aggressive conduct and the availability of
alternative means of securing desired goals. The
likelihood that aggression will be punished, the
nature, severity, and duration of the aversive

consequences, and the time elapsing between
aggressive actions and negative outcomes also
determine the suppressive power of punishment.
Additionally, the level of instigation to aggression

and the characteristics of the prohibitive agents
influence how aggressors will respond to being
punished. By considering these different
influences one can predict, with some accuracy,
what the effects of punishing aggression are likely

to be in given circumstances.
Elimination of Aggression Through Punishment
Under conditions in which alternative means
are made available for people to get what they

want, aggressive modes of behavior that carry a
high risk of punishment are rapidly discarded.
Findings reviewed in the concluding chapter
demonstrate that selective reward of prosocial
solutions combined with penalties for aggressive

activities achieves enduring reductions in
aggression without any adverse effects.
Societies rely heavily upon deterrent threats in
their efforts to curtail behavior designated as
criminal. Punishments are widely administered by

enforcement agencies on the grounds that they
discourage transgressors from committing future
crimes and deter others should they be tempted
by the sway of circumstances to engage in criminal

conduct. Obviously if punishments for
transgressive behavior were severe and certain,
they would have strong deterrent effects. In

reality, only about half of the criminal offenses
that have been committed are ever reported by
the victims (Ennis, 1967), and the chance of
apprehension may be relatively low for the crimes
that come to the attention of police. The deterrent
potential of legal sanctions will vary depending on
the risk of the threatened consequences.
The restraining effects of legal sanctions are
difficult to evaluate because fluctuations in
baseline crime rates arising from many
contributory factors obscure the influence of
threatened penalties. When severe punishments

are instituted during periods of widespread
infractions, as is often the case, predictable
declines in contagious transgressions may be
erroneously credited to the stricter control

measures. Sometimes drops in offenses following
the institution of more severe penalties turn out to
be no greater than past fluctuations when

compared over an extended time series (Campbell,
1969). Nor do high crime and recidivism rates
necessarily refute the deterrent value of
punishment, because the forbidden behavior
might be much higher without the legal
prohibitions. In neither case does one know what
the crime rates would be if the threatened

consequences were absent. Only by instituting and
removing the penalties several times, or by
comparisons with nonregulated situations, can the
effects of legal sanctions be estimated. Zimring
(1971) argues convincingly that legal and

enforcement agencies bear some responsibility for
measuring what effects their punishment
prescriptions are having on people. When officials
are granted punishment powers without being
required to evaluate the results of their practices,

aversive measures are more likely to be used
vindictively than in ways that contribute to

rehabilitative or preventive purposes.
Legal threats probably have greatest deterrent
force for the more advantaged segments of the
population. Individuals who have developed
legitimate options that are reasonably rewarding
are less attracted to criminal alternatives. They
not only face weaker inducement to illegal

conduct, but the punishment costs are more
severe for them. After people have achieved a
financial and reputational stake in their
community, they fear the ruinous consequences of
stigmatization and loss of standing as well as the

hardships of imprisonment. Members of
delinquent subcultures, or those relegated to an
impoverished existence, do not risk status losses
for engaging in criminal activities. In addition,
high-status law-abiders are more likely to have

adopted the moral standards of the larger society
so that they are deterred from criminal conduct by

self-reprobation, quite apart from the fear of
external punishment. Finally, the deterrent
efficacy of legal threats is weakened in high

delinquency areas where residents see offenses
committed repeatedly without detection. The
chances of escaping punishment look pretty good.
The fact that threatened penalties work best
for those who least need them does not mean that
they do not serve a restraining function. Unlawful

behavior and breaches of social prohibitions do
occur among supposedly law-abiding citizens.
Transgression rates would undoubtedly rise if

negative sanctions were not in effect. However,
control measures that punish forbidden conduct
but offer no positive alternatives have serious
limitations.
Selective inhibition and refinement of aggression

through punishment.
Aggression control through punishment
becomes more problematic when aggressive

actions are socially or tangibly rewarded, while
alternate means of securing desired outcomes are
either unavailable, less effective in producing
results, or not within the capabilities of the
aggressor. Here punishment must be applied in
considerable force to outweigh the benefits of

aggression. But even then it achieves, at best, only
temporary partial control, with the risk of

aggravating the problem.
When aggressive behavior that brings some
rewards is consistently and severely punished, it is
inhibited in the threatening situation. Functional

aggression, however, tends to reappear after the
punishment has been discontinued, and is readily
performed in settings in which the chance of
punishment is low (Bandura, 1960; Bandura and
Walters, 1959). The inhibitory power of

punishment is affected not only by prevailing
conditions of reinforcement but also by those
associated with aggressive behavior in the past.
Deur and Parke (1970) report that children who
had been consistently rewarded for hitting
responses later reduced such behavior more when
it was either ignored or consistently punished

than did children whose aggression had
previously been treated inconsistently by being
rewarded and punished or rewarded and ignored.
Comparison of the long-term effects of the two
mixed patterns of reinforcements revealed that

persistence of aggressive responding under
unfavorable conditions of reinforcement was due
mainly to the prior history of reward. Children
who had come to expect that their aggressive
behavior would always be rewarded were quick to

abandon such actions when they no longer paid
off. By contrast, children who learned that the

benefits of aggression are obtained at the risk of
some negative outcomes were not easily
discouraged by nonreward or censure.
Evidence that punishment does not always
reduce aggressive behavior has led some writers
to characterize repetitive offenders as
“psychopaths” who lack the capacity to learn from

experience. This apparent learning deficiency
might be readily refuted if information were
available concerning the outcomes experienced
for every transgression. The occasions on which
antisocial aggression is punished are well

publicized, but its successes remain unknown.
Antisocial aggression is most likely to persist
when the rewards it produces outweigh the
inhibitory effects of occasional punishment.
Given some level of reward for aggression, the

behavior of a person who is primarily deterred by
fear is largely governed by his estimation of the

probability of being caught and of the severity of
the punishment that may follow. Most repetitive
delinquents therefore do not commit offenses that
bring extreme penalties, nor do they ordinarily
carry out their antisocial acts under circumstances
in which the chance of being apprehended is high.
The accuracy with which probable consequences

are assessed depends to a large extent on the
aggressor’s ability to profit from experience and to
evaluate the reality circumstances at any given
moment. Habitual offenders may more readily
perform hazardous actions because they have

altered their techniques on the basis of prior
failures, thus gaining new confidence that they can
get what they want with reduced risk of
punishment. They often err by overestimating
their chances of success (Claster, 1967). Failures

normally occur because of unforeseen events that
the aggressor could not have easily predicted even

after careful assessment of the situation. Given
limited options, the human mind is sufficiently
facile to benefit from the successful examples of

history but to discount its negative lessons.
Persons who have few alternatives that bring
some payoffs will be slow to abandon them even
though the behavior results in occasional adverse

consequences. Under these conditions,
punishment is likely to lead offenders either to

adopt safer forms of antisocial aggression or to
change their methods to avoid punishment on
future occasions. An excellent example of how
punishment results in refining rather than

eliminating antisocial behavior that has functional
value is provided in the autobiography of a
talented habitual offender.
My prison surroundings have been

completely a life apart, something so far
away that at times it was my real
circumstances that seemed so fantastic. In
between I went over jobs which I had pulled
off and mentally surveyed them to see how
they could be improved upon. Then I went
over my mistakes again, and learned how
they had occurred and let me down. So you
see there was always plenty for me to do
when I lay on the board with no occupation
but thinking. And plan future jobs. Oh yes, if
a survey could be taken it would be proved
that most of the big criminal jobs, and
thousands of small ones, are planned in gaol.
Planned to the last detail because there is not
sufficient alternative interest to occupy
prisoners’ minds. (Hill, 1955, p. 39)
Had this individual as ingeniously pursued a more
socially acceptable vocation, he would
undoubtedly have been admired for his flexibility

and his ability to profit from failure experiences.
Modeling of aggressive tactics through

punishment.
Punishment is not only precarious as an
external inhibitor of intermittently rewarded

aggression, but its frequent use can inadvertently
promote aggressive modes of behavior. In

exercising punitive control, prohibitive agents
model aggressive styles of behavior not unlike

those they wish to discourage in others. Recipients
may, on later occasions, adopt similar aggressive
solutions in coping with the problems confronting
them. Consistent with modeling theory, Hoffman

(1960) reports that mothers who forced
compliance through coercive means had children
who likewise resorted to aggressive tactics in

controlling the behavior of their peers. Numerous
field studies have shown that parents of
aggressive boys rely heavily on punitive methods

of control, whereas parents of nonaggressive
children favor nonpunitive verbal forms of
influence (Bandura, 1960; Bandura and Walters,

1959; Glueck and Glueck, 1950). Although the
direction of causal relationships cannot be

unequivocally established from correlational data,
it is clear from controlled studies that aggressive

modeling breeds aggression.
Punitive treatment, in addition to furnishing

aggressive examples of behavior, may reduce
restraints over aggression in observers.
Disinhibitory effects are especially apt to occur
when aggressors in positions of influence

convincingly justify their punitive actions. If
empowered authorities can get what they want by
aggressive means, the reasoning goes, why should
such behavior not be acceptable for others in the
pursuit of their own goals. Through modeling

influences, institutionally sanctioned aggression,
when misused in the service of behavior control,
can thus promote the learning and legitimation of
aggressive tactics. Moreover, authorities who
exploit their coercive power to suppress

warranted grievances create, by their actions, a
cynical contempt for official condemnations of

violence.
Capital punishment represents societal
modeling of violence in the most brutal form,
ostensibly for the purpose of homicide control.
The moral and legal justifications offered for
putting to death a prisoner against whom society
is already protected by his imprisonment are

essentially disguises for vengeance. Studies
comparing murder rates in different areas that
vary in their use of capital punishment and in the
same area before and after the death penalty was
abolished provide no evidence that presence of

the death penalty reduces homicide rates (Bedau,
1967). It would appear that legal executions
achieve mainly a lessened regard for human life.
In States that impose the death penalty only a

fraction of those who have committed capital
crimes are in fact executed. Discretionary power is
exercised throughout the criminal process by

prosecutors, judges and juries. As a result, the
poor and disfavored minorities are more likely to
be sentenced to death than the more advantaged
for similar crimes. Execution rates of those
committed to death row show the operation of
selective influences in the commutation of death
sentences (Johnson, 1957; Wolfgang, Kelly and

Nolde, 1962). Thus, legal executions not only
indicate how society regards human life but
exposes the inequities in the administration of its
system of justice.
Escalation of aggression through punishment.
There are certain conditions under which
neither threatened nor actual punishment deters
people from behaving aggressively, at least in the

short run. As might be expected, low aggressors
inhibit hurtful actions when recipients can
retaliate. In contrast, high aggressors do not
temper their punitiveness under threat of
counterattack (Edwards, 1967). In fact, the more
self-assertive the opponent, the more aggressively
they are inclined to treat him (Peterson, 1971).
The different consequences anticipated by high
and low aggressors is most likely the operative
factor producing the diverse effects of threatened

reprisal. Individuals who recurrently engage in
aggressive behavior have experienced some
success in controlling people through force. Strong
attacks are designed to show potential challengers

that the aggressor is a formidable person whose
wishes should be heeded. Counterattacks merely
evoke more forceful demonstration of power to
gain eventual acquiescence. Given other options,
people who have had little success in wielding

coercive power are easily dissuaded from
escalating the aggressive interchange by

expectations of defeat.
The use of punishment as a control technique

carries greater risks when applied to group
aggression. Seeing others engage in aggressive

conduct weakens the inhibitory effects of
threatened retaliation (Baron, 1971b). Protest
tactics, furthermore, have some legitimacy as an
instrument of social change. After nonviolent

means have proved ineffective, severe punishment
of militant actions used to achieve laudable goals
is likely to escalate collective aggression. These
particular conditions are especially salient in
campus protests and violence. Searle (1968) and

Spiegel (1970) describe a common sequence of
reciprocal escalation of aggression that was
repeated with notable regularity on many college
campuses, both in the United States and abroad. In
the initial stage, a small group of students whose

grievances have been slighted confronts a
university administration with a set of demands

relating to a fundamental issue (for example, war,
free speech, racism, institutional policies) with
which most nonprotesters sympathize, even

though they are not moved to action. The demands
are presented in a coercive style, often
accompanied by provocative actions or by
violations of campus rules that invite punishment.
Authorities refuse to give in under such
pressure for a variety of reasons. Acquiescence to

blatant or unlawful force will lose them the
respect and support of other factions without
whose backing they can no longer function

effectively. They also fear that, however laudable
the goals, granting demands under duress rewards
violent tactics so that physical force will become a
favored strategy rather than a method of last
resort. Some of the ultimatums cannot be met

within the specified time limits because decision-
making procedures are unwieldy and universities
may lack either the resources or the power to
modify the causes of certain grievances. Partial
rejection rather than full compliance is the
predictable outcome. Protesters, in turn, become
more resistant to negotiation because not only
issues but their reputations in the face of
perfunctory or unyielding response are now at

stake. The net result is that some of the protesters,
usually the most vocal leaders, are suspended or
otherwise disciplined for campus rule violations.
Punitive measures escalate the level of
aggression to the second stage. The original issue

is now transformed so that the malevolence of the
institution itself becomes the primary issue.
Punishing individuals for protesting against evil
practices, regardless of whether the protest is in

good faith or to generate student revolt, mobilizes
sympathizers to massive action against the
university. The confrontation creates a strong

sense of community in which provocative rhetoric
is selectively rewarded, while concerns voiced
over the morality of violence are silenced through
justification or disregard. Restraints over
aggressive actions are weakened by indicting the
university for, among other things, scientific
corruption, contributing knowledge and

technology to pernicious societal practices, and
self-serving ethical bias that favors existing
systems causing social ills. Faculty members with
similar sentiments give ideological and social
sanction to forceful action. Additional directives

are issued to the university to restructure its
activities and decision-making; amnesty is
demanded for individuals charged with disruptive
actions. When these nonnegotiable demands are
refused, as is invariably the case, protesters seize

buildings, destroy facilities, or embark on other
large-scale disruptions of university functions.

Strong political and public pressure is then
brought to bear on beleaguered administrators to
put a halt to the unlawful behavior.
Lacking adequate internal mechanisms for
dealing with major confrontations, university
authorities calf police onto campus, which elevates
the conflict to the third, and final, stage. The sight

of mass arrests, often carried out in a brutal
manner, “radicalizes” many of those who had not
publicly committed themselves one way or
another, whereupon they too join the fray. The
university is essentially shut down amid open

hostilities between opposing factions, vigorous
petitioning, emergency faculty meetings, and
outpouring of grievances against the university.
Few college presidents survive a major campus
uprising without being seriously discredited.
After the predictable pattern became apparent,
activists began to select tactics calculated to

provoke authorities to punitive actions.
Universities, in turn, adopted new
countermeasures, principally in the form of quasi-
judicial procedures and court injunctions.
Administrators controlled disruptive activities
without provoking more serious forms of
aggression by reliance on legal deterrents rather

than on personally administered punishments.
Aggression directed at social injustices in the
society at large is frequently escalated and spread
by the punitive actions of those in positions of
authority. Violent response to extralegal protest

provides manifest demonstrations of how coercive
power can be misused by the social system itself.
Protection of dissenters from encroachment of
their civil rights by agents of control and the

integrity of authoritative agencies become matters
of concern in such struggles.
Disinhibition of Aggression by Reducing the Risk of

Punishment
When individuals are deterred from behaving
aggressively mainly by fear of negative
consequences, conditions that reduce anticipated
risk of punishment weaken restraints over
aggressive responding. This process, designated as

disinhibition in social learning theory, has also
been studied to some extent by social
psychologists as “deindividuation” (Festinger,
Pepitone, and Newcomb, 1952; Zimbardo, 1969).

Although both concepts refer to a similar process,
they differ in that disinhibition encompasses a
wide range of determinants, whereas
deindividuation is largely confined to restraint
reduction arising from loss of individuality by

immersion in a group.
Disinhibitory conditions may take many forms,

some of which are graphically described by
Zimbardo (1969). By decreasing fear of detection

and punishment, anonymity is highly conducive to
transgressive behavior. It is not uncommon for

otherwise law-abiding people to perform socially
prohibited activities without too much resistance
when there is little chance they can be identified.
Restraints over aggression are further weakened

by the actions of a group. Some of the contagious
effects are doubtless attributable to the modeling
influence of seeing others engage in antisocial
behavior without experiencing any adverse
consequences. Moreover, in a joint performance

no one feels personally responsible for the effects
of the collective action. There is an additional
safety in numbers. Since huge crowds cannot be
easily punished, even a feeling of shared
responsibility for cruel deeds does not have much

inhibitory power. This is especially evident in
censurable societal practices, when individuals

facilely admit guilt for their contributions by their
support or inaction, knowing full well that such
confessions serve mainly to placate critics.

Confessions of collective guilt would decrease
markedly if there were any chance of being
punished for acknowledged complicity. Indeed,
structuring the problem as a societal act

implicating everyone safeguards against accepting
personal responsibility for one’s own conduct.
Zimbardo (1969) examined the combined
effects of anonymity and distributed responsibility
on aggressive behavior. Female college students

administered painful shocks to a victim in groups
of four, in situations in which all members need
not have acted punitively to fulfill the demands of
the experiment. Participants who were never
recognized by name and wore hoods over their

heads shocked other girls much longer than those
who were clearly identified by name tags and had
no disguises. Anonymity heightened punitiveness
toward charitable and obnoxious victims alike.
Results of subsequent experiments suggest
that probable consequences rather than
concealment per se is the critical disinhibitory
factor. Belgian soldiers aggressing in groups and

female students performing alone were
significantly less punitive when they wore hooded
disguises than when they were publicly identified.
The conflicting findings were explained by

Zimbardo on the basis that concealment made
aggressors feel conspicuous and self-conscious. An
alternative, though not incompatible,
interpretation is in terms of anticipated
reprobation. Sly aggressors are usually judged

more severely than those willing to accept the
consequences of their actions without being

secretive. A single aggressor can escape neither
self-reproof nor social censure for underhanded

cruel actions by concealing her face during the
performance. Similarly, a group of soldiers

behaving cruelly in disguise during an official
assignment nevertheless remain known and
subject to disciplinary measures. In their studies
of interracial aggression, Donnerstein et al. (1972)

further corroborate how anonymity and freedom
from reprisal heighten interpersonal punitiveness.
Most of the conditions described earlier that
obscure or redefine the relationship between
one’s actions and its effects to lessen self-

condemnation reduce the chance of external
punishment as well. People who are deterred by
external threats can be persuaded to aggress when
a recognized authority is willing to accept the
consequences for their actions, when extenuating

circumstances can be invented, or when
reprehensible behavior is redefined in ways that
make it appear socially acceptable.
Dehumanization of potential victims likewise
disinhibits externally restrained aggression, since
punishment is not only less likely but is much
weaker for aggressing toward devalued people
than toward revered ones. In everyday situations,
disinhibitory influences ordinarily increase

aggressiveness by reducing both self-reprimands
and external threats.
REINFORCEMENT INFLUENCES IN
COLLECTIVE AGGRESSION
Psychological explanations of collective
aggression generally invoke historical causes or

personality traits rooted in the individual. In
contrast, specific instigational and reinforcement
influences, which function as powerful
contemporary determinants of behavior, are

either assigned an incidental role or neglected
altogether. Differences between theories in the
relative emphasis they place on the conditions that

produce protest or on the types of individuals who
become protesters are reflected in studies of
student activism. This phenomenon has special
significance for theories of aggression because,
unlike most mass movements, it represents revolt
by advantaged rather than by underprivileged

segments of society.
Individual Determinants of Collective Action
Research in this area, guided by the
characterological tradition, has been largely
preoccupied with delineating the attributes and

backgrounds of students who are disposed to
protest activities. Given appropriate conditions,
protest behavior can be elicited in almost anyone.

However, as numerous survey studies have
demonstrated, some students are provoked to
forceful actions more easily than others (Astin and
Bayer, 1971; Flacks, 1967; Keniston, 1968). For

the most part, militant students come
predominantly from upper-status professional
homes. They are bright, and generally do well
scholastically. However, they do not view the
traditional options available as providing them the
satisfactions in life they seek. Their interests are

predominantly in the social sciences and
humanities rather than in technically oriented
pursuits. On the social side, they tend to be
sensitive to arbitrary authority, to social injustice,
and to hypocrisy.
Contrary to popular belief, student activists are

largely modeling their behavior after parental
values rather than rebelling against them (Flacks,
1967; Keniston, 1968). In comparative studies of

political orientation, for example, nonactivists and
their fathers typically endorse conservative
political positions, whereas activists and their

fathers tend to espouse liberal views. Attitudes of
activists in other areas are likewise correlated
with those of their parents. Flacks (1967)
concisely summarizes the major differences in
familial orientations as follows: “Whereas
nonactivists and their parents tend to express
conventional orientations toward achievement,

material success, sexual morality, and religion, the
activists and their parents tend to place greater
stress on involvement in intellectual and esthetic
pursuits, humanitarian concerns, opportunity for
self expression, and tend to de-emphasize or

positively disvalue personal achievement,
conventional morality and conventional
religiosity.” (P. 68)
In other words, there exists about as wide a

divergence in beliefs among parents as among
their offspring, though the latter tend to assume
somewhat stronger positions on issues. A similar

diversity also exists in the ideologies of college
faculty. Thus, the view that student activism
represents intergenerational conflict is disputed
by substantial evidence showing that, within a
society, beliefs are not distributed like a layer cake
in homogeneous layers; rather, each age layer
contains analogous degrees of heterogeneity. For

this reason, when youth eventually becomes the
Establishment and must deal with the issues of
their time, the customary factional conflicts are
reproduced.
Generational similarities in value orientation

tend to be obscured to some extent by behavioral
disparities arising, paradoxically, from successful
modeling. The behavior of parents who espouse
“liberal” or “radical” views is not always congruent

with their ideals. They adhere, though often
grudgingly, to conventional authority. Most of
them pursue traditional occupations in

competitive ways that amass wealth. Though
permissive toward self-expression, their own life
style does not stray far from conventional social
norms. In the face of conflicting influences arising
from the multiple responsibilities of adulthood,
they are often forced to compromise their values.
And it is not uncommon for highly principled

people to express verbal outrage but to take no
action to rectify injustices. Students who have
adopted magnanimous principles are easily
disillusioned and angered when the social
practices of elders fail to live up to their professed

beliefs. Such discrepancies, which inevitably occur,
typically elicit accusations of hypocrisy and of
complicity with the established order for self-
interest. Adults, of course, do not hold a monopoly
on hypocrisy. Public critics of student activism are

equally skilled in exposing self-righteous tyrants
in student movements who often preach a

different morality from what they practice.
Coupled with the liberalist origin, there is
some evidence (Flacks, 1967) that students who
are disposed to militant action have received
greater parental reinforcement for self-expression
conducive to the development of strong self-
assertive behavior. As a rule, those who come from

advantaged backgrounds are more likely to
experience early success in realizing their wishes
than the less privileged who lack the resources for
controlling their environment. People with a
strong sense of personal efficacy are especially

prone to take action against conditions they do not
like. There is another aspect of privilege that
encourages oppositional response. To the extent
that high status affords some protection against
the negative consequences of one’s behavior,

students of elite backgrounds will more readily
risk hazardous action (Powers, 1971).
Obviously not all students who participate in
forceful protest necessarily fit the pattern
described above. Nor do most students sharing the
same general background and ideological
perspective become militants. In point of fact,
proportionately few students actively participate
in demonstrations. The vast majority, between 90

and 95 percent, either remain uncommitted
onlookers or oppose the protesters (Keniston,
1967; Peterson, 1966). Campus confrontations,
depending on the issues in dispute and the
manner in which grievances are handled, can

recruit new supporters of diverse social learning
backgrounds and attributes. Some of the attitudes
and behavior of activists are shaped by repeated
exposure to militant leaders or through direct
participation in demonstrations. In his intensive

study of student activists, Keniston (1968)
documents the role of modeling in the process of

radicalization. Novices drawn into the movement
learn the specific tactics, rhetoric, and ideology to
a large extent by emulating more experienced and

committed figures.
The time, the place, and the form of collective
aggression can be best understood by considering
both the characteristics of potential protesters and

the presence of aggression-promoting influences.
The latter conditions are especially important

because, if sufficiently strong, they can override
the residual effects of prior learning in
determining how one behaves.
The immediate instigators of protest behavior

were specific events of a provocative nature
related either to military activities or to
discriminatory racial practices. Although campus
demonstrations were initiated by such factors,

inept or precipitous counterresponses by
university authorities often generated a sequence
of events culminating in violent confrontations.
The poor quality and impersonality of higher
education were widely cited as principal causes of
student unrest. However, evidence from
comparative studies did not lend much support for
this popular interpretation. Student protesters
were no more dissatisfied with their educational

experiences than nonactivists (Somers, 1965).
Moreover, student activism was more prevalent in
universities noted for their academic excellence
and progressiveness than in those of lesser caliber
(Kahn and Bowers, 1970; Foster and Long, 1970).
Public critics were quick to attribute

oppositional activities on such campuses to
greater leniency toward student protest.
Traditional disciplinary procedures were

adequate for adjudicating petty infractions, but ill-
equipped for handling large-scale disruptions,
especially when strengthened by nonstudent

recruits. Countermeasures were rapidly devised
and students were in fact penalized for prohibited
activities (Keniston and Lerner, 1971). In addition
to initial indecisiveness, exercise of restraint in
disciplinary actions so as not to escalate peaceful
dissent to violent confrontation was viewed by
some as laxness. It should also be recognized that

universities of high academic stature selectively
recruit large numbers of talented students who
have low thresholds for inequities and moral
compromise in others, if not always in themselves.
Kahn and Bowers (1970) provide some evidence,

however, that high rates of activism in prestigious
institutions are not due simply to the kinds of
students they attract. High-quality universities
tend to foster social and political protest in
students of varied backgrounds. Kahn and Bowers

identify several conditions that may be conducive
to developing activist subcultures. Leading

universities, charged with educating gifted
students, many of whom will become the decision-
makers and social reformers of the future, are

more inclined to encourage critical scrutiny of
traditional systems than to defend them.
Principled protest is therefore not only accepted,
but supported in varying degrees by questioning

faculty members. Through scientific discoveries
and consultation, the faculty of prestigious

universities contribute to the course of social life
and national policies. When this knowledge is put
to bad use, influential universities are more
vulnerable to student attack than those having few

ties with industrial and governmental agencies.
After a visible activist subculture has evolved, new
students sharing similar orientations are drawn to

the university, thus perpetuating the subculture.
Modeling influences, another contemporary

determinant, not only define the critical
instigators at any given time, but shape the
specific style and tactics of collective aggression.
Demonstrations at a few salient universities
established whether military and industrial
recruiting, classified research, campus officer
training, ethnic studies, or some other issue
became the principal object of protest on other

campuses throughout the country. Equally striking
was the rapid contagion of the style of aggression.
The forms became progressively more violent
over time, beginning with peaceful sit-ins, and
graduating to combative disruptions of university

functions, to mobile trashing of buildings, and
eventually to firebombing. There is another aspect
of modeling influences that has received little
notice, though it may constitute a strong
motivating force. Reformers have a militant

reputation to uphold. Failure to take forceful
actions when others in similar settings are

protesting threatens loss of status. Widespread
modeling may thus compel activists to espouse
aggressive measures as a means of retaining their

integrity and stature in the eyes of their
supporters.
Modeling influences can substantially alter the

personality correlates of activism over time. Those
who set a style of behavior usually differ in
characteristics from later adopters. Initially,

certain types of students in prominent schools
may establish the mode of protest. Like most
forms of prestigious modeling, the pattern is

eventually adopted by all kinds of students in
diverse institutions, especially if the actions
produce some benefits. When modeling influences
override the effects of personal and immediate
situational determinants, different correlates will

emerge for participants in early and late stages in
the diffusion of aggression (Kahn and Bowers,
1970; Mankoff and Flacks, 1971).
Analysis of reinforcement supports for
collective aggression requires separate
consideration of the consequences of individual
and of collective response. At the individual level,
the reinforcements for participating in aggressive

activities are varied and complex, and they can
change even for the same person in the course of
continuing activism. Some of the protesters, who
adhere to high social and ethical principles,

become distressed by institutional practices that
exploit and oppress disadvantaged people. In such
cases, coercive actions are sustained, even in the
face of punishing consequences, by self-approval
for upholding one’s convictions and by

expectations that continued pressure may
eventually produce humane social reforms.
It would be naive to assume that all militant
actions, even by idealists, are in the service of

altruistic goals. Public-spirited reformers are often
changed by the social recognition and power their

activities bring them. Discontent is easily exploited
for self-serving interests without genuine concern
for the plight of others. Moreover, after a protest
movement has been launched, it creates a

powerful subculture that supplies a new source of
reinforcement supports for aggressive actions.
Participants are united into a cohesive community
that impinges on every aspect of their lives. Here
they experience close personal relationships that

are rarely matched in the usual course of events.
They share their money and resources in the best
equalitarian tradition; they educate and entertain
themselves; and they develop a strong sense of
camaraderie in battle through mutual protection

and legal support should they get “busted” by the
police.
Valor in mock combat or open fighting can
bring group recognition and status rewards, as in
other subcultures valuing aggressive
accomplishments. Faithful participants therefore
strive, in word and deed, to display the qualities
prized by the group. On the occasions when
victors forcibly discomfit their adversaries, they

are strongly reinforced by a feeling of potency,
particularly if their past efforts were slighted or
brought them a sense of helplessness (Powers,
1971). Although of much less import, it should
nevertheless be recognized that campus protest

activities bring welcome relief from the boredom
of academic routines. Hence, springtime, when
outdoor attractions hold greater interest than
textbooks, is an especially opportune time for
mobilizing collective action. In sum, militant

protest may be instigated by adverse social
conditions, but it is sustained from day to day

largely by the interpersonal reinforcements
derived from the activist subculture. Indeed, for
some of the participants, the broader social issues

in dispute are poorly understood or, at best, of
secondary concern. Even for those more
ideologically inclined, the political determinants
usually fade into the background as the

subcultural influences assume major reinforcing
functions. Groups seeking vast revolutionary

changes in an affluent society are unlikely to
achieve any quick successes for their efforts, nor
are the prospects of long-range benefits
particularly encouraging. Nevertheless, violent

actions can be maintained after hope of tangible
results is abandoned by social rewards from an
estranged cohesive group, especially if the acts are

carried out in ways that minimize risk of
detection.
The diverse personal influences that may lead
one to adopt violent courses of action are vividly
documented in the tragic case of Diana Oughton
(Powers, 1971). This sensitive girl, who came from
a privileged conservative background, was
progressively shaped, by events and by selective
association with militant peers, from dedicated
social service to revolutionary violence. Her

disillusionment with existing social systems began
in Philadelphia ghettos and intensified among the
destitute Indians of Guatemala, where she served
as a volunteer in a Quaker service program. Seeing
at first hand the devastating human costs of severe

privation, Diana—who as a child was repeatedly
taunted for her family’s affluence—developed a
growing dislike for American extravagance and
foreign influence. Her tireless efforts to improve
living conditions in the remote village were

negated by friends grown cynical about the
prospects for peaceful social change. They argued

that the magnitude of the problems and the
indifference of self-seeking politicians demanded
revolutionary measures to achieve sweeping

reforms. These experiences provided the
ideological justification for militant action.
Upon her return to the United States, Diana

became deeply involved in the Students for a
Democratic Society, which at the time was
splintering into opposing ideological camps. She

was readily drawn into the militant Weatherman
faction through attachment to her boyfriend, who
served as one of the national officers of the

organization. His open admiration of Bernadine
Dohrn’s toughness presented additional
requirements for callousness. After issuing their
manifesto, the Weathermen set out to precipitate
the revolution by terrorism in the streets. Analysis

of the antecedent influences illustrates how
development into radicalism, like most other life
styles, is determined by a series of fortuitous
events, not by studied design. Long-term
prediction of social functioning is difficult because
adventitious branching influences are often more
decisive than personal attributes in shaping the
course of life.
Transformation of a normally compassionate

girl into a tough street fighter was not achieved
through humanitarian concerns and revolutionary
rhetoric alone. To extinguish any residual self-
censure and revulsion over ruthless behavior, the
Weathermen created small collectives where they

eradicated their “bourgeois morality” with a
vengeance. The brutalizing process, carried out in
barricaded houses, is described in gruesome detail
by Franks and Powers (1970).
Inside they lived a 24-hour existence of

intense political discussion, marked by a
complete abandonment of all the bourgeois
amenities of their largely middle class
childhoods. Clothes were strewn
everywhere, food rotted on unwashed plates,
milk turned sour in half-empty containers,
toilets jammed, flies and cockroaches
swarmed in kitchens filled with encrusted
spoons and spilled food…. When the
collectives needed money for bail or for
buying guns and, later, explosives, and
sometimes simply as a matter of discipline,
the members would go without food for
days. On other occasions they would stay
awake for two days or even longer to harden
themselves for life in the “Red Army.” … On
at least one occasion, partly from genuine
hunger and partly to instill in themselves a
kind of savagery, a collective killed, skinned,
and ate a tomcat. … The military aspects of
the training—karate, target shooting,
practice in street fighting and, later, the
making of bombs —suffered in the chaotic
atmosphere of the collectives where
everyone was always overtired and
underfed. (P. 27)
Though estranged from her parents, Diana
could rely on them to protect her against the legal
consequences of her violent actions. They were
always quick to bail her out of legal and financial

difficulties and, as the activities grew more violent,
she sought assurance that they would come to her
aid should she be apprehended for serious
offenses. When destructive rampages and street
battles with police failed to produce the heralded

revolution, the Weathermen resorted to
underground bombing attacks. Diana’s brief
career as a revolutionary ended with tragic
suddenness in a townhouse bomb factory, where
an antipersonnel explosive intended to help

destroy the society site had come to hate took her
life instead.
Consequences of Collective Action
The actions of individuals, however varied
their reinforcement determinants, produce a
collective response that creates another set of
consequences which can have widespread and

enduring effects for protesters and
nonparticipants alike. Changes in social practices
that resist individual effort are attainable through
the force of numbers. One can distinguish
circumstances in which group action is likely to
gain tangible results from those in which it not

only proves ineffective, but may even retard
beneficial reforms.
When legitimate demands and constructive
efforts to produce needed changes are repeatedly
thwarted by persons who benefit from the

prevailing system, this usually evokes more
intense disruptive actions that cannot be ignored.
Conditions favorable to change occur when
forceful action is taken against social practices

that lack sufficient justification to withstand
concerted protest. Once they are revealed, their
defenders receive little public support for

continuing them. Warranted changes may be
temporarily blocked by suppressive
countermeasures, at high costs to the society and
at the risk of enlarging the oppositional activities.
However, given high justification for protest and
the strong coercive power of mass supporters,
continued aggressive behavior eventually

succeeds in securing desired goals.
Coercive group behavior can be powerfully
reinforced by the enduring changes it produces in
institutional contingencies that affect the lives of
many people. Collective aggression can, for

example, remove disliked aversive controls. It can
restore lawful rights and privileges. It can force
authorities to relinquish some of their power and
to introduce organizational changes that make life

more rewarding for those in less advantaged
positions. The more compliant members, who tend
to look with disfavor upon militant tactics,

eventually profit from the nonconformists’’
protests.
Because the consequences of collective actions
are varied and removed in time, their effectiveness
is difficult to verify. It is not uncommon to find
advocates of conflicting tactics defining the same

incidents as successes and as failures. In situations
where protest activities partly serve to boost
personal stature and self-esteem, demands are
often issued in a form that assures apparent

success even though the victors may gain little in
the way of tangible results for their efforts. Usually
some of the conditions demanded already exist as
part of the informal practices of the system, while
others would be granted upon firm challenge

without requiring incensed confrontation. But
authorities do not ordinarily give in to major

demands that appear unreasonable at the time,
and they can therefore rally support to their side.

Protesters can nonetheless claim victory by
pointing to concessions on issues that could

probably have been obtained without massive
expenditure of time and resources in combative
clashes.
Fundamental changes that are resisted by the
established leadership are attainable by coercive
measures when applied concertedly,

unremittingly, and in less violent forms that enlist
rather than alienate supporters. Unfortunately,
far-reaching consequences are invariably delayed.

It takes time to institute organizational changes
embodying the desired reforms. Their
implementation may be further postponed by
defeated factions through appeals to higher
agencies in an effort to restore some of the power

and privileges they have lost. After appeal
processes have been exhausted, new practices are
still forestalled while effective mechanisms are
devised for their enforcement. The long delay
between action and noticeable results can give rise
to disillusionment and a sense of failure, even
though changes of long-term significance are
eventually achieved. The power gained through
these initial reforms increases the success of more

forceful tactics that gain little unless backed up
with coercive potential. To complicate assessment
of effects, protesters may clearly lose critical
battles and yet, through their actions, create a
repercussion of influences that change the system

in the desired directions long after their coercive
pressures have ceased. Moreover, authorities in
comparable situations, fearing similar
confrontation, may initiate self-protective reforms
on the basis of observed costly battles in which no

clear victors emerged.
Forceful aggression, being more aversive,

tends to produce faster, though often more
superficial, victories. Since behavior is more
powerfully controlled by its immediate than by its
delayed consequences, prevailing conditions of
reinforcement generally favor militant action over
restrained forceful tactics. It is therefore easier to
get people to aggress than to adhere stoically to

nonviolent resistance when their efforts are
slighted or derided without any foreseeable
changes. Under conditions in which immediate
effects of group action are distressing but benefits
are considerably delayed, powerful intragroup

reinforcement and self-reward are required to
sustain members until sufficient changes accrue to
assume the reinforcing function. Reliance on
collective aggression as a preferred method of
achieving social change will decrease not by

decrying violence, but by ensuring that nonviolent
methods produce faster results. Though most

people condemn coercive means, they are not
averse to using them when their own interests are
threatened.
Group force is less successful when social
practices violate the professed values of the
society but secure popular support because they
are sufficiently beneficial to a majority of its

members. Many enterprises that create jobs,
profits, or privileges for a sizable segment of the
population flourish under majority preferences at
the expense of minorities and even long-term
costs to all. Given sizable public support for

inequitable practices, repressive measures
succeed in curbing challenges to the existing
order. Such systems are usually resistant to
coercive influences for change from within unless
augmented by pressures from without. It is

doubtful, for example, that the civil rights
movement in the South would have achieved the

results it did without the influx of large numbers
of protesters from the North who, being outsiders
possessing some power, were not as vulnerable to

punitive countercontrol.
Collective aggression is least effective when
existing practices are not entirely satisfactory but
have sufficient justification to gain the backing of

different interest groups. Lacking both coercive
power and adequate grounds for disruptive

protest, forcible actions tend to discredit the
protesters rather than enlist supporters for their
cause in the numbers required to challenge the
system. Such miscalculation of probable

consequences sometimes arises through
indiscriminate modeling. Seeing people get what
they want through forceful action prompts others

to adopt similar tactics without realizing that
conditions favorable for change may be lacking in

their case.
The conditions identified as conducive to social
change through coercive action have been shown
by Gurr (1970b) to be significant predictors of the
magnitude of civil disorder in Western nations.
Gurr examined three sets of determinants. The
first factor is the level of social discontent, defined
in terms of economic declines, new oppressive

restrictions, and persisting discriminatory
practices that prevent people from improving
their life circumstances. Not all discontented
people aggress. Modeling and reinforcement
conditions largely determine whether the

discontent will take aggressive or some other
behavioral expression.
The second factor—reflecting the combined
influence of historical example and sanctions—is

the degree of social justification for using force
tactically to achieve desired reforms. In some
political systems mass protests and coups are

regarded as an acceptable modus operandi of
change, whereas in other nations such tactics lack
functional value because they are widely
condemned. Acceptability of forcible means was
coded by Gurr in terms of legitimacy of the
political system and the frequency with which
previous civil disorder produced favorable

changes in policies or in regimes. The assumption
here is that the past successes of aggressive tactics
legitimate their use.
The third, and final, determinant is the balance
of coercive power between the regime and the

dissident groups. Coercive capacity was assessed
by the ability of contenders to enlist and to
organize support on their side. Loyalty of the
military and police forces, economic influence, and

political and labor support defined the regime’s
coercive power. The size and organization of
dissenting groups, as well as their military, labor,

and foreign support, served as the criteria of
dissident coercive power.
In the correlational analysis, social discontent
was found to be unrelated to disruptive protest,
but positively associated with more violent forms
of civil disorders, such as organized rebellion,

justification of forcible tactics correlated highly (r
= .71) with civil disorder. Countries that have
unpopular governments and traditionally embrace
militant counterresponses experience frequent

public protest, even under low levels of
dissatisfaction. In other words, when aggression is
a common style of life, people are quick to engage
in disruptive actions though not burdened with
hardships. The data further disclose that

widespread violent strife, involving terrorism and
rebellion, can be provoked by intense grievances

regardless of the activist traditions of the country.
The balance of coercive power favorable to

dissenters also appeared as an important correlate
(r = .74) of all forms of protest. Dissident power,

which increases the likelihood of success,
enhances the prospects for aggressive challenges
of the established system. Although regime
support from political and labor sources mitigates

against all forms of violence, military coercive
force is more effective in suppressing rebellion

than in reducing lesser protest activities.
Taken together, the three sets of determinants
account for approximately 70 percent of the

variation in civil disorders. However, the relative
contribution of the component influences differs
depending on the severity of the protest activity.
Widespread marked discontent and dissident
coercive power emerged as the important

determinants of revolutionary violence. In less
extreme forms of collective aggression,
endorsement of forcible tactics and coercive
power are equally strong determinants, whereas
variations in discontent have less effect.
Decline of Contagious Aggression
Earlier it was shown that new modes of
aggression spread in a contagious fashion but,
after being widely adopted, tend to be discarded. A

number of reinforcement influences can
contribute to abrupt decline of contagious
aggression. Foremost among these factors is the
development of effective countermeasures. For
example, as hijacking escalated, airlines instituted

behavioral screening procedures along with
magnetometric devices for detecting ferrous
metal. As additional deterrents to air piracy, it was
widely publicized that concealed armed guards

rode unspecified flights. News of hijackers killed
aloft by security guards and the growing

reluctance of nations to grant asylum further
reinforced restraints against contemplated
hijacking. The incidence of both domestic and

international air piracy dropped markedly.
College administrators similarly improved
their methods of countercontrol through repeated
experience in dealing with provocative actions.

Eventually they adopted the position that
disruptive tactics were no longer an acceptable

means of settling disputes. Unlawful actions were
promptly curtailed by invoking legal restraining
orders that barred protesters from the campus or
from participating in activities forbidden by court

injunctions. Violation of institutional rules by
students and faculty were also firmly discouraged
by college judicial bodies empowered to impose

severe penalties, if necessary. Disruptive behavior
now carried a high risk of punishment. Just as

coercive tactics were widely modeled, so were the
methods of countercontrol. It is difficult to
estimate how much of the decrement in

contagious aggression is attributable to
disciplinary countermeasures and how much to
reduction in aggression-promoting conditions.
Deescalation of the war, reduction of student
grievances by institutional reforms, and increasing
weariness over perennial confrontations were

undoubtedly contributing factors.
The discrepancy between anticipated and
experienced consequences plays an important role
in determining the future course of aggression.

Direct observation and media portrayals of mass
aggression are largely confined to dramatic
confrontation episodes, not to the long intervening
hours of boredom and fatigue. What is seen of the
group’s functioning generally conveys the sense of

solidarity and principled dedication rather than
the stress of pressures for conformity, the feeling
of isolation that accompanies estrangement from
the larger community, and the discouragement
arising from failure to achieve fundamental
changes. Participants who are drawn into protest
activities mainly by their apparent excitement may
drop out after experiencing at first hand the full
consequences of the endeavor. Functional leaders,

who share a sense of purpose and camaraderie,
derive adequate reinforcement supports to sustain
their behavior; the lower ranks, however,
experience high turnover, making it difficult to
sustain the same group action with constantly

changing recruits, some of whom vie for
leadership of a different style. Certain forms of
collective violence produce consequences in such
devastating proportions that they prevent
repetition of similar actions. Burning a ghetto area

to the ground precludes a repeat performance.
Group aggression, like other collective

phenomena shaped by modeling influences, often
follows the transitory cyclic pattern of fads. When
a new style of behavior first appears, it wins social
recognition and thus becomes the “in” thing to do.
As its popularity spreads, however, it rapidly loses
its positive value. What was once highly
fashionable becomes trite. After a practice grows

stale through overuse, it is discarded in favor of
another pattern that undergoes a similar rise and
decline in popularity. There are additional effects
of stereotyped imitation that shorten the life span
of a particular mode of aggression. When the same

rhetoric and tactics are used repeatedly, protest
activities take on the quality of staged productions
rather than genuine expressions of principle. As
leaders reduce their credibility by stereotyped
appeals, their exhortations tend to be viewed as

manipulative and increasingly resisted.
Unresponsiveness to trite rhetoric is often

misread as indifference to the issues. Even among
members of activist groups, some of the more
idealistically inclined may eventually become

disillusioned by the contriving of power plays that
appear to be based more on strategic than on
moral considerations.
The course of collective aggression is often
misjudged on the basis of the linear projection
error. Expecting a continuous heightening of

aggression, control agencies resort to needlessly
extreme counterattacks at the point at which the
activities may begin to decline, thus prolonging the

strife for a time. The eventual cessation of
aggressive activities may be erroneously credited
to the punitive countermeasures when, in fact,
they temporarily bolstered a waning phenomenon.
chapter five
Modification and Control

of Agression
The social learning principles that account for
the development of aggressive behavior apply
equally to its modification. In recent years learning
principles have been used widely to promote
diverse changes in psychological functioning,
including aggression (Bandura, 1969a; Kanfer and

Phillips, 1970). Social learning approaches differ
from traditional intrapsychically oriented

techniques in several important ways. Since
aggression is largely under situational, cognitive,
and reinforcement control, these are the events to
which treatment addresses itself, rather than to
traits, to presumptive drive forces, or to historical
causes.
Discourses on treatment of aggressive
disorders often fail to distinguish between

historical and contemporary determinants of
behavior. As a consequence, potentially powerful
influences are often misapplied with disappointing

results. This issue is well illustrated in views
regarding the essential conditions for modifying
delinquent patterns. It is commonly assumed that
because delinquents come predominantly from
rejecting familial backgrounds, they require a
great deal of unconditional love to remedy their

problems. As Ackerman (1944) astutely observed
years ago, delinquents repeatedly experience
interpersonal difficulties not because they were

unloved as children, but because these adverse
conditions created hostile, distrustful conduct that
currently prevents them from achieving
rewarding relationships, even though there are

plenty of friendly people in their immediate
environment. To take but one example, a person
who has learned to expect rejection is inclined to
test expressions of friendliness and interest from
others by making difficult demands and otherwise

subjecting potential friends to provocative
challenges. Alienating love tests eventually evoke
rejection from most people, producing recurrent
reinforcement for guardedness and hostility. Thus,
antagonistic behavior, due to its negative

character, creates the very conditions likely to
perpetuate it long after the original determinants

have disappeared. The causes of behavior,
including aggression, reside in the here and now,
not in the past. Favorable changes are therefore

best achieved not by replenishing lost love, but by
helping people develop new ways of relating that
make them more lovable. Although interest and
affection are important ingredients in fostering

behaviors conducive to mutual friendship, it will
be shown later that beneficial results are rarely

achieved by indiscriminate and unguided interest,
however well intentioned.
Change programs based upon social learning

principles differ from those relying heavily on
conversational methods in the content, the locus,
and the agents of treatment. With regard to
content, treatment procedures are mainly applied
to the actual problem behaviors requiring

modification, instead of to verbal reports of
troubles. Change agents therefore devote their
efforts to altering the social conditions producing
aggressive behavior rather than conversing about
them, as so often happens in approaches using the
interview as the basic vehicle for modifying
behavior.
To enhance successful results, treatment is
typically carried out in the natural settings in

which aggressive problems arise. It may be
conducted in the home, in schools, in work
situations, or in the larger community, depending
on the source of the critical determinants. The
same general principle applies even more

importantly in modifying subcultural patterns of
aggression. Group problems demand group
solutions. The most fundamental and enduring
changes in aggression are accomplished by

altering the social instigators and reinforcement
practices prevailing in the deviant subculture.
However, owing to the individualistic bias of

psychological treatments and to the difficulties in
changing the contingency structure of groups, the
common practice is to remove individual
transgressors from their usual environment and
subject them to competing social influences.
Should the offenders adopt the institutional
standards of conduct, they not only lose their

original basis for self-evaluation but, since the new
styles of behavior conflict with what is valued in
their community, they also lose the means of
gaining the approval and respect of their peers.
For these reasons, delinquents ordinarily resist

institutional influences and revert to their
customary patterns of behavior upon their return
to the community. From a social learning
perspective, subcultural delinquents must
develop, through provision of suitable role models

and valued incentives, new modes of behavior that
produce better payoffs than do delinquent

pursuits if more constructive patterns are to be
established and normatively sanctioned.
The third factor that partly determines the
success of a change program is concerned with
who implements the corrective practices. The
most effective treatments are generally carried out
under close professional supervision by persons

who have intensive contact with the aggressor and
can therefore serve as powerful change agents.
Numerous applications will be cited later showing
that, given adequate training, supervision, and
reinforcing feedback, parents can effectively

eliminate the injurious conduct of their children;
teachers can change aggressively disruptive
students into productive learners; nurses can help
violently assaultive patients to gain control over
dangerous behavior; and peers can regulate the

level of aggressiveness in their social interactions.
The potential influence of such persons derives

from their dose association with the aggressing
individuals so that they can exercise substantial
control over the very conditions that govern both

aggressive and prosocial conduct. Unless
influences promoting aggression are altered,
changes produced by interventionists, whether
professional or otherwise, may not endure for

long.
By treating the actual problems in the contexts

within which they arise with influential members
in those settings, social learning procedures are
ideally suited for achieving enduring changes in

psychological functioning. The further one departs
from these optimal conditions, the weaker the
results are likely to be.
Another distinguishing feature of a social

learning approach is the specification of explicit
goals coupled with objective evaluation of the
results remedial practices are producing. Perhaps
because psychological interventions rarely have
immediate spectacular consequences, a casual
attitude toward subjecting people to untested
programs has developed. New approaches are
promoted enthusiastically, and it is not until the
methods have been applied for some time by a

coterie of advocates that objective tests of efficacy
are conducted, if at all. By couching objectives in
high-sounding terms and defining results as
alterations in inner states, high success rates are

claimed for methods that accomplish no
demonstrable changes in interpersonal
functioning (Bandura, 1969a). Now that the
behavioral consequences of social practices are
being measured systematically, some methods

widely assumed to have beneficial value have been
found to actually produce deleterious effects. As

the system of accountability is extended to
psychological services and financial support is

made contingent upon tangible results, ineffectual
methods and the weak theories from which they

are derived should be retired more rapidly.
The initial task in a social learning program
(Bandura, 1969a) is to delineate the desired
changes in performance terms. Unless goals for

action are clearly stated in demonstrable form,
there is no adequate guide either for selecting

optimal conditions or for determining whether the
objectives have in fact been attained. Behavioral
specification of objectives in no way restricts the
range of goals. Change agents may legitimately

pursue abstract objectives, provided they indicate
how the effects their programs supposedly
produce will be manifested. How will people

behave when they are claimed to be self-
actualized, internally integrated, self-accepted,

personally reconstructed, homeostatically
equilibrated, or emotionally matured? When goals
are stated as abstract virtues devoid of

recognizable manifestations, recipients have no
basis for knowing how they will be changed or
whether the agents have delivered what they
promised.
The second step involves a causal analysis of
the troublesome behavior to identify the
antecedent conditions that produce it and the

consequences that maintain it. After ascertaining
the determinants, the changes required to achieve
desired outcomes are instituted. Complex

behavior is best modified by arranging the
requisite learning experiences into a stepwise
sequence that ensures continuous progress.
Poorly organized change programs, as evidenced
by haphazard and inadequate gradation of

experiences, will produce discouraging results,
however valid the principles guiding the social
practices.
Objective assessment of results constitutes
another important part of responsible
interventions. By continually monitoring the
effects produced by a given program, it becomes
readily apparent when the methods are

succeeding, when they are failing, and when they
must be altered to increase their power. This self-
corrective feature not only provides a safeguard
against perpetuating ineffective procedures, but

influences the rate of change itself. Behavior is
generally modified in small incremental steps, so
that individuals may not fully recognize the
progress they are making. Objective feedback of
progress reinforces participants’ efforts and

thereby accelerates successive improvement,
whereas the same interventions prove weaker

when their effects go unmeasured. As outcomes
are monitored during the course of intervention,

necessary adjustments can be made to optimize
results.
Value Judgments in the Modification and Control of
Aggression
In designing change programs it is important
to distinguish between judgments in the empirical

domain relating to strategies for altering
aggressiveness and those in the value domain that
are principally concerned with whether aggressive

actions ought to be supported or eliminated in
given situations. It is appropriate for change
agents to select the means by which particular
outcomes can be best achieved, but recipients of

interventions should play a major role in
determining the directions in which their behavior

is to be modified (Bandura, 1969a). When
beneficiaries select the goals they wish to achieve,

ethical questions frequently raised about the
manipulation and control of human behavior
become pseudo issues.
The general principle stated above is readily
applicable to instances of individual aggression in
which people seek help with their social behavior.
Some highly submissive individuals choose to
increase their assertiveness and ability to cope

with situations requiring force. Normally passive
individuals may overinhibit warranted aggressive
reactions to disparagement or exploitation only to
respond violently under cumulative provocation.
Learning to deal forcefully with lesser affronts

reduces the likelihood of more explosive episodes.
Still others aggress only too readily and seek ways
of controlling their injurious behavior when it gets
them into trouble.
Change agents are by no means value free.

Occasionally individuals may select goals that
change agents have no desire to promote because

the intended outcomes conflict with their values.
For this reason, they do not sell their services to
felons who seek to improve their mugging skills,
nor are they likely to desensitize assaulters to
expressions of suffering to enable them to aggress
without undue distress. On the other hand,
counselors may serve those in power in devising

better ways of aggressing against certain people
when such practices are institutionally sanctioned.
Value conflicts arise when the rewards of
aggression far exceed its costs for the user. Under
these circumstances successful aggressors, or

groups valuing aggressive accomplishment, resist
pressures to modify their behavior from those
who either suffer the negative consequences of
their actions or object to them on moral grounds.

A society has the right to protect itself from
injurious transgressors by imposing legal
sanctions, though it is not granted the

psychological or legal power to compel prosocial
behavior from them. By the same token,
individuals are entitled to adequate protection
against transgressions of coercive power by the
social system. Agencies of social control generally
try to discourage the use of aggression by
increasing penalties and the surveillance of such

conduct. Individuals thus have the choice of
behaving aggressively and taking the
consequences or adopting more socially
acceptable means of getting what they want. From
a social learning perspective, the most effective

way of diminishing aggression without raising
moral dilemmas is to increase the functional value
of prosocial alternatives. By developing rewarding
options, aggressiveness is reduced through
enhancement of individuals’ freedom of choice.
Disputes over the morality of aggression and
its control arise most often in relation to collective

action directed against established social systems.
In such instances, not only is the legitimacy of the
actions contested, but even the victims and
aggressors are difficult to identify. Social
institutions can, through use of coercive power,
control the behavior of dissident members
without having to resort to openly aggressive

measures. Those in privileged positions are not
easily persuaded to make needed social changes
that conflict with their self-interest. As long as
groups that believe the system is biased against
them accept their subordinated position without

unauthorized protest, peaceable order is
maintained. However, when dissidents take
oppositional collective action in extralegal ways
against inadequate or harmful practices of the
system, enforcement agencies respond with

aggressive force likely to provoke physical
violence. Indeed, the counterattacks of authorities

who wield power are often more violent than the
actions of the protesters (Grimshaw, 1969; Marx,
1970b).
Aggressors tend to extol their own violence as
principled behavior but condemn that of their
opponents as morally reprehensible. The
moralizing rhetoric of protagonists usually goes as

follows: Challengers define their coercive actions
as justifiable forms of protest designed to bring
about social change in a corrupt or oppressive
system. Established authorities, in turn, condemn
disruptive activism as impatient resort to

aggressive solutions when alternative avenues of
influence are open to rectify legitimate grievances.
They argue further that many protesters use
aggression for self-aggrandizement rather than as
a selfless means to laudable goals.
Dissidents rebut the democracy argument with
the countercharge that the institutionalized

procedures of influence, while adequate in theory,
do not work in practice because they are
controlled by partisan functionaries of the
inequitable system. Authorities, in turn, respond
that if the enforcement system, or its laws, are
indeed unjust and traditional means to modify
them have failed, the moral solution is nonviolent

resistance including, if necessary, acts of civil
disobedience designed to get the policies judicially
rescinded. However, disobeying laws under these
specified conditions can be justified only if
transgressors do so publicly and willingly accept

the consequences of their unlawful conduct. In this
way specific unjust laws can be challenged while
maintaining respect for the judicial process itself.
It is presumably the suffering endured by
aggrieved protesters that shakes the moral

complacency of compassionate citizens, if not the
oppressors themselves. In theory, this will

mobilize the massive support required to force
political agencies to institute warranted reforms.
However, to demand amnesty for one’s unlawful

conduct not only defeats the purpose of civil
disobedience, but is morally wrong. If individuals
do not have to accept responsibility for their
actions, violent tactics and threats of force will be

readily used whenever a grievance arises. It is
further argued that illegal defiance of the rules of a

representative society fosters contempt for the
principle of democratic authority. Anarchy would
flourish in a climate in which individuals acted on
the basis of private moral principles and

considered coercive tactics acceptable when they
disliked majority decisions. Almost any form of
conduct can be justified in terms of private

conscience. An additional moral objection levied
against the use of force is that it corrupts the user

and dehumanizes the victims.
Challengers refute such moral arguments by
appeal to a higher level of morality derived from
communal concerns that transcend the immediate
political system. Their constituencies are
expanded to include all people, both at home and
aboard, victimized either directly or indirectly by
injurious institutional practices. From the

protesters’ perspective they are acting under a
moral imperative to stop inhumane treatment of
victims who are outside the system and therefore
have no way of modifying its destructive policies.
Even the disadvantaged members within the

system are powerless to influence it in any
significant way. Some are disenfranchised, most
have no voice in decision-making, and even the
judicial procedures are partially administered so
that efforts to remedy legitimate grievances are

repeatedly thwarted. Not only is one not obliged to
obey authorities lacking legitimacy and laws

enacted either undemocratically or selfishly, so
the reasoning goes, but one is morally right to
disobey them. Although nonviolent resistance

does not jeopardize democratic institutions, it is
relatively ineffective when leaders who are not
highly principled secure widespread support from
a populace that benefits sufficiently from existing

policies to remain morally unmoved by seeing a
small faction acting in a civilly disobedient

manner. When self-interest conflicts with common
welfare, self-interest usually prevails. The majority
principle, selfishly applied, thus is claimed to work
against disadvantaged minorities. Submitting to

the punitive consequences of disruptive protest,
the challengers argue, places institutional
procedures above the welfare of humans and

simply allows the system to perpetuate its
exploitation of the disadvantaged. They further

contend that they cannot afford the protracted
time and expenses required to exhaust legal
remedies.
When moral and democratic arguments fail to
dissuade people from using illegal tactics, force is
questioned on practical grounds. People in
privileged positions and those who consider
aggression to be morally wrong, whatever its

source, contend that aggression is tactically
ineffective. Challenging groups, on the other hand,
contest this empirical assertion by downgrading
the value of nonaggressive solutions and pointing
to instances in which others successfully obtain

what they want through aggressive means.
As the preceding discussion shows, one can
easily marshal moral reasons either for or against
the use of collective aggression to force desired

changes. When institutional authorities and
dissidents alike act as aggressors, it is a debatable
question which of the contenders must change its

ways. In some instances, collective aggression can
be effectively reduced only by correcting injustices
in the social system and pressuring advantaged
members to subordinate their interests to the
common good. Other instances require
discouraging dissidents from resorting to
aggressive tactics before alternative means of

influence have been adequately tested. When a
malfunctioning authoritarian system collides with
aggressive malcontents who ascribe moral
superiority to their own personal principles, then
both adversary groups are in need of change.
Traditional prescriptions for the modification

of aggression are typically stated in broad general
terms offering little guidance on the specific
courses of action that ought to be taken. According

to intrapsychically oriented practitioners, one
must find the personal causes of aggression and
interpret them to the client. Environmentalists, on

the other hand, advise that the society must be
changed. What is lacking is a detailed specification
of the procedures best suited for altering
aggression arising from different causes. The
following sections of this chapter discuss how
social learning principles can be used to alter
conditions fostering injurious modes of behavior.
MODELING PRINCIPLES
The power of modeling influences can be
utilized to modify, as well as to transmit,

aggressive patterns of behavior. People will
persist in ineffective conduct when they have not
learned other ways of handling situational
demands. Even partially successful behavior is

often maintained and valued, not because it is
highly preferred, but because better alternatives

are lacking. Aggression can therefore be best
reduced at the individual level by developing more
effective means of coping with interpersonal
problems. There are many different ways of
implementing modeling principles, some of which
are more effective than others.
The method that has yielded the most
impressive results with diverse problems contains
three major components (Bandura, 1972). First,

alternative modes of response are repeatedly
modeled, preferably by several people who

demonstrate how the new style of behavior can be
used in dealing with a variety of aggression-
provoking situations. Second, learners are
provided with necessary guidance and ample

opportunities to practice the modeled behavior
under favorable conditions until they perform it
skillfully and spontaneously. The latter procedures

are ideally suited for developing new social skills,
but they are unlikely to be adopted unless they

produce rewarding consequences. Arrangement of
success experiences, particularly for initial efforts
at behaving differently, constitutes the third

component in this powerful composite method.
After new skills have become functional through
repeated reinforcement, they are habitually used
even though they may not always succeed. Given
adequate demonstration, guided practice, and
success experiences, this method is almost certain
to produce favorable results. The potency of

reinforced modeling derives largely from the fact
that individuals learn and practice effective modes

of behavior under conditions similar to those they
typically face in everyday interactions. An
additional advantage of this form of treatment is
that a broad range of nonprofessional resource
people can be enlisted to serve as role models.
Modeling procedures are used in some

traditional treatment approaches, but their
potential value is negated by enactment of the
wrong behaviors. In efforts to promote cathartic
discharge and insight into neurotic patterns,
clients are prompted to reenact their faulty
patterns of behavior and to revivify the emotional
reactions engendered by their inadequate
functioning. Observing one’s negative behavior
and its likely consequences may temporarily

reduce its use, but it is insufficient to produce
constructive changes. What most people need is
not the insight that they are behaving
inadequately, but the means to learn more
successful ways of behaving. It is understandable

that many clients who find themselves in
treatments oriented toward exposing deficiencies
leave after a few sessions, and those who remain
spend much of their time defending themselves
against negative interpretations of their habitual

conduct. The tendency to accentuate the negatives
is even more prevalent in the modification of

aggressive behavior disorders in young children.
They are typically urged to perform in doll play
and other expressive media negative assaultive

reactions toward parents, teachers, siblings, and
peers that, if transferred to actual relationships,
would further exacerbate their problems.
Chittenden (1942) shows how substantial
benefits can be derived from modeling and
practicing positive solutions to recurrent

interpersonal conflicts. Children who were
excessively domineering and hyperaggressive
observed and discussed a series of portrayed

interactions in each of which dolls, representing
preschool children, exhibited alternately an
aggressive and a cooperative solution to
interpersonal conflicts that children frequently
encounter. In addition to modeling these

alternative modes of response, the consequences
of aggression were depicted as unpleasant and
those of cooperativeness as rewarding. In one of
the modeled situations, for example, two boys
engage in a fight over the possession of a wagon;
during the struggle the wagon is broken, and both
boys end up unhappy. By contrast, the cooperative
alternative showed the boys enjoying themselves
as they take turns playing with the wagon.
Children for whom the different reactions and
consequences were modeled reduced their
dominative aggressiveness in competitive
situations, whereas similarly hyperaggressive
children who received no treatment continued in

their aggressive ways. The differential modeling of
coping styles also decreased domination and
increased cooperativeness as assessed from
behavior observations in the nursery school made

prior to treatment, immediately after treatment,
and one month later (Figure 5.1).
This program relied primarily upon modeling
influences. After desired patterns of behavior are
transmitted through some form of modeling,
whether they will be retained or discarded
FIGURE 5.1 Amount of cooperative and aggressive
domineering behavior displayed by hyperaggressive children
before and after receiving a modeling treatment in which
cooperative coping styles were favored over aggressive ones.
(Plotted from data by Chittenden, 1942)
depends on the results they produce. Hence, it
may be necessary to arrange favorable
consequences to support newly acquired response

patterns until they become well established. An
approach combining modeling with reinforcement
is well suited for achieving enduring changes in
social behavior.
Reduction of Aggression Through Participant

Modeling
Symbolic modeling, in which desired response
patterns are demonstrated concretely through
play activities, may be adequate for modifying the
behavior of young children. In most cases, social
skills are best acquired through repeated trials of

behavioral demonstration, enactment, and
performance feedback until a high level of

proficiency is attained.
As Toch (1969) has shown, people who have a
propensity for getting into fights are usually
deficient in verbal and other social skills for
coping with provocative situations. Socially adept

individuals are able to defuse potentially explosive
situations through pacifying moves, such as face-
saving actions, friendly persuasion, and humor.
Verbal skills not only reduce actions productive of

violence, they can also help preserve one's
integrity and self-esteem in embarrassing
situations without having to dispose of
antagonists physically. Social class differences in
physical aggressiveness partly result because

members of higher strata are taught to talk rather
than to fight their way out of difficult situations.
Socially and verbally unskilled persons, having
limited means for handling discord, are likely to

become physically aggressive on slight
provocation, especially in contexts where violent

conduct is viewed favorably. Assaultive people can
therefore profit greatly from a treatment program
that teaches them nonviolent techniques for
handling interpersonal conflicts. By enlarging

their repertory of skills, aggressors achieve
greater freedom in meeting present and future
problems. The propensity to violence can be
further reduced by changing the valuation of
combativeness to make it signify ineptitude rather

than manliness.
In the development of social skills, new modes
of behavior are first modeled. Nonviolent ways of
handling antagonists can be effectively

demonstrated through role-playing simulating
everyday experiences. Actual demonstrations may
be supplemented with videotaped and filmed

modeling of successful responses to common
problems. To produce skills that can be applied to
many situations, the treatment should emphasize
general strategies rather than specific responses.
Abstraction of a general style from individual
examples is achieved by showing models handling
diverse problems in similar ways.
After nonaggressive modes of response have
been sufficiently portrayed, learners practice them
while being variously insulted, goaded, thwarted,
and threatened in role-playing exchanges. Social
learning is retarded when performers fail to

observe thoroughly the responses they are
making, as is usually the case. Performance
feedback, which provides information for
corrective action, is thus required to realize the

full benefits of behavior rehearsal. Reports of
onlookers may be of some help, but the explicit
feedback provided by audio and videotape

recordings is ideally suited for this purpose.
Through immediate videotape playback learners
can observe their own performances, including
their attitudes, tones of voice, and accompanying
mannerisms. Successful accomplishments are
noted, while aspects of their performance
requiring further improvement are pointed out.

Troublesome situations are then reenacted, with
additional modeling if necessary, until they are
handled satisfactorily. Through role reversal,
participants can experience how their typical
conduct affects others, and the greater

understanding may perhaps reduce the
aggression-arousing potential of antagonistic
reactions in social interchanges.
Gittelman (1965) illustrates how modeling

combined with behavioral enactment can be
adopted for modifying aggressive behavior in
older children. They first describe everyday

situations that typically provoke them to
aggression and belligerence. A hierarchy of
irritating situations is then constructed, ranging
from mild annoyances to enraging provocations.
The participants enact these progressively
aggravating situations and practice effective
nonviolent means of coping with them. When

certain events serve as unusually powerful
elicitors of anger arousal, a stimulus neutralization
procedure may also prove helpful. Herrell (1971)
successfully treated a chronically assaultive adult
who responded with uncontrollable anger to

commands by pairing imagined scenes in which he
was being told what to do with relaxation.
Following the treatment the client, whose
explosive reactions to directions from parents,
teachers, police, and military personnel repeatedly

got him into difficulty, was able to accept
reasonable orders without becoming enraged or

combative. Desensitization methods, which have
proved effective in eliminating anxiety reactions,
warrant further exploration as a means of

neutralizing aggression elicitors.
Individuals may be reluctant to discard
aggressive modes of response as long as such
behavior remains personally valued. Cultivation of

social competence may have to be accompanied by
devaluation of assaultiveness. In altering a

person’s basis for self-evaluation one must, of
course, consider the ethical issues involved.
Anyone who sets out to modify standards of self-
reinforcement should be prepared to teach new

styles of behavior that serve as more self-
satisfying alternatives. The self-evaluative
consequences of aggressive behavior can be

altered to some extent through the judgments
expressed and modeled by change agents. It might

be pointed out, for example, that the risk of
permanent physical injury is an exorbitant price to
pay for fleeting glory or a temporary boost in self-

esteem; that it is much smarter to dispose of an
antagonist by one’s wits than by one’s fists; and
that hasty fighters become easily manipulable
after others discover how to provoke them to
foolhardy acts. An assaultive person is apt to
exercise better control over his behavior if he also

knows which things easily goad him into violent
action. This knowledge is achieved through
behavioral analyses of the social determinants of

aggression in any given case.
Neither preaching nor example will be fully

convincing unless backed up by corroborating
rewarding experiences. Success endows activities
with positive value (Breer and Locke, 1965). A
number of methods can be used to make the

control rather than the expression of aggression a
source of self-pride. In one approach, described by
Mahoney (1971), hyperaggressive boys each take
turns being harassed and are rewarded in terms of
how well they maintain their equanimity. One
could similarly devise confrontation situations in
which aggression-prone individuals are rewarded
for tactful pacifying moves and adroitness of
repartee.
When people are deliberately taught new ways
of behaving, there may be a tendency to view the
resultant changes as somehow artificial. Any new
skill, regardless of how it was acquired, is initially
performed awkwardly and self-consciously. With

repeated use, however, the actions are executed
spontaneously as part of one’s natural style of
behavior. But participant modeling does more
than teach skills. If people behave in new ways,

eventually their attitudes change in the direction
of their actions. Indeed, numerous studies have
shown that one of the most effective methods for

altering attitudes and values is by producing a
change in behavior (Bandura, 1969a; Elms, 1969).
Negative self-attitudes partly result from
behavioral deficits and are repeatedly reinforced
by failure experiences in handling common
situations. No amount of self-exploration in

therapeutic conversations will create esteem-
producing vocational skills or social competencies
required to support realistic positive self-
evaluations. Here the primary concern must be

with self-development rather than self-
examination. With achievement of competencies
that bring success, people revise their estimates of
themselves in more favorable directions.
Participant modeling, though primarily addressed

to the acquisition of skills, improves self-
evaluations as well. Without the requisite

competencies people can be talked into thinking
well of themselves, but positive self-evaluations

are unlikely to survive for long in the face of
disconfirming failure experiences.
Development of Assertiveness by Participant
Modeling
Some forms of physical aggression result
paradoxically from a lack of self-assertiveness.

One can easily call to mind obsequious individuals
who invite maltreatment through their passivity,
only to respond explosively after being subjected

to repeated humiliating affronts. These are the
timid, indecisive people who cannot express their
legitimate rights; who are disregarded, exploited,
and victimized; and who harbor resentment rather

than seek redress for justified grievances. People
who are unable to behave assertively will suffer

considerable aversive control by others, some of
which may be of a sufficiently provocative

character to force violent response. Under high
states of emotional arousal, the risk of hazardous
behavior is increased because anticipated
response consequences, which ordinarily deter
reckless actions, are given little consideration.
Venting anger and annoyance alone has little
lasting benefit. Rather, a submissive person needs
to develop assertive responses that enable him to

control his social environment in ways that stop
humiliating treatment. The modeling principles

used to foster nonviolent responsiveness apply
equally to teaching people how to behave
assertively. Assertive modes of response must be
modeled for the submissive. They must practice

assertive styles with feedback that rewards their
successes and corrects their errors. They must
learn when and where assertiveness is

appropriate. And they need positive reinforcement
for asserting themselves in their daily interactions.
Some efforts have been made to assess the
relative contributions to treatment outcomes of
the various components in participant modeling.
Friedman (1972), for example, compared the
effectiveness of different forms of modeling with
and without guided practice for increasing
assertive behavior in passive college students.
Different groups had assertive response to

provocation modeled in the form of either verbal
descriptions or actual behavioral demonstrations,
or they relied on reconstructed examples of
assertiveness as guides for their behavior. Half the
students in each of these treatments also received

guided practice in behaving assertively toward an
irritating associate. Untreated controls remained
passive in the face of provocation. Behavioral
modeling supplemented with guided practice
proved to be the most powerful treatment; it

produced approximately a triple increase in
assertive behavior. The remaining procedures,

which doubled the students’ level of assertive
responding, were found to be equally effective.
The above study set out to produce changes in
a circumscribed range of assertive behavior. In
formal treatment applications, of course, a number
of assertive skills would be modeled and
practiced. Depending on individual needs, these

might include such things as complaining about
inadequate service, returning purchases, refusing
arbitrary or unreasonable demands, responding to
unfair criticism, making rightful claims to goods
and facilities, defending one’s position in the face

of opposition, and in other ways standing up for
one’s rights.
Fear of revealing one’s inadequacies creates
initial reluctance to engage in behavior rehearsals

even under simulated conditions. There are
several ways in which such resistances can be
reduced. First, participant modeling is structured

in a nonthreatening manner aimed at fostering
new competencies and confidence rather than
exposing weaknesses. Concern over poor
performance is further decreased by noting that in
all successful skill learning initial efforts are
awkward; it is progressive improvement with
practice rather than instant proficiency that is

expected. In an optimistic, forward-looking
program, people more readily accept and profit
from their mistakes. Second, prior modeling
provides a helpful guide for new styles of
behavior. Needless failure is avoided because

participants do not have to grope around for
appropriate responses. Third, assertion tasks are
graduated, beginning with relatively easy
performances. As anxiety is extinguished through
repeated practice, progressively more threatening

encounters demanding more assertive actions are
introduced.
Transfer of assertiveness from the training
situation to the natural environment should be an
integral part of the program rather than being left
to fortuitous circumstances. A transfer program
might proceed as follows: After the person has
overcome his timidity, he accompanies the model
on field excursions in which he witnesses further

demonstrations of how to handle situations calling
for assertive action. The change agent then
reduces his level of participation to background
support and guidance as the client tries his newly
acquired skills under conditions likely to produce

favorable results. By careful selection of
encounters of increasing difficulty, the assertion
requirements can be adjusted to the client’s
capabilities at any given time and so bolster his
sense of confidence. As a final step, the client

would be assigned a series of performance tasks
requiring assertiveness to carry out on his own.
Reduction of Aggression by Law Enforcement

Agencies
Toch’s (1969) studies of individual aggression
provide ample evidence that violent incidents
often are partly created by police through their
inept handling of people who have brushes with
the law. The contribution of police violence to

collective aggression is equally well documented
in analyses of past and more recent civil
disturbances (Kerner et al., 1968; Marx, 1970b).
Urban riots are almost invariably set off by an
incident involving harsh or humiliating police

action against a ghetto resident. Spectators,
sensitized to maltreatment by police harassment
and unredressed grievances, retaliate against
police and ghetto merchants singled out for their
exploitive practices. Police forces under weak

command control who are brought into a rampant
situation are readily susceptible to aggression

contagion. The anonymity afforded by concealed
identity and the cover of darkness further reduces
restraints against their behaving cruelly. When

police violence gets out of hand, indiscriminate
counterfire and brutal attacks on rioters and
bystanders alike aggravate the riotous activities
the police were sent in to quell.
Because police must deal with individuals
prone to aggressive action and because misuse of

coercive power can easily turn innocuous
incidents into violent episodes, prescriptions for
the curtailment of violence must address

themselves as well to provocative aggression by
law enforcement agencies. The most effective way
of controlling unwarranted police aggression
depends on whether it arises from individual
ineptitude or social sanction.
At the individual level, police who recurrently
provoke violent altercations respond to suspected
transgressors with precipitous force that invites
defiance from those who consider such blatant
display of authority as humiliating and unjust
harassment. The frontal approach eventually
provokes fighting in belligerent transgressors
(Toch, 1969). Policemen who are prone to

provocative aggressive actions must be taught
how to cope more constructively with potentially
dangerous situations. This requires tactful
routines for determining whether there is any

cause for police action. Given that a transgression
has occurred, the law can be better enforced,
without risk of inciting assaultive reactions, by
pacifying moves than by excessive application of
force. Not only is the probability of police-

generated violence reduced when aggressive
measures are used as last resorts, but judicious

police conduct reinforces responsiveness to their
authority that generalizes beyond the immediate

event.
The modeling procedures outlined earlier for

training in nonviolence are equally well suited for
police training. Skill in handling transgressors
cannot be ensured by lectures and trial and error
experiences on the street. Indeed, since most

people acquiesce to police authority, forceful
means are likely to be reinforced and selectively
developed if police are left to their own devices.
Desired styles of conduct can be reliably
established through demonstration and reinforced

practice conducted initially under simulated
conditions and later in field situations. After adept
response patterns have been adequately modeled
by exemplary police officers, recruits are given

ample opportunity to perfect them through
enactment of difficult situations they will
encounter on duty. Videotape techniques provide

a convenient means of both illustrating alternative
courses of action and recording for analysis the
conduct of trainees in simulated situations as they
are called upon to secure identification and
information from suspects, to mediate disputes,
and to pacify, disarm, and arrest offenders. Since
police are most prone to misuse force when their

authority is challenged, special attention should be
given to constructive ways of handling recalcitrant
individuals to reduce the likelihood of escalating
aggressive exchanges. During the performance
analysis, adept responses are socially rewarded,

while offensive and overaggressive ones are firmly
discouraged. The process of demonstration,
enactment, and corrective feedback is continued
until the desired practices are performed skillfully
and spontaneously. In order to add meaning to the

formal instruction, recruits might accompany
experienced officers into the field as observers

during the period.
Police training must equip recruits with more
than technical and interpersonal skills. They also
require supervised practice in how to exercise
discretionary power. A policeman on the beat
must judge whether a given conduct performed
under involved circumstances violates any one of

many statutes. Judgmental errors are fraught with
danger because it is disputable accusations that
provoke the challenges to police authority most
likely to end in violence. Police require detailed
regulatory guidelines defining the types of

behaviors that are forbidden by law. In training for
judgmental accuracy, recruits might respond to
descriptions of problematic situations and be
informed, in each instance, whether they
succeeded or erred in identifying transgressions.

Offense recognition training should be continued
until each recruit understands the range of

conduct encompassed by the legal codes he is
expected to enforce.
In addition to establishing social and
judgmental skills, a police training program must
instill sensitivity to the civil liberties of citizens.
Judicial systems are designed to safeguard rights
as well as to control crime. The balanced view of

law enforcement is conveyed not by lofty rhetoric,
but by the operating practices of the police
department. Values of justice must be embodied in
concrete guidelines for responsible conduct and
procedural rules that make explicit what laws do

not permit policeman to do. Unless behavioral
guidelines are backed up by sanctions, they will
have no lasting impact on police practices in the
field. The behavioral enactments and judgmental
tests used in training sessions should include

incidents that raise constitutional issues. In this
way, recruits can be sensitized to the conflicts they

will repeatedly experience between the efficiency
of disallowed methods and citizen rights.
The judicial aspect of training deserves serious
consideration in view of evidence that policemen
are disposed to make arrests partly on the basis of
their own moral codes (Reiss, 1971). In effect,
officers make a judgment of guilt influenced by

such factors as mitigating circumstances, the
public visibility of the offense, the officers’ own
personal standards of reprehensibility, and the
attitudes of the offender. Many offenders are not
arrested for unlawful conduct, whereas

disrespectful individuals, or those whose
transgressions violate the policeman’s sense of
justice, run a high risk of ending up in jail. Training
and supervisory effort should be aimed at
reducing the partiality of initial determinations of

criminality.
After neophytes have demonstrated sufficient

competence in the preparatory instruction, they
need supervised field practice. Through modeling
and differential reinforcement by supervisory
officers, recruits are aided in handling actual
incidents proficiently while performing their
assigned duties. Since police quickly learn that
they can extract obedience more easily through

force than through tact, periodic monitoring of
their performances is necessary to safeguard
against abuse of their coercive power. Changes in
laws and in societal standards of conduct may
necessitate retraining from time to time.

Considering the personal hazards and serious
social consequences of police actions, it is
remarkable that few police departments combine
comprehensive recruit training with supervised
field experience and periodic performance checks.

A police agency that adequately trains and
oversees its officers and that views harsh practices

disapprovingly should not be plagued by
provocative police practices.
Police aggression is much more difficult to
control when such behavior is organizationally
condoned. Because regional enforcement agencies
are linked to local politics through appointment of
the police chief, police forces tend to serve as

partisan instruments of power interest groups. As
Lohman succinctly puts it, “The police function to
support and enforce the interests of the dominant
political, social, and economic interests of the
town" (Neiderhoffer, 1967). It is not uncommon

for police to enforce the law differentially against
rich and poor, against minorities and those in the
majority, against superiors and subordinates, and
against those who follow conventional and
unorthodox life styles (Tieger, 1971). That police

do not function impartially is most clearly
revealed in their handling of collective protests

against dominant interest groups. When
dissidents challenge the traditional power system,
as in the case of civil rights workers and antiwar

demonstrators, police often become assaultive in
the guise of curbing lawlessness (Marx, 1970b).
Police attacks on unpopular targets may not only
be viewed permissively within the police system,

but may receive widespread public support
(Blumenthal et al., 1972; Gamson and McEvoy,

1970). Societal antipathy toward dissenters
obviously does not foster police restraint in the
use of coercive power.
From time to time investigators report

evidence of other forms of police misconduct
which shows that enforcement agencies do not
police themselves very well. Observers who have
accompanied patrolmen on their daily rounds

report that, although most perform their official
duties properly, a substantial number engage in
criminal conduct or behave abusively without
provocation (Reiss, 1971). That some police
transgress is hardly surprising considering they
work under conditions that tax both restraint and
virtue. Discretionary power is easily misused,
especially when there are no witnesses to the
social confrontations. Should the propriety of

injurious actions be questioned in any given
instance, it can always be justified as needed to
apprehend defiant wrongdoers or to restore
order. As an additional self-protection, the
seriousness of arresting charges can be scaled to

excuse unwarranted assaults. In situations of
physical jeopardy, it is of course difficult to act
with restraint. Apart from inducements to
violence, there are many entrepreneurs of
unlawful activities who are eager to provide

policemen with attractive income supplements for
protective services. Those who shun illegal

subsidies have ready access to merchandise that
can be pilfered with little chance of detection. It is
to the credit of most policemen that they maintain

their integrity under the circumstances.
Nevertheless, transgression rates are sufficiently
serious to require better regulatory mechanisms.
A citizen feedback system can function as a
significant deterrent to police misconduct. Unless
one knows how the clients on the street are being

treated, corrective measures are unlikely to be
instituted. Although to err is understandable, the
lack of effective means by which citizens can

control police misconduct is inexcusable in an
open society. It is amply documented that police
review procedures, where they do exist, serve
more to discourage than to adjudicate grievances
(Chevigny, 1969; Task Force Report, 1967). To

begin with, people are dissuaded by fear of
retaliation because they must lodge complaints
through the offending agency. Should they choose
to pursue the matter, the grievance procedures are
inadequately publicized. Of the complainants who
eventually find out where to go, many are
discouraged, either by intimidation or
bureaucratic harassment, from filing their
complaints. The record of adjudication of

registered charges is perhaps the most serious
indictment of the internal control system. Despite
numerous charges and objective evidence that
police infractions are common, in virtually all
cases the policemen are exonerated of any

misbehavior. Granted that many allegations are
unwarranted and that corroborating evidence is
sometimes difficult to obtain, nevertheless it
strains credibility that patrolmen are that
exemplary and complainants are that malicious.
It is not difficult to find reasons why policemen
are averse to indicting themselves for

delinquencies. Police transgressions are
frequently witnessed by other patrolmen. Informal
codes of the police subculture, however, prohibit
informing on, or testifying against, fellow officers.
The threat of peer sanctions is strengthened by
promotion policies that tend to affect the quality of
enforcement service as well. Advanced positions

are filled by elevation within the ranks of the
department, not by open competition for the best
candidates. Barring lateral entry of personnel
breeds conformity throughout the administrative
hierarchy. Members of a closed system can ill

afford to jeopardize their welfare by revealing
infractions that many of their associates—and line
officers, for that matter —may themselves have
committed. Although police officials try to ferret
out corruption in their ranks, few are apt to get

exercised because of overzealous use of coercive
force. On the contrary, policemen generally resent

the legal restrictions imposed on the methods they
can employ. In the private police ethic, the use of
violent means to apprehend offenders is not

viewed as a transgression. They are thus not about
to indict fellow officers for roughing up
argumentative suspects or infringing on the rights
of dissidents. With law enforcement a major

political issue, police administrators are quick to
defend their officers in order to protect the public

image of the department.
Overaggressive police behavior that is
organizationally sanctioned is unlikely to be

changed much by education, attractive salaries, or
community relations programs. When police
conduct is inadequately regulated from within,
external influences are required to modify the
policies of enforcement agencies. Responsible

police conduct is best assured by holding
patrolmen and their supervisory staff accountable
for their actions, not only to vested interest
groups, but to representatives of diverse segments
of the community as well. However, insiders
vigorously oppose having their actions reviewed
by outsiders. And those most subject to police
abuse unfortunately lack the political strength to
institute monitoring procedures. Attempts to

moderate police behavior therefore usually take
the form of retaliatory threats and an occasional
public expose of police brutality that goes
unpunished.
Efforts to improve police practices, like other

endeavors at social reform, require a multifaceted
approach. At the organizational level, changes in
promotion policies allowing appointment of
officials from the outside as well as from internal

ranks can partially check perpetuation of police
subcultures that subvert the purposes of the law.
Corrective actions require an apparatus for

assessing the performance of police departments
by citizens outside the organization. If a
monitoring system is to have much impact on
police practices it must perform a broader
function than merely processing complaints of
police abuses. As Chevigny (1969) demonstrates
convincingly, the individual case approach has

limited value. Abusive police behavior is shielded
by standard charges of disorderly conduct,
resisting arrest, and felonious assault. Because
allegations are difficult to prove, the unwieldy
hearings rarely produce any disciplinary action.

But although individual charges of misconduct can
easily be dismissed, the accumulated complaints, if
carefully evaluated, can reveal the types of
transgressive behavior condoned by police mores
and by the leading offenders in the system.

Observers who have studied the problem in depth
(Chevigny, 1969; Reiss, 1971) are of the opinion

that complaint data should be used to change the
regulations of police departments in ways that
curtail future abuses. Since the credibility of

complainants can always be challenged, such
information might supplement broader periodic
assessments that include interviews to secure
citizens’ reactions to the police services they are

receiving. Community surveys should measure
commendable performances as well as grievances

arising from partiality or inaction in police work,
from abuses of coercive power, and from
controversial enforcement procedures. Public
disclosure of current practices, along with annual

trend data for different precincts, can exert
pressure on police administrators to improve their
field operations.
Judicial reviews provide an additional (though

admittedly limited) measure of police
performance that might be used in a formal way to
correct unauthorized practices (Task Force
Report, 1967). Courts can furnish a record of
instances in which evidence was excluded because
police behaved illegally. A high incidence of
exclusionary rulings suggests that the
departmental guidelines, the discipline, or both
require administrative attention. Police, of course,

exercise informal control over conduct that does
not violate legal codes and is therefore not subject
to judicial review. When the values shared by
residents in a given neighborhood conflict with the
standards of morality espoused by the policing

agency, disputes over personally disliked behavior
tend to get transformed into crimes against
patrolmen. Some of the value clashes could be
diminished by having community representatives
participate in the formation of policies for police

conduct toward the public they are serving.
It would be a mistake to assume that police

abuses arise solely from malfunctions in
enforcement agencies. A society gets the type of
policing it wants. If a majority of citizens feel
sufficiently threatened by social protest or deviant
styles of behavior, they will condone unauthorized
means of controlling the troublesome individuals.
However, police can be less easily used as

instruments of coercion if their performance is
subject to public scrutiny than if it remains
unexamined.
Contribution of the Mass Media
Various survey studies have shown American

television to be a rich cultural source of violence.
Gerbner (1972a, b), who has conducted the most
sophisticated yearly analyses of televised violence,

reports a stable violence rate of more than seven
assaults per hour, with approximately 80 percent
of dramatic programs containing some form of
violent activity. Seasonal comparisons disclose

some fluctuation in violence rates for the different
networks and program categories, but no evidence
of overall decline in this respect.
Evaluation of the impact of the mass media on
individual and collective aggression has been
impeded by fear that, should a relationship be

demonstrated, such findings might prompt
governmental regulation or even censorship of

program content. As a result, many social
scientists were quick to proclaim the
innocuousness of televised influences long before
there were any pertinent empirical data, and some

continued to defend its impotence in the face of
cumulative evidence to the contrary.
One prevalent disclaimer is couched in terms
of multiple determination of behavior. Aggression,

it is said, does not arise from any single source. To
indict television as the cause of aggression is to
find a convenient scapegoat for a complex
problem. The force of this argument is usually
bolstered by noting parenthetically that at one
time some people (without supporting evidence, it
should be added) imputed the roots of brutal
aggressive acts to comic books. Since no behavior
is ever completely controlled by a single

psychological determinant, this type of argument
applies equally to any other influence that might
be invoked. It could be similarly contended, with
justification, that broken homes, rejecting parents,
impoverishment, oppression, and provocative

police are not individually the sole and invariable
causes of aggression. One does not advance
understanding of media effects by posing
nonsensical causal assertions that call for negative
answers. Platitudes to the effect that televised

influences operate through a variety of mediating
factors are widely quoted in the literature in this

field as incisive causal analysis. Obviously,
portrayal of violent methods is neither a sufficient
nor a necessary condition of aggression. The same

can be said for every other determinant of
aggression.
Closely related to the multiple cause disclaimer

is the argument of secondary influence. Here, the
possibility that modeled aggression might have
some effect on viewers is acknowledged, but it is

regarded as a relatively minor factor. The criteria
by which such judgments are made unfortunately
remain unstated. Subjective weightings of

determinants actually have little meaning because
the same stimulus can have weak or powerful
effects for different individuals, and even for the
same individual on different occasions, depending
on the presence of other aggression inducements.

The value of averaged influences calls to mind the
apocryphal nonswimmer who drowned while
crossing a river that averaged 4 feet in depth. To
assign a general value to televised influences
simply obscures their role in causal processes. A
combination of influences is typically required to
produce a violent act that otherwise would not
have occurred if any of the contributing elements
were missing. When behavior is evoked only by a

particular set of coexisting conditions,
determinants that prove ineffective alone or even
in partial conjunction each play a decisive role in
the multiform cause. As we have already observed,
media influences have differential impact on

different forms of individual and collective
aggression.
Another common misconception is that
aggressive modeling affects only disturbed people.

This belief rests on the unsupported foundation
that aggression is a manifestation of an emotional
disorder. Since the presence of an “emotional

disorder’’ is partly inferred from a person’s violent
actions, traditional explanations have a built-in
circularity that reinforces such beliefs. Even a
casual review of value judgments and social causes
of aggression quickly dispels the notion that
aggression is sick behavior generated by an
internal pathology. A social system gets its

members to aggress, whatever their personal
makeup, by legitimizing, modeling, and
sanctioning such behavior, not by inducing
emotional disorders in them. For this reason, one
would not contend that broadcasters make heavy

use of violence because they are emotionally
disturbed, have a weak sense of reality, or are
homicidally predisposed. Aggression is typically
used because it is successful in getting people
what they want. Although the industry is quick to

attribute aggressive modeling to defects in
children and parents, it doubtless ascribes

increases in consumer behavior to the power of
the medium rather than to pathologies in the
viewers.
A popular variant on the dispositional theme is
that only individuals already predisposed to
aggression are susceptible to televised influences.
In this view, predisposition is misconstrued as a

property inherent in the individual when, in fact, it
is a multifaceted determinant. First, dispositions
take the form of response capabilities. People who
have learned aggressive styles of conduct are
more disposed to behave aggressively than those

who lack such responses. Second, predisposition
reflects self-control functions. Individuals who
expect to derive rewards or self-satisfaction from
combativeness are more inclined to aggress than
those who anticipate punishing or self-

condemning consequences. The third and most
important feature of dispositions is concerned

with how the behavior is socially sanctioned in the
environment. People are more disposed to act
aggressively in settings where such conduct is well

received than in situations in which it is
prohibited. Considering the fact that dispositional
determinants operate through multiple interacting
influences, it is misleading to depict aggression as

primarily under internal dispositional control.
In evaluating the assumption of restrictive

influence, one must distinguish learning from the
performance effects of televised violence.
Obviously, one does not have to be emotionally

disturbed or hyperaggressive to learn aggressive
tactics observationally. Mild-mannered children
learn as much from aggressive example as do
aggressive children. Girls, who ordinarily are less
disposed to physical aggression than boys,

nevertheless model patterns of aggression equally
well, though they are less inclined to put them into
practice unless given incentives to do so. To the
extent that aggressive people spend more time
watching televised violence, they have more
opportunities to be affected by it. Under these
circumstances, however, the effects of personal
aggressiveness are mediated more through
variations in self-exposure to violence than

through differential susceptibility to such
contents.
Aggression proneness is better reflected by
spontaneous performance of modeled acts of
aggression, but even here the differences are

smaller than might be expected. Although modeled
brutality more readily prompts authoritarians
(Epstein, 1966) and those who have a history of
assaultiveness or criminal offenses (Wolfe and

Baron, 1971; Hartmann, 1969) to behave
punitively, it can also increase cruelty in those
who are less prone to harshness in their daily

interactions. Contrary to what is frequently
asserted, nondisturbed children are just as likely
to imitate modeled forms of aggression as are
emotionally disturbed children who have been
hospitalized for severe “personality” disorders
(Walters and Willows, 1968). Through the
teaching and disinhibitory functions of symbolic

modeling, television can create as well as activate
“dispositions” to aggression.
It is often claimed that displays of violence may
affect children because they are less able to
foresee the full consequences of their actions, but

that adults, who possess a greater reality sense,
are relatively immune to such influences. This
opinion is also refuted by numerous studies
demonstrating that intelligent and otherwise

considerate college students behave more
punitively after seeing others act aggressively than
if they have had no exposure to such modeling

influences. Adult response to modeled violence is
further shown in the rapid spread of new styles of
collective aggression to people in widely dispersed
areas who learn about them mainly through
detailed media coverage. Nor are adults entirely
unresponsive to fictional violence. It is adult
rather than child extortionists who threaten to

blow up airliners with altitude-triggered bombs in
a manner copied from a televised plot. There is
every indication that the actions of adults are
shaped by example just as much as those of
children. Fashion and taste industries flourish on

the basis of adolescent and adult modeling.
Like many other issues in aggression, the
impact of televised violence on viewers is better
understood in terms of social influence than

characterological factors. Although it is true that
the psychological attributes of viewers can affect
the likelihood that modeled aggression may be

performed on future occasions, social conditions
that increase the permissibility and functional
value of aggression easily override the effects of
personal dispositions. For this reason, it is
primarily types of social inducements rather than
types of people that should be examined in
predicting who will put into practice what has

been learned from television.
The familial antidote argument also appears
with some regularity in commentaries on mass
media effects. By providing a warm, secure home
and wholesome peers, parents, it is claimed, can

neutralize any potential negative impact of
repeated exposure to human brutality.
Fortunately, the medical profession never adopted
an analogous stance in promoting the physical

health of citizens. Had parents been told by
specialists in preventive medicine that the best
way to counteract any ill effects of contaminated

water and milk supplies delivered to their
households was to raise healthy children, the
competence of such authorities would have been
seriously questioned. If parents cannot create a
counterenvironment, which is no easy task, they
are told they should control what their children
watch on television. To be most effective, parental

monitoring requires advance information on the
quantity of carnage contained in given programs.
Parents would be greatly aided in this function if
networks and TV guides reported the violence
rates in the various offerings.
The antidote recommendations, even if they

could be easily implemented, would not
necessarily provide adequate safeguards against
imitative assaultiveness. Aggressive acts, like

other forms of behavior, are not exclusively under
parental control. Children may refrain from
imitative aggression in the presence of adults who

disapprove of such actions, but perform them
when the adults are no longer around (Hicks,
1968a). Interpersonal aggression is a social not a
private matter. Nonviewers can be secondarily
affected by television either by emulating the
conduct of influenced viewers or as victims of
their heightened aggression. In the controlled field

studies, it is nonviewers who are physically
assaulted by those exposed to aggressive models
of behavior. In the case of adults, it would be far-
fetched, for example, to claim that modeled
extortion behavior toward airlines is mediated

through maternal influences. Moreover, under
certain conditions, parental warmth and
compassion may facilitate rather than counteract
aggressive modeling in their offspring. Students
raised in principled, humanitarian homes were the

most disposed to follow models of coercive action
in attempts to rectify social injustices.
Disavowals of the aggression-promoting
effects of televised modeling are usually
accompanied by enumeration of the positive
effects of displaying brutality. The cathartic
function is listed as if this were a documented fact
rather than a questionable belief. Even if there
were evidence to show that a distinct aggressive

drive existed and that watching someone commit
physical violence upon others drained it, the
wisdom of saturating the airwaves with displays
of violence would still be questionable, if only
because of their instructional function. Such

sluicing operations would achieve, at most,
temporary drainage at the price of teaching new
techniques of cruelty, devaluing life by treating
human beings like disposable commodities, and
sanctioning violent solutions to interpersonal

problems. A social influence must be judged on the
basis of the multiple effects it can have on people.

Concern with empirically demonstrable effects
should not completely overshadow the morality of
using cruelty to entertain people.
Material reviewed extensively in preceding
chapters reveals that many vigorously debated
issues are reasonably well clarified by controlled
studies and corroborated by recurring modeling

incidents in everyday life. The cumulative
evidence provides answers to certain questions
about the effects of television and increases the
validity of probabilistic judgments of other effects.
Given the abundant evidence for observational

learning, there is no longer any justification for
equivocating about whether children or adults
learn techniques of aggression from televised
models. People who watch commercial television
for any period of time will learn a number of

aggressive tactics and countless methods of
murder. Television is a superb tutor. People do

not, of course, perform everything they learn.
Substantial progress has been made in identifying
some of the factors that interact with modeling

influences to foster or restrain aggressive actions.
Individuals who have been exposed to displays of
violence are more likely to act aggressively and
with greater intensity than those who have not

when they are subjected to provocation or
emotional arousal; when they have received prior

training in aggression; when adequate justification
for aggressive conduct is provided; when threat of
retaliation is reduced; when the injurious
consequences of aggression are not fully evident;

and when aggressive behavior has functional
value in securing material rewards, social
recognition, a boost in self-esteem, or successful

retaliation against disliked antagonists.
Television does not control the aggression

instigators experienced by its viewers or the
reinforcement supports their environment
provides for such conduct. Rather, it influences
their aggressive tendencies primarily by teaching
them how to aggress and by the way in which it
portrays the functional value of coercive behavior.
If violence were presented as a morally
reprehensible and costly way of achieving desired

goals, viewers might be discouraged from
resorting to similar tactics should the temptation
arise. Analyses of program contents show,
however, that television is more inclined to
condone than to devalue violent solutions.
Depicting punishing consequences for violence

can reinforce aggression restraints in viewers,
although it is an undependable inhibitor. Observed
punishment may suppress, but does not erase,

what viewers have learned. They may put this
knowledge into practice on future occasions if
sufficiently provoked or if the behavior is

necessary to get what they want. Dramatic
requirements further reduce the inhibitory
potential of portrayed punishment. In the
development of a plot villainous characters
appropriate important resources and amass
attractive rewards through variegated series of
aggressive exploits. If they never succeeded, it

would be difficult to sustain a plot. Thus, for
example, through physical coercion protagonists
gain control over grazing lands, gold mines, steers,
nightclubs, attractive belles, the constabulary,
corporations, and prominent figures. Aggressors

are not usually punished—and then only once—
until the end of the program, whereas their
antisocial life styles prove highly profitable over
an extended period. The message is not that
“crime does not pay,” but rather that the wages of

violent sin are pretty good except for an occasional
mishap.
It has been demonstrated that immediate
rewards exert greater influence on behavior than
delayed punishment. Moreover, observed
successes weaken the inhibitory power of
witnessed punishments wherever they may occur
temporally (Rosekrans and Hartup, 1967). When
commercials are inserted before the ethical

ending, as is generally the case, young children fail
to grasp the relationship between actions and
outcomes and are therefore not dissuaded from
adopting aggressive solutions (Collins, 1973).
Finally, in some instances depicted punishment

can sustain adherence to aggressive methods
through its informative value. To the extent that
observers profit from the mistakes of others, they
may act on the belief that, with slight modification
of tactics, they can gain the benefits of aggression

without suffering the costs. For these various
reasons, it is better not to teach violent methods

than to teach them and try to inhibit their use.
Affixing an ethical ending to a succession of
aggressive triumphs in no way removes the

enduring learning effects.
There has been much speculation but
comparatively little research on the emotional and
attitudinal effects of extended exposure to

violence. Concerns have been voiced that viewing
brutality day after day and week after week

eventually produces callousness and insensitivity
to cruelty. Repeated exposure to displays of
violence can gradually extinguish emotional
responsiveness. The power of filmed influences to

diminish fears and behavioral inhibitions is most
convincingly demonstrated in therapeutic
applications of modeling. Phobic children lose

some of their fears (Bandura and Menlove, 1968)
and seclusive children overcome their social

inhibitions by observing venturesome examples
(O’Connor, 1969, 1972). As in other areas of
functioning, adult inhibitions are susceptible to

modification by modeling influences as are those
of children. Adults who have suffered from severe
phobias of long standing lose their fears and
develop more favorable attitudes toward things
they formerly abhorred after watching scenes of
others interacting freely with threatening objects

(Bandura, Blanchard, and Ritter, 1969). Results of
these and other studies (Bandura, 1971a) attest to
the positive potential of televised influences, were

the medium to be used more constructively.
Of special interest are the behavioral effects of

removing fears and aversions. Aggression anxiety
inhibits a person from engaging in abhorrent
activities. After aversions have been extinguished,
he can more easily commit such acts, but whether

or not he ever will do so depends on the social
influences that impinge on him. This issue has
some bearing on controversies over the
cumulative effects of daily portrayals of combat
activities in television news. Some people
maintain that such scenes elicit shock reactions
that condition aversion to war. Wertham (1966), a
vigorous crusader against displays of violence in
any form, contends that as a result of daily viewing

of fragmented military scenes, people eventually
become accustomed to warfare and indifferent to
the suffering it causes. It therefore produces
receptivity to war rather than revulsion against it.
He argues further that most movie versions of

warfare, with their glorification of patriotic
violence, heroism, and adventure, are essentially
war commercials. Through repeated exposure
people can become emotionally habituated to
depictions of horrific carnage (Hess, 1968). It

remains to be demonstrated, however, whether
the emotional desensitization leads them to accept

or to abhor war. Emotional reactions and
behavioral preferences are not perfectly
correlated. That is, one can fear things he likes and

deplore things he does not fear.
Ordinarily one might expect a widespread
public outcry against corporations that provide
continuous instruction in methods of murder for

all ages and sanction aggression as a means of
attaining goals or settling disputes. In reality, most

of the viewing public have not only grown
accustomed to a heavy dose of brutality on
television, but having no comparative alternatives,
accept what they are shown. Apart from its

entertainment value, television serves as a most
welcome baby sitter when parents want their
children out of the way. If exposing children to

human cruelty is the price for this function, then
most preoccupied parents are willing to pay it,

especially if they are reassured that it has no
influence. Audience preferences partly shaped by
television then serve as justification for similar

offerings that produce further habituation to
violence.
In contrast to the ready acceptance of televised
violence by American viewers, representatives of
other nations respond with incredulity to

disclaimers that televised modeling has no
influence and voice concern over the export of

American violence abroad (Lejins, 1968). The
amazement and uneasiness shown by other
nations is well grounded, considering the massive
evidence that the behavior of models can shape

diverse classes of behavior including linguistic
styles, moral judgments, self-gratification patterns,
altruism, cognitive skills, inhibitions, emotional

responsiveness, transgressive behavior, attitudes,
tastes, and preferences, as well as aggressive

modes of response (Bandura, 1969a, 1971 d;
Campbell, 1961; Flanders, 1968). The networks’
unwavering uncertainty about the effects their

programs have on people is captured in the story
once told of a patient who consulted a psychiatrist
concerning doubts over his wife’s fidelity. During
the next visit the patient reported seeing his wife
accompany an amorous man into a hotel room,
whereupon they promptly disrobed and jumped

into bed. At that moment the lights were turned
off, leaving the husband at the keyhole still feeling

uncertain about his wife’s fidelity.
In selling commercial time to sponsors

television executives undoubtedly argue for the
power of television to influence the behavior of
viewers with the same strong conviction with
which they disclaim its capability to affect
aggressive responsiveness. Their public

evaluations of research concerning social behavior
often convey the impression that only
demonstration of instant direct effects on viewers
would constitute convincing proof that televised
modeling can be a contributing influence to
aggression. Evidence that exposure to aggressive
models of behavior increases aggressive
tendencies does not mean that after watching
protagonists use aggressive methods, viewers

leave the television set and assault the first person
they meet, any more than smoking a package of
cigarettes will produce instant cancer in everyone.
Since most behavior is under the multiple
control of interacting influences, the coexistence of

several factors is generally required to produce a
given response. Multiple determination applies
equally to television influence on consumer
behavior. A well-endowed blonde entreats the

viewing audience to “join the Dodge rebellion”
while seductively enumerating the virtues of this
brand of automobile. The commercial vendors

obviously do not expect every viewer to dash off
immediately to his nearest Dodge dealer. Most
people will not instantly act on what they have
learned because they do not need a car, are
satisfied with the one they have, lack the money,
or have other priorities at the moment. The idea
has been planted, however, and given appropriate

conditions some of the viewers will act in
accordance with the televised directive on some
future occasion. A rise in sales above a reliable
baseline establishes television as a determinant,
although the particular combinations of factors

that prompt action in different individuals are not
predictable in advance. The television industry
justifiably sells commercial time even though it
cannot specify beforehand exactly who will be
influenced, at what time, and in what setting.

Indeed, most policy decisions are made on the
basis of group incidence rates rather than precise

individual prediction.
The abundant evidence reviewed in this book
and elsewhere (Goranson, 1970; Liebert, 1972;
Maccoby, 1964; Siegel, 1970) supports the
conclusion that televised aggressive modeling,
through its instruction and sanction of aggressive
methods, lowers the threshold of aggressive

response and shapes its form. Exposure to
televised violence increases the likelihood that
some viewers will behave aggressively in the face
of other inducements. As televised modeling is
studied further in conjunction with other

aggression-promoting conditions, greater
accuracy will be achieved in predicting for which
individuals and under what circumstances
televised influences play a determining role in the
occurrence of aggression.
Given the capability of television to influence
behavior and the marketability of conflict, the use

of brutal content is a proper matter of social
concern. A sizable fraction of viewers from time to
time register complaints against how the medium
is used. Some of the criticisms represent
moralizing and elitist preferences for what people
should be taught to enjoy. Other critics are less
concerned about negative effects than by the fact

that pressures to maximize profits have deprived
viewers of stimulating and creative experiences
that develop the positive potentialities of man. In
many instances the critical indictments reflect
general concern over the commercial exploitation

of violence. Members of minority groups, who find
themselves victimized by being repeatedly
portrayed in derogatory stereotypes, experience
the effects of television more personally and
pervasively. They have compelling reasons to seek

change.
Television practices are under powerful

advertiser control. As a commercial business,
television networks must compete successfully to
protect the financial interests of their
stockholders. Most advertisers favor the
competitor who can deliver the largest audiences
at the cheapest rates. This in itself would not
seriously prejudice programming against viewers’

interests were it not for the contiguity problem.
Advertisers are reluctant to purchase time
adjacent to programs that attract small audiences
because viewers tend to watch whatever comes on
the channel next without much switching.

Therefore sponsors willing to support quality
programs that appeal to more limited audiences
do not have equal access to prime viewing time
even if they have the money to buy it.
When constitutional principles are used to

defend television programming, the matter of
restricted access to television is never raised.

Freedom of expression combined with the right of
public access to the medium would ensure a much
broader expression of ideas on television than
now exists. Exorbitant production costs further
increase the force of economic control and curtail
experimentation with varied dramatic contents.
What catches on is widely copied, with the result

that different networks counterprogram similar
contents, leaving the viewer with few choices.
Imitative proliferation of Westerns and crime
series provide additional testimony to the
influence of profitable aggressive modeling.
Without effective means of counterinfluence

both the medium and the general public are
subject to the abuses of excessive
commercialization. There are four different ways

in which the public can attempt to influence the
operating practices of the television industry. The
most popular—but minimally effective—method

is to appeal to governmental agencies to control
the commercial marketing of violence. Call for
federal action produces few results for a variety of
reasons. Constitutional guarantees of free
communication restrict the government from
regulating the content of television programming.
Even if the government had such restrictive

power, it would rarely be used because most
people are, by tradition, opposed to having official
agencies prescribing what they can watch.
Although government cannot regulate program
content, it can take action against grossly

irresponsible performance through the licensing
power of the Federal Communications
Commission. There is little danger, however, that a
television station would ever lose its license from
commercial overuse of violence, whatever its

effects. Not only would such rulings provoke
endless semantic disputes over what is violence,

but they would be challenged on constitutional
grounds. Being subject to political pressures,
regulatory agencies of government are usually

heavily influenced by the very industries they are
supposed to oversee. As militant consumer
advocates have repeatedly shown, federal
regulatory bodies are more likely to serve

corporate than public interests.
In sum, broadcasting is a powerful industry

subject to limited governmental checks in matters
of program content. Official criticisms of the
manner in which the medium is used elicit charges

of impending governmental censorship in the
image of notorious political dictatorships. Most
countries have adopted some reasonable
guidelines regarding the portrayal of brutality and
some have set aside a portion of their broadcast

time for different interest groups to present their
views without producing any collapse of their
democratic institutions. Indeed, increased public
activism has raised the issue of corporate
censorship. By restricting access to television time
(Barron, 1967) and excluding things that conflict
with sponsor interests, network officials exercise
selective control over what people will be shown.
Without some regulatory guidelines the

commercial pressures for screening what will be
telecast might well increase with conglomerate
ownership of communications media by powerful
interest groups. Johnson (1967), an outspoken
FCC commissioner, was probably accurate in

characterizing many of the disputes in this area as
more concerned with profitable speech than with
free speech. It is through concerted legal action
that the public will probably eventually gain some
control over access to television time.
Although the government has neither the
sanction nor the desire to police television

programming, it cannot ignore public protests.
Governmental efforts to reduce the incidence of
violence on commercial television usually take the
form of congressional hearings. Like television
reruns, at periodic intervals representatives of the
broadcast industry are brought before a
congressional committee, where they argue over

various charges and documented evidence of
exploitation of violence. Like standardized serials,
congressional investigations have essentially the
same protagonists, playing the same stylized roles,
with the same outcomes.
After the networks are commended for serving

the public well in many ways, their record of
violence is presented along with testimony by
researchers on the psychological effects of such

influences. Broadcasters challenge the criteria for
judging violence; industry researchers dispute the
empirical evidence. The influence of the broadcast

industry is brought to bear, if necessary, even on
officially commissioned evaluations of research
findings (Paisley, 1972). Some examples of
television practices are usually presented at the
hearings; this prompts network executives to
register strong complaints against excerpting
scenes for display on the grounds that the

aggressive actions are distorted when shown out
of context. At one of the hearings (U.S. Senate,
1963) this proved to be an impolitic complaint,
because the industry itself makes heavy use of
excerpts in the form of “teasers” (opening scenes

of things to come in the program), “trailers”
(previews of the action to be shown next week),
and “promotionals” (clips shown throughout the
broadcast day). The committee proceeded to show
a collection of trailers that represented a gory

catalog of human atrocities. People were
bludgeoned to death, machine-gunned, attacked

brutally with knives, meat cleavers, pitchforks,
and stilettos, stomped, drowned, burned, whipped,
beaten viciously, and ruthlessly tortured. When

asked to comment on this carnage, the president
of one of the networks staunchly maintained that
the segments constituted scenes of “physical
action,” not violence, to the utter disbelief of his

questioners. The hearing continued in this vein
with much time spent quibbling about definitions

of violence and categorization of programs. The
testimony of producers, supplemented with
industry memos documenting instances in which
violence was callously exploited, could not be as

easily dismissed semantically.
At the conclusion of such hearings, the
presiding chairman predictably issues threats to
the effect that if the industry fails to improve its

procedures of self-regulation, federal controls will
have to be imposed. Broadcasters, in turn, reaffirm
that they have not been misbehaving; besides, the
objectionable practices are a thing of the past, and
they shall continue to abide by the industry code
of good conduct. The reformed self-assessment is
typically accompanied by righteous denunciations
of the evils of governmental censorship, as though
freedom carried no responsibility. In reality, the

industry fears the adverse publicity more than the
threat of censorship. Since no one bothers to
maintain a check on violence rates in programs
after hearings have concluded, there is little
reason to discontinue the customary profitable

practices. After a time public pressure on Congress
again mounts, prompting another congressional
rerun. The commotion generated by pseudo
threats serves to evade a discussion of
constructive changes that do not involve

infringement of free expression. As a public
service, some amount of broadcasting time could

be set aside for children’s programming without
having advertisers using youngsters to pressure
parents to buy products that they either don’t

want or don’t need. Suitable funding arrangements
would then be developed. Freed from the
pressures of commercialism, producers could
devote their talents to creating entertaining

quality programs for children.
A second method for promoting social

responsibility in television programming relies on
the industry’s system of self-regulation. In this
approach discontented viewers are urged to write

letters to networks and their sponsors threatening
withdrawal of patronage if they continue to
exploit themes of violence. A commercial business
is highly responsive to economic threats, but the
correspondence approach has little effect because

the letters do not get written and the sponsors’
products are rarely boycotted.
However well-intentioned media practitioners
might be, when left to internal regulation, profits
will dictate content. This is forcefully revealed in
the operating principle stated by the president of a
television network to the director of his news
division: “They say to me [he meant the system,

not any specific individual], ‘Take your soiled little
hands, get the ratings, and make as much money
as you can’; they say to you, ‘Take your lily white
hands, do your best, go the high road and bring us

prestige’ ” (Friendly, 1968, pp. xi-xii). Networks
supervise the production process, but in the face
of strong financial pressures and weak public
sanctions, the operating guidelines are quite
pliable. Monitors screen mainly the upper limits of

violence, and even these are negotiable. Appeals to
superiors in the system can overrule objections of

the overseers. Some excesses in brutality get by
because of the prohibitive costs of reshooting

scenes. Righteous codes are fine for public
relations, but neither the performance record nor

frank statements by industry personnel (Baldwin
and Lewis, 1972) inspire confidence that the
broadcasters’ self-regulatory practices are
especially conducive to wholesome programming.
The third approach to corporate influence
rests on the well-established principle that socially

responsible behavior is produced through
performance consequences. From time to time
people knowledgeable in mass communications

have proposed establishment of a private institute
designed to report annually on the performance of
the broadcasting industry and, whenever
appropriate, to propose constructive reforms. A
respected private body might gain some token

concessions, but it is unrealistic to expect a global
annual report to alter television practices in any
significant way. The consequences, should any
result, are too weak and remote.
Private individuals can promote socially
responsible behavior in corporate enterprises
through the power of publicity and constructive
alternatives. If a change effort is to have much

effect, it must combine continuing performance
assessment with appropriate consequences. A
company will behave differently when its
performance is publicly evaluated than if no one

examines what it is doing.
A public violence-monitoring service is the first

step in fostering social responsibility in
commercial television programming. A service of
this type might be funded by foundation or private
sources to preserve its independence from
government and industry control. It would

attempt to reduce some of the excesses of
commercial control of program content by

publicized assessments, since it would have no
regulatory or enforcement power. Gerbner

(1972a, b) has developed a sensitive method for
measuring the amount of violence shown on
television through content analysis of its output.
The standard coding system records the number

and kinds of violent actions, and provides an index
of the rate of violent episodes per hour of
broadcasting. The manner in which aggression is
portrayed may be even more important than the
amount presented. These qualitative features are

measured in terms of the frequency with which
righteous and villainous characters use violent
tactics, and the percentage of leading characters
involved in killing either as killers or as victims.
Other ratable aspects of violence include its

instigators, its rewarding and punishing
consequences, and the types of people who are

repeatedly victimized. The prevalence, rate, and
role component scores are combined into an
overall violence index. A multidimensional index,

derived from depicted patterns of violent
relationships, is much more difficult to circumvent
than a simple prevalence or rate measure.
At periodic intervals a trained staff would
conduct systematic analyses of the commercial
fare. Otherwise responsible individuals feel less

concerned over the production and marketing of
violence for profit if their contribution to the total
enterprise is not singled out. Analysis of media

content should therefore restore accountability for
violence wherever possible. This is achieved by
computing separate violence rates for the different
networks, sponsors, and programs.
The violence ratings, as well as seasonal
trends, could be disseminated widely through TV
guides, PTA publications, and—ideally—periodic
televised reports to the public. Analyses of
television practices should be concerned with
more than just violence. The goal of improving the
quality of television programming is better served
by providing examples of creative uses of the
medium that viewers find enjoyable. It is through

comparison with positive alternatives that the
commercialization of violence is most forcefully
revealed. Reports to the public also provide a
suitable context in which to examine the broader

issue of television as a cultural transmitter and
innovator of attitudes and patterns of behavior.
Performance evaluation is a straightforward
matter, but the promotion of consequences
becomes a more sensitive issue. A violence rating
service is likely to have greatest impact if
conducted responsibly without assuming a strong

advocacy position. The evidence of what viewers
are being shown would be presented to them to

use as they might wish without infringing on
anybody’s rights. As consumer advocates have

demonstrated, disclosure of objectionable
practices is, in itself, likely to mobilize some
collective actions by citizen groups. Publicized
violence ratings are most likely to achieve results

through their effects on the principal force in the
system, namely, the financial subsidizers. Since
corporate sponsors are sensitive about their
public image, publication of their violence ratings
would prompt them to demand from producers

engrossing drama without gratuitous use of
violence. A business that employs violence to sell
breakfast food to children, for example, could not
afford to maintain that practice for long if it were
widely and repeatedly publicized.
Any public efforts to diminish commercial
control of television practices are easily

discredited on misleading grounds. The most
commonly used argument is that public influence
will banish all violence from television, leaving a
bland fare in its place. Fear-arousing appeals of
this sort distort the nature of the public concern. It
is not portrayal of conflict that people object to;
rather, it is the unrelenting use of physical

assaultiveness to convey human discord and
fixation on violent solutions as the preferred
remedies. Themes of conflict can, in fact, serve to
counteract assaultive modes of conduct. Modeling
prosocial solutions to conflict provides viewers

with constructive options that tend to reduce their
dependence on combative tactics (Chittenden,
1942).
The entertainment value of televised programs

need in no way suffer by reducing violence as a
salable commodity. Quite the contrary. Producers
generally prefer creating nonviolent programs to

those oriented around violent superheroes. Some
acknowledge that limitations on routine use of
violence are more likely to prod than to curtail
creativity: “Sometimes these restrictions are
challenging. How can I show tension in a fresh way
without a punch in the gut?” (Baldwin and Lewis,
1972, p. 331). The more inventive producers,

whose efforts deserve full support, welcome the
creative challenge when network directives allow
them to ease up on manufacturing violent
contents: “You get sick doing that crap. We had to
do it for two years because that is what the

network wanted. … Now we are using our brains
coming up with ways of doing things that do not
rely on smashing, hitting, and hanging. Now we
are enjoying our work” (Cantor, 1972, p. 266).
Analyzing bad programs is not necessarily going to

produce good ones. A violence monitoring system
may be required to create the opportunities for

better offerings, but other vehicles are needed to
bring this about.
Greater progress is achieved by rewarding
desirable practices than by curtailing
objectionable ones. Public efforts to improve the
quality of television programming should not be
limited to negative sanctions. The fourth, and most

affirmative, approach relies on the power of
positive example. Programs that use violence as
the principal means for capturing audiences would
be put out of business by providing people with
more interesting alternatives. The influence of

successful example is nowhere better illustrated
than in children's programming, which has
aroused the greatest public criticism. For years
parents complained about the serial dramas and
weekend commercial fare served their children,

but did little to discourage them from watching it.
The appearance of Sesame Street on

noncommercial stations demonstrated that a
program that is instructive as well as entertaining
can attract large child audiences with parental

endorsement.
A viable public broadcasting system, free of
commercial pressures, is perhaps the best means
of improving and diversifying television offerings.

Within this context, a program development unit
might be created for the purpose of testing new

ideas as well as assessing the impact of programs,
especially on young viewers. Close collaboration
between producers and social scientists could
serve the dual purpose of expanding the quality of

television programming and advancing
understanding of communication influences. The
adoption of new formats by the commercial

networks could be accelerated by displaying
successful prototypes in the reports to the public

mentioned earlier. Commercial networks can
derive other benefits from a good public
broadcasting service. It can function not only as a

place for testing ideas, but for creating new
interests which commercial television is
understandably reluctant to undertake because of
the initial financial losses. Under severe economic
competition, programs are created for instant

commercial success.
From time to time economic proposals have

been offered on how to provide public
broadcasting with adequate funding for facilities
and professional talent. The most imaginative plan

was developed by the Ford Foundation (Friendly,
1968). Satellites will replace land lines as the
mode of transmission with considerable savings.
Ford proposed that a nonprofit corporation
operate the communications satellite service and

reserve some of the revenues from commercial
users to support public broadcasting. This plan
was challenged by private carrier corporations
that claim the right to operate a multipurpose
satellite service. Whether any of the revenue from
a communications satellite will be earmarked for
noncommercial television remains to be seen.
There are other major developments in the
offing that can alter the manner in which the

medium is used. Technological innovations in the
form of cable television and laser and satellite
telecasting will permit simultaneous broadcasting
of varied programs to audiences with diverse
interests. The technical capacity for diversified

programming can free viewers from the tyranny of
the common denominator. Computer retrieval
systems, wherein viewers can select programs
from videotape libraries, will add another new

dimension to broadcasting services. These
systems, of course, will not be free of commercial
pressures to maximize profits at the expense of

public interests. If cost per thousand viewers
continues to dictate program production, and
there is every reason to believe it will, cable
television is likely to duplicate the practices of the
current system. The net result will be many
channels ritualistically repeating similar formula
shows, old movies, and reruns that, being cheaper,

make more money than more entertaining
programs requiring higher production
expenditures. Whether the technical changes bring
significant reforms in programming partly
depends on whether the public is granted some

right of access to the facilities and how actively it
is exercised. A watchful public can increase the
likelihood that new opportunities will change
television from a mediocre entertainer and
teacher of aggressive life styles to an instrument of

human enlightenment.
Television Journalism
A high proportion of newscasting time is

devoted to coverage of violent events in one form
or another. Both local stations and national
networks undoubtedly vary in the emphasis they
give to violent incidents. Over an extended period
Singer (1970) compared the proportion of news

items of an aggressive nature reported by an
American network and a Canadian network that
has full access to the same content and frequently
rebroadcasts a substantial amount of American
material. American newscasts carried

approximately twice as many aggressive items.
Even when war reports were deleted from the
data, American newscasts continued to offer
substantially more aggressive items (36 percent)
and devoted more time to them (40 percent) than

its Canadian counterpart, for which the
corresponding figures were 18 percent and 19

percent, respectively. Whether the heavy
American emphasis on violence mirrors the
cultural reality or the predilection of the medium

itself cannot be determined from this study. A
comparative analysis of the proportion of news
items devoted to violence by the national
networks and by local stations within the same

city would provide some indication of
broadcasting company preferences.
Heated debates have raged over the partiality
of broadcast journalism and its social and political
impact. Considering the central role played by

television in public affairs, surprisingly little study
has been made of these issues. Some researchers
have analyzed the social influences operating
within the media systems in determining the
selection and presentation of news contents

(Breed, 1955; Warner, 1971). The effects of the
media output, however, remain largely
unexamined. Reports of ordinary crimes rarely
arouse controversy, except perhaps for their
disproportionate emphasis as newsworthy items
in the brief time allotted to television news.
Rather, it is in conflicts for power between
contending groups, where the medium can
function as an instrument of influence, that

charges of partisan treatment arise.
There are several ways in which accounts of
violence might produce social effects. Detailed
reports of activities have instructional potential.
Previous analyses of aggression contagion, though

complicated by many determinants, lend some
substance to the view that mass media coverage of
civil disturbances can shape the form of collective
aggression and inadvertently contribute to its

spread. Observers can learn from newscasts,
among other things, how to firebomb with
Molotov cocktails, how to conduct a sniper siege,

and the steps required to hijack airliners
successfully. Although the printed media
disseminate information about acts of violence,
television has the greater potential to influence
behavior because it portrays events more vividly
and concretely. Furthermore, most people learn
about newsworthy incidents from watching

television rather than from reading magazines or
newspapers (Schramm, Lyle, and Parker, 1961).
The power of pictures is confirmed by Barrow and
Westley (1960), who found that young viewers
learned more from television than from an

equivalent radio version of the news events.
Media reports of violent episodes not only
reveal the use of destructive weapons and
aggressive tactics, but can also influence

aggression restraints by their portrayal of the
consequences of violence. In order to hold
viewers’ attention, the visual displays

accompanying the news reports are selected to
entertain as well as to inform. The outcomes of
given courses of actions are easily misrepresented
when graphic effects are favored over common,
but less interesting, consequences. Thus, showing
people running off with appliances and cases of
liquor from looted stores in a jovial atmosphere is

more likely to promote aggression in viewers
living under similar circumstances than showing
the terror and suffering caused by massive
destruction of one’s neighborhood.
Another way of affecting aggressive restraints

in viewers is by the degree of justification given to
violent activities in the narration of news stories.
Meyer (1971) showed angered adults a filmed
news item in which a wounded North Vietnamese

prisoner was stabbed to death by a South
Vietnamese soldier. Subjects who watched the
newsfilm with a voiceover narration that justified

the execution subsequently behaved more
punitively than those who observed the same
news event with either no commentary or verbal
portrayal of the stabbing as unjustified. These
results suggest that reporters’ descriptions of
newsworthy violence can influence viewers’
readiness to behave aggressively.
Televised reports of violence, depending on
what they accent, may in some instances affect the
course of collective aggression for better or for
worse through their sanctioning potential.
Because of the scarcity of newscast time, only a

few items from a vast array of events can tie
presented. Selective processes are similarly
involved even in reporting the violent episodes
chosen, since only a few aspects of the sequence of

events can be introduced. One can highlight the
social conditions provoking strife, the specific
tactics of aggression used, the immediate or

delayed consequences accruing to aggressors, the
injurious effects to the victims, the
pronouncements of leading protagonists, or the
types of countermeasures employed by control
agents. From knowledge of social influence
processes one would expect the potential
consequences of media presentations to vary

depending on what aspects are selected for public
display and how they are characterized in
accompanying commentaries. Ball-Rokeach
(1972) attaches special significance to media
transmission of evaluative reactions concerning

violent events. The reason for this is that relatively
few viewers experience sufficient inducements to
emulate portrayed violent conduct, but the
modeled justifications and evaluative responses
can help to mobilize public support for policy

actions that have widespread social and political
consequences.
If disruptive actions are depicted as
lawlessness and the coverage selectively
emphasizes the revolutionary rhetoric of
extremists and the injuries suffered by the targets,
most viewers will deplore the protest activities
and demand punitive controls. In contrast, reports
that attend to the causes of aggression and

characterize the group action as a means of
securing needed reforms will produce a more
sympathetic response in viewers toward
improving social conditions. Visual documentation
of police practices can partly determine the sides

viewers will take in a social conflict. Early in the
civil rights movement, for example, the sight of
policemen brutally attacking nonviolent
protesters probably did more to arouse a hitherto
complacent populace than all the moral appeals or

democracy arguments combined.
The foregoing presumptions follow plausibly

from existing knowledge, but they require
experimental verification with studies offering
differential reports of the same aggressive
incidents. The behavioral and attitudinal
consequences of media practices could be
assessed by varying the amount of attention given
to causes, tactics, consequences, and

countercontrols in the coverage of newsworthy
events. Should the predicted relationships be
confirmed, such findings would sensitize
television journalists to the specific effect that
differential emphasis of news content is likely to

have on the viewing public.
The discussion thus far has dealt with how the
mass media may influence the course of
aggression through selective or incomplete

portrayal of newsworthy violence. Informal
observations suggest that the presence of
reporters and television cameras at the scene of

aggressive altercations can exert some effect on
the incidents themselves. Cameras tend to invite
and to give greater attention to the provocative
rhetoric of extremists than to less colorful
moderates. It has been shown that selective
attention to certain members of a group, who
ordinarily received little recognition, confers

leadership status on them (Hastorf, 1965). To the
extent that such an influence process is operative
in picking engrossing news shots, television may
contribute to the emergence of a leader from
among the various contenders. In a continuing

conflict, selective editing of performances having
dramatic appeal may distort the nature or purpose
of the contending groups in ways that promote
divisive stereotyping and angry polarization.
Acrimonious retorts performed for television

audiences would further prolong the aggressive
struggle. Although claims of misrepresentation

abound, no one has systematically compared the
actual events and what is chosen for television
news coverage.
With camera crews around, challengers are apt
to become more exhibitionistic and police more
restrained in their behavior. Indeed, the
inhospitable reception that reporters typically

receive from police at the scene of a conflict
suggests that the presence of cameras reduces the
incidence of unauthorized police aggression. It
would be of considerable interest to measure
changes in the behavior of contending groups

when television camera crews are present and
when they are absent.
The preceding discussion concentrated on how
media output can affect public opinion and action.

Because of its influence potential, the
communication system itself is subject to constant
pressure from various interest groups within the

society. A complete analysis of communication
processes must encompass the reciprocal
influence between transmitters and recipients.
Conflicts over coverage of newsworthy events
inevitably arise from a multitude of sources. Since
television is a competitive industry, concern over
advertising revenue partly dictates how the

medium will be used (Friendly, 1968). Political
officials apply leverage in efforts to sway the
reporting of domestic and foreign events in their
favor. Additionally, there are the pressures from
the viewing public that criticize the industry for

slanted treatment of the news. Conservatives view
newscasters as advocates of liberal causes,
whereas dissidents contend that television, being
controlled by privileged groups, avoids
controversial issues that conflict with their

corporate or political interests. Finally, the
newsmen, who become the unenviable targets of

the opposing pressures, have to contend with their
own ideological commitments in choosing what to
show the public.
Television journalism undoubtedly serves as
an instrument in aiding social reforms or
reinforcing prevailing customs by how it presents
controversial issues. Few serious students of

journalism espouse the illusion of neutrality. In
commenting on the matter of balanced reporting,
the distinguished broadcast journalist Edward R.
Murrow, rightly pointed out that one cannot do
equal justice to contending points of view because

rarely are they equally justifiable. The different
segments of society are better served by striving
for communication pluralism rather than
neutrality. As long as there are alternative
channels of information to which the public has

access, divergent views can be adequately aired.
A number of conditions, some organizational

and others social, limit the pluralistic possibilities
of television journalism. Media ownership is
concentrated in the higher ranks of society.
Although newsmen strive for fairness, diverse
interests would obviously receive greater
exposure if the media were under more varied
proprietorship. Relatively little time is devoted to

news coverage because it produces a low financial
yield. Even with the best of intentions, time
pressures preclude the type of analytic reporting
required to explain the interplay of forces creating
and maintaining violent conflicts. As a result,

complex issues are essentially reduced to
headlines with fragmented pictorial
accompaniment interspersed with numerous
commercials. Public affairs programs treat
selected topics in greater depth, but this format is

only sparingly used. Although television coverage
is not intended to substitute for the depth and

range of newspaper reports, most people do rely
on television as their main source of information
about matters that affect their lives.
In addition to organizational constraints, those
who espouse unpopular views encounter social
barriers to gaining a satisfactory hearing on
television. This is because societal reinforcements

are heavily weighted in favor of neglecting
dissident groups. The television industry can enjoy
public support by airing matters that fall within
the bounds of popular sentiment, whereas
providing a forum for dissenters brings

condemnation and economic threats. Such
differential reinforcements tend to shape
broadcast policies in conservative directions. This
is not to say that dissident factions are rarely
heard. By resorting to coercive and outlandish

behavior they command attention that they would
not otherwise receive. Thwarted access to the

media thus selectively reinforces disturbing forms
of conduct that capture attention at the expense of
a sympathetic hearing.
Changes in the communication industry should
be directed at reducing inequities in access to the
means by which the public can be informed and
influenced. Diversified control of television

facilities is one way of creating greater access
opportunities for diverse aggregations in the
society. This goal can also be served, though less
completely, by officially designating a block of
time on television as a forum for minority or

controversial views. Here spokesmen for different
interest groups could freely express the
aspirations, the felt grievances, and the remedies
sought by their constituencies. Since the public
forum would be conducted by the participants

themselves rather than by industry personnel, the
networks are afforded protection against criticism

for expression of dissenting views.
Good television journalism would require
increasing the depth of news coverage and
sacrificing some of the exciting features of
violence. Knowing the causes of actions can reduce
vindictiveness; seeing their long-term
consequences can discourage ill-judged adoptions

of the behavior. Unfortunately, such moderating
influences are not only more difficult to portray,
but have less audience appeal than the actions
themselves. Brief newscasts do not provide time to
dwell on causes, nor can the aftermath of violence,

which remains in the distant future, be reported at
the time when it would have the greatest impact
on potential adopters and controllers alike. Given
the demonstrated power of example, the news
media must exercise caution in displaying detailed

aggressive techniques and their immediate
rewards if they are to safeguard against promoting

the very things they are reporting.
DIFFERENTIAL REINFORCEMENT
Since aggression is partly under external
reinforcement control, injurious behavior can Ire
successfully modified by altering its customary
consequences. The standard procedure begins

with a baseline measurement of aggressive acts as
they occur under the usual social practices. A
change program is then instituted in which
constructive behavior is positively reinforced
while rewards for aggressive conduct are withheld

or withdrawn. The findings generally show that
aggressive behavior continues at a high rate
during baseline conditions but declines when it is
no longer positively reinforced. In order to
confirm that the aggressive conduct is in fact

supported by its consequences, the original
reinforcement practices are reinstated,

whereupon aggressive responding typically
reappears. In the final phase of the program the
beneficial contingencies are reintroduced so that

aggression is again rendered ineffective through
nonreward, whereas constructive modes of
response are strengthened. Demonstration of
replicative control by varying response

consequences is essential in research for
establishing the influence of outcomes, but in

social applications it is unnecessary, if not unwise,
to reinstate aggression temporarily after it has
been reduced.
In implementing reinforcement practices one

must select suitable reinforcers for alternative
styles of behavior. The types of incentives likely to
be most appropriate and effective are largely
determined by the developmental level at which

people are functioning. As people change, the type,
amount, and source of reinforcement supporting
their behavior are gradually modified. Hence, the
incentives used initially in change programs may
differ considerably from those that ultimately
assume rewarding functions.
Some of the criticisms that have been levied
against reinforcement practices fail to recognize
the complexity and developmental changes in the

consequences that influence human behavior. At
the lowest developmental levels behavior is

primarily responsive to tangible consequences. As
physically rewarding experiences are repeatedly
associated with expressions of interest and
approval and punishments with disapproval,

social reactions themselves acquire reinforcing
properties.
A number of symbolic reinforcers other than
social reactions take on reinforcing functions.

Credits attainable through designated
performances, which may be exchanged for
countless things that people want, serve as
dependable generalized rewards. This type of
incentive system essentially involves negotiating a
social contract with an individual to help him
bring his aggressive behavior under control or to
develop other skills that displace aggression
because they are more functional (Cohen and

Filipczak, 1971; Colman, 1971; Steffy et al., 1969).
Ordinarily, concrete rewards are employed in
initial stages with individuals who are not
reinforceable in other ways and would gain little
from treatment efforts. Many of the contentious

reactions of hyperaggressive individuals to task
requirements, which are often attributed in
clinical theory (Redl and Wineman, 1951) to high
impulsivity, weak ego control, and low frustration

tolerance, may in fact result from inadequate
incentives. Aversive actions become an effective
means of avoiding or turning off environmental

demands that bring little or no reward. Thus, for
example, belligerent boys who remained
unmotivated by praise worked productively when
their efforts were concretely rewarded and
continued to do so even after the tangible
reinforcers were progressively reduced and
eventually discontinued altogether (Levin and

Simmons, 1962). Another illustration of the
change of reinforcement supports for behavior
during the course of treatment is provided by
Wahler (1968), who assisted parents in modifying
extreme antagonism in their children. An initial

program in which parents ignored defiance and
commended cooperativeness proved ineffective. A
subsequent reinforcement system combining
parental approval with credits exchangeable for
valued items produced dramatic and enduring

increases in cooperativeness. Thereafter, the
credits were gradually eliminated and cooperative

behavior was maintained by expressions of
appreciation alone.
At the highest level of functioning people
regulate their own behavior by self-evaluative and
other self-administered consequences. Several
researchers (Goodlet and Goodlet, 1969; Bolstad
and Johnson, 1972) have shown that

aggressiveness can be eliminated by having boys
reward themselves for reductions in aggressive
conduct as well as or better than by having adults
evaluate and reward their behavioral
improvements. Tangible self-reinforcement and

symbolic rewards can be used to maintain new
patterns of behavior until they become a source of
personal satisfaction.
Reinforcement influences have been

extensively employed with notable success in
modifying hyperaggressive behavior in children.
The case reports involve physically assaultive

children who repeatedly disrupt ongoing activities
and, when others do not accede to their wishes,
become abusive or display violent temper
tantrums until they get what they want. When
disciplined, they respond with aggressive defiance
that discourages adults from attempting corrective
action.
Treatment programs are generally
implemented (under professional supervision) by
those who regularly deal with the problem
behavior, because it is their reactions that largely
determine the direction and future course of

change, however induced. In a well-designed
program a behavioral analysis is first conducted to
identify the social conditions that maintain
aggressive conduct. Change agents are then given

detailed instructions on how they must alter their
characteristic ways of reacting to the child’s
behavior to achieve desired outcomes. This

usually involves a reversal of reinforcement
practices. Whereas previously coercive actions
commanded attention and compliance and
prosocial behavior received little notice, they are
now advised to ignore or, if necessary, to disallow
aggression and to reward constructive behavior.
Prescriptive instructions alone are likely to

produce limited results unless combined with
other aids that facilitate adoption of remedial
measures. Change agents may come to understand
principles for modifying behavior but remain at a
loss as to how to implement what they have

learned. What they need is guided practice with
corrective feedback that rewards their successes
and corrects their mistakes. Remedial measures
must not only be sketched out in concrete actions,

but preferably modeled through videotape, role
playing, or, if feasible, in actual demonstrations of
how to deal with problematic behaviors. After the

procedures have been sufficiently demonstrated,
change agents are supervised until they attain
proficiency in handling problems without external
directions (Johnson and Brown, 1969; Russo,
1964; Wolf, Risly, and Mees, 1964).
Detailed instruction combined with
demonstration and guided practice are well suited

for altering reinforcement practices, but unless
adequately rewarded, the efforts of change agents
are unlikely to endure very long. This problem is
especially critical in initial stages of treatment
when withdrawal of rewards for aggression may

temporarily increase abusive responses in efforts
to restore the former benefits of coercive
behavior. During this period social support is
required. Later, reductions in disturbing behavior

begin to provide natural rewards for change
agents, until eventually the interpersonal patterns
become self-sustaining through reciprocal

satisfactions.
Given adequate training along the lines
discussed above, the behavior of parents, teachers,
and other influential people is much more
changeworthy than traditional theories would
lead one to believe. Any remedial program,

whether directed at change agents or at the clients
themselves, will encounter problems from time to
time. When disappointing results are obtained,
there is a tendency to attribute difficulties to

pathologies in the change agents and to lapse into
protracted conversations concerning their
dysfunctional behavior. The remedies usually lie
elsewhere. Sometimes change agents do not
consistently apply what they have learned because

it is bothersome to do so. They adopt the remedial
practices, however, when social or tangible

rewards are made contingent upon improvements
in their performance (Ayllon and Azrin, 1964;

Bernal, Williams, Miller, and Reagor, 1972). At
other times change agents cannot alter their

behavior sufficiently because of disruptive
influences impinging upon them. Here, successful
results require an action program directed at the
interfering conditions as well.
Hawkins, Peterson, Schweid, and Bijou (1966)
provide one of many illustrations of home-based

modifications of aggressive behavior. The case
involved an assaultive, belligerent boy who,
though young in years, effectively controlled the

household by aversive means. After completing a
baseline measurement of aggressive behavior, the
mother was instructed to go about her household
activities. Whenever her son acted reprehensibly,
the observer signaled the mother either to tell him

to stop or to place him in his room for a brief time.
In contrast, when he behaved commendably the
mother was encouraged to express her interest
and approval. As summarized in Figure 5.2, the
new reinforcement practices produced a marked
decline in aggressiveness. In the next phase the
FIGURE 5.2 Number of 10-second periods in which the boy
behaved aggressively during each one-hour session. (Hawkins
et al., 1966)
mother was asked to resume her customary habit
of chastising antagonistic actions while ignoring
desirable ones. Having experienced success, she

found it difficult to revert to her former style. The
mother not only maintained the favorable changes
without any further guidance, but the boy became
more considerate and affectionate, a marked
contrast with his former domineering control. As

Hawkins points out, a major benefit of enlisting
parents as therapists for their own children is that
the skills they acquire can serve them well in
meeting future developmental problems.
There is ample evidence that selective

reinforcement reduces aggressiveness, but it
remains unclear whether it is the consequences

associated with aggression, with prosocial
behavior, or both sources of influence that
contribute to the effectiveness of this approach. In
some studies reporting high success aggressive
behavior is consistently ignored while considerate
conduct is rewarded with social approval and
attention (Bernal et al., 1972; Johnson and Brown,

1969; Russo, 1964; Scott, Burton, and Yarrow,
1967). In other cases prosocial behavior is
positively reinforced, but aggression is punished
by removing the child from the situation for short
periods immediately following the act or by some

other negative consequences (Bernal et al., 1968;
Hawkins et al., 1966; Sloane, Johnston, and Bijou,
1967; Wolf, Risley, and Mees, 1964; Zeilberger,
Sampen, and Sloane, 1968). Still others (Staats and
Butterfield, 1965) report that assaultive

delinquents ceased behaving destructively when
they acquired new competencies, even though no

effort was made to alter the consequences for
aggressive acts. To isolate the therapeutic
ingredients would require a comparative study in

which aggressive responses are ignored, punished,
or not singled out for special treatment and,
within each of these conditions, prosocial
alternatives are either fostered or not.
Approaches to the modification of child
behavior disorders have undergone many changes

over the years. Intrapsychically oriented
therapists, who assumed that personality patterns
are permanently established in the first few years

of life, practiced an exclusively child-centered
treatment. Therapists analyzed children’s social
behavior and play patterns in their offices while
the parents assisted passively by not interfering in
the endeavor. With mounting evidence that social

behavior is changeable at all ages and that parents
exert a continuing influence on their children, a
collaborative form of treatment came into vogue.
Growing out of the psychodynamic tradition, this
approach regarded the child’s troublesome
behavior as a symptom of the parents’ internal
pathology. Family members were segregated in
treatment, with the child’s therapist serving
essentially as a baby sitter who played with the

detached symptom amid sand boxes, doll families,
finger paints, and an arsenal of aggression-
eliciting toys, while the adult’s therapist delved
into the mother’s intrapsychic life. Parents’
requests for advice on how to alter their children’s

problems were ignored, evaded, or disapproved as
resistance to self-exploration. Remedial action was
thus deferred for self-analysis of dubious value.
Disappointing results with these time-

consuming endeavors eventually gave rise to
family therapy in which all family members were
seen together. The focus on family

interrelationships was a move in the right
direction, but a joint interview technique does not
necessarily guarantee better outcomes. If the time
is spent mainly in ventilating complaints,
assigning blame, and analyzing how each member
contributes to familial malfunctioning, there is
little reason to expect that improvements will

occur with any consistency. If, on the other hand,
family members receive guidance in rewarding
ways of relating to each other, the prospects of
favorable outcomes are good.
A number of studies have already been cited in

which parents were taught child management
skills for specific aggressive problems. Patterson
and his colleagues (Patterson and Cobb, 1971;
Patterson and Reid, 1970) have been evolving

procedures for teaching parents how to deal with
a wide range of problems exhibited by different
family members. Initially, parents familiarize

themselves with principles of learning that help
them understand how social behavior is
established and modified. Following this brief
didactic instruction, parents select problem
behaviors they wish to modify. They are then
assigned a special time during each day to record
the incidence of these behaviors, the various

effects they produce, and the family members who
provide the consequences. This observational
phase is designed both to sensitize parents to
social influence processes and to reinforce the set
that they will serve as active agents rather than as

passive recipients of treatment. Once they have
learned to observe interpersonal contingencies
accurately, they are taught, by demonstration and
guided practice, constructive ways of handling
their children's behavior. Family problems are

modified in this manner usually one at a time.
When necessary, new reinforcement practices

are also introduced in classroom settings and peer
groups. Influence over deviant behavior in
extrafamilial situations is typically achieved by a
peer contingency procedure in which the child and
his peers initially earn valued rewards for positive
conduct on his part. Concrete rewards are then
gradually reduced until eventually the child’s

behavior is maintained by natural social reactions
of teachers and peers.
In one project reported by Patterson, Ray, and
Shaw (1968), parents reduced the frequency with
which they rewarded coercive behavior from an

average rate of 35 percent during the baseline
period to 10 percent during the intervention
program. Modification of familial reinforcement
practices not only lowered the families’ output of

aggressive behavior, but increased the level and
reciprocity of positive interactions within the
family system. Results of additional studies with

families presenting severe problems of social
aggression reaffirm the efficacy of this type of
approach (Patterson, Cobb, and Ray, 1972).
Through a combination of instruction,
demonstration, and guided practice parents can
acquire proficient skills for changing problem
behavior. Difficulties sometimes arise, however, in

their implementation of requisite practices. Efforts
to enhance the effectiveness of this method might
do well to concentrate on the development of
parental incentive systems to ensure that parents
apply the methods they have learned more widely.
Teachers must often deal with children whose

assaultive and disruptive behavior seriously
intrudes on the activities of others. In some
instances these are domineering children who

have been repeatedly reinforced for controlling
their social environment through coercive means.
More often, however, the children resort to

aggressive ways of securing attention when they
lack alternative means of gaining acceptance.
Negative attention-forcing behavior is highly
resistant to change because, to a child of limited
skills who desires the recognition of others, even
reprimands serve as rewards. Moreover, in group
situations aversive behaviors that alienate

teachers are frequently reinforced by the
amusement of peers from whom attention is
sought. Once cast in the role of class buffoon, one
loses opportunities to develop his potentialities
unless the detrimental contingencies are reversed.
Madsen and his associates (1968) illustrate

how reprimands may reinforce the very
troublesome behaviors they are intended to
diminish. When teachers admonished students for

disruptive activities, the number of transgressors
promptly increased. In subsequent phases of the
study the number of disrupters declined when

teachers reduced their reprimands and rose again
when they resumed their admonishments. The
chronic problem was solved by ignoring annoying
actions and praising engrossment in learning
activities.
Many interventions intended as punishments
actually serve as positive reinforcers that maintain

troublesome behavior. Self-defeating practices
usually go unnoticed by those who use them since
they see the immediate results but rarely assess
the full or long-term effects their behavior may
have on others. Detrimental social systems are

unknowingly created and mutually sustained
because aversive behavior is rewarded by the
attention it commands, while the ineffective
control techniques are reinforced by their success

in temporarily checking disturbing performances.
In the preceding example, the teachers
inadvertently rewarded annoying conduct in

students who, in turn, rewarded the teacher’s
shouting by their momentary compliance, with the
net result that both sets of performances were
mutually escalated.
In modifying disturbing aggression in school
settings professional personnel serve as
consultants to teachers who are given the

knowledge and skills with which to solve the
problems confronting them. School-based

treatments generally achieve reductions in
aggressive and defiant classroom behavior by
having teachers ignore disruptive activities and
attend positively to educational pursuits (Becker,

Thomas, and Carnine, 1969; Goodlet, Goodlet, and
Dredge, 1970).
Whenever one or more children are singled out
for special treatment in a social context, one must

be sensitive to how other members may be
affected by seeing another’s actions differentially
reinforced. Some teachers fear that increased
attention to the behavior of a problem child will
prompt others to behave in a troublesome way to
secure similar solicitous interest. Such adverse
effects would occur if the increased attention were
given noncontingently or if it were gained through
abusive behavior. When annoying conduct is

ignored and constructive actions are commended,
the treatment, in fact, provides others with
positive models for emulation. Nothing is gained
by behaving aggressively. Consistent with these
expectations, research has shown that differential

reinforcement of one child either has no effect on
others (Ward and Baker, 1968) or leads them to
adopt productive activities that they see well
received (Reppucci and Reiss, 1970).
Extinction of Aggression
When reinforcement for a learned response is
withheld, the behavior continues to be performed

for a time, sometimes even briefly increasing in
rate, depending on its prior reinforcement history.
Under consistent nonreward, however, the
behavior declines and is eventually discarded. The
use of extinction in multiple procedures has
already been touched upon in the previous

section. Ordinarily, a method combining
nonreward of negative behavior with positive
reinforcement of competing patterns is more
effective than extinction alone. Some of the
problems that may arise in the use of extinction

procedures, either singly or in conjunction with
rewarded alternatives, merit discussion.
When the rewards gained through coercive
behavior are first withheld, it is often intensified in

an effort to restore former benefits. If the more
aversive tactics should force compliant reactions,
the results are worse than if no intervention had

been attempted. In the latter instance the
aggressor is not only rewarded for using more
extreme measures but learns that persistence
eventually pays off, thus increasing the resistance
of aggression to extinction. A psychological
principle, no matter how powerful, will yield weak
or variable results if it is not applied consistently.
The importance of uniform social practice in
implementing a change program based on
extinction is revealed in Williams’ (1959)
treatment of aggressively demanding behavior in a
young boy. The child had required special care

during a severe illness and, upon his recovery,
continued to demand the parents’ undivided
attention with intense tantrums and crying spells,
especially at bedtime. The parents were instructed

to put their son to bed in a leisurely, amicable way,
but after completing the bedtime routine, to
ignore his screaming and raging. The duration of

the unrewarded tantrum behavior dropped
markedly and ceased altogether within a few days.
The child no longer created a scene at bedtime, but
instead played happily until he dropped off to
sleep. A week later an aunt reinstated the tantrum
behavior by complying with the child's demands
to remain in the room after being put to bed. A

second extinction series produced complete and
enduring remission of the tantrums.
A child’s aversive control of his parents, if
intermittently reinforced, is likely to generalize to
other areas of behavior and to other people. As the

above study shows, after alienating coercive
behaviors are removed, the familial atmosphere
changes from one of recurrent drawn-out battles
to reciprocally rewarding interactions. This case

illustrates several other interesting issues.
Psychotherapy cannot be separated from the
experiences of everyday life. No amount of

insightful conversation will have much effect
unless the reinforcement practices of influential
people in the social system change. There are no
psychological methods that can render a person
permanently insensitive to the consequences of
his actions. Nor would such an outcome be
desirable, even if it were possible, because

behavior that remained immune to its effects
would be grossly maladaptive. A remedial
intervention can correct a reciprocally reinforcing
system gone awry, but whether induced changes
generalize and endure depends upon the

prevailing reinforcement practices in different
social contexts. To continue with the preceding
example, if parents ignore tantrums but others
reward them, tantrums will occur to some extent
unless others also alter their way of reacting to

them. If, for whatever reason, the behavior of
certain members of the social environment cannot

be changed, significant though less generalized
reduction in aggression can still be achieved.
Given conflicting contingencies, individuals

eventually distinguish between situations in which
their aggressive actions succeed or fail and behave
accordingly.
Some forms of aggression are more amenable
to extinction treatment than others. One can
render verbal attacks, annoying actions, and

oppositional response ineffective simply by
choosing not to respond to them. But a tough
aggressor can get what he wants by physical force

regardless of the interest or cooperation shown by
his victims. Brown and Elliott (1965) report
findings that have a bearing on this issue. They
instructed teachers to ignore aggression and to
reward cooperative behavior in an effort to reduce

the combativeness of 27 boys in a nursery school
class. Under these reinforcement practices both
physical and verbal aggression declined. After the
program was discontinued, physical aggression
showed some recovery during a follow-up period.
The authors attributed the recurrence to teachers’
difficulty in refraining from intervening and
interacting with the boys when they assaulted
others. Although these results are explainable in

terms of differential reactions of teachers, it is also
probably true that physical aggression was
intermittently reinforced by its success in
extracting compliance from peers, apart from the
teachers’ interest. A second application of the

reinforcement program produced additional
reductions in physical and verbal assaults.
In cases in which aggressive behavior is
injurious or seriously infringes on the

prerogatives of others, a simple extinction
procedure may not be feasible. Pinkston, Reese, Le
Blanc, and Baer (1972) tested an extinction

treatment that protects the victim from harm. It
consists essentially of ignoring the aggressor while
attending to the victim by shielding him from
further attack and arranging interesting activities
for him. An explosively violent boy who proved
unresponsive to reasoning, imposed limits, and
reprimands promptly reduced his physical attacks

when this type of treatment was instituted.
Hurtful conduct generally alienates one from
others. In keeping with expectation, elimination of
injurious behavior alone enhanced the boy’s
positive relationship with his peers.
Protective extinction may be feasible with

young children but it would be difficult to apply
either to older aggressors or to problems in which
an entire group is victimized. In such instances

aggressive actions are typically controlled by
negative sanctions while other coping styles are
being developed.
The behavior exhibited during the course of
extinction is primarily determined by the response
options available to the individual. As dominant
aggressive responses prove ineffective, a person
will try alternative courses of action that have
been successful in the past. No special problems

are created by the use of extinction alone when
the available alternatives are constructive. If,
however, the responses in an individual’s
repertoire are largely negative, a change agent

may be faced with the arduous task of eliminating
a succession of detrimental tendencies. This
problem can be avoided by combining extinction
with modeling and reinforcement procedures that
foster alternative means of securing desired

outcomes. Individuals undergoing extinction alone
learn what no longer works but remain uncertain

about, or unskilled in, better options, whereas the
combined use of extinction and reinforced

modeling provides positive guidance. These
considerations apply equally to punishment

control of aggression, which is discussed next.
Punishing consequences are widely applied to
people in all walks of life in efforts to reduce or to
eliminate troublesome behavior. In many

instances aversive control is favored because it is
reinforcing to the punisher rather than beneficial

to the recipient. A method that is quick and
effective in stopping disturbing behavior will be
sufficiently rewarding to the user to be retained
even though it may have minimal or adverse long-

term effects for all concerned.
In professional circles two major criticisms
were levied against the use of punishment to
modify behavior. It was argued, more from belief
than from fact, that negative sanctions are

relatively ineffective and that furthermore, they
produce undesirable side effects. However, as
empirical evidence accumulated on how negative

consequences influence behavior, blanket
rejection gave way to qualified acceptance. Many
of the unfavorable properties ascribed to
punishment are not necessarily inherent in the

methods themselves, but result from the faulty
manner in which they are usually applied. In
everyday use punishments are typically excessive,
ill-timed, erratic, and administered vengefully

without providing any positive direction. Results
of numerous studies show that under certain
conditions detrimental behavior can be effectively
modified by negative consequences and that
adverse side effects need not occur.
Negative consequences may involve either
withdrawal of rewards or administration of

aversive effects. Aversive control through physical
punishment and other painful outcomes is seldom

employed in judicious change programs and then
only to control seriously injurious behavior

requiring active interventions. Withdrawal of
rewards is more commonly used on occasions
when negative sanctions are needed. In the latter
case, the response costs of aggression may take

the form of monetary penalties, temporary
forfeiture of privileges and possessions, or brief
removal from rewarding situations for assaultive
or destructive actions.
The adjunctive use of punishment by reward

withdrawal has certain advantages over physical
aversive procedures. As previously shown, painful
treatment may arouse counteraggression or
avoidance of the punishers, thus weakening their

potential influence. In contrast, interventions that
chiefly involve removal of valued things not only
generate weaker emotional disturbances, but

foster contact with the disciplinarian who can
reinstate the forfeited rewards. If restoration of
positive reinforcers is made contingent upon
constructive actions, rapid behavioral changes
may in fact result. In addition, reward withdrawal
techniques provide more acceptable and feasible
alternatives to physical punishment, which most

change agents are understandably loath to use.
Punishment is rarely employed as a sole
method of modifying aggressive behavior. Rather,
it is applied in conjunction with positive influences
conducive to the development of more rewarding

styles of conduct. In other words, in effective
intervention one does not punish without offering
better alternatives. For aggressors who have
already developed other social skills that are

potentially reinforceable, negative consequences
can produce rapid, lasting changes. As long as the
aversive outcomes of aggression override its

benefits, it will not be utilized. If more effective
options are practiced during this period, the
behavioral changes initiated by punishment are
likely to endure after the negative sanctions have
been discontinued. For persons who have
perfected mainly aggressive ways of achieving
desired results, punishment alone produces, at

best, selective inhibition. Disapproved actions will
be suppressed in the punisher’s presence but
readily performed toward others and at times and
places where the probability of adverse
consequences is low. In treating aggressors with

skill deficiencies, punishment serves as a
temporary inhibitor of injurious behavior while
rewardable competencies are being established
and strengthened.
Reduction of Aggression by Reward Withdrawal
Reinforcement withdrawal has proved to be an
effective means of managing deleterious behavior

that impedes the aggressor’s own development or
seriously infringes on the well-being of others. In
work with children short periods of social
exclusion are typically used as negative
consequences for transgressive behavior. The
child is told in advance that whenever he

misbehaves he will be sent to another room for a
specified time. This is done immediately, naturally,
and in a firm but nonhostile manner. To avoid
arousing needless anger, only brief exclusion
periods of 10 to 15 minutes are employed, and the

timeout room is designed as a neutral rather than
an aversive place. The source of influence should
be the temporary loss of rewards rather than
devaluation or vindictiveness, which are likely to
create more serious problems than those the

intervention was intended to remedy. If, during
the exclusion interval, the person continues to act

aggressively, the separation period is extended
until after the behavior subsides. Under this type
of contingency self-control is readily established.
Social attention accompanying a disciplinary
intervention may unintentionally reinforce the
preceding transgressive actions. Change agents
therefore minimize social and verbal interaction

as much as possible while the negative sanction is
applied. They do not argue, moralize, or offer
apologetic explanations. When prearranged
contingencies are applied in a consistent,
nonpunitive fashion, a person is more likely to

regard the interventions as natural consequences
of his behavior than as arbitrary or malevolent
treatment by others. The effectiveness of a timeout
procedure is, of course, determined by the
rewarding value of the situation from which one is

withdrawn.
Judicious use of reinforcement withdrawal as

part of a broad program is illustrated in the
treatment of an autistic boy by Wolf, Risley, and
Mees (1964). In addition to grossly retarded social
and verbal development, the boy exhibited violent
temper tantrums during which he would inflict
serious injury on himself. After a tantrum he was
badly bruised and bleeding, and refused to sleep at

night, forcing one or both parents to remain by his
bed. Sedatives, tranquilizers, and physical
restraints were applied without success. When it
became clear that refusal to wear eyeglasses
(necessitated by removal of cataractal lenses)

might result in ultimate blindness, behavior
therapists were invited by the hospital staff to
devise a treatment program for him. Ward
attendants, and later the parents, carried out the
prescribed procedures under the guidance of the

consultants.
Most of the serious recurrent behaviors, which

obstructed any remedial efforts, were eliminated
by combined use of extinction and brief
separation. In modifying the tantrum behavior, for
example, whenever the boy slapped himself and
whined, he was placed in his room for ten minutes
or until the disturbing action ceased. Under this
contingency tantrums gradually declined and

eventually disappeared. Eating problems, in which
the boy grabbed food from others, threw it about
the room, or ate with his fingers, were rapidly
eliminated in a similar manner. The attendants
simply removed him from the dining room for the

remainder of his meal for snatching or tossing
food after a warning, and withdrew his plate for a
few minutes whenever he ate with his fingers.
Normal sleeping patterns were established within
a week by not acceding to his demanding

behavior. Finally, destructive discarding of glasses
ceased within five days when the boy was placed

briefly in his room for throwing his eyeglasses, or
if tantrums developed, until they abated.
The foregoing account describes the punishing
consequences for destructive actions. In the total
program, positive incentives were also extensively
used to get the boy to wear glasses, and he was
assisted and rewarded for positive

responsiveness. Prior to treatment, he lacked any
communicative skills. These were gradually
established by reinforced modeling. The
development of rewarding competencies
undoubtedly contributed to the effectiveness of

mild punishment.
As the boy’s social functioning improved,
contacts with the family and home were
progressively increased. At first the parents

visited the hospital and observed how tantrums
and bedtime struggles were handled by the
attendants. Subsequently the parents made

several weekly visits, during which an attendant
observed and instructed them in their handling of
their son. Then the boy began short home visits
accompanied by an attendant, followed by
progressively longer visits. Initially, the child
almost succeeded in reinstating control over the
parents by aversive tantrums, but with the

attendant’s corrective guidance coercion no longer
worked. After discharge, his extreme behavior
problems ceased, he had become increasingly
verbal, and the family interactions were
considerably more enjoyable. The reinforcement

practices were applied for several years in a
nursery school setting where the boy made
sufficient progress to enroll in public school
(Risley and Wolf, 1966; Wolf et al., 1967). Social
withdrawal procedures were used occasionally in

early phases to eliminate tantrum behavior,
exhibited when he was asked to perform

instructional tasks, and to control hurtful acts
toward other children.
Following the general rule that problems are
best treated in the environment in which they
occur, hyperaggressive behavior has been
successively eliminated in children by their
parents (Allison and Allison, 1971; Zeilberger,

Sampen, and Sloane, 1968), nursery school
teachers (Clark, Rowbury, Baer, and Baer, 1970;
Sloane, Johnston, and Bijou, 1967), and hospital
attendants (Bostow and Bailey, 1969), by
rewarding friendliness and punishing harmful

actions with brief social withdrawal. Figure 5.3
presents illustrative documentation of results
from treatments employing contingency reversals.
Similar interventions have yielded favorable
outcomes with older children when ignoring

troublesome behavior failed to produce results
(Sibley, Abbott, and Cooper, 1969; Sulzer, Mayer,

and Cody, 1968). As injurious behavior is
eliminated, positive interactions with peers
become reciprocally rewarding so that aggression

is rarely resorted to even after the negative
FIGURE 5.3 Rate of aggressive responding during treatment
periods when positive responses were rewarded and
injurious actions were punished by brief social exclusion.
The upper figure shows the reductions in physical
aggression achieved by a mother in her 5-year-old son
(Zeilberger, Sampen, and Sloane, Jr., 1968); the lower
figure presents the results for the treatment of an
assaultive adult by hospital personnel (Bostow and Bailey,
1969).
sanctions have been discontinued (LeBlanc, Busby,
and Thomson, 1971).
A variant of reward withdrawal is reported by
Keirsey (1969) for handling seriously disruptive
or injurious behavior that has resisted concerted

remedial efforts. The method essentially involves a
behavior contract that specifies clearly the
distressing actions requiring control, the

responsibilities that each participant must
assume, and the consequences that transgressions
will produce. In a joint meeting including the child,
his parents, his teacher, and the principal, the
supervising counselor enlists the child’s help to

control, as best he can, behavior that disturbs all
concerned. Whenever his aggressive actions
exceed acceptable limits, the teacher must ask him
to leave the school for the remainder of the day.
She is further instructed not to threaten, coax,
urge, or scold the child, nor to engage in
persuasive attempts to alter his behavior, all of
which could function as attentional rewards for
transgressing. Rather, the prearranged sanctions

are applied immediately and consistently so that
the child looks upon them as natural, predictable
consequences of his misdeeds. When the child
arrives home, the parents are told to keep him at
home until the time he normally returns from

school, but otherwise to refrain from scolding or
applying other disciplinary measures.
High rates of success are reported for this
method in reducing chronically disruptive

behavior (Blackham and Silberman, 1971; Keirsey,
1969). However, no quantitative results are
presented, nor have the conditions under which

this method is most efficacious been
systematically examined. An exclusion procedure
will prove ineffective, or may even augment
deviant responsiveness, if the situation from
which the individual has been removed is
unrewarding or unpleasant. Apparently this
problem has rarely arisen, which suggests that

even children who experience serious difficulties
in the classroom find attendance at school more
attractive than sitting alone at home. Results of
studies previously cited suggest that short periods
of exclusion might work as well as, or even better

than, prolonged suspension from classroom
activities. When misdeeds produce brief reward
withdrawal, children have repeated opportunities
to cope successfully with situational demands and
thus to improve their level of self-control. The

corrective power of even the extended exclusion
method could probably be increased by rewarding

achievements as well as punishing transgressions.
In most of the cases discussed thus far the
favorable changes induced by parents or teachers
are likely to be supported in peer relations. When
peer influences conflict with the behaviors
promoted by the staff, as in correctional
institutions, negative sanctions achieve, at best,

only temporary control over aggressive behavior.
Tyler and his associates (Burchard and Tyler,
1965; Tyler and Brown, 1967) greatly reduced
aggressive behavior in delinquent boys by means
of brief social exclusion without any threats,

invectives, sermonizing or negotiations for a
second chance. However, when the negative
sanctions were removed, the rate of offenses
increased rapidly. After the punishment
contingency was again reinstated, disruptive

behavior subsided with equal rapidity and
remained at a low level. In a subsequent study,

Brown and Tyler (1968) succeeded in dethroning
a tough delinquent who exercised powerful
intimidating control over his peers by excluding

him whenever he or his cohorts forced
subservience from weaker members. The costs of
autocratic delinquent control proved too high to
continue it. Since he was held accountable for any

trouble that might develop in the cottage, he not
only ceased promoting disruptive behavior but

tried to dissuade others from it. Given a powerful
delinquent system, however, punishment may
downgrade one domineering leader only to
elevate another one without restructuring the

peer reinforcement system to any significant
degree.
Aversive control can be an effective method for
managing aggressive behavior in residential

settings, but it is unlikely to have much
rehabilitative value in correctional institutions as
they are constituted at present. If negative
sanctions applied by staff members are strong
enough to outweigh contravening peer influences,
socially conforming behavior may be achieved and
sustained as long as the institutional sanctions
remain in effect. When aversive controls are
removed, however, peer reinforcement practices

usually reinstate deviant patterns of behavior. It
will be shown later that modification of antisocial
aggression requires fundamental changes in the
treatment systems themselves.
Reduction of Aggression by Aversive Consequences
Physically painful consequences are

sometimes applied to behavior so destructive to
oneself or to others that it calls for drastic

measures. Such behavior is most frequently found
among institutionalized retardates and those
suffering from gross behavior disorders. Ludwig,
Marx, Hill, and Browning (1969), for example,

curbed potentially homicidal behavior in a
schizophrenic woman who terrorized fellow
patients and staff alike by surprise vicious attacks
upon them. After she had received every variety of
psychiatric therapy to no avail, an effort was made
to control her violent actions through aversive

consequences. The aggressive episodes generally
followed a common sequence: When asked to do
something, the patient would become surly and
resistive. If pressed further, she would
successively cry, complain angrily that she was

being persecuted and abused, and then threaten to
harm or to kill the staff. After voicing threats she
would calm down, causing the staff to lower their
guard, whereupon she would strike out without
warning. In the treatment program physical

assaults, aggressive threats, and accusatory
verbalizations were successively eliminated by

administering a brief shock via a faradic
stimulator upon their occurrence. As her violence
was brought under control, the quality of her

interactions changed from paranoid
combativeness to a more warm and friendly
responsiveness. Her general level of functioning
also improved substantially. In many respects, the

treatment achieved what the patient perceived its
purpose to be: “You’re trying to make a human

being out of me.” Unfortunately, sometimes this
requires painful interventions.
Results of a study by Agras (1967) suggest that

violent behavior may, in some instances, be
controllable by eliminating its ideational
precursors. A chronic schizophrenic had to be
physically restrained because of uncontrollable
tendencies to smash any glass in sight. He

participated in a series of sessions during which
he visualized himself breaking glass, whereupon
he was administered a painful shock. As the
treatment progressed, the latency of the
destructive imagery increased and eventually he
lost all urge to smash glass. A follow-up study
revealed that, except for one minor incident, the
destructive behavior never appeared again.
One of the most perplexing and dangerous

disorders involves self-injurious behavior. In its
more extreme forms, children pummel their faces
repeatedly, bang their heads forcibly against hard
or sharp objects, tear and bite off pieces of flesh
from their bodies, or exhibit some other type of

self-mutilating behavior. Because of the serious
risk of permanent physical injury, such children
are usually kept in continuous physical restraints.
The maintaining conditions of self-injurious

behavior are not fully understood, but it is
amenable to control through variation of
reinforcing consequences. In a series of studies
Lovaas and his colleagues (Bucher and Lovaas,
1968; Lovaas, Freitag, Gold, and Kassorla, 1965;
Lovaas and Simmons, 1969) have shown that
neither social deprivation nor social satiation have
any appreciable effect upon self-mutilating
responses. Contingent rewards, however,

increased such actions, especially if given only
periodically, as typically happens in understaffed
institutions. Thus, self-injurious responses
increase when they elicit solicitous attention and

reassuring comments from adults, whereas such
behavior is promptly reduced to its original
baseline level after the positive reactions are
discontinued. Self-injurious responses are also
readily cued off by events signifying withdrawal of

social reinforcement for other behaviors.
Experiments with primates reported by Schaefer

(1970) shed further light on reinforcement control
of self-inflicted injuries. He was able to establish

and to maintain a high rate of head-hitting by
rewarding such performances with food initially

for each self-hit and later for every fifth response.
Moreover, by rewarding self-hits only when the
experimenter emitted sympathetic remarks, but
not in the absence of these words, self-hitting was

brought under verbal stimulus control. Self-hitting
could now be promptly evoked by the verbal cues

alone.
Although self-injurious behavior may be partly
maintained by its social consequences, the

evidence is conflicting on this point. Subtle
changes in social reinforcement, as when an adult
remains with a child without attending to him
when lie hits himself, does not seem to affect the
rate of self-injurious acts (Lovaas et al., 1965;

Risley, 1968). Withdrawal of social contact for a
brief period reduces self-injurious behavior, but
occasionally children are unaffected even by these
more powerful consequences (Risley, 1968).
Whatever the controlling variables might be in any
given case, preliminary findings indicate that
chronic and intractable self-injurious behavior can
be successfully eliminated with beneficial results
by either ignoring or punishing it.
Lovaas and Simmons (1969) showed, with
children who had been in continuous restraints,
that if self-injurious behavior produces no social
effects, it decreases gradually and eventually
disappears. In some cases, however, when self-

mutilating behavior is first placed on extinction it
may temporarily increase in intensity, which could
be potentially dangerous. Extinctions therefore
cannot be safely employed unless adequate

safeguards are available (Harmatz and
Rasmussen, 1969). Self-destructive responses can
be promptly and completely eliminated by

contingent application of a few painful shocks. In
one of several cases reported by Bucher and
Lovaas (1968), a seven-year-old schizophrenic
boy who had been self-injurious since the age of
two performed approximately 3,000 self-
pummeling responses during a period of 90
minutes when his physical restraints were

removed. This behavior was almost totally
eliminated in four sessions by the use of 12
contingent shocks. Self-destructive behavior that
had persisted over a period of six years was also
rapidly and durably removed in a schizophrenic

after she received a total of 15 shocks for beating
her head. In each case reported, contingent
aversive consequences not only removed self-
mutilating behavior, but the children decreased
their whining and became more responsive to the

therapists. Uniformly favorable results have been
obtained with other retardates and autistic

children who for years suffered severe self-
inflicted injuries. Cessation of self-injurious
behavior is often specific to the situation in which

punishment is administered. Generalized
reductions can be achieved, however, by a few
punishments or verbal reprimands in other
settings in which the responses occur.
Brief social withdrawal made contingent on
self-destructive behavior is sometimes sufficient

to remove it permanently. Hamilton, Stephens, and
Allen (1967) were consistently successful in
eliminating injurious aggressive and self-

destructive behavior in severely retarded
adolescents by social exclusion. In each instance,
the individual was physically confined to a chair in
a timeout area for a fixed period following an
injurious act. In one case, a girl beat her head and

back against the wall a total of 35,906 times
during four 6-hour observations. While the
timeout contingency was in effect, head-banging
precipitously dropped to a negligible level of 7, 2,
0, 1, 0 for five successive weeks and never
reappeared during nine months of follow-up
study. Of considerable import, both from a clinical
and an ethical standpoint, is that after the
continual self-mutilating behavior was eliminated,

the children participated with evident enjoyment
in daily social and recreational activities.
It is noteworthy that brief programs of
contingent punishment can remove self-
destructive behavior of long standing and improve

social functioning as well. After self-injurious
behavior is eliminated, children generally become
more socially responsive and display increased
imitativeness, thus enabling them to acquire new

patterns of behavior. If further studies support
this conclusion, there will remain no justification
for having children mutilating themselves or

spending their early years uselessly in physical
restraints.
Several factors may account for spontaneous
improvements in social behavior following
reduction of aggression. One explanation is in
terms of the self-enhancing effects of increased

freedom of action. When aggressors discard
provocative activities that consume much time,
they begin to make contact with potentially
rewarding aspects of the environment that were

always available and that automatically reinforce
emergent patterns. A person’s environment
expands markedly after he has been freed from
the restraints imposed by his aggressiveness
toward himself or others. Not only is contact with

positive features of the existing environment
increased, but reduction in violent conduct

changes the nature of the social environment.
Others become more friendly when they are no

longer being frightened or angered. Positive
responsiveness, in turn, promotes cordial

counterreactions. Once a reciprocally reinforcing
interaction is initiated between nonviolent
behavior and environmental rewards, widespread
changes may result even though amicable conduct

was never deliberately cultivated.
Another issue requiring comment is why

relatively mild punishment used therapeutically
can achieve lasting results, whereas even drastic
punishment often fails to suppress aggressive

conduct or even aggravates it. The discrepancy
may be explained in several ways. In treatment,
unpleasant consequences are used constructively
to improve social functioning rather than
vindictively or devaluatively. When aversive

procedures are essential for alleviating
detrimental conditions, they are generally viewed
as an unpleasant though necessary part of
treatment, comparable to the painful routines of
physical medicine. If the therapeutic intent is
clearly conveyed in the context of positive
alternatives, negative sanctions will have fewer
adverse side effects than when punishment is
applied hostilely merely to suppress troublesome

behavior with high functional value for the
performer. The same punishing consequences may
therefore be strongly resented or willingly
accepted depending upon the perceived intent of
the agent, and whether the sanctions are applied

mainly for his own convenience or for the benefit
of the recipient. If the initially painful
ministrations of dentists and surgeons were seen
as punishment for developing maladies, these
practitioners would be feared and hated rather

than sought out.
Undesirable emotional effects can be

substantially reduced by arranging in advance
explicit contractual contingencies that clearly
define the broad limits of permissible and
punishable behavior. Prearranged contingencies
implemented in a responsible fashion will more
likely be regarded as natural consequences of
transgression than as arbitrary interpersonal

assaults.
The power of punishment to create
dysfunctional fears or defenses depends on the
total context of experiences in which it appears.
Occasional painful experiences occurring among

many neutral or rewarding ones are generally
ineffective in creating inappropriate fears
(Rescorla, 1969). If change agents did little more
than punish aggressive behavior, they would

undoubtedly take on frightening or angering
qualities. In well-designed treatment programs
they, in fact, are highly supportive and rewarding

so that a few circumscribed sanctions are unlikely
to jeopardize the usually positive relationship.
Punishment by an otherwise warm, rewarding
person is more effective in reducing prohibited
behavior than that by a nonnurturant one (Parke,
1969; Parke and Walters, 1967).
A third possible explanation concerns the

manner in which negative consequences are
applied. Ordinarily the contingency, certainty, and
immediacy of punishment are more important
than its severity. In therapeutic practice
consequences are administered immediately and

consistently for clearly designated transgressions.
If generalized reduction in prohibited behavior is
desired, the negative sanctions are applied in
varied contexts by different individuals. In a

comparison of certainty and severity of
punishment, Pendergrass (1971) found that brief
seclusion for each interpersonal attack was more

effective in eliminating assaultive behavior than
long seclusion applied periodically. Even weak
punishment used in conjunction with rewarded
alternatives can produce enduring positive
changes where severe punishment alone fails. An
additional factor contributing to the relative
success of remedial sanctions is that rewards for

aggressive behavior are removed as much as
possible, whereas in usual efforts at social control
punishment is simply superimposed upon
rewards that remain operative. Any of these
features may account for variations in the

corrective power of punishing consequences.
MODIFICATION OF ANTISOCIAL
AGGRESSION
Most youngsters who repeatedly engage in
delinquent conduct eventually appear in

correctional institutions. Such centers generally
provide adequate custodial care, but they are
severely limited by their structure and

institutional practices in creating the kinds of
experiences conducive to constructive personality
change. Most are admittedly oriented toward
custody and make no pretense at remediation.
However, Polsky (1962) among others, documents
how even well-intentioned treatment centers can

become agencies for antisocial learning.
A sizable group of delinquents is typically
assigned to a living unit with one or two adult
supervisors. Within this social context a powerful
peer subculture is formed containing status

hierarchies based upon delinquent values and
skills. Social ranking and control are achieved
largely through intimidation and physical
domination. Tough boys and their cohorts, having

gained power through forceful means, bully and
exploit weaker members who become subservient.
Their victimization is usually facilitated by

devaluing them as “queers” or subhumans.
Newcomers are subjected to similar humiliating
treatment from which they can escape by
successful aggressive challenges of higher ranking
members. Under a reign of force victors perfect
aggressive styles of behavior, while the
downgraded are demoralized.
The peer system receives support from various
sources. The prevailing code of conduct effectively
shapes and controls its members in directions that
often subvert the treatment efforts of the staff. A
contingency analysis by Buehler, Patterson, and

Furniss (1966), for example, revealed that
delinquents generously rewarded antisocial
conduct but disapproved of actions that either
deviated from their own norms or conformed to

institutional standards. Nor does the residential
staff escape the boys’ coercive control. Being
greatly outnumbered, supervisors need the

cooperation of peer leaders to maintain order in
residential units for which they bear
responsibility. Supervisors are therefore inclined
to patronize the delinquent leadership by
accommodating to their reinforcement practices
(Polsky, 1962). At periodic intervals the
professional staff engage the boys in therapeutic

conversations, while in their daily cottage life they
are mutually reinforcing the very attitudes and
behaviors that led to their incarceration.
Upon their return to society many continue a
life of aggression that, sooner or later, sends them

to more callous institutions in the correctional
hierarchy. Eventually they end up serving long
terms in walled prisons under subhuman living
conditions. During this confinement prisoners are

repeatedly subjected to degrading and brutal
treatment at the hands of their keepers or fellow
inmates.
Almost everyone acknowledges that the
present correctional practices are antiquated. High
recidivism rates attest to the fact that they do not
accomplish the purposes by which they are
justified. If anything, the dehumanizing prison
experiences create obstacles to constructive

change. It is incongruous that a society remains
willing to underwrite, at enormous human and
financial costs, penal systems that are admitted
failures. Although the need for drastic reforms is

repeatedly voiced by insiders and outsiders alike,
the corrosive practices remain. Responsible
indictments produce inaction perhaps because,
despite the avowed aims, the public is more
interested in punishing than in rehabilitating

transgressors. One can point to evidence that
public figures are applauded by large segments of

the population more enthusiastically for
advocating punitive controls than for humane

remedial measures, which get criticized as
“coddling.” The public costs of punitiveness are

too remote and obscure to be seen as doing more
social harm than good.
Another explanation is that dysfunctional

systems persist because they are not held
accountable for the quality of their services.
Agency practices tend to be adopted for the

convenience of those who run the place at the
expense of client benefits. As long as operating
funds are not linked to accomplishments, there is

little challenge for change. Popular sentiment
regarding criminal offenders reinforces inaction
even by dedicated personnel with the best of
intentions. The remedial gains that accrue from
lifting punitive restraints receive little notice,

whereas occasional failures are widely publicized,
often with political repercussions. This is hardly
the climate conducive to innovative practice.
Rather, it requires special effort to introduce
reforms that carry some risk when efficient,
though nonrehabilitative, custody suffices. Major
accomplishments in the correctional field require
a system of accountability that rewards program
improvement. External audits of correctional

practices and their degree of success would
provide motivating incentives to devise more
promising approaches.
It is difficult to alter huge malfunctioning
agencies by internal modification alone because

many of the functionaries in the system either
have a vested interest in the existing procedures
or are fearful of trying unproven alternatives.
Agencies can be changed much faster by devising

successful programs on a limited scale outside the
system, and then using the power of superior
alternatives as the instrument of influence. If

officials are held answerable for the results of
their treatments, ineffective practices cannot be
long defended after new ones have been tried and
proved superior.
Amelioration of antisocial aggression requires
changing not only correctional systems, but also
weak psychological theories that prescribe the

remedial measures. Interventions must be
reoriented from probing for internal

malfunctioning to social treatment. By altering
social practices in the family and school and
changing community influences that foster
aggressive behavior, the development of

assaultive personalities can be largely prevented.
Tharp and Wetzel (1969) provide many examples
of the techniques of consultation and intervention

in the natural environment. Serious forms of
aggression that may initially require residential

treatment will, of course, occasionally arise.
Home-style remedial programs based on
learning principles are being developed as an
alternative system to existing correctional
facilities. Functioning units within the community
contract with the court to treat juvenile offenders
who would otherwise be sent to reformatories. In
the prototype program (Phillips, 1968; Phillips,

Wolf, and Fixsen, 1972), a pair of house parents,
skilled in methods of behavior modification,
resides with a small group of delinquent boys in a
home located in the community. Beneficial
changes are fostered initially through a system of

immediate tangible rewards and later by more
intrinsic and natural consequences. A positive
learning environment is created in which boys
earn points for designated social behavior, self-
care, and academic activities that enhance their

competencies, and lose points for transgressive
actions. Various treats and privileges are secured

with the points, first on a daily basis and later on a
weekly arrangement.
Contingency reversals demonstrate the
necessity of concrete rewards in early phases of
the remedial program. Boys adopt and maintain
constructive patterns of behavior when they result
in point consequences linked to deferred rewards,

whereas they quickly revert to their old patterns
when all privileges are made available
noncontingently (Phillips, 1968). Home-based
reinforcement is likewise highly effective in
improving class-room performance (Bailey, Wolf,

and Phillips, 1970). In the absence of special
incentives boys spent about 65 percent of their
time disrupting classroom activities and only 35
percent of their time studying. In contrast, when
they earned privileges in the home for working

productively in school and lost privileges for
disrupting the place, they studied better than 90

percent of the time, a notable achievement for
students formerly regarded as unmotivated and
uneducable.
Contingency systems can vary on several
dimensions, depending on whether performance
tasks are assigned to individuals or if the group
decides on the distribution of labor; whether

reinforcements are given for individual
performance or for the group’s combined
achievements; whether rewards and privileges are
tied to specific performances or to levels of
functioning encompassing multiple competencies;

whether the system rewards desired
performances, punishes transgressions, or
arranges both types of consequences; and whether
the contingencies are implemented by the staff or
by peer members themselves.
Researchers have begun to examine how
different organizational practices affect group

functioning and satisfaction. The merits of a given
system are measured against four standards. It
should be effective in achieving changes in
psychological functioning. It should be preferred
by the participants. It should be practical in that
participants assume responsibility for regulating
their activities with minimal adult supervision.

And it should be educational by developing
competencies that will be useful for the
participants in the future. Phillips, Wolf, and
Fixsen (1972) found that a system combining
group assignment of tasks with group

consequences was of limited effectiveness. When
performance by any member benefits the entire
group, individual members come to expect
somebody else to do the work, with the result that
both individual and group performances

progressively deteriorate. Individual assignment
with group consequences produced somewhat

better results, as did completely individualized
systems in which adults assigned tasks to each boy
and rewarded his individual accomplishments.
One limitation of an exclusively adult-
implemented program is that it places the burden
for managing conduct on the staff, which tends to
create antagonistic positions. Greater progress can

be achieved if peers are enlisted to further the
values and goals of the treatment program. In this
regard, Phillips, Wolf, and Fixsen (1972) report
that the contingency system which proved most
effective and which was clearly preferred was one

in which the boys periodically elected a peer
leader who assigned required tasks to them
individually and rewarded or penalized them
according to their performance. To ensure that the
leader used his delegated power constructively, he

earned or lost points depending on how well those
he supervised carried out their obligations. Apart

from its favorable short-term results, a self-
governing reinforcement system has greater
educational potential because it provides

experience in responsible and cooperative social
living. As a general rule, contingency systems
combining rewards and negative sanctions were
not only more effective but, surprisingly, were

liked better by the boys than ones relying solely
on rewards. An environment in which individuals

can engage in transgressive behavior with
immunity apparently holds lesser satisfactions for
its members.
There is no single optimal system because

different goals and variations in the functioning
levels of groups require different reinforcement
practices. A completely individualized system is
well suited for creating autonomous, self-

determined people. If, on the other hand, one
wishes to promote shared responsibility and
contribution to common goals, then this objective
can be best accomplished by instituting
reinforcement contingencies on a group basis.
Here, individual benefits depend upon the level of
group performance; conversely, with shared
accountability censurable conduct by any given
member produces negative consequences for the

entire group. By having people share in the
consequences of their decisions and actions, their
level of social responsibility and involvement is
increased. The negative by-products of excessive
individualism or collectiveness can be minimized

by including both individual and group-oriented
reinforcement. In a double contingency system, a
given member’s outcomes are determined by both
the extent of his own contribution and the group's
overall attainments.
At higher levels of functioning, rewards and
privileges are usually tied to ranks rather than to

individual performances. In areas requiring
specialized skills social ranking is conducted
formally, but in many situations members are
differentiated in status informally. In applying
hierarchical contingency systems, successive
phases that require progressively higher levels of
functioning in a variety of activities are devised. As

individuals progress through these sequential
steps by developing the requisite skills, they
receive increased rewards and freedom of action.
In the study cited earlier, Martin and her
associates (1968) found that a hierarchical reward

structure was superior to one in which specific
performances were individually reinforced.
Indeed, the latter arrangement produced much
wrangling over unfairness because, in an effort to
support progress by youngsters at different

competence levels, they received comparable
rewards for differing achievements.
Programs aimed at increasing social and
intellectual competencies in persons lacking self-
regulatory skills may initially require immediate
reinforcement of specific performances. However,
as progress is achieved, individuals are promoted
to phase systems analogous to those existing in
community life. In the treatment approach

developed by Phillips, Wolf, and Fixsen (1972),
after youngsters increase their competence
sufficiently, they are advanced to a higher system
in which all privileges are given free as long as
they exhibit responsible behavior. Should they

revert to reprehensible conduct, they are returned
to the more concrete system until they are better
able to manage increased independence. When
members reach the stage at which they can
maintain self-directive behavior under natural

consequences, they return home or, if that
possibility does not exist, are placed in a foster

home. Parents receive some training in
constructive ways of relating to their children
before they are discharged.
If the promise of home-style treatment is
confirmed by follow-up studies, it will gain
acceptance as a more effective and humane
alternative to incarceration. This approach holds

additional advantages. In reorienting life styles
one is dealing with morals and values. Although
heterogeneous communities have many shared
values, different segments of the population favor
different conduct norms and ideals. Because of

their small size and flexibility, home-style,
community-based programs allow for the
modification of antisocial aggression within
multivalued frameworks.
Institutional Remedial Systems
Most correctional institutions are conducted
on a contingent-punishment, noncontingent-
reward basis. That is, the residents obtain
whatever rewards and privileges the place has to
offer as long as they comply with the house rules,
but the privileges are promptly withdrawn for
uncooperative or disruptive behavior. Thus, the
threat of punishment is ever present, whereas the

positive incentives for behavior change, though
available, are poorly managed. Under such
circumstances, most participants comply
halfheartedly with the minimum demands of the
institution in order to avoid penalties for any

breach of the rules.
In an effective program of change,
reinforcement contingencies should be arranged
to create competencies for future use, rather than

to extract minimal compliance with situational
demands. All residents receive humane treatment
on a noncontingent basis, but they can earn valued

rewards and privileges for progress in acquiring
educational, vocational, and social skills that
create effective alternatives to antisocial conduct.
A remedial program so structured can be managed
without resort to those punitive measures that are
commonly employed in penal institutions. By
placing the residents’ living conditions on a self-

determining basis, the system not only provides
positive incentives for constructive change but
renders nonfunctional peer control practices that
subvert remedial efforts. When contingencies are
arranged so that constructive behavior secures

good results and antisocial conduct has limited
payoff, little is to be gained from outwitting staff
members or provoking them by crisis-producing
transgression.
Competence Training
People are unlikely to adopt or to adhere to
socially acceptable life styles if they lack legitimate

means of gaining the things that are highly valued
by the culture. Because of adverse learning
histories and wasted years of incarceration, most
inmates in correctional institutions are
handicapped by gross deficits in educational and
vocational skills. As a consequence, those who lack

the necessities of life or who value costly
possessions and the benefits of high status are
forced to resort to deviant means of getting them.
Regardless of whether other objectives may be
selected in rehabilitative programs, little progress

can be made unless habitual offenders develop
broadly useful skills that enable them to
participate successfully in the larger society. The
program devised by Cohen and Filipczak (1971)
demonstrates how a custodial environment can be

transformed into one that develops human
potentialities by restructuring institutional

reinforcement practices. This program focused
primarily upon development of basic intellectual
skills required for expanding one’s capabilities.

Although the learning environment was created
within the context of a penal institution, the
procedures developed have general applicability.
From the reformatory Cohen and Filipczak
selected a group of delinquent boys who had few
means for maintaining their subsistence other

than through antisocial activities, and they failed
even at these. A positive incentive system based
on a point economy was employed to foster

acquisition of language facility, reading,
mathematics, and other functional skills. Money
points earned for learning achievements in self-
instructional courses could be used to purchase
preferred accommodations and a wide range of

services and commodities.
Most coercive sanctions imposed on inmates in
prisons were eliminated. Rather, heavy reliance
was placed on positive incentives, individual
initiative, and self-determination. Consistent with
contingency systems in daily life in the society,
boys paid for their private rooms and selected
meal plans that varied in choice of foods. They also

used point currency to rent recreational items and
private offices, and to purchase phone privileges,
admission to entertainment activities, and
merchandise available in either the project store

or through mail order catalogs. To help the boys
learn to manage their finances, a banking system
was included so that they could establish savings
accounts with interest rates and negotiate loans to
tide them over critical periods. Those who

presented an inadequate work record were told
that they constituted poor loan risks and were not

granted loans until they established a better credit
rating. Within this environment the boys were

provided with considerable freedom: They
determined their own study and bedtime routines;

they selected their own leisure-time activities and
planned their outside programs; they aided in
planning menus; and they had open mailing and
visiting privileges.
A boy was free to choose not to pursue any of
the self-instructional activities. If, however, his

points fell below a minimum level, he lost his
private room, was served the regular institutional
food, and enjoyed none of the extra luxuries

earned by his peers. This problem rarely arose. In
an environment designed to reward learning,
delinquent boys, who had received few prior
accolades for scholarship worked productively at
self-managed educational activities. They studied

conscientiously in their spare time, gained more
than two grade levels on standard achievement
tests within an eight-month period, and
substantially raised their intellectual level as
measured by intelligence tests. By making benefits
attainable through their own actions, rather than
by management fiat, the boys had few reasons to
create disciplinary problems. If they lacked things
they wanted, they had mainly themselves to

blame.
The foregoing contingency system fostered
intellectual development and, although favorable
changes in attitudes and social behavior were
noted, no concerted effort was made to increase

interpersonal competencies. Contingencies can, of
course, be easily extended to include social skills
and self-discipline which supplemented with
proficiency in a selected occupation, would

remove the major personal barriers to a prosocial
mode of life. Sarason and Ganzer (1969) report
preliminary studies that make use of modeling

procedures with delinquents to develop social and
other skills requisite for effective vocational and
interpersonal functioning. Models demonstrate
and then the boys practice proficient ways of
handling common problem situations such as
coping with authorities, negative peer influences,
vocational demands, situations requiring self-

control, and a variety of social predicaments.
Delinquents who have the benefit of constructive
modeling achieve greater changes in attitudes and
behavior than boys who do not receive such
training.
Deviant behavior can be modified in

custodially oriented facilities through differential
reinforcement. When good conduct is imposed
from without, the resultant conforming behavior is

likely to persist as long as the institutional
sanctions remain in effect. The residents may, in
fact, come to behave irreproachably or even to

perform obligingly whatever behavior is expected
of them in order to improve their institutional
subsistence and to expedite their release. A
beneficent incentive system may thus extract
considerable prosocial behavior from delinquents,
but they may revert to their usually antisocial
conduct in situations in which it is permissively

viewed. Therapeutically oriented facilities that
arrange outcomes on a self-determining basis
promote responsible rather than compliant
behavior.
The competency training approach assumes

that by increasing a person’s options, a prosocial
way of life will prove more functional and
supplant antisocial styles of behavior. Certainly a
person who has a number of effective alternatives

will be less likely to revert to illegitimate activities
than if he is unemployable except for menial jobs
and is unskilled at handling interpersonal

problems. Competency training, though essential,
is generally insufficient of itself. In the
demonstration project conducted by Cohen and
Filipczak (1971), for example, delinquents who
participated in the incentive program displayed
substantially lower recidivist rates than boys who
received the regular penal treatment in the same

institution, but over a period of years, the
differences between the groups decreased. A
person who has pursued a delinquent career for
many years will find it difficult to alter his social
orientation, especially if he can gain recognition

easily by antisocial conduct while alternative
avenues entail unpleasant responsibilities with
uncertain or deferred rewards. A treatment
program is likely to achieve greater success if, in
addition to cultivating competencies, it changes

association preferences and reduces susceptibility
to the influence of delinquent peers.
Development of Self-regulatory Functions
The sway of external reinforcement can be
diminished by establishing countervailing self-
reinforcement functions. When new standards for
self-evaluation are adopted as self-regulative
influences, people adhere to prosocial styles of

behavior because they derive self-satisfaction
from them; conversely, they refrain from
antisocial conduct because it produces self-
devaluative consequences.
Social learning theory suggests several

procedures for developing self-regulatory
functions. First, valued patterns of behavior and
standards for self-evaluation should be
exemplified by change agents. Second, an explicit

set of conduct guides linked with a graded system
of incentives should be instituted, such that
progressive adoption of more advanced behavior

brings increased freedom, privileges, and access to
rewarding activities. Participants will adopt new
forms of behavior on the basis of their utilitarian
value. However, provision of exemplary models
and positive inducements alone may not be
sufficient to create new self-directing capabilities
in persons whose behavior has been governed by

dissocial standards.
Conditions that provide explicit training in self-
reinforcement should be arranged. This is
achieved by gradually transferring evaluative and
reinforcing functions from change agents to the

individual himself. At the first transition stage the
person judges whether his behavior warrants
reward, but others still serve as the reinforcing
agents. After accurate self-evaluation is

established, the reinforcing function is likewise
transferred so that the individual both evaluates
his own behavior and reinforces himself

accordingly. Concrete rewards are gradually
reduced as the person’s behavior is increasingly
regulated by social, self-evaluative, and symbolic
consequences. The ultimate aim of training in self-
determination is to produce a level of functioning
at which participants regulate their own actions
with minimum external constraints and artificial

inducements.
Another means of instilling self-regulatory
functions is to provide ample opportunities for
participants to perform role behaviors with peers
that are ordinarily carried out by the supervisory

staff. Specifically, this entails delegating
progressively more of the standard-setting,
evaluative, and reinforcement functions to
members of the group as they progress in the

program. Thus, with staff guidance, peers play an
active role in implementing the treatment goals. In
a system containing shared responsibility, games

designed to outwit the staff become pointless. In
order to enhance participants' willingness to
adopt role behaviors that carry bothersome
responsibilities, increased benefits must be
associated with these higher positions.
Values and preferences are changed more
effectively when members participate actively in

decision-making, reward their own
accomplishments, and discipline their
transgressions than when standards of conduct
are imposed by administrative authority on
covertly resistant individuals. Research evidence

further suggests that a person who takes on the
task of reinforcing the behavior of others is likely
to alter his own standards of self-reinforcement in
the practiced direction. When peers assume

responsible roles in their subcommunity it not
only fosters changes in themselves, but advances
the development of newcomers and subordinates

by providing success models for valued conduct.
Since peers are closer socially and more available
to serve as guides in daily life, they are likely to be
emulated to a greater extent than staff members.
Additionally, there is less risk of being ostracized
for adopting behavioral standards that are
modeled by peers and staff alike.
Change in Association Preferences
Just as self-regulatory functions are socially
transferable through modeling and reinforcement,
they are also extinguishable unless given sufficient
social support. In any given community an

assortment of subgroups, which differ widely in
their standards of conduct, are potentially
available to individuals. The groups with which

one chooses to affiliate largely determine the role
models and sanctions to which one will be
exposed and, consequently, the direction in which
one’s behavior will be further modified. Several

researchers have documented how selective
association shapes the course of one’s life
(Bandura and Walters, 1959; Ellis and Lane, 1963,
Krauss, 1964). Attention to factors governing the
selection of reference groups is of critical
importance, especially when newly adopted

patterns conflict with the behavior reinforced by
former associates.
New competencies and self-evaluative
standards can exert decisive influence on whether
participants gravitate toward groups that

reinforce prosocial or antisocial pursuits.
Moreover, given increased options, which reduce
dependence on any single group, and incompatible
self-evaluative reactions, a person is better able to

resist pressures for conformity to antisocial
values. When the advocated conduct violates self-
prescribed codes, the individual may attempt to

alter the value placed on the activities, may remain
a marginal member, or, if the rewards for
membership are insufficient, may discontinue his
association with the group.
If delinquents are to affiliate with prosocial
groups, they must possess the means for securing
the acceptance and recognition necessary to

sustain their active involvement. Otherwise, they
will be unable to fulfill the new roles and will

either eventually withdraw from or be rejected by
their new associates. Many rehabilitative
programs concentrate on changing offenders'
behavior in ways that lose them the satisfactions

associated with a deviant career without
providing alternative means of obtaining
gratifications. An effective remedial program

should establish the entry skills and provide
whatever initial support and guidance is needed to

succeed in the new milieu. After the person has
developed a rewarding stake in the system, he is
unlikely to revert to antisocial behavior, especially

if it carries punishment costs, as it does for those
who do poorly at it.
CHANGES IN SOCIAL SYSTEMS
It would be short-sighted to concentrate solely
on the casualties of detrimental social systems
without changing the conditions that foster

widespread aggression. In the preceding
discussion guidelines were presented for changing

the practices of subsystems that can affect the
level of aggression within a society. The present
section is mainly concerned with the use of
collective action as an instrument of social change.

Interventions at the social systems level require
group action against the individuals and
institutions contributing to social problems.
Collective action usually fails, however, because it
is poorly organized, because efforts are dissipated

on too many different problems or on factional
disputes over leadership, and the group pressure
is not sustained. Consequently, the sufferers of

malfunctioning systems come to display much
discontent and cynicism but little social action.
Social change must be approached in a number
of different ways. At the neighborhood level,

residents can reduce the amount of destructive
behavior that plagues their community from
within by instituting sanctions against such
conduct. As we have already seen, aggression
stems from its functional value as well as from

aversive instigating conditions. People who lack
power cannot easily change adverse social
conditions, but the reinforcements that their
community provides for aggressive conduct lie

within their sphere of influence. Accordant
community functioning is something that is
maintained to a large extent by the general

expectations and reinforcing reactions of residents
toward each other rather than imposed from
without by enforcement agencies. Most aggressive
offenses in fact lend themselves more to informal
control than to legal sanctions, which are applied
too infrequently to serve as reliable deterrents.
Community control of aggression through shared

enforcement of sanctions is illustrated by
Christopherson and Risley (1972). They assisted
residents of a low income housing project in
curbing vandalism by having them articulate a set
of conduct codes and impose consequences on

destructive offenders. Other problems of common
concern can be similarly alleviated by building
into social relationships a sense of community
through neighborhood reinforcement practices.
Institutions designed to serve the community

must be held more accountable for their practices.
Given our present knowledge, educational

systems, for example, should not be turning out
sizable numbers of students so lacking in basic
skills that their choices of livelihood are
essentially restricted to menial pursuits, some
form of relief, or a life of crime. Results reported
earlier and elsewhere (Bandura, 1969a)
demonstrate that methods exist for creating

learning environments that can transform
academic failure into success. If students are to
acquire requisite skills for participating
successfully in the society, communities must
institute a system of accountability in which

financial support is tied to demonstrated
educational effectiveness. When educational
agencies are held answerable for their results,
instructional systems rather than students are
failed.
The system of accountability should be
extended to other community services as well. In

public agencies that enjoy monopolies over given
functions, the practices that evolve are more likely
to serve the interests and convenience of the staff
than to maximize benefits for their clientele.
Internal rewards for conformity to operating
procedures mitigate against external efforts to
produce change. Improvements in social services

are achieved more rapidly by rewarding superior
accomplishments of alternative systems than by
imposing negative sanctions for poor performance
by a single system. Many community services
could be placed under a voucher system in which

citizens are given certificates to purchase the
service from any one of several authorized
sources. Providing options creates strong
incentives for organizations to increase the
effectiveness of their operations, for otherwise

they lose their patronage. When functions are best
conducted through a single agency, bureaucratic

barriers to change can be lessened by creating a
semi-autonomous unit under public jurisdiction
empowered to devise better ways of doing the job.

Procedures for measuring how well an agency is
meeting stipulated performance standards should
be an integral part of the system. Performance
assessments that have budgetary or staff

consequences provide additional inducements to
improve the quality of services.
Maltreatment is most prevalent where marked
imbalances of power exist. In seeking
improvements in their life situation, neglected and

disadvantaged people are faced with the difficult
task of altering the practices of those who wield
power and do not wish to relinquish it. In such
situations change can be pursued through the
organized strength of numbers.
The first task in the agenda for social change is
to gain broad participation of people in the
collective endeavor. Organizational methods vary
depending on whose support is sought and
whether influence will be exerted indirectly
through moral persuasion or by use of social
power. Exponents of direct influence by power
tactics depend mainly on the efforts of the

aggrieved. The work of Saul Alinsky (1971), an
astute community organizer, exemplifies
successful mobilization of subordinated groups for
action against inequitable treatment. The

organizational efforts in this approach are aimed
at building community self-determination.
Organizers enter the situation as consultants by
invitation only. The residents play a major role in
selecting common goals for which broad support

is enlisted by appeal to their self-interest. To
ensure that local power bases survive over time,

indigenous leadership is developed along with
serviceable problem-solving mechanisms.
Reformers seeking basic changes in the society
at large rely on the force of intermediary

influences, which they attempt to mobilize largely
through persuasive appeals (Zaltman, Kotler, and
Kaufman, 1972). By arousing sympathetic reaction
in people with greater power to create rewarding

and punishing outcomes for officials, public
pressure is brought to bear on offending
institutions to change their policies. Whatever
direct actions are taken are designed more to
dramatize nationally the grievances and need for

reform than to solve a local problem expeditiously.
Tactical power is thus sacrificed for public
credibility. In this process of mobilization
communications media play a prominent role.
Over time the change efforts of subordinated

people are extinguished, with the result that most
become apathetic though discontented with their

way of life. A change program must therefore help
them learn how to exercise control over their
social environment. They need to practice skills in
decision-making, learn influence tactics and how
to tailor them to changing circumstances, develop
disciplined ways of reacting to counterpressure,
and allay their fear of adversaries. Skill acquisition

is best promoted through group role-playing in
simulated conflicts with experienced leaders
acting as models. Having established
competencies, the next task is to build a sense of
efficacy through success in alleviating everyday

problems. By selecting goals that can be realized
with concerted effort, participants gain confidence
for tackling tougher issues.
If apathy reflected only past failures in

effecting change, training for constructive action
would pose no special difficulties. Subordinated
groups, however, are deterred from challenging

injustices by fear of punishment and economic
reprisals. Why risk dangers for uncertain benefits?
Inhibitions of this sort are overcome by the
example of success. Seeing others like themselves
improve their living conditions by organized effort
provides compelling demonstrations that they too
have something to gain from contesting

unwarranted practices. After a group develops its
capacity for social action, its bargaining efficacy is
enhanced so that it has less need to carry out
protest tactics. This is because group pressure,
successfully applied, makes threats credible.

Consequently, in future disputes threats of
collective action alone are usually sufficient to
produce accommodating changes.
The political system is a major agency of social

change. People improve their society through
reform legislation. They rely on the sanctions of
judicial and administrative agencies to enforce

laws that affect their everyday life. The
governmental apparatus, however, is often
subverted from its public function by the pressure
of vested interests. Legislative procedures built
into the system enable influential entrepreneurs
to use the powers of government to resist social
changes that conflict with their self-interests.

Efforts to improve the functioning of society must
also be directed at governmental practices to
make them serve the public more equitably.
Gardner (1972) created Common Cause as a
public mechanism for exerting pressure on

Congress and state legislatures. A major thrust of
this citizens’ lobby is focused on congressional
reforms. Numerous changes are sought in
congressional committee operations that currently

make it easy to slight the public interest. These
include abolishing the seniority system, reducing
excessive power of committee chairmen, and

ending secrecy in the conduct of legislative
business by making committee votes public.
Common Cause worked successfully for a law
requiring disclosure of political expenditures so
that the public can be informed of the financial
interests backing various candidates. As a further
step in reducing congressional control by special

moneyed interests, Common Cause is advocating
lobbying laws to monitor who is spending how
much for what purposes. In addition to curtailing
entrenched private sources of congressional
control, this citizens’ organization exerts its

influence on legislative actions in promoting new
solutions to a broad range of social problems. In
some of these areas inequitable or potentially
detrimental practices are sufficiently beneficial to
large segments of the population to gain popular

support. Efforts to legislate social change are
therefore subjected to conflicting sources of public

and private control.
The law can be used to preserve existing
practices or as an instrument of social change.
Legal means are heavily relied upon by people to
secure their rights and to advance their welfare. It
is by converting social demands into formal codes
backed by sanctions that enduring reforms, which

reduce the likelihood of civil strife, are achieved.
In some discriminatory practices, individuals may
be willing to change if everyone else does so but
not individually because they would be singled out
for attack. Personal threats supporting adherence

to customs that have adverse effects on others are
greatly reduced when modifications in behavior
can be attributed to legal mandates (Zimring and
Hawkins, 1971). Under these conditions, laws
serve to accelerate widespread social change.
Passing laws does not necessarily ensure
adequate enforcement of them. Efforts to

eliminate inequitable practices are resisted by
entrenched interest groups and elements of the
general public. When legal prohibitions are
instituted against activities that have influential
backing, it is not uncommon for the laws to be
reinterpreted, circumvented, or poorly
administered (Friedman, 1967). Further litigation

is therefore required to refine statutes into
effective operating rules and to guarantee their
active enforcement. In addition to inducing and
monitoring the process of change, legal
mechanisms can uphold a society’s capacity for

constructive change by protecting dissenters from
unauthorized coercion.
Like so many other problems confronting man,
there is no single grand design for lowering the

level of destructiveness within a society. It
requires both individual corrective effort and
group action aimed at changing the practices of

social systems. Since aggression is not an
inevitable or unchangeable aspect of man but a
product of aggression promoting conditions
operating within a society, man has the power to
reduce his level of aggressiveness. Whether this
capability is used wisely or destructively is
another matter.
REFERENCES
Ackerman, N. W. Psychotherapy and “giving love."

Psychiatry, 1944, 7, 129-37.
Adler, L. L., and Adler, H. E. Age as a factor in

observational learning in puppies. American
Dachshund, March 1968, 13-14.
Agras, W. S. Behavior therapy in the management of

chronic schizophrenia. American Journal of
Psychiatry, 1967, 124, 240-43.
Alinsky, S. D. Rules for radicals. New York: Random

House, 1971.
Alland, A., Jr. The human imperative. New York:

Columbia University Press, 1972.
Allison, T. S., and Allison, S. L. Time-out from

reinforcement: Effect on sibling aggression.
Psychological Record, 1971, 21, 81-86.
Amsel, A. The role of frustrative nonreward in
noncontinuous reward situations. Psychological
Bulletin, 1958, 55, 102-19.
Andrus, B. C. The infamous of Nuremberg. London:

Fravin, 1969.
Ardrey, R. The territorial imperative. New York:

Atheneum, 1966.
Aronfreed, J. The problem of imitation. In L. P. Lipsitt

and H. W. Reese (Eds.), Advances in child
development and behavior. Vol. IV. New York:
Academic Press, 1969. Pp. 210-319.
Astin, A. W., and Bayer, A. E. Antecedents and

consequents of disruptive campus protests.
Measurement and Evaluation in Guidance, 1971, 4,
18-30.
Ax, A. F. The physiological differentiation between

fear and anger in humans. Psychosomatic
Medicine, 1953, 15, 433-42.
Ayllon, T., and Azrin, N. H. Reinforcement and

instructions with mental patients. Journal of the
Experimental Analysis of Behavior, 1964, 7, 327-
31.
Azrin, N. H. Pain and aggression. Psychology Today,

1967, 1, 27-33.
Azrin, N. H. Punishment of elicited aggression. Journal
of the Experimental Analysis of Behavior, 1970, 14,
7-10.
Azrin, N. H, Hake, D. F., and Hutchinson, R. R.

Elicitation of aggression by a physical blow.
Journal of the Experimental Analysis of Behavior,
1965, 8, 55-57.
Azrin, N. H. and Holz, W. C. Punishment. In W. K. Honig

(Ed.), Operant behavior. New York: Appleton-
Century-Crofts, 1966. Pp. 380-447.
Azrin, N. H., and Hutchinson, R. R. Conditioning of the

aggressive behavior of pigeons by a fixed-interval
schedule of reinforcement. Journal of the
402.
Azrin, N. H., Hutchinson, R. R., and Hake, D. F. Pain-

induced fighting in the squirrel monkey. Journal
620.
Azrin, N. H., Hutchinson, R. R., and Hake, D. F. Attack,

avoidance, and escape reactions to aversive
shock. Journal of the Experimental Analysis of
Behavior, 1967, 10, 131-48.
Azrin, N. H., Hutchinson, R. R., and McLaughlin, R. The
opportunity for aggression as an operant
reinforcer during aversive stimulation. Journal of
the Experimental Analysis of Behavior, 1965, 8,
171-80.
Azrin, N. H., Rubin, H. B., and Hutchinson, R. R. Biting

attack by rats in response to aversive shock.
Journal of the Experimental Analysis of Behavior,
1968, 11, 633-39.
Azrin, N. H. Ulrich, R. E., Hutchinson, R. R., and

Norman, D. G. Effect of shock duration on shock-
induced fighting. Journal of the Experimental
Analysis of Behavior, 1964, 7, 9-11.
Bach-y-Rita, G., Lion, J. R., Climent, C. E., and Ervin, F.

R. Episodic dyscontrol: A study of 130 violent
patients. American Journal of Psychiatry, 1971,
127, 1473-78.
Baenninger, R., and Grossman, J. C. Some effects of

punishment on pain-elicited aggression. Journal
1017-22.
Bailey, J. S., Wolf, M. M., and Phillips, E. L. Home-based

reinforcement and the modification of pre-
delinquents’ classroom behavior. Journal of
Applied Behavior Analysis, 1970, 3, 223-33.
Baker, J. W., II, and Schaie, K. W. Effects of aggressing
“alone” or “with another” on physiological and
psychological arousal. Journal of Personality and
Social Psychology, 1969, 12, 80-96.
Baldwin, T. F., and Lewis, C. Violence in television: The

industry looks at itself. In G. A. Comstock and E. A.
Rubinstein (Eds.), Television and social behavior.
Vol. 1. Media content and control. Washington,
D.C.: Government Printing Office, 1972. Pp. 290-
373.
Ball-Rokeach, S. J. Review of Television and Aggression,

by S. Feshbach and R. D. Singer. Public Opinion
Quarterly, 1971, 35, 501-4.
Ball-Rokeach, S. J. The legitimation of violence. In J. F.

Short, Jr., and M. E. Wolfgang (Eds.), Collective
violence. Chicago: Aldine-Atherton, 1972.
Bandura, A. Relationship of family patterns to child

behavior disorders. Progress Report, 1960,
Stanford University, Project No. M-1734, United
States Public Health Service.
Bandura, A. Social learning through imitation. In M. R.

Jones (Ed.), Nebraska symposium on motivation:
1962. Lincoln: University of Nebraska Press,
1962. Pp. 211-69.
Bandura, A. Influence of models' reinforcement
contingencies on the acquisition of imitative
responses. Journal of Personality and Social
Psychology, 1965a, 1, 589-95.
Bandura, A. Vicarious processes: A case of no-trial

learning. In L. Berkowitz (Ed.), Advances in
experimental social psychology. Vol. II. New York:
Academic Press, 1965b. Pp. 1-55.
Bandura, A. Principles of behavior modification. New

York: Holt, Rinehart and Winston, 1969a.
Bandura, A. Social-learning theory of identificatory

processes. In D. A. Goslin (Ed.), Handbook of
socialization theory and research. Chicago: Rand
McNally, 1969b. Pp. 213-62.
Bandura, A. Psychotherapy based upon modeling

principles. In A. E. Bergin and S. L. Garfield (Eds.),
Handbook of psychotherapy and behavior change.
New York: Wiley, 1971a. Pp. 653-708.
Bandura, A. Vicarious and self-reinforcement

processes. In R. Glaser (Ed.), The nature of
reinforcement. New York: Academic Press, 1971b.
Pp. 228-78.
Bandura, A. Social learning theory. New York: General

Learning Press, 1971c.
Bandura, A. (Ed.), Psychological modeling. Chicago:
Aldine-Atherton, 197 Id.
Bandura, A. The process and practice of participant

modeling treatment. Unpublished manuscript.
Stanford University, 1972.
Bandura, A., Blanchard, E. B., and Ritter, B. The

relative efficacy of desensitization and modeling
approaches for inducing behavioral, affective, and
attitudinal changes. Journal of Personality and
Bandura, A., Grusec, J. E., and Menlove, F. L.

Observational learning as a function of
symbolization and incentive set. Child
Development, 1966, 37, 499-506.
Bandura, A., and Huston, A. C. Identification as a

process of incidental learning. Journal of
Abnormal and Social Psychology, 1961, 63, 311-
18.
Bandura, A., and Jeffery, R. W. Role of symbolic coding

and rehearsal processes in observational
learning. Journal of Personality and Social
Psychology, 1972, in press.
Bandura, A., Lipsher, D. H., and Miller, P. E.

Psychotherapists’ approach-avoidance reactions
to patients’ expressions of hostility. Journal of
Consulting Psychology, 1960, 24, 1-8.
Bandura, A., and McDonald, F. J. The influence of social

reinforcement and the behavior of models in
shaping children’s moral judgments. Journal of
81.
Bandura, A., and Menlove, F. L. Factors determining

vicarious extinction of avoidance behavior
through symbolic modeling. Journal of Personality
and Social Psychology, 1968, 8, 99-108.
Bandura, A., and Mischel, W. Modification of self-

imposed delay of reward through exposure to
live and symbolic models. Journal of Personality
Bandura, A., and Rosenthal, T. L. Vicarious classical

conditioning as a function of arousal level. Journal
of Personality and Social Psychology, 1966, 3, 54-
62.
Bandura, A., Ross, D., and Ross, S. A. Transmission of

aggression through imitation of aggressive
models. Journal of Abnormal and Social
Psychology, 1961, 63, 575-82.
Bandura, A., Ross, D., and Ross, S. A. Imitation of film-
mediated aggressive models. Journal of Abnormal
and Social Psychology, 1963a, 66, 3-11.
Bandura, A., Ross, D., and Ross, S. A. A comparative

test of the status envy, social power, and
secondary reinforcement theories of
identificatory learning. Journal of Abnormal and
Social Psychology, 1963b, 67, 527-34.
Bandura, A., Ross, D., and Ross, S. A. Vicarious

reinforcement and imitative learning. Journal of
Abnormal and Social Psychology, 1963c, 67, 601-
7.
Bandura, A., and Walters, R. H. Adolescent aggression.

New York: Ronald Press, 1959.
Bandura, A., and Walters, R. H. Social learning and

personality development. New York: Holt,
Rinehart and Winston, 1963.
Barker, R., Dembo, T., and Lewin, K. Frustration and

regression: An experiment with young children.
University of Iowa Studies in Child Welfare, 1941,
18 (Whole No. 386).
Barnett, S. A. Attack and defense in animal societies.

In C. D. Clemente and D. B. Lindsley (Eds.),
Aggression and defense. Los Angeles: University of
California Press, 1967. Pp. 35-56.
Baron, R. A. Reducing the influence of an aggressive
model: The restraining effects of discrepant
modeling cues. Journal of Personality and Social
Psychology, 1971a, 20, 240-45.
Baron, R. A. Exposure to an aggressive model and

apparent probability of retaliation from the
victim as determinants of adult aggressive
behavior. Journal of Experimental Social
Psychology, 1971b, 7, 343-55.
Baron, R. A. Magnitude of victim’s pain cues and level

of prior anger arousal as determinants of adult
aggressive behavior. Journal of Personality and
Social Psychology, 1971c, 17, 236-43.
Baron, R. A. Aggression as a function of magnitude of

victim’s pain cues, level of prior anger arousal,
and aggressor-victim similarity. Journal of
Personality and Social Psychology, 1971 d, 18, 48-
54.
Baron, R. A., and Kepner, C. R. Model’s behavior and

attraction toward the model as determinants of
adult aggressive behavior. Journal of Personality
Barron, J. Access to the press—a new first amendment

right. Harvard Law Review, 1967, 80, 1641-78.
Barrow, L. C., Jr., and Westley, B. H. “Exploring the
news”: An experiment on the relative
effectiveness of radio and TV versions of a
children’s news program. In W. Schramm (Ed.),
The impact of educational television. Urbana:
University of Illinois Press, 1960. Pp. 143-50.
Bateson, G. The Naven. Stanford, Calif.: Stanford

University Press, 1936.
Bateson, G. The frustration-aggression hypothesis and

culture. Psychological Review, 1941, 48, 350-55.
Bateson, G. (Ed.). Perceval’s narrative: A patient’s

account of his psychosis, 1830-1832. Stanford,
Calif.: Stanford University Press, 1961.
Beach, F. A. It’s all in your mind. Psychology Today,

1969, 3, 33-35.
Becker, W. C., Thomas, D. R., and Carnine, D. Reducing

behavior problems: An operant conditioning guide
for teachers. Urbana, Ill.: Educational Resources
Information Center Clearinghouse on Early
Childhood Education, 1969.
Bedau, H. A. The death penalty in America. Garden

City, N.Y.: Anchor Books, 1967.
Berger, S. M. Conditioning through vicarious

instigation. Psychological Review, 1962, 69, 450-
66.
Berkowitz, L. The expression and reduction of

hostility. Psychological Bulletin, 1958, 55, 257-83.
Berkowitz, L. Aggression: A social psychological

analysis. New York: McGraw Hill, 1962.
Berkowitz, L. The concept of aggressive drive: Some

additional considerations. In L. Berkowitz (Ed.),
Advances in experimental social psychology. Vol. 2.
New York: Academic Press, 1965a. Pp. 301-29.
Berkowitz, L. Some aspects of observed aggression.

Journal of Personality and Social Psychology,
1965b, 2, 359-69.
Berkowitz, L. The contagion of violence: An S-R

mediational analysis of some effects of observed
aggression. In W. J. Arnold and M. M. Page (Eds.),
Nebraska symposium on motivation, 1970.
Lincoln: University of Nebraska Press, 1970. Pp.
95-135.
Berkowitz, L. The “weapons effect,” deviant

characteristics, and the myth of the compliant
subject. Journal of Personality and Social
Psychology, 1971, 20, 332-38.
Berkowitz, L. Words and symbols as stimuli to

aggressive responses. In J. F. Knutson (Ed.),
Control of aggression: Implications from basic
research. Chicago: Aldine-Atherton, 1972 (in
press).
Berkowitz, L., and Geen, R. G. Film violence and the

cue properties of available targets. Journal of
Personality and Social Psychology, 1966, 3, 525-
30.
Berkowitz, L., and Geen, R. G. Stimulus qualities of the

target of aggression: A further study. Journal of
68.
Berkowitz, L., and Knurek, D. A. Label-mediated

hostility generalization. Journal of Personality and
Berkowitz, L., and Le Page, A. Weapons as aggression-

eliciting stimuli. Journal of Personality and Social
Psychology, 1967, 7, 202-7.
Berkowitz, L., and Macaulay, J. The contagion of

violence. In H. Hirsch and C. Leiden (Eds.),
Political micro-violence. New York: Harper & Row,
1972, in press.
Bernal, M. E., Duryee, J. S., Pruett, H. L., and Burns, B. J.

Behavior modification and the brat syndrome.
Journal of Consulting and Clinical Psychology,
1968, 32, 447-55.
Bernal, M. E., Williams, D. E., Miller, W. H., and Reagor,

P. A. The use of videotape feedback and operant
learning principles in training parents in
management of deviant children. In R. D. Rubin
(Ed.), Advances in behavior therapy. Vol. III. New
York: Academic Press, 1972. Pp. 19-31.
Bernard, V., Ottenberg, P., and Redl, F.

Dehumanization: A composite psychological
defense in relation to modern war. In M.
Schwebel (Ed.), Behavioral science and human
survival. Palo Alto, Calif.: Science and Behavior
Books, 1965. Pp. 64-82.
Bettelheim, B. Individual and mass behavior in

extreme situations. Journal of Abnormal and
Bevan, W., Daves, W. F., and Levy, G. W. The relation of

castration, androgen therapy and pre test
experience to competitive aggression in male C57
BL/10 mice. Animal Behaviour, 1960, 8, 6-12.
Binder, A., McConnell, D., and Sjoholm, N. A. Verbal

conditioning as a function of experimenter
characteristics. Journal of Abnormal and Social
Psychology, 1957, 55, 309-14.
Blackburn, R. Personality in relation to extreme
aggression in psychiatric offenders. British
Journal of Psychiatry, 1968, 114, 821-28.
Blackham, G. J., and Silberman, A. Modification of child

behavior. Belmont, Calif.: Wadsworth, 1971.
Blixt, S., and Ley, R. Force-contingent reinforcement in

instrumental conditioning and extinction in
children: A test of the frustration-drive
hypothesis. Journal of Comparative and
Physiological Psychology, 1969, 69, 267-72.
Blumenthal, M., Kahn, R. L., Andrews, F. M., and Head,

K. B. Justifying violence: Attitudes of American
men. Ann Arbor, Mich.: Institute for Social
Research, 1972.
Blumer, H. Social problems as collective behavior.

Social Problems, 1971, 18, 298-305.
Bolles, R. C. Theory of motivation. New York: Harper &

Row, 1967.
Bolstad, O. D., and Johnson, S. M. Self-regulation in the

modification of disruptive classroom behavior.
Journal of Applied Behavior Analysis, 1972.
Bostow, D. E., and Bailey, J. S. Modification of severe

disruptive and aggressive behavior using brief
timeout and reinforcement procedures. Journal of
Bowen, D. R., Bowen, E. R., Gawser, S. R., and Masotti,

L. H. Deprivation, mobility, and orientation
toward protest of the urban poor. American
Behavioral Scientist, 1968, 11, 20-24.
Bramel, D., Taub, B., and Blum, B. An observer’s

reaction to the suffering of his enemy. Journal of
92.
Breed, W. Social control in the newsroom: A

functional analysis. Social Forces, 1955, 33, 326-
35.
Breer, P. E., and Locke, E. A. Task experience as a

source of attitudes. Homewood, Ill.: Dorsey, 1965.
Brock, T. C., and Buss, A. H. Dissonance, aggression,

and evaluation of pain. Journal of Abnormal and
Brock, T. C., and Buss, A. H. Effects of justification for

aggression and communication with the victim
on postaggression dissonance. Journal of
12.
Brown, G. D., and Tyler, V. O., Jr. Time out from

reinforcement: A technique for dethroning the
“duke” of an institutionalized delinquent group.
Journal of Child Psychology and Psychiatry and
Allied Disciplines, 1968, 9, 203-11.
Brown, L. Television: The business behind the box. New

York: Harcourt Brace Jovanovich, 1971.
Brown, P., and Elliott, R. Control of aggression in a

nursery school class. Journal of Experimental
Child Psychology, 1965, 2, 103-7.
Bucher, B., and Lovaas, O. I. Use of aversive

stimulation in behavior modification. In M. R.
Jones (Ed.), Miami symposium on the prediction of
behavior, 1967: Aversive stimulation. Coral Gables,
Fla.: University of Miami Press, 1968. Pp. 77-145.
Buehler, R. E., Patterson, G. R., and Furniss, J. M. The

reinforcement of behavior in institutional
settings. Behaviour Research and Therapy, 1966,
4, 157-67.
Burchard, J., and Tyler, V. O., Jr. The modification of

delinquent behavior through operant
conditioning. Behaviour Research and Therapy,
1965, 2, 245-50.
Buss, A. H. The psychology of aggression. New York:

Wiley, 1961.
Buss, A. H. Physical aggression in relation to different
frustrations. Journal of Abnormal and Social
Psychology, 1963, 67, 1-7.
Buss, A. H. Instrumentality of aggression, feedback,

and frustration as determinants of physical
aggression. Journal of Personality and Social
Psychology, 1966a, 3, 153-62.
Buss, A. H. The effect of harm on subsequent

aggression. Journal of Experimental Research in
Personality, 1966b, 1, 249-55.
Buss, A. H., Booker, A., and Buss, E. Firing a weapon

and aggression. Journal of Personality and Social
Psychology, 1972, 22, 296-302.
Buss, A. H., and Durkee, A. Conditioning of hostile

verbalizations in a situation resembling a clinical
interview. Journal of Consulting Psychology, 1958,
22, 415-18.
Calhoun, J. B. The study of wild animals under

controlled conditions. Annals of the New York
Academy of Sciences, 1950, 51, 1113-22.
Calhoun, J. B. Population density and social pathology.

Scientific American, 1962, 206, 139-50.
Campbell, B. A., and Church, R. M. Punishment and

aversive behavior. New York: Appleton-Century-
Crofts, 1969.
Campbell, D. T. Conformity in psychology’s theories of
acquired behavioral dispositions. In I. A. Berg and
B. M. Bass (Eds.), Conformity and deviation. New
York: Harper, 1961. Pp. 101-42.
Campbell, D. T. Reforms as experiments. American

Psychologist, 1969, 24, 409-429.
Cantor, M. G. The role of the producer in choosing

children’s television content. In G. A. Comstock
and E. A. Rubinstein (Eds.), Television and social
behavior. Vol. 1. Media content and control.
Washington, D.C.: Government Printing Office,
1972. Pp. 259-89.
Caplan, N. S. The new ghetto man: A review of recent

empirical studies. Journal of Social Issues, 1970,
26, 59-73.
Caplan, N. S., and Paige, J. M. A study of ghetto rioters.

Scientific American, 1968, 219 (2), 15-21.
Carrighar, S. War is not in our genes. In M. F. A.

Montagu (Ed.), Man and aggression. New York:
Oxford University Press, 1968. Pp. 37-50.
Carthy, J. D., and Ebling, F. J. (Eds.). The natural history

of aggression. New York: Academic Press, 1964.
Chaffee, S. H., and McLeod, J. M. Adolescents, parents,

and television violence. Paper presented at the
meeting of the American Psychological
Association, Washington, D.C., September 1971.
Chagnon, N. Yanomamö: The fierce people. New York:

Holt, Rinehart and Winston, 1968.
Chevigny, P. Police power: Police abuses in New York

City. New York: Pantheon Books, 1969.
Chittenden, G. E. An experimental study in measuring

and modifying assertive behavior in young
children. Monographs of the Society for Research
in Child Development, 1942, VII, No. 1 (Serial No.
31).
Christophersen, E. R., and Risley, T. R. Community

controlled sanctions in an urban poverty area.
Unpublished manuscript, University of Kansas,
1972.
Christy, P. R., Gelfand, D. M., and Hartmann, D. P.

Effects of competition-induced frustration on two
classes of modeled behavior. Developmental
Psychology, 1971, 5, 104-11.
Chu, G. C., and Schramm, W. Learning from television.

Stanford University: Institute for Communication
Research, 1967.
Clark, G. R., Telfer, M. A., Baker, D., and Rosen, M. Sex

chromosomes, crime, and psychosis. American
Journal of Psychiatry, 1972, in press.
Clark, H. B., Rowbury, T„ Baer, A. M., and Baer, D. M.

Control of a preschooler’s disruptive behavior
through the use of time-out. Paper presented at
the meeting of the American Psychological
Association, Miami, September 1970.
Claster, D. S. Comparison of risk perception between

delinquents and nondelinquents. Journal of
Criminal Law, Criminology, & Police Science, 1967,
58, 80-86.
Clemente, C. D., and Lindsley, D. B. (Eds.), Aggression

and defense: Neural mechanisms and social
patterns. Los Angeles: University of California
Press, 1967.
Cleveland Press. Untouchable gang. May 19, 1961, p. 1.
Cline, V. B., Croft, R. G., and Courrier, S. The

desensitization of children to television violence.
Unpublished manuscript, University of Utah,
1972.
Cohen, A. R. Social norms, arbitrariness of frustration,

and status of the agent of frustration in the
frustration-aggression hypothesis. Journal of
26.
Cohen, H. L., and Filipczak, J. A new learning
environment. San Francisco: Jossey-Bass, 1971.
Cole, J. M., and Litchfield, P. M. Stimulus control of

schedule-induced aggression in the pigeon.
Psychonomic Science, 1969, 17, 152-53.
Cole, J. M., and Parker, B. K. Schedule-induced

aggression: Access to an attackable target bird as
a positive reinforcer. Psychonomic Science, 1971,
22, 33-35.
Collins, W. A. The effect of temporal separation

between motivation, aggression and
consequences: A developmental study.
Developmental Psychology, 1973, in press.
Colman, A. D. Planned environment in psychiatric

treatment: A manual for design. Springfield, Ill.:
Charles C Thomas, 1971.
Conner, R. L., and Levine, S. Hormonal influences on

aggressive behaviour. In S. Garattini and E. B. Sigg
(Eds.), Aggressive behaviour. New York: Wiley,
1969. Pp. 150-63.
Corning, P. A., and Corning, C. H. Toward a general

theory of violent aggression. Social Science
Information, 1972, 11, 7-35.
Cousins, N. The time-trap. Saturday Review of
Literature, Dec. 24, 1949, 20.
Cowan, P. A., and Walters, R. H. Studies of

reinforcement of aggression: I. Effects of
scheduling. Child Development, 1963, 34, 543-51.
Crawford, T., and Naditch, M. Relative deprivation,

powerlessness, and militancy: The psychology of
social protest. Psychiatry, 1970, 33, 208-23.
Creer, T. L., Hitzing, E. W., and Schaeffer, R. W.

Classical conditioning of reflexive fighting.
Crook, J. H. The nature and function of territorial

aggression. In M. F. A. Montagu (Ed.), Man and
aggression. New York: Oxford University Press,
1968. Pp. 141-78.
Currie, E., and Skolnick, J. H. A critical note on

conceptions of collective behavior. Annals of the
American Academy of Political and Social Science,
1970, 391, 34-45.
Das, J. P., and Nanda, P. C. Mediated transfer of

attitudes. Journal of Abnormal and Social
Psychology, 1963, 66, 12-16.
Davies, J. C. Toward a theory of revolution. American

Sociological Review, 1962, 27, 5-19.
Davies, J. C. The J-curve of rising and declining
satisfactions as a cause of some great revolutions
and a contained rebellion. In H. D. Graham and T.
R. Gurr (Eds.), Violence in America: Historical and
comparative perspectives. Vol. II. Washington,
D.C.: U.S. Government Printing Office, 1969. Pp.
547-76.
Davitz, J. R. The effects of previous training on

postfrustration behavior. Journal of Abnormal and
de Charms, R., and Wilkins, E. J. Some effects of verbal

expression of hostility. Journal of Abnormal and
Delgado, J. M. R. Cerebral heterostimulation in a

monkey colony. Science, 1963, 141, 161-63.
Delgado, J. M. R. Social rank and radio-stimulated

aggressiveness in monkeys. Journal of Nervous
and Mental Disease, 1967, 144, 383-90.
Delgado, J. M. R., Vernon, M., Sweet, W., Ervin, F. R.,

Weiss, G., Bach-y-Rita, G., and Hagiwara, R.
Intracerebral radio stimulation and recording in
completely free patients. Journal of Nervous and
Mental Disease, 1968, 147, 329-40.
Denenberg, V. H. The mother as motivator. In W. J.
Arnold and M. M. Page (Eds.), Nebraska
symposium on motivation: 1970. Lincoln:
University of Nebraska Press, 1970. Pp. 69-93.
Deur, J. L., and Parke, R. D. Effects of inconsistent

punishment on aggression in children.
Developmental Psychology, 1970, 2, 403-11.
Ditrichs, R., Simon, S., and Greene, B. Effect of

vicarious scheduling on the verbal conditioning
of hostility in children. Journal of Personality and
Dollard, J., Doob, L. W., Miller, N. E., Mowrer, O. H., and

Sears, R. R. Frustration and aggression. New
Haven, Conn.: Yale University Press, 1939.
Dominick, J. R., and Greenberg, B. S. Attitudes toward

violence: The interaction of TV exposure, family
attitudes, and social class. In G. A. Comstock and
E. A. Rubinstein (Eds.), Television and social
behavior. Vol. 3. Television and adolescent
aggressiveness. Washington, D.C.: Government
Printing Office, 1972. Pp. 314-35.
Donnerstein, E., Donnerstein, M., Simon, S., and

Ditrichs, R. Variables in interracial aggression:
Anonymity, expected retaliation, and a riot.
Journal of Personality and Social Psychology, 1972,
22, 236-45.
Doob, A. N., and Wood, L. Catharsis and aggression:
Effects of annoyance and retaliation on
Dreyer, P. I., and Church, R. M. Reinforcement of

shock-induced fighting. Psychonomic Science,
1970, 25, 147-48.
Dubanoski, R. A., and Parton, D. A. Imitative

aggression in children as a function of observing
a human model. Developmental Psychology, 1971,
4, 489.
Duffy, E. Activation and behavior. New York: Wiley,

1962.
Eaton, J. W., and Weil, R. J. Culture and mental

disorders. New York: Free Press, 1955.
Edwards, N. L. Aggressive expression under threat of

retaliation. Unpublished doctoral dissertation,
University of Iowa, 1967.
Ehrhardt, A. A., Epstein, R., and Money, J. Fetal

androgens and female gender identity in the
early-treated adrenogenital syndrome. Johns
Hopkins Medical Journal, 1968, 122, 160-67.
Ellis, G. T., and Sekyra, F. The effect of aggressive

cartoons on the behavior of first grade children.
Journal of Psychology, 1972, 81, 37-43.
Ellis, R. A., and Lane, W. C. Structural supports for

upward mobility. American Sociological Review,
1963, 28, 743-56.
Ellison, C. D., and Flynn, J. P. Organized aggressive

behavior in cats after surgical isolation of the
hypothalamus. Archives of Italian Biology, 1968,
106, 1-20.
Elms, A. C. (Ed.). Role playing, reward, and attitude

change. New York: Van Nostrand Reinhold, 1969.
Ennis, P. H. Crime, victims, and the police. Trans-

action, 1967, 4, 36-44.
Epstein, R. Aggression toward outgroups as a function

of authoritarianism and imitation of aggressive
models. Journal of Personality and Social
Psychology, 1966, 3, 574-79.
Eron, L. D. Relationship of TV viewing habits and

aggressive behavior in children. Journal of
96.
Eron, L. D., Huesmann, L. R., Lefkowitz, M. M., and

Walder, L. O. Does television violence cause
aggression? American Psychologist, 1972, 27,
253-63.
Farris, H. E., Fullmer, W. H., and Ulrich, R. E. Extinction
of classically conditioned aggression: Results
from two procedures. Proceedings of the 78th
annual convention of the American Psychological
Association. Washington, D.C.: American
Psychological Association, Pp. 775-76.
Farris, H. E., Gideon, B. E„ and Ulrich, R. E. Classical

conditioning of aggression: A developmental
study. Psychological Record, 1970, 20, 63-68.
Feshbach, S. The drive-reducing function of fantasy

behavior. Journal of Abnormal and Social
Psychology, 1955, 50, 3-11.
Feshbach, S. The catharsis hypothesis and some

consequences of interaction with aggressive and
neutral play objects. Journal of Personality, 1956,
24, 449-62.
Feshbach, S. The stimulating versus cathartic effects

of a vicarious aggressive activity. Journal of
85.
Feshbach, S. The function of aggression and the

regulation of aggressive drive. Psychological
Review, 1964, 71, 257-72.
Feshbach, S. Aggression. In P. H. Mussen (Ed.),
Carmichael’s manual of child psychology, Vol. II.
New York: Wiley, 1970. Pp. 159-259.
Feshbach, S. Reality and fantasy in filmed violence. In

J. P. Murray, E. A. Rubinstein, and G. A. Comstock
(Eds.), Television and social behavior. Vol. 2.
Television and social learning. Washington, D.C.:
Government Printing Office, 1972. Pp. 318-45.
Feshbach, S., and Singer, R. D. Television and

aggression: An experimental field study. San
Francisco: Jossey-Bass, 1971.
Feshbach, S., Stiles, W. B„ and Bitter, E. The

reinforcing effect of witnessing aggression.
Journal of Experimental Research in Personality,
1967, 2, 133-39.
Festinger, L. A theory of social comparison processes.

Human Relations, 1954,7,117-40.
Festinger, L., Pepitone, A., and Newcomb, T. Some

consequences of de-individuation in a group.
Journal of Abnormal and Social Psychology, 1952,
47, 382-89.
Fisher, G. Discriminating violence emanating from

over-controlled versus under-controlled
aggressivity. British Journal of Clinical and Social
Psychology, 1970, 9, 54-59.
Flacks, R. The liberated generation: An exploration of
the roots of student protest. Journal of Social
Issues, 1967, 23, 52-75.
Flanders, J. P. A review of research on imitative

behavior. Psychological Bulletin, 1968, 69, 316-
37.
Flory, R. K. Attack behavior in a multiple fixed-ratio

schedule of reinforcement. Psychonomic Science,
1969a, 16, 156-57.
Flory, R. K. Attack behavior as a function of minimum

inter-food interval. Journal of the Experimental
Analysis of Behavior, 1969b, 12, 825-28.
Ford, C. S., and Beach, F. A. Patterns of sexual behavior.

New York: Harper & Row, 1951.
Foster, J., and Long, D. Protest! Student activism in

America. New York: Morrow, 1970.
Franks, L., and Powers, T., of UPI. Profile of a terrorist.

Palo Alto Times, Sept. 17, 1970, pp. 26-28.
Fredericson, E. Competition: The effects of infantile

experience upon adult behavior. Journal of
Abnormal and Social Psychology, 1951, 46, 406-9.
Fredericson, E., and Birnbaum, E. A. Competitive

fighting between mice with different hereditary
backgrounds. Journal of Genetic Psychology, 1954,
85, 271-80.
Freeman, E. Effects of aggressive expression after

frustration on performance: A test of the
catharsis hypothesis. Unpublished doctoral
dissertation, Stanford University, 1962.
Freud, A. The ego and the mechanisms of defence. Tr.

by Cecil Baines. New York: International
Universities Press, 1946.
Freud, S. A general introduction to psycho-analysis.

New York: Boni & Liveright, 1920.
Freud, S. Beyond the pleasure principle. London:

International Psychoanalytic Press, 1922.
Freud, S. The ego and id. London: Hogarth, 1923.
Freud, S. New introductory lectures on psycho-analysis.

New York: Morton, 1933.
Freud, S. Why war? In J. Strachey (Ed.), Collected

papers. Vol. V. London: Hogarth Press, 1950. Pp.
273-87.
Friedman, L. M. Legal rules and the process of social

change. Stanford Law Review, 1967, 19, 786-840.
Friedman, P. H. The effects of modeling, roleplaying

and participation on behavior change. In B. A.
Maher (Ed.), Progress in experimental personality
research. Vol. VI. New York: Academic Press,
1972. Pp. 41-81.
Friendly, F. W. Due to circumstances beyond our

control. New York: Vintage Books, 1968.
Gambaro, S., and Rabin, A. I. Diastolic blood pressure

responses following direct and displaced
aggression after anger arousal in high- and low-
guilt subjects. Journal of Personality and Social
Psychology, 12, 87-94.
Gamson, W. A., and McEvoy, J. Police violence and its

public support. Annals of the American Academy
of Political and Social Science, 391, 97-110.
Garattini, S., and Sigg, E. B. (Eds.), Aggressive

behaviour. New York: Wiley, 1969.
Gardner, J. W. In common cause. New York: Norton,

1972.
Gardner, R., and Heider, K. G. Gardens of war: Life and

death in the New Guinea stone age. New York:
Random House, 1969.
Geen, R. G. Effects of frustration, attack, and prior

training in aggressiveness upon aggressive
behavior. Journal of Personality and Social
Psychology, 1968, 9, 316-21.
Geen, R. G. Perceived suffering of the victim as an
inhibitor of attack-induced aggression. Journal of
Geen, R. G., and Berkowitz, L. Name-mediated

aggressive cue properties. Journal of Personality,
1966, 34, 456-65.
Geen, R. G., and Berkowitz, L. Some conditions

facilitating the occurrence of aggression after the
observation of violence. Journal of Personality,
1967, 35, 666-76.
Geen, R. G., and O’Neal, E. C. Activation of cue-elicited

aggression by general arousal. Journal of
92.
Geen, R. G., and Pigg, R. Acquisition of an aggressive

response and its generalization to verbal
behavior. Journal of Personality and Social
Psychology, 1970, 15, 165-70.
Geen, R. G., and Stonner, D. Effects of aggressiveness

habit strength on behavior in the presence of
aggression-related stimuli. Journal of Personality
Geen, R. G., and Stonner, D. Context effects in observed

violence. Journal of Personality and Social
Psychology, 1972, in press.
Gentry, W. D. Fixed-ratio schedule-induced

aggression. Journal of the Experimental Analysis of
Behavior, 1968, 11, 813-17.
Gentry, W. D. Effects of frustration, attack, and prior

aggressive training on overt aggression and
vascular processes. Journal of Personality and
Gentry, W. D., and Schaeffer, R. W. The effect of FR

response requirement on aggressive behavior in
rats. Psychonomic Science, 1969, 14, 236-38.
Gerbner, G. Violence in television drama: Trends and

symbolic functions. In G. A. Comstock and E. A.
Vol. 1. Content and control. Washington, D.C.:
Government Printing Office, 1972a. Pp. 28-187.
Gerbner, G. The violence index: A rating of various

aspects of dramatic violence on prime-time
network television 1967 through 1970.
Unpublished manuscript, University of
Pennsylvania, 1972b.
Gerst, M. S. Symbolic coding processes in

observational learning. Journal of Personality and
Gillespie, W. H. Aggression and instinct theory.
International Journal of Psycho-analysis, 1971, 52,
155-160.
Ginsburg, B., and Allee, W. C. Some effects of

conditioning on social dominance and
subordination in inbred strains of mice.
Physiological Zoology, 1942, 15, 485-506.
Gittleman, M. Behavior rehearsal as a technique in

child treatment. Journal of Child Psychology and
Psychiatry, 1965, 6, 251-55.
Glueck, S., and Glueck, E. Unraveling juvenile

delinquency. Cambridge, Mass.: Harvard
Goldfrank, E. S. Historic change and social character: A

study of the Teton Dakota. American
Anthropologist, 1943, 45, 67-83.
Goldfrank, E. S. Socialization, personality, and the

structure of Pueblo society (with particular
reference to Hopi and Zuni). American
Anthropologist, 1945, 47, 516-39.
Goldman, J. R. The relation of certain therapist

variables to the handling of psychotherapeutic
events. Unpublished doctoral dissertation,
Stanford University, 1961.
Goodlet, G, R., and Goodlet, M. M. Efficiency of self-
monitored and externally imposed schedules of
reinforcement in controlling disruptive behavior.
Unpublished manuscript, University of Guelph,
1969.
Goodlet, G. R„ Goodlet, M. M„ and Dredge, K.

Modification of disruptive behavior of two young
children and follow-up one year later. Journal of
School Psychology, 1970, 8, 60-63.
Goodwin, C. The social organization of the Western

Apache. Chicago: University of Chicago Press,
1942.
Goranson, R. E. Media violence and aggressive

behavior: A review of experimental research. In
L. Berkowitz (Ed.), Advances in experimental
social psychology. Vol. 5. New York: Academic
Press, 1970. Pp. 1-31.
Grimshaw, A. D. Violence: A sociological perspective.

George Washington Law Review, 1969, 37, 816-
34.
Grimshaw, A. D. Three views of urban violence: Civil

disturbance, racial revolt, class assault. American
Behavioral Scientist, 1970, 11, 2-7.
Grusec, J. E. Demand characteristics of the modeling
experiment: Altruism as a function of age, and
Psychology, 1972, 22, 139-48.
Gulevich, G. D., and Bourne, P. G. Mental illness and

violence. In N. Daniels, M. F. Gilula, and F.
Ochberg (Eds.), Violence and the struggle for
existence. Boston: Little, Brown, 1970. Pp. 309-26.
Gurr, R. T. Why men rebel. Princeton, N.J.: Princeton

University Press, 1970a.
Gurr, R. T. Sources of rebellion in Western societies:

Some quantitative evidence. Annals of the
1970b, 391, 128-44.
Hallie, P. O. Justification and rebellion. In N. Sanford

and C. Comstock (Eds.), Sanctions for evil. San
Francisco: Jossey-Bass, 1971. Pp. 247-63.
Halloran, J. D., and Croll, P. Television programs in

Great Britain: Content and control. In G. A.
Comstock and E. A. Rubinstein (Eds.), Television
and social behavior. Vol. 1. Media content and
control. Washington, D.C.: Government Printing
Office, 1972. Pp. 415-92.
Hamburg, D. A. A combined biological and

psychosocial approach to the study of
behavioural development. In A. Ambrose (Ed.),
Stimulation in early childhood. New York:
Academic Press, 1969. Pp. 269-77.
Hamburg, D. A. Recent evidence on the evolution of

aggressive behavior. Engineering and Science,
1970, 23, 15-24.
Hamilton, J., Stephens, L., and Allen, P. Controlling

aggressive and destructive behavior in severely
retarded institutionalized residents. American
Journal of Mental Deficiency, 1967, 71, 852-56.
Haner, C. F., and Brown, P. A. Clarification of the

instigation to action concept in the frustration-
aggression hypothesis. Journal of Abnormal and
Hanratty, M. A., Liebert, R. M., Morris, L. W., and

Fernandez, L. E. Imitation of film-mediated
aggression against live and inanimate victims.
Proceedings of the 77th Annual Convention of the
American Psychological Association. Washington,
D.C.: American Psychological Association, 1969.
Pp. 457-58.
Hanratty, M. A., O’Neal, E. C., and Sulzer, J. L. Effect of

frustration upon imitation of aggression. Journal
34.
Hapkiewicz, W. G., and Stone, R. D. The effect of age,
sex, and model characteristics on children’s
aggressive and cooperative play. Unpublished
manuscript, Michigan State University, 1972.
Harmatz, M. G., and Rasmussen, W. A. A behavior

modification approach to head banging. Mental
Hygiene, 1969, 53, 590-93.
Harris, M. B. Field studies of modeled aggression.

Journal of Social Psychology, 1972.
Hartley, R. E. A review and evaluation of recent studies

on the impact of violence. New York: Office of
Social Research, Columbia Broadcasting System,
1964.
Hartmann, D. P. Influence of symbolically modeled

instrumental aggression and pain cues on
Hastorf, A. H. The “reinforcement” of individual

actions in a group situation. In L. Krasner and L.
P. Ullmann (Eds.), Research in behavior
modification. New York: Holt, Rinehart &
Winston, 1965. Pp. 268-84.
Hawkins, R. P., Peterson, R. F., Schweid, E., and Bijou,

S. W. Behavior therapy in the home: Amelioration
of problem parent-child relations with the parent
in a therapeutic role. Journal of Experimental
Child Psychology, 1966, 4, 99-107.
Helm, B., Bonoma, T. V., and Tedeschi, J. T. Reciprocity

for harm done. Journal of Social Psychology, 1972,
87, 89-98.
Herrell, J. M. Use of systematic desensitization to

eliminate inappropriate anger. Proceedings of the
79th Annual Convention of the American
Psychological Association. Washington, D.C.:
American Psychological Association, 1971. Pp.
431-32.
Hess, E. H. The relationship between imprinting and

motivation. In M. R. Jones (Ed.), Nebraska
symposium on motivation. 1959. Lincoln:
University of Nebraska Press, 1969. Pp. 44-77.
Hess, E. H. Pupillometric assessment. In J. M. Shlien

(Ed.), Research in psychotherapy. Washington,
D.C.: American Psychological Association, 1968.
Pp. 573-83.
Hicks, D. J. Imitation and retention of film-mediated

aggressive peer and adult models. Journal of
100.
Hicks, D. J. Effects of co-observer’s sanctions and adult
presence on imitative aggression. Child
Development, 1968a, 39, 303-9.
Hicks, D. J. Short and long-term retention of affectively

varied modeled behavior. Psychonomic Science,
1968b, 11,369-70.
Hicks, D. J. Girls’ attitudes toward modeled behaviors

and the content of imitative private play. Child
Development, 1971, 42, 139-47.
Hill, B. Boss of Britain’s underworld. London: Naldreth

Press, 1955.
Himmelweit, H. T., Oppenheim, A. N., and Vince, P.

Television and the child. London: Oxford
Hinde, R. A. Ethological models and the concept of

“drive.” British Journal of Philosophical Science,
1956, 6, 321-31.
Hinde, R. A. Energy models of motivation. Symposium

on social experimental biology, 1960, 14, 199-213.
Hoelle, C. The effects of modeling and reinforcement

on aggressive behavior in elementary school
boys. Dissertation Abstracts, 1969, 29B, 3483-84.
Hoffman, M. L. Power assertion by the parent and its
impact on the child. Child Development, 1960, 31,
129-43.
Hogan, J. A. Fighting and reinforcement in the Siamese

fighting fish (Betta splendens). Journal of
Comparative and Physiological Psychology, 1967,
64, 356-59.
Hokanson, J. E. Psychophysiological evaluation of the

catharsis hypothesis. In E. I. Megargee and J. E.
Hokanson (Eds.), The dynamics of aggression.
New York: Harper & Row, 1970. Pp. 74-86.
Hokanson, J. E., and Edelman, R. Effects of three social

responses on vascular processes. Journal of
47.
Hokanson, J. E., Willers, K. R., and Koropsak, E. The

modification of autonomic responses during
aggressive interchange. Journal of Personality,
1968, 36, 386-404.
Holmes, D. S. Effects of overt aggression on level of

physiological arousal. Journal of Personality and
Holmes, D. S. Dimensions of projection. Psychological

Bulletin, 1968, 69, 248-68.
Holton, R. B. Amplitude of an instrumental response
following the cessation of reward. Child
Development, 1961, 32, 107-16.
Hoyt, J. L. Effect of media violence “justification” on

aggression. Journal of Broadcasting, 1970, 14,
455-64.
Hunt, J. McV., Cole, M. W., and Reis, E. E. S. Situational

cues distinguishing anger, fear, and sorrow.
American Journal of Psychology, 1958, 71, 136-51.
Hutchinson, R. R., Azrin, N. H., and Hunt, G. M. Attack

produced by intermittent reinforcement of a
concurrent operant response. Journal of the
95.
Hutchinson, R. R„ Ulrich, R. E., and Azrin, N. H. Effects

of age and related factors on the pain-aggression
reaction. Journal of Comparative and Physiological
Psychology, 1965, 59, 365-69.
Iglitzin, L. B. Violence and American democracy.

Journal of Social Issues, 1970, 26, 165-86.
Insko, C. A., and Oakes, W. F. Awareness and the

“conditioning” of attitudes. Journal of Personality
Jacobs, P. A., Brunton, M., and Melville, M. M.

Aggressive behavior, mental sub-normality and
the XYY male. Nature, 1965, 208, 1351-52.
Jegard, S., and Walters, R. H. A study of some

determinants of aggression in young children.
Child Development, 1960, 31, 739-47.
Johnson, A. M., and Szurek, S. A. The genesis of

antisocial acting out in children and adults. The
Psychoanalytic Quarterly, 1952, 21, 323-43.
Johnson, E. H. Selective factors in capital punishment.

Social Forces, 1957, 36, 165-69.
Johnson, N. How to talk back to your television set.

Boston: Little, Brown, 1967.
Johnson, S. M., and Brown, R. A. Producing behaviour

change in parents of disturbed children. Journal
of Child Psychology and Psychiatry, 1969, 10, 107-
21.
Kahn, M. The physiology of catharsis. Journal of

86.
Kahn, M. W. The effect of severe defeat at various age

levels on the aggressive behavior of mice. Journal
of Genetic Psychology, 1951, 79, 117-30.
Kahn, R. M., and Bowers, W. J. The social context of the

rank-and-file student activist: A test of four
hypotheses. Sociology of Education, 1970, 43, 38-
55.
Kanfer, F. H. Vicarious human reinforcement: A

glimpse into the black box. In L. Krasner and L. P.
Ullmann (Eds.), Research in behavior modification.
New York: Holt, Rinehart and Winston, 1965. Pp.
244-67.
Kanfer, F. H., and Phillips, J. S. Learning foundations of

behavior therapy. New York: Wiley, 1970.
Kaufmann, H., and Feshbach, S. The influence of

antiaggressive communications upon the
response to provocation. Journal of Personality,
1963, 31, 428-44.
Keirsey, D. W. Systematic exclusion: Eliminating

chronic classroom disruptions. In J. D. Krumboltz
and C. E. Thoresen (Eds.) Behavioral counseling:
Cases and Techniques. New York: Holt, Rinehart &
Winston, 1969. Pp. 89-113.
Keniston, K. The sources of student dissent. Journal of

Social Issues, 23, 108-37.
Keniston, K. Young radicals. New York: Harcourt Brace

Jovanovich, 1968.
Keniston, K. Student activism, moral development,

and morality. American Journal of
Orthopsychiatry, 1970, 40, 577-92.
Keniston, K., and Lerner, M. Campus characteristics

and campus unrest. Annals of the American
Academy of Political and Social Science, 1971, 395,
39-53.
Kenny, D. T. An experimental test of the catharsis

theory of aggression. Unpublished doctoral
dissertation, University of Washington, 1952.
Kerner, O., et al. Report of the National Advisory

Commission on Civil Disorders. New York: Bantam
Books, 1968.
Kessler, S., and Moos, R. H. The XYY karyotype and

criminality: A review. Journal of Psychiatric
Research, 1970, 7, 153-70.
Klein, M. W., and Crawford, L. Y. Groups, gangs and

cohesiveness. Journal of Research in Crime and
Delinquency, 1967, 4, 63-75.
Klopfer, P. H. Behavioral aspects of ecology. Englewood

Cliffs, N.J.: Prentice-Hall, 1962.
Kniveton, B. H., and Stephenson, G. M. The effect of

pre-experience on imitation of an aggressive film
model. British Journal of Social and Clinical
Psychology, 1970, 9, 31-36.
Knutson, J. F. Aggression during the fixed-ratio and
extinction components of a multiple schedule of
reinforcement. Journal of the Experimental
Analysis of Behavior, 1970, 13, 221-31.
Knutson, J. F. The effects of shocking one member of a

pair of rats. Psychonomic Science, 1971, 22, 265-
66.
Kobasigawa, A. Observation of failure in another

person as a determinant of amplitude and speed
of a simple motor response. Journal of Personality
Konecni, V. J., and Doob, A. N. Catharsis through

displacement of aggression. Journal of Personality
Krauss, I. Sources of educational aspirations among

working-class youth. American Sociological
Review, 1964, 29, 867-79.
Kuhn, D. Z., Madsen, C. H., Jr., and Becker, W. C. Effects

of exposure to an aggressive model and
“frustration” on children’s aggressive behavior.
Child Development, 1967, 38, 739-45.
Kuo, Z. Y. The genesis of the cat’s responses to the rat.

Journal of Comparative Psychology, 1930, 11, 1-
35.
Kuo, Z. Y. Further study on the behavior of cats
toward rats. Journal of Comparative Psychology,
1938, 25, 1—8.
Kuo, Z. Y. Studies on the basic factors in animal

fighting: VII. Interspecies coexistence in
mammals. Journal of Genetic Psychology, 1960, 97,
211-25.
Lagerspetz, K. M. J. Studies on the aggressive behavior

of mice. Helsinki: Suomalainen Tiedeakatemia,
1964.
Lagerspetz, K. M. J. Aggression and aggressiveness in

laboratory mice. In S. Garattini and E. B. Sigg
(Eds.), Aggressive behaviour. New York: Wiley,
1969. Pp. 77-85.
Lantis, M. Alaskan Eskimo cultural values. Polar Notes,

1959, 1, 35-48.
Larder, D. L. Effect of aggressive story content on

nonverbal play behavior. Psychological Reports,
1962, 11, 14.
Larsen, O. N., Gray, L. N., and Fortis, J. G. Achieving

goals through violence on television. In O. N.
Larsen (Ed.), Violence and the mass media. New
York: Harper & Row, 1968. Pp. 97-111.
LeBlanc, J. M., Busby, K., and Thomson, C. The
functions of time-out for changing the aggressive
behaviors of a preschool child: A multiple
baseline analysis. Paper presented at the meeting
of the American Psychological Association,
Washington, D.C., September 1971.
Legrand, R. Successful aggression as the reinforcer for

runway behavior of mice. Psychonomic Science,
1970, 20, 303-5.
Lehrman, D. S. A critique of Konrad Lorenz’s theory of

instinctive behavior. Quarterly Review of Biology,
1953, 28, 337-63.
Leifer, A. D., and Roberts, D. F. Children’s responses to

television violence. In J. P. Murray, E. A.
Rubinstein, and G. A. Comstock (Eds.), Television
and social behavior. Vol. 2. Television and social
learning. Washington, D.C.: Government Printing
Office, 1972. Pp. 43-180.
Leitenberg, H. Is time-out from positive reinforcement

an aversive event? A review of the experimental
evidence. Psychological Bulletin, 1965, 64, 428-
41.
Lejins, P. J. International opinions on American media

violence. In O. N. Larsen (Ed.), Violence and the
mass media. New York: Harper & Row, 1968. Pp.
168-72.
Lerner, M. J. Observer’s evaluation of a victim: Justice,
guilt, and veridical perception. Journal of
35.
Levin, G. R., and Simmons, J. J. Response to food and

praise by emotionally disturbed boys.
Psychological Reports, 1962, 11, 539-46.
Levy, R. I. On getting angry in the Society Islands. In

W. Caudill and T.-Y. Lin (Eds.), Mental health
research in Asia and the Pacific. Honolulu: East-
West Center Press, 1969. Pp. 358-80.
Lieberson, S., and Silverman, A. R. The precipitants

and underlying conditions of race riots. American
Liebert, R. M. Television and social learning: Some

relationships between viewing violence and
behaving aggressively. In J. P. Murray, E. A.
Rubinstein, and G. A. Comstock (Eds.), Television
and social behavior. Vol. 2. Television and social
learning. Washington, D.C.: Government Printing
Office, 1972. Pp. 1-42.
Liebert, R. M., and Baron, R. A. Short-term effects of

televised aggression on children's aggressive
behavior. In J. P. Murray, E. A. Rubinstein, and G.
A. Comstock (Eds.), Television and social behavior.
Vol. 2. Television and social learning. Washington,
D.C.: Government Printing Office, 1972. Pp. 181-
201.
Liebert, R. M., Sobol, M. D., and Davidson, E. S.

Catharsis of aggression among institutionalized
boys: Fact or artifact. In G. A. Comstock, E. A.
Rubinstein, and J. P. Murray (Eds.), Television and
social behavior. Vol. 5. Television effects: Further
explorations. Washington, D.C.: Government
Linton, R. The Comanche. In A. Kardiner (Ed.), The

psychological frontiers of society. New York:
Columbia University Press, 1945. Pp. 47-80.
Lipset, S. M. University students and politics in

underdeveloped countries. Comparative
Education Review, 1966, 10, 132-62.
Lockhart, W. B. The report of the commission on

obscenity and pornography. New York: Bantam
Books, 1970.
Loew, C. A. Acquisition of a hostile attitude and its

relationship to aggressive behavior. Journal of
41.
Logan, F. A., and Boice, R. Aggressive behaviors of
paired rodents in an avoidance context.
Behaviour, 1969, 34, 161-83.
Longstreth, L. E. Distance to goal and reinforcement

schedule as determinants of human instrumental
behavior. Proceedings of the 74th Annual
Convention of the American Psychological
Association. Washington, D.C.: American
Psychological Association, 1966. Pp. 39-40.
Lorenz, K. On aggression. New York: Harcourt Brace

Jovanovich, 1966.
LoSciuto, L. A national inventory of television viewing

behavior. In E. A. Rubinstein, G. A. Comstock, and
J. P. Murray (Eds.), Television and social behavior.
Vol. 4. Television in day-to-day life: Patterns of use.
1972. Pp. 33-86.
Lovaas, O. I. Effect of exposure to symbolic aggression

on aggressive behavior. Child Development, 1961,
32, 37-44.
Lovaas, O. I., Freitag, G., Gold, V. J., and Kassorla, I. C.

Experimental studies in childhood schizophrenia:
Analysis of self-destructive behavior. Journal of
Experimental Child Psychology, 1965, 2, 67-84.
Lovaas, O. I., and Simmons, J. Q. Manipulation of self-
destruction in three retarded children. Journal of
Luchins, A. S., and Luchins, E. H. Imitation by rote and

by understanding. Journal of Social Psychology,
1961, 54, 175-97.
Ludwig, A. M., Marx, A. J., Hill, P. A., and Browning, R.

M. The control of violent behavior through
faradic shock. Journal of Nervous and Mental
Disease, 1969, 148, 624-37.
Lyon, D. O., and Ozolins, D. Pavlovian conditioning of

shock-elicited aggression: A discrimination
procedure. Journal of Experimental Analysis of
Behavior, 1970, 13, 325-31.
Maccoby, E. E. The effects of mass media. In N. L.

Hoffman and L. W. Hoffman (Eds.), Review of child
development research. Vol. 1. New York: Russell
Sage Foundation, 1964. Pp. 323-48.
Maccoby, E. E., and Wilson, W. C. Identification and

observational learning from films. Journal of
Abnormal and Social Psychology, 1957, 55, 76-87.
Mackay, C. Extraordinary popular delusions and the

madness of crowds. Boston: Page, 1932.
McCord, W., and Howard, J. Negro opinions in three

riot cities. American Behavioral Scientist, 1968,
11, 24-27.
McCord, W., McCord, J., and Zola, 1. K. Origins of crime:

A new evaluation of the Cambridge-Somerville
Youth Study. New York: Columbia University
Press, 1959.
McDougall, W. An introduction to social psychology.

London: Methuen, 1931.
McIntyre, J., and Teevan, J. Television violence and

deviant behavior. In G. A. Comstock and E. A.
Vol. 3. Television and adolescent aggressiveness.
1972. Pp. 383-435.
McMains, M. J., and Liebert, R. M. Influence of

discrepancies between successively modeled self-
reward criteria on the adoption of a self-imposed
standard. Journal of Personality and Social
Psychology, 1968, 8, 166-71.
McPhail, C. Civil disorder participation: A critical

examination of recent research. American
Madsen, C. H., Jr. Nurturance and modeling in

preschoolers. Child Development, 1968, 39, 221-
36.
Madsen, C. H., Jr., Becker, W. C., Thomas, D. R., Koser,
L., and Plager, E. An analysis of the reinforcing
function of “sit-down” commands. In R. K. Parker
(Ed.), Readings in educational psychology. New
York: Allyn & Bacon, 1968. Pp. 265-78.
Mahoney, M. J. A residential program in behavior

modification. Paper presented at the Fifth Annual
Meeting of the Association for the Advancement
of Behavior Therapy, Washington, D.C.,
September 1971.
Maier, S. F., Seligman, M. E. P., and Solomon, R. L.

Pavlovian fear conditioning and learned
helplessness: Effects on escape and avoidance
behavior of (a) the CU-US contingency and (b) the
independence of the US and voluntary
responding. In B. A. Campbell and R. M. Church
(Eds.), Punishment and aversive behavior. New
York: Appleton-Century-Crofts, 1969. Pp. 299-
342.
Mallick, S. K., and McCandless, B. R. A study of

catharsis of aggression. Journal of Personality and
Mankoff, M., and Flacks, R. The changing social base of

the American student movement. Annals of the
1971, 395, 54-67.
Mann, J., Sidman, J., and Starr, S. Effects of erotic films
on the sexual behavior of married couples. In
Technical report of the Commission on Obscenity
and Pornography. Vol. 8. Erotica and social
behavior. Washington, D.C.: U.S. Government
Mark, V. H., and Ervin, F. R. Violence and the brain.

New York: Harper & Row, 1970.
Marmor, J. Psychoanalytic therapy as an educational

process: Common denominators in the
therapeutic approaches of different
psychoanalytic “schools.” In J. H. Masserman
(Ed.), Science and psychoanalysis. Vol. 5.
Psychoanalytic education. New York: Grune &
Stratton, 1962. Pp. 286-99.
Martin, M., Burkholder, R., Rosenthal, T. L., Tharp, R.

G., and Thorne, G. L. Programming behavior
change and reintegration into school milieux of
extreme adolescent deviates. Behaviour Research
and Therapy, 1968, 6, 371-83.
Marx, G. T. Issueless riots. Annals of the American

Academy of Political and Social Science, 1970a,
391, 21-33.
Marx, G. T. Civil disorders and the agents of social

control. Journal of Social Issues, 1970b, 26, 19-58.
Marx, M. H. Some relations between frustration and
drive. In M. R. Jones (Ed.), Nebraska symposium
on motivation: 1956. Lincoln, Neb.: University of
Nebraska Press, 1956. Pp. 92-130.
Maslow, A. H. Deprivation, threat, and frustration.

Psychological Review, 1941, 48, 364-66.
Mead, M. Sex and temperament in three savage tribes.

New York: Morrow, 1935.
Megargee, E. I. Undercontrolled and overcontrolled

personality types in extreme antisocial
aggression. Psychological Monographs, 1966, 80,
No. 3 (Whole No. 611).
Megargee, E. I., and Mendelsohn, G. A. A cross-

validation of twelve MMPI indices of hostility and
control. Journal of Abnormal and Social
Psychology, 1962, 65, 431-38.
Meyer, T. P. Some effects of real newsfilm violence on

the behavior of viewers. Journal of Broadcasting,
1971, 15, 275-85.
Meyer, T. P. Effects of viewing justified and unjustified

real film violence on aggressive behavior. Journal
29.
Milgram, S. Behavioral study of obedience. Journal of
78.
Milgram, S. Group pressure and action against a

person. Journal of Abnormal and Social
Psychology, 1964, 69, 137-43.
Milgram, S. Liberating effects of group pressure.

Journal of Personality and Social Psychology,
1965a, 1, 127-34.
Milgram, S. Some conditions of obedience and

disobedience to authority. Human Relations,
1965b, 18, 57-76.
Miller, N. E. The frustration-aggression hypothesis.

Miller, N. E. Liberalization of basic S-R concepts:

Extensions to conflict behavior, motivation, and
social learning. In S. Koch (Ed.), Psychology: A
study of a science. New York: McGraw-Hill, 1959.
Pp. 196-292.
Miller, N. E. Learning of visceral and glandular

responses. Science, 1969, 163, 434-45.
Miller, N. E., and Dollard, J. Social learning and

imitation. New Haven, Conn.: Yale University
Press, 1941.
Miller, W. B. Lower class culture as a generating
milieu of gang delinquency. Journal of Social
Issues, 1958, 14, 5-19.
Miller, W. B., Geertz, H., and Cutter, H. S. G. Aggression

in a boys’ streetcorner group. Psychiatry, 1961,
24, 282-98.
Mischel, W. Personality and assessment. New York:

Wiley, 1968.
Mischel, W., and Baker, N. Cognitive stimulus

transformations in self-control. Unpublished
manuscript, Stanford University, 1972.
Mischel, W., Ebbesen, E. B., and Zeiss, A. R. Cognitive

and attentional mechanisms in delay of
gratification. Journal of Personality and Social
Psychology, 1972, 21, 204-18.
Mischel, W., and Liebert, R. M. Effects of discrepancies

between observed and imposed reward criteria
on their acquisition and transmission. Journal of
Personality and Social Psychology, 1966, 3, 45-53.
Mischel, W., and Liebert, R. M. The role of power in the

adoption of self-reward patterns. Child
Development, 1967, 38, 673-83.
Montagu, M. F, A. (Ed.), Man and aggression. New

York: Oxford University Press, 1968.
Moser, D. Screams, slaps and love. Life, May 7, 1965,
90A-101.
Munroe, R. L. Schools of psychoanalytic thought. New

York: Dryden, 1955.
Murphy, J. T. Federal Aviation Administration.

Personal communication, 1971.
Myer, J. S., and White, R. T. Aggressive motivation in

the rat. Animal Behaviour, 1965, 13, 430-33.
Neiderhoffer, A. Behind the shield. New York:

Doubleday, 1967.
Nelsen, E. A. Social reinforcement for expression vs.

suppression of aggression. Merill-Palmer
Quarterly, 1969, 15, 259-78.
Nelson, J. D., Gelfand, D. M., and Hartmann, D. P.

Children’s aggression following competition and
exposure to an aggressive model. Child
Development, 1969, 40, 1085-97.
The New York Times. Veteran kills 12 in mad rampage

on Camden street. September 7, 1949, p. 1.
The New York Times. Sniper in Texas U. tower kills 12.

August 2, 1966a, p. 1.
The New York Times. Sniper told psychiatrist he had

thought of going to tower and “shooting people.”
August 3, 1966b, p. 20.
The New York Times. Friends of Whitman reveal he

often broke law. August 8, 1966c, p. 19.
The New York Times. Youth, 18, slays 4 women and

child in beauty school. November 13, 1966d, p. 1.
The New York Times. TV show blamed by F.A.A. for

rise in bomb hoax calls. December 21, 1966e, p.
79.
Nisbett, R. E., and Schachter, S. Cognitive manipulation

of pain. Journal of Experimental Social Psychology,
1966, 2, 227-36.
O’Connor, R. D. Modification of social withdrawal

through symbolic modeling. Journal of Applied
Behavior Analysis, 1969, 2, 15-22.
O’Connor, R. D. Relative efficacy of modeling, shaping,

and the combined procedures for modification of
social withdrawal. Journal of Abnormal
Psychology, 1972, 79, 327-34.
Osborn, D. K., and Endsley, R. C. Emotional reactions

of young children to TV violence. Child
Development, 1971, 42, 321-31.
Ostrow, M., and Ostrow, M. Bilaterally synchronous

paroxysmal slow activity in the
electroencephalograms of non-epileptics. Journal
of Nervous and Mental Disease, 1946, 103, 346-58.
Owen, D. R. The 47, XYY male: A review. Psychological

Bulletin, 1972, 77, 209-33.
Packer, H. L. The limits of the criminal sanction.

Stanford, Calif.: Stanford University Press, 1968.
Page, M. M., and Scheidt, R. J. The elusive weapons

effect: Demand awareness, evaluation
apprehension, and slightly sophisticated subjects.
Journal of Personality and Social Psychology, 1971,
20, 304-18.
Paisley, M. B. Social policy research and the realities of

the system: Violence done to TV research.
Unpublished manuscript, Stanford University,
1972.
Panksepp, J. Aggression elicited by electrical

stimulation of the hypothalamus in albino rats.
Physiology and Behavior, 1971, 6, 321-30.
Panksepp, J., and Trowill, J. Electrically induced

affective attack from the hypothalamus of the
albino rat. Psychonomic Science, 1969, 16, 118-19.
Palo Alto Times. Jetliner bomb hoax backfires. July 17,

1971, p. 1.
Palo Alto Times. Hijackers flee to jungle, Cuba. May 6,
1972, p. 1.
Parke, R. D. Effectiveness of punishment as an

interaction of intensity, timing, agent nurturance,
and cognitive structuring. Child Development,
1969, 40, 213-35.
Parke, R. D., Berkowitz, L., Leyens, J., West, S., and

Sebastian, R. The effects of repeated exposure to
movie violence on aggressive behavior in juvenile
delinquent boys: A field experimental approach.
Unpublished manuscript, University of
Wisconsin, 1972.
Parke, R. D., Ewall., W., and Slaby, R. G. Hostile and

helpful verbalizations as regulators of nonverbal
Psychology, 1972, 23, 243-48.
Parke, R. D., and Walters, R. H. Some factors

influencing the efficacy of punishment training
for inducing response inhibition. Monographs of
the Society for Research in Child Development,
1967, Vol. XXXII, No. 1 (Serial No. 109).
Parke, R. D., Wiederholt, C., and Slaby, R. G. The effect

of exposure to a model’s aggressive
verbalizations on the observer’s motor
aggression. Unpublished manuscript, University
of Wisconsin, 1972.
Pastore, N. The role of arbitrariness in the frustration-

aggression hypothesis. Journal of Abnormal and
Patterson, G. R., and Cobb, J. A. A dyadic analysis of

aggressive behaviors. In J. P. Hill (Ed.), Minnesota
symposia on child psychology, Vol. 4. Minneapolis,
Minn.: University of Minnesota Press, 1971.
Patterson, G. R., Cobb, J. A., and Ray, R. S. A social

engineering technology for retraining the families
of aggressive boys. In H. E. Adams and I. P. Unikel
(Eds.), Issues and trends in behavior therapy.
Springfield, Ill.: Charles C Thomas, 1972, in press.
Patterson, G. R., Littman, R. A., and Bricker, W.

Assertive behavior in children: A step toward a
theory of aggression. Monographs of the Society
for Research in Child Development, 1967, 32, No. 5
(Serial No. 113).
Patterson, G. R., Ludwig, M., and Sonoda, B.

Reinforcement of aggression in children.
Unpublished manuscript, University of Oregon,
1961.
Patterson, G. R., Ray, R. S., and Shaw, D. Direct

intervention in families of deviant children.
Unpublished manuscript, University of Oregon,
1968.
Patterson, G. R., and Reid, J. B. Reciprocity and

coercion: Two facets of social systems. In C.
Neuringer and J. L. Michael (Eds.), Behavior
modification in clinical psychology. New York:
Appleton-Century-Crofts, 1970. Pp. 133-77.
Pendergrass, V. E. Effect of length of time-out from

positive reinforcement and schedule of
application in suppression of aggressive
behavior. Psychological Record, 1971, 21, 75-80.
Penney, R. K. The effects of non-reinforcement of

response strength as a function of number of
previous reinforcements. Canadian Journal of
Psychology, 1960, 14, 206-15.
Peterson, R. A. Aggression level as a function of

expected retaliation and aggression level of
target and aggressor. Developmental Psychology,
1971, 5, 161-66.
Peterson, R. E. The scope of organized student protest

in 1964-65. Princeton, N.J.: Educational Testing
Service, 1966.
Pettigrew, T. F. Actual gains and psychological losses:

The Negro American protest. Journal of Negro
Education, 1963, 32, 493-506.
Pettigrew, T. F. Social evaluation theory:

Convergences and applications. In D. Levine (Ed.),
Nebraska symposium on motivation: 1967.
Lincoln: University of Nebraska Press, 1967. Pp.
241-311.
Phillips, E. L. Achievement place: Token

reinforcement procedures in a home-style
rehabilitation setting for “pre-delinquent” boys.
Journal of Applied Behavior Analysis, 1968, 1, 213-
23.
Phillips, E. L., Wolf, M. M., and Fixsen, D. L. An

experimental analysis of governmental systems
at achievement place, a group home for pre-
delinquent boys. Journal of Applied Behavior
Analysis, 1972.
Piliavin, I., Hardyck, J., and Vadum, A. Reactions to the

victim in a just or non-just world. Paper
presented at the meeting of the Society of
Experimental Social Psychology, Bethesda,
Maryland, August 1967.
Pinkston, E. M., Reese, N. H., LeBlanc, J. M., and Baer, D.

M. Independent control of a preschool child’s
aggression and peer interaction by contingent
teacher attention. Journal of Applied Behavioral
Analysis, 1972, in press.
Pisano, R., and Taylor, S. P. Reduction of physical

aggression: The effects of four strategies. Journal
of Personality and Social Psychology, 1971, 19,
237-42.
Polsky, H. W. Cottage six: The social system of

delinquent boys in residential treatment. New
York: Russell Sage Foundation, 1962.
Porro, C. R. Effects of the observation of a model's

affective responses to her own transgression on
resistance to temptation in children. Dissertation
Abstracts, 1968, 28, 3064.
Portland (Me.) Press-Herald, It looked easy on TV, says

the man held in killing here. November 28, 1963,
p. 1.
Powell. D. A., and Creer, T. L. Interaction of

developmental and environmental variables in
shock-elicited aggression. Journal of Comparative
and Physiological Psychology, 1969, 69, 219-25.
Powers, P. C., and Geen, R. G. Effects of the behavior

and the perceived arousal of a model on
instrumental aggression. Journal of Personality
Powers, T. Diana: The making of a terrorist. Boston:
Houghton Mifflin, 1971.
Premack, D. Reinforcement theory. In D. Levine (Ed.),

Nebraska symposium on motivation: 1965.
Lincoln, Neb.: University of Nebraska Press, 1965.
Pp. 123-80.
Prentice, N. M. The influence of live and symbolic

modeling on promoting moral judgments of
adolescent juvenile delinquents. Journal of
Abnormal Psychology, 1972 (in press).
Price, W. H., and Whatmore, P. B. Behaviour disorders

and pattern of crime among XYY males identified
at a maximum security hospital. British Medical
Journal, 1967, 1, 533-36.
Rachman, S. The effects of psychotherapy. Oxford, Eng.:

Pergamon, 1971.
Radke-Yarrow, M., Trager, H., and Miller, J. The role of

parents in the development of children’s ethnic
attitudes. Child Development, 1952, 23, 13-53.
Ransford, H. E. Isolation, powerlessness, and violence:

A study of attitudes and participation in the
Watts riot. American Journal of Sociology, 1968,
73, 581-91.
Raush, H. L. Interaction sequences. Journal of

99.
Redl, F., and Wineman, D. Children who hate: The

disorganization and breakdown of behavior
controls. Glencoe, Ill.: Free Press, 1951.
Rees, J. R. The shaping of psychiatry by war. New York:

Norton, 1945.
Reich, P., and Hepps, R. B. Homicide during a

psychosis induced by LSD. Journal of the
American Medical Association, 1972, 219, 869-71.
Reichard, G. A. Social life. In F. Boas (Ed.), General

anthropology. Boston: Heath, 1938. Pp. 409-86.
Reifler, C. B., Howard, J., Lipton, M. A., Liptzin, M. B.,

and Widmann, D. E. Pornography: An
experimental study of effects. American Journal of
Psychiatry, 1971, 128, 575-82.
Reiss, A. J., Jr. The police and the public. New Haven,
Conn.: Yale University Press, 1971.
Reppucci, N. D., and Reiss, S. Effects of operant

treatment with disruptive and normal
elementary school children. Paper presented at
the meeting of the American Psychological
Association, Miami, September 1970.
Rescorla, R. A. Pavlovian conditioned inhibition.
Psychological Bulletin, 1969, 72, 77-94.
Reynierse, J. H. Submissive postures during shock-

elicited aggression. Animal Behaviour, 1971, 19,
102-7.
Reynolds, G. S., Catania, A. C., and Skinner, B. F.

Conditioned and unconditioned aggression in
pigeons. Journal of the Experimental Analysis of
Behavior, 1963, 6, 73-74.
Risley, T. R. The effects and side effects of punishing

the autistic behaviors of a deviant child. Journal
of Applied Behavior Analysis, 1968, 1, 21-34.
Risley, T. R., and Wolf, M. M. Experimental

manipulation of autistic behaviors and
generalization into the home. In R. Ulrich, T.
Stachnik, and J. Mabry (Eds.), Control of human
behavior. Vol. 1. Glenview, Ill.: Scott, Foresman,
1966. Pp. 193-98.
Roberts, W. W., Steinberg, M. L., and Means, L. W.

Hypothalamic mechanisms for sexual, aggressive,
and other motivational behaviors in the opossum,
Didelphis virginiana. Journal of Comparative and
Physiological Psychology, 1967, 64, 1-15.
Robinson, J. P., and Bachman, J. G. Television viewing

habits and aggression. In G. A. Comstock and E. A.
Vol. 3. Television and adolescent aggressiveness.
1972. Pp. 372-82.
Rosekrans, M. A., and Hartup, W. W. Imitative

influences of consistent and inconsistent
response consequences to a model on aggressive
behavior in children. Journal of Personality and
Rosenbaum, M. E., and de Charms, R. Direct and

vicarious reduction of hostility. Journal of
11.
Rosenhan, D., Frederick, F., and Burrowes, A.

Preaching' and practicing: Effects of channel
discrepancy on norm internalization. Child
Development, 1968, 39, 291-301.
Rosenzweig, S. An outline of frustration theory. In J.

McV. Hunt (Ed.), Personality and the behavior
disorders. Vol. I. New York: Ronald Press, 1944.
Pp. 379-88.
Ross, L., Rodin, J., and Zimbardo, P. G. Toward an

attribution therapy: The reduction of fear
through induced cognitive-emotional
misattribution. Journal of Personality and Social
Psychology, 1969, 12, 279-88.
Rothballer, A. B. Aggression, defense and

neurohumors. In C. D. Clemente and D. B.
Lindsley (Eds.), Aggression and defense. Berkeley,
Calif.: University of California Press, 1967. Pp.
135-50.
Russo, S. Adaptations in behavioural therapy with

children. Behaviour Research and Therapy, 1964,
2, 43-47.
San Francisco Chronicle. “James Dean” knifing in South

City. March 1, 1961, p. 6.
San Francisco Chronicle. Klan rally—hate and fears.

May 13, 1963, p. 17.
San Francisco Chronicle. A “club” for teen sadists.

November 26, 1964, p. 3.
San Francisco Chronicle. A badge for killing Reds. June

11, 1970, p. 23.
San Francisco Chronicle. Tate killings—“Right thing to

do.’’ February 11, 1971a, p. 21.
San Francisco Chronicle. Calley says killing was order

of day. February 24, 1971b, p. 1.
San Francisco Chronicle. Bomb hoax costs airline

$560,000. May 27, 1971c, p. 1.
San Francisco Chronicle. BOAC knew the plot—and
foiled the plotter. August 4, 197Id, p. 1.
San Francisco Chronicle. Hijacker’s slick parachute

escape. November 26, 1971e, p. 1.
Sandler, J„ and Quagliano, J. Punishment in a signal

avoidance situation. Paper presented at the
meeting of the Southeastern Psychological
Association, Gatlinburg, Tennessee, 1964.
Sarason, I. G. and Ganzer, V. J. Social influence

techniques in clinical and community psychology.
In C. D. Spielberger (Ed.), Current topics in clinical
and community psychology. New York: Academic
Press, 1969. Pp. 1-66.
Sarnoff, I. Identification with the aggressor: Some

personality correlates of antisemitism among
Jews. Journal of Personality, 1951, 20, 199-218.
Saul, L. J., Rome, H., and Leuser, E. Desensitization of

combat fatigue patients. American Journal of
Psychiatry, 1946, 102, 476-78.
Savitsky, J. C., Rogers, R. W., Izard, C. E„ and Liebert, R.

M. Role of frustration and anger in the imitation
of filmed aggression against a human victim.
Psychological Reports, 1972, 29, 807-10.
Schachter, J. Pain, fear, and anger in hypertensives
and normotensives: A psychophysiological study.
Psychosomatic Medicine, 1957, 19, 17-29.
Schachter, S., and Singer, J. E. Cognitive, social, and

physiological determinants of emotional state.
Schaefer, H. H. Self-injurious behavior: Shaping “head-

banging” in monkeys. Journal of Applied Behavior
Analysis, 1970, 3, 111-16.
Schenkel, R. Submission: Its features and functions in

the wolf and dog. American Zoologist, 1967, 7,
319-29.
Schmidt, G., and Sigusch, V. Sex differences in

responses to psychosexual stimulation by flints
and slides. Journal of Sex Research, 1970, 6, 268-
83.
Schramm, W., Lyle, J., and Parker, E. B. Television in the

lives of our children. Stanford, Calif.: Stanford
Schwartz, G. E. Voluntary control of human

cardiovascular integration and differentiation
through feedback and reward. Science, 1972, 175,
90-93.
Schwartz, L. A. Group psychotherapy in the war

neuroses. American Journal of Psychiatry, 1945,
101, 498-500.
Scott, J. P. An experimental test of the theory that

social behavior determines social organization.
Science, 1944, 99, 42-43.
Scott, J. P. Hostility and aggression. In B. Wolman (Ed.)

Handbook of genetic psychology. Englewood Cliffs,
N.J.: Prentice-Hall, 1972.
Scott, J. P., and Marston, M. V. Nonadaptive behavior

resulting from a series of defeats in fighting mice.
48, 417-28.
Scott, P. M„ Burton, R. V., and Yarrow, M. R. Social

reinforcement under natural conditions. Child
Development, 1967, 38, 53-63.
Searle, J. R. A foolproof scenario for student revolts.

The New York Times Magazine, December 29,
1968, p. 4.
Searle, L. V. The organization of hereditary maze-

brightness and maze-dullness. Genetic Psychology
Monographs, 1949, 39, 279-325.
Sears, D. O., and McConahay, J. B. Participation in the

Los Angeles riot. Social Problems, 1969, 17, 3-20.
Sears, R. R. Non-aggressive reactions to frustration.
Sears, R. R„ Maccoby, E. E., and Levin, H. Patterns of

child rearing. Evanston, Ill.: Row, Peterson, 1957.
Sears, R. R., Whiting, J. W. M., Nowlis, V., and Sears, P.

S. Some child-rearing antecedents of aggression
and dependency in young children. Genetic
Psychology Monographs, 1953, 47, 135-234.
Shah, S. A. Report on the XYY chromosomal

abnormality. Washington, D.C.: Government
Printing Office, 1970.
Short, J. F., Jr. (Ed.). Gang delinquency and delinquent

subcultures. New York: Harper & Row, 1968.
Short, J. F., Jr., and Strodtbeck, F. L. Why gangs fight.

Trans-action, 1964, 1 (6), 25-29.
Shuntich, R. J. and Taylor, S. P. The effects of alcohol

on human physical aggression. Journal of
Experimental Research in Personality, 1972, 6, 34-
38.
Shupe, L. M. Alcohol and crime: A study of the urine

alcohol concentration found in 882 persons
arrested during or immediately after the
commission of a felony. Journal of Criminal Law,
Criminology and Police Science, 1954, 44, 661-64.
Sibley, S. A., Abbott, M. S., and Cooper, B. P.
Modification of the classroom behavior of a
disadvantaged kindergarten boy by social
reinforcement and isolation. Journal of
Experimental Child Psychology, 1969, 7, 203-19.
Siegel, A. E. Film-mediated fantasy aggression and

strength of aggressive drive. Child Development,
1956, 27, 365-78.
Siegel, A. E. The influence of violence in the mass

media upon children’s role expectation. Child
Development, 1958, 29, 35-56.
Siegel, A. E. Violence in the mass media. In D. N.

Daniels, M. F. Gilula, and F. M. Ochberg (Eds.),
Violence and the struggle for existence. Boston:
Little, Brown, 1970. Pp. 193-239.
Silver, L. B., Dublin, C. C., and Lourie, R. S. Does

violence breed violence? Contributions from a
study of the child abuse syndrome. American
Journal of Psychiatry, 1969, 126,404-7.
Simkins, L. Effects of examiner attitudes and type of

reinforcement on the conditioning of hostile
verbs. Journal of Personality, 1961, 29, 380-95.
Singer, B. D. Violence, protest and war in television

news: The U.S. and Canada compared. Public
Opinion Quarterly, 1970-71, 34, 611-16.
Skeyhill, T. (Ed.). Sergeant York: His own life story and

war diary. Garden City, N.Y.: Doubleday, Doran,
1928.
Sloane, H. N„ Johnston, M. K., and Bijou, S. W.

Successive modification of aggressive behaviour
and aggressive fantasy play by management of
contingencies. Journal of Child Psychology and
Psychiatry and Allied Disciplines, 1967, 8, 217-26.
Snow, C. P. Either-or. Progressive, 1961, 25 (2), 24-25.
Somers, R. H. The mainsprings of rebellion: A survey

of Berkeley students in November, 1964. In S. M.
Lipset and S. S. Wolin (Eds.), The Berkeley student
revolt: Facts and interpretations. Garden City,
N.Y.: Doubleday, 1965, Pp. 530-57.
Sorenson, E. R. The evolving Fore: A study of

socialization and cultural change in the New
Guinea highlands. Unpublished doctoral
dissertation, Stanford University, 1971.
Sosa, J. N. Vascular effects of frustration on passive

and aggressive members of a clinical population.
Unpublished master’s thesis, Florida State
University, 1968.
Southwick, C. H. An experimental study of intragroup
agonistic behavior in rhesus monkeys (Macaca
mulatta). Behaviour, 1967, 28, 182-209.
Spelt, D. K. The conditioning of the human fetus in

utero. Journal of Experimental Psychology, 1948,
38, 375-76.
Spence, K. W. Anxiety (drive) level and performance

in eyelid conditioning. Psychological Bulletin,
1964, 61, 129-39.
Spiegel, J. P. Campus disorders: A transactional

approach. The Psychoanalytic Review, 1970, 57,
472-504.
Staats, A. W., and Butterfield, W. H. Treatment of

nonreading in a culturally deprived juvenile
delinquent: An application of reinforcement
principles. Child Development, 1965, 36, 925-42.
Staats, A. W., and Staats, C. K. Attitudes established by

classical conditioning. Journal of Abnormal and
Staats, C. K., and Staats, A. W. Meaning established by

classical conditioning. Journal of Experimental
Psychology, 1957, 54, 74-80.
Stachnik, T. J., Ulrich, R., and Mabry, J. H.

Reinforcement of intra- and inter-species
aggression with intracranial stimulation.
American Zoologist, 1966, 6, 663-68.
Staples, F. R., and Walters, R. H. Influence of positive

reinforcement of aggression on subjects differing
in initial aggressive level. Journal of Consulting
Psychology, 1964, 28, 547-52.
Steffy, R. A., Hart, J., Crow, M., Torney, D., and Marlett,
N. Operant behaviour modification techniques
applied to a ward of severely regressed and
aggressive patients. Canadian Psychiatric
Association Journal, 1969, 14, 59-67.
Stein, A. H.. Friedrich, L. K., and Vondracek, F.

Television content and young children’s
behavior. In J. P. Murray, E. A. Rubinstein, and G.
A. Comstock (Eds.), Television and social behavior.
Vol. 2. Television and social learning. Washington,
D.C.: Government Printing Office, 1972, Pp. 202-
317.
Steuer, F. B., Applefield, J. M., and Smith, R. Televised

aggression and the interpersonal aggression of
preschool children. Journal of Experimental Child
Psychology, 1971, 11, 442-47.
Stevenson, H. W. Studies of racial awareness in young

children. In W. W. Hartup and N. L. Smothergill
(Eds.), The young child: Review of research.
Washington, D.C.: National Association for the
Education of Young Children, 1967. Pp. 206-13.
Stone, L. J., and Hokanson, J. E. Arousal reduction via

self-punitive behavior. Journal of Personality and
Sulzer, B., Mayer, G. R., and Cody, J. J. Assisting

teachers with managing classroom behavioral
problems. Elementary School Guidance and
Counseling, 1968, 3, 40-48.
Tannenbaum, P. H. Studies in film- and television-

mediated arousal and aggression: A progress
report. In G. A. Comstock, E. A. Rubinstein, and J.
P. Murray (Eds.), Television and social behavior.
Vol. 5. Television effects: Further explorations.
1972. Pp. 309-50.
Tannenbaum, P. H., and Gaer, E. P. Mood change as a

function of stress of protagonist and degree of
identification in a film-viewing situation. Journal
16.
Task force report: The police. Washington, D.C.: U.S.

Government Printing Office, 1967.
Taylor, S. P. Aggressive behavior and physiological

arousal as a function of provocation and the
tendency to inhibit aggression. Journal of
Personality, 1967, 35, 297-310.
Tharp, R. G., and Wetzel, R. J. Behavior modification in

the natural environment. New York: Academic
Press, 1969.
Thelen, M. H., and Soltz, W. The effect of vicarious

reinforcement on imitation in two social racial
groups. Child Development, 1969, 40, 879-87.
Thibaut, J. W., and Coules, J. The role of

communication in the reduction of interpersonal
hostility. Journal of Abnormal and Social
Psychology, 1952, 47, 770-777.
Thompson, T. Aggressive behavior of Siamese fighting

fish. In S. Garattini and E. B. Sigg (Eds.),
Aggressive behaviour. New York: Wiley, Pp. 15-31.
Thompson, T., and Bloom, W. Aggressive behavior and

extinction-induced response-rate increase.
Tieger, J. PI. Police discretion and discriminatory

enforcement. Duke Law Journal, 1971, 717-43.
Tinbergen, N. The study of instinct. London: Oxford

Toch, H. Violent men. Chicago: Aldine, 1969.
Toch, H., and Schulte, R. Readiness to perceive
violence as a result of police training. British
Journal of Psychology, 1961, 52, 389-93.
Tomlinson, T. L. The development of a riot ideology

among urban Negroes. American Behavioral
Scientist, 1968, 11, 27-31.
Turnbull, C. M. The forest people. New York: Simon

and Schuster, 1961,
Turner, R. H. The public perception of protest.

American Sociological Review, 1969, 34, 815-31.
Tyler, V. O., Jr., and Brown, G. D. The use of swift, brief

isolation as a group control device for
institutionalized delinquents. Behavior Research
and Therapy, 1967, 5, 1-9.
Ulrich, R. E. Pain as a cause of aggression. American

Zoologist, 1966, 6, 643-62.
Ulrich, R. E. The experimental analysis of aggression.

Unpublished manuscript, Western Michigan
University, 1967.
Ulrich, R. E., and Azrin, N. H. Reflexive fighting in

response to aversive stimulation. Journal of the
Experimental Analysis of Behavior, 5, 511-20.
Ulrich, R. E., and Flavell, J. E. Human aggression. In C.
Neuringer and J. L. Michael (Eds.), Behavior
modification in clinical psychology. New York:
Appleton-Century-Crofts, 1970. Pp. 105-32.
Ulrich, R. E., Hutchinson, R. R„ and Azrin, N. H. Pain-

elicited aggression, Psychological Record, 1965,
15, 111-26.
Ulrich, R. E., Johnston, M., Richardson, J., and Wolff, P.

The operant conditioning of fighting behavior in
rats. Psychological Record, 13, 465-70.
Ulrich, R. E., Wolfe, M., and Dulaney, S. Punishment of

shock-induced aggression. Journal of the
Experimental Analysis of Behavior, 1969, 12,
1009-15.
U.S. Senate Hearings, Subcommittee to Investigate

Juvenile Delinquency. Effects on young people of
violence and crime portrayed on television.
Washington, D.C.: U.S. Government Printing
Office, 1963.
van Lawick-Goodall, J. Some aspects of aggressive

behavior in a group of free-living chimpanzees.
International Social Science Journal, 1971, 23, 89-
97.
Vantress, F. E., and Williams, C. B. The effect of the

presence of the provocator and the opportunity
to counteraggress on systolic blood pressure. The
Journal of General Psychology, 1972, 86, 63-68.
Vernon, W., and Ulrich, R. Classical conditioning of

pain-elicited aggression. Science, 1966, 152, 668-
69.
Wahler, R. G. Behavior therapy for oppositional

children: Love is not enough. Paper read at
Eastern Psychological Association meeting,
Washington, D.C., April 1968.
Waldman, D. M., and Baron, R. A. Aggression as a

function of exposure and similarity to a
nonaggressive model. Psychonomic Science, 1971,
23, 381-83.
Walters, R. H., Bowen, N. V., and Parke, R. D. Influence

of looking behavior of a social model on
subsequent looking behavior of observers of the
model. Perceptual & Motor Skills, 1964, 18, 469-
83.
Walters, R. H., and Brown, M. Studies of reinforcement

of aggression. III. Transfer of responses to an
interpersonal situation. Child Development, 1963,
34, 563-71.
Walters, R. H., and Llewellyn-Thomas, E.

Enhancement of punitiveness by visual and
audiovisual displays. Canadian Journal of
Psychology, 1963, 17, 244-55.
Walters, R. H., Parke, R. D„ and Cane, V. A. Timing of

punishment and the observation of consequences
to others as determinants of response inhibition.
Journal of Experimental Child Psychology, 1965, 2,
10-30.
Walters, R. H., and Willlows, D. C. Imitative behavior

of disturbed and nondisturbed children following
exposure to aggressive and nonaggressive
models. Child Development, 1968, 39, 79-89.
Ward, M. H., and Baker, B. L. Reinforcement therapy in

the classroom. Journal of Applied Behavior
Analysis, 1968, I, 323-28.
Warner, M. Organizational content and control of

policy in the television newsroom: A participant
observation study. British Journal of Sociology,
1971, 22, 283-94.
Washington Post. Youth theft ring cracked. January 30,

1971, p. Bl.
Wechsberg, J. (Ed.). The murderers among us. New

York: McGraw-Hill, 1967.
Weisz, A. E., and Taylor, R. L. American Presidential

assassinations. In D. N. Daniels, M. F. Gilula, and F.
Ochberg (Eds.), Violence and the struggle for
existence. Boston: Little, Brown, 1970. Pp. 291-
307.
Welch, A. S., and Welch, B. L. Reduction of

norepinephrine in the lower brainstem by
psychological stimulus. Proceedings of the
National Academy of Sciences, 1968, 60, 478-81.
Wells, W. D. Television and aggression: Replication of

an experimental field study. Unpublished
manuscript, University of Chicago, 1971.
Wertham, F. Is TV hardening us to the war in

Vietnam? The New York Times, December 4, 1966,
p. D23.
Wheeler, L. Toward a theory of behavioral contagion.

Wheeler, L., and Caggiula, A. R. The contagion of

aggression. Journal of Experimental Social
Psychology, 1966, 2, 1-10.
Wheeler, L., and Levine, L. Observer-model similarity

in the contagion of aggression. Sociometry, 1967,
30, 41-49.
Wheeler, L., and Smith, S. Censure of the model in the

contagion of aggression. Journal of Personality
Whiting, J. W. M. Becoming a Kwoma. New Haven,
Conn.: Yale University Press, 1941.
Whiting, J. W. M., and Child, I. L. Child training and

personality. New Haven, Conn.: Yale University
Press, 1953.
Whyte, W. F. Street corner society. 2nd ed. Chicago:

University of Chicago Press, 1955.
Williams, C. D. The elimination of tantrum behavior by

extinction procedures. Journal of Abnormal and
Social Psychology, 1959, 59, 269.
Wolf, M. M., Risley, T., Johnston, M. K., Harris, F., and

Allen, E. Application of operant conditioning
procedures to the behavior problems of an
autistic child: A follow-up and extension.
Behaviour Research and Therapy, 1967, 5, 103-11.
Wolf, M. M„ Risley, T. R., and Mees, H. Application of

operant conditioning procedures to the
behaviour problems of an autistic child.
Behaviour Research and Therapy, 1964, 1, 305-12.
Wolfe, B. M., and Baron, R. A. Laboratory aggression

related to aggression in naturalistic social
situations: Effects of an aggressive model on the
behavior of college student and prisoner
observers. Psychonomic Science, 1971, 24, 193-
94.
Wolfe, M., Ulrich, R., and Dulaney, S. Fighting and

escape reaction in paired rats. Psychological
Record, 1971, 21, 59-68.
Wolfgang, M. E., and Ferracuti, F. The subculture of

violence. London: Tavistock, 1967.
Wolfgang, M. E., Kelly, A. and Nolde, H. C. Comparison

of the executed and the commuted among
admissions to death row. Journal of Criminal Law,
Criminology, and Political Science, 1962, 53, 301-
11.
Worchel, P. Catharsis and the relief of hostility.

55, 238-43.
Wright, G. O. Projection and displacement: A cross-

cultural study of folk-tale aggression. Journal of
28.
Wright, M. E. Constructiveness of play as affected by

group organization and frustration. Character
and Personality, 1942, 11, 40-49.
Wright, M. E. The influence of frustration upon the

social relations of young children. Character and
Personality, 1943, 12, 111-22.
Yablonsky, L. The violent gang. New York: Macmillan,
1962.
Zajonc, R. B. Attitudinal effects of mere exposure.

Journal of Personality and Social Psychology
Monograph Supplement, 1968, 9, (2), Part 2. Pp.
1-27.
Zaltman, G., Kotler, P., and Kaufman, I. (Eds.), Creating

social change. New York: Holt, Rinehart and
Winston, 1972.
Zeilberger, J., Sampen, S. E., and Sloane, H. N., Jr.

Modification of a child’s problem behaviors in the
home with the mother as therapist. Journal of
Zillmann, D. Excitation transfer in communication-

mediated aggressive behavior. Journal of
Experimental Social Psychology, 1971, 7, 419-34.
Zillmann, D., and Johnson, R. C. Motivated

aggressiveness perpetuated by exposure to
aggressive films and reduced by exposure to non-
aggressive films. Journal of Research in
Personality, 1973, 7, 261-76.
Zimbardo, P. G. The human choice: Individuation,

reason, and order versus deindividuation,
impulse, and chaos. Nebraska symposium on
motivation, 1969. Lincoln, Neb.: University of
Nebraska, 1969. Pp. 237-309.
Zimring, F. E. Perspectives on deterrence. Washington,

D.C.: Government Printing Office, 1971.
Zimring, F., and Hawkins, G. The legal threat as an

instrument of social change. Journal of Social
Issues, 1971, 27, 33-48.

AGGRESSION - A Social Learning Analysis Albert Bandura PDF

Uploaded by

Copyright:

Available Formats

AGGRESSION - A Social Learning Analysis Albert Bandura PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

AGGRESSION - A Social Learning Analysis Albert Bandura PDF

Uploaded by

Copyright:

Available Formats

a social learning analysis

e-Book 2015 International Psychotherapy

All Rights Reserved

This e-book contains material protected under

Created in the United States of America

THEORETICAL ANALYSES OF AGGRESSION

DIFFERENTIAL IMPLICATIONS OF DRIVE AND

LEARNING THROUGH PRACTICE

LEARNING OF AGGRESSION UNDER NATURALLY

INSTRUCTIONAL CONTROL OF AGGRESSION

PUNISHMENT CONTROL OF AGGRESSION

REINFORCEMENT INFLUENCES IN COLLECTIVE

Modification and Control of Agression

PUNISHMENT CONTROL OF AGGRESSION

MODIFICATION OF ANTISOCIAL AGGRESSION

CHANGES IN SOCIAL SYSTEMS

This book is concerned with why man

addressing this issue, despite the great deal of

attention that has already been devoted to it.

Although aggression pervades our lives, few

its causes or to devise constructive ways of

reducing the level of societal violence. Man’s

technological capacity for massive destruction has

now developed to the point where he can no

longer continue to settle conflicts by destructive

means. In addition to the threats posed to survival,

developments make it all the more necessary to

understand the determinants of injurious conduct.

Theories of human behavior have been

undergoing drastic change in the last decade.

Interest in traditional approaches that depict

impelled by drive forces declined as their

deficiencies became apparent. New perspectives

human behavior. In this book I have attempted to

formulate a social learning theory of aggression

sufficiently broad in scope to integrate evidence

on all facets of aggression, whether individual or

better basis for explaining, predicting, and

explanatory power of social learning theory.

The worth of a psychological theory should be

the modification procedures that it produces. A

sizable portion of this book is therefore devoted to

demonstrating how social learning principles can

be applied individually and at the social systems

not require violent solutions. Some forceful

measures have functional value in promoting

The exercise of social power as an instrument of

No account of aggression can be complete

without discussing the social labeling and ethics of

conduct can be rapidly and drastically altered by

determinants of human aggression are best sought

in man’s social practices.

Many people have contributed to this book in

take this opportunity to acknowledge my

National Institute of Mental Health and the John

Simon Guggenheim Foundation greatly aided my

work. This project was begun during my residence

as a Fellow at the Center for Advanced Study in the