WWI

r
w
1

MM”)!

'L‘HE EFFECTS G5 VERBAL ENSHUCHONS
0N iMITATWE AGGiRfiSSiGé‘é

Wiilliliflh

1
w

W

 

—lw_;
I-bO
(0010)

Thesis {or ”W Dogma of DH. D.
MKCHEGAN STATE {INE'V'ERSITY
Gilbert Wiliiams DeRath
1963

THESIS

LIBRARY

Michigan State
University

 

This is to certify that the

thesis entitled

The Effects of Verbal Instructions
on Irnitative Aggression

presented by

Gilbert Williams De Rath

has been accepted towards fulfillment
of the requirements for

Ph, D. ._ degree in—ES¥.chongy

m/{My

Major, professor/w

Date / 5’15“; [yé k3

0-169

 

 

ABSTRACT

THE EFFECTS OF VERBAL INSTRUCTIONS
ON IMITATIVE AGGRESSION

by Gilbert Williams De Rath

Sixty nursery school boys were randomly assigned to three con-
ditions. Subjects in one experimental group individually observed a
five-minute film sequence of an aggressive model; a second experi-
mental group received, during the film presentation, prohibitive
verbal instructions regarding specific aggressive acts which the film
model directed at an inflated clown. A control group did not see the
film and received no instructions of any kind. Half of the subjects in
each-of the groups were randomly as signed to a neutral female experi-
menter in a doll play test situation which followed mild frustration
designed to‘elicit aggressive behavior. Subjects were observed in doll
play for a 20-minute period during which ratings were made of their
behavior at five-second intervals. Over half of the subjects were rated
by two independent raters in order to determine rater reliability.
-Ratings were made on 27 different behaviors which were broadly. classi-
fied as Imitative Aggression, Partially Imitative Aggression, Non-
Imitative Aggression, Non-Aggression, and Imitative Verbal Aggression.
The latter category was eliminated because of poor sound transmission
and the lack of intelligibility in the speech of these subjects.

It was predicted that boys exposed to an aggressive model would,
following a frustrating experience, reproduce the aggressive behaviors

of the model and would differ in this respect both from boys prohibited

Gilbert Williams De Rath

from performing such acts and boys not exposed to the model.

A second hypothesis was that boys who were verbally prohibited from
performing the aggressive actions of the model would inhibit these acts
in subsequent doll play. The third hypothesis concerned the generali-
zation of effects of exposure to an aggressive model and the generali—
zation of prohibiting instructions. It was expected that following
frustration, subjects not receiving inhibiting instructions would display
more aggression in the presence of the frustrator than in the presence
of a more neutral figure of a different sex. Finally, it was expected
that the effects of verbal prohibition on imitating of the model's aggre-
sive behavior would be greater in the presence of the person invoking
the prohibition than in the presence of a neutral experimenter.

The responses scored involved highly specific concrete classes
of behavior and yielded high inter-scorer reliabilities, the rank
correlation coefficients being in the . 903.

Subjects seeing the film without instructions clearly imitated the
aggressive behaviors of the model following the mild frustration experi-
ence. These differences were generally highly significant. Those
subjects who were prohibited during the presentation-of the film were
inhibited in imitating aggressive behaviors displayed earlier by the
model.

As a test of generalization of the effects of exposure to the film
and as a test of the generalization of prohibitive instruction, half of the
subjects in each group were exposed to a "neutral" experimenter in
doll play. The highly significant differences between the experimental
groups on imitative aggression remains, indicating that both-effects
readily generalize. Subjects who viewed the film without prohibitive

instructions displayed more aggression, both imitative and non-imitative,

Gilbert Williams De Rath

in the presence of the experimenter'whofrustrated them. The difference
‘ is highly significant for non-irnitative aggression,“ but only approaches
significance for imitative aggression.

It was expected that the effects of verbal prohibitions would be
greater in the presence of the person invoking the prohibition. However,
the differences here were not significant, both groups receiving pro-
hibitive instructions showing little imitative behavior.

The results were discussed in terms of a learning theory of dis-
placement rather than other theories utilizing such concepts as "identifi-

cation with the aggressor" or "defensive identification. "

Approved MM

Major Prfis sor

Date ’5/yfv/T /7é37

THE EFFECTS OF VERBALINSTRUCTIONS
ON IMITATIVE AGGRESSION

BY

Gilbert Williams De Rath

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCT OR OF PHILOSOPHY

Department of Psychology

1963

G 52qu 5L
7/8/64

To Gib and Rick

ii

ACKNOWLEDGEMENT

. Appreciation-is accorded to all those who were helpful in the
design, data gathering, and writing of this dissertation. Foremost
I wish to thank themembers of my committee: - Dr. Charles Hanley,
.Chairman, Dr. Norman Abeles, Dr. Alfred G. Dietze, and Dr. Bill
L. Kell. Appreciation is also. expressed to Dr. Clarence L.. Winder
who served as chairman during the early planning of the disserataion
and who maintained a helpful interest while away from Michigan State
University.

Grateful appreciation is due the following:

Miss Roseann Bulgarella, Miss Martha G. Andrews, and Miss
SandraCotter who served as raters and whose enthusiasm was always
encouraging.

Miss Betty Garlick, Miss Mariella Aikman, Mrs. Vera Borosage,
and Mrs. JoAnn ‘Lifshin, teachers at Spartan Cooperative‘Nursery
School, who took a personal interest in preparing and providing the
children whoserved as subjects. To the children themselves I want to
express my thanks for their help.

Dr. Robert E. McMichael who provided the transparent mirror
‘andtfurniture used in the experiment.

Dr. Albert Bandura of Stanford University who'provided the film
and sound track and whose studies were used as a model [for the research.

Dr.'. Alice C. Thorpe, Chairman of the Home, Management and Child
Development Department, who- was very helpful in obtaining 'Space for
the. experiment and giving permission to use the Spartan Cooperative
Nursery School facilities.

Mrs. Gloria De Rathlwho assisted as a rater, typist, and proof-
reader. Her continued support and understanding is gratefully

acknowledged.

***********

iii

TABLE OF CONTENTS

INTRODUCTION 0 O O O O O 0 O O O I O O I O O

Theories of Aggression . . . . . . . . .
The Frustration-Aggression Hypothesis

Displacement of Aggression. . .
Imitative Aggression . . . . . .
Hypotheses. . . . ... . . . . .

MET HOD O O I O O O O O O O O O O O O 0

subjects 0 O O O I O O I O O O O 0
Experimental Design . . . . . .
Experimental Conditions . . . .

Prohibitive Instructions . .
'Aggression Arousal. . . . . . .
D011 Play. O O O O O O O O I O O O O O O 0

Behavior Measures ...........
RESULTSo'eoooooo0000000000000

Rater'Reliability . . . . . . . . . . . . . .
Imitation of the Model's Behavior . . . . .
Effects of Verbal Instructions. . . . . . .
Generalization of Imitative Aggression . .
Generalization of Prohibitive Instructions

DISCUSSION 0 O O O O O O O O O O O O O O O O O 0

SUMMARY 0 O O O O O O O O O O O O O O . O O O O O 0

REFERENCES 0 O O O O O O O O O O O O O O O O O I

APPENDIC ES 0 O O O O O O O O O O O O O O O O O 0

iv

Page

ONIU‘UJH

12

12
12
13
14
15
18
22

26
26
28
30
32
33
34
41
44

48

LIST OF TABLES

‘ Page

Allocation of SubjectstoExperimental Groups . . . . . .

A Rank CorrelationCoefficients for Raters A, B, and C

on the Three Response Categories . . . . . . . . . . . .

Mean Scores for the Experimental and Control Subjects
on the'Three Major Response Categories . . . . . . . .

Significance of Differences (H-Test) Between Experi-
mental Groups and Control Groups in the Expression of
Imitative and Non-Imitative Aggression . . . . . . . . .

Significance of Differences Between Groups in the
Expression of Aggression (Mann-Whitney Test) . . . . .

Significance of Differences in Expression of Aggression
and Experimenters in Doll Play. . . . . . . . . . . . . .

13

27

29

29

30

32

LIST OF FIGURES

FIGURE Page
1. Room used in frustration situation . . . . . . . . . . . 16
2. Room used in frustration situation . . . . . . . . . . . l6
3. Room used in frustration situation . . . . . . . . . . . l7
4. Room used in frustration situation . . . . . . . . . . . 17
5. Room used in final doll play session . . . . . . . . . . 20
6. Room used in final doll play session . . . . . . . . . . 20
7. Room used in final doll play session . . . . . . . . . . 211‘
8. Room used in final doll play session . . . . . . . . . . 21

vi

LIST OF APPENDICES

APPENDIX Page

A. Raw Scores of Subjects Seeing the Film Without
Prohibitive Instructions. . . . . . . . . . . . . . . 49

B. Raw Scores of Subjects Seeing the Film With
"Prohibitive Instructions. . . . . . . . . . . . . . . 50

C. Raw Scores of Control Group. . . . . . . . . . . . 51

D. Total Scores and Ranks on the Three‘Major Scor-
ing Categories Assigned by Rater A and his
co-Rateroooo'ooooooo...ooooooooo 52

E. Total Scores and Ranks on the Three Major Scor-
ing Categories Assigned by Rater B and his
co-Rateroeoooooocoococoon-coo. 53

F. Total Scores and Ranks on the Three Major Scor-
ing Categories Assigned by Rater C and his
CO'Rater00000000000000.0000... 54

G. Total Scores and Ranks on the Three Major Scor-
ing Categories Assigned by Combined Raters A,
B, and C and Their Co-Raters D and E . . . . . . 55

THE EFFECTS OF VERBAL INSTRUCTIONS
ON IM‘ITATIVE AGGRESSION

INTRODUCTION

This is a study of the effects of verbal instructions onimodeled
aggressive behavior in nursery school boys. Studies in the area of
aggression are important because of the critical role it plays in the
process of socialization. The handling of aggression in children is a
major problem area forvmany parents and is often the most frequent
reasonfor referral in agencies treating children with behavior problems.

Recently several studies have appeared which show that iInitation
and modeling play important roles in the acquisition and expression of
aggression. Other investigations suggest that the way to handle
aggression is to xmake it abundantly clear through verbal means that the
behavior in question is undesirable. . The experimental procedure from
the studies of modeling and aggression is combined with verbal instruc-

tions in the present investigation.

Theorie s of Aggres sion

 

Aggression has given rise to-a tremendous volume of theory
particularly since Freud, impressed by the sadistic and masochistic
elements he found in all neurosis and by the events of World War I,
viewed aggression as "an innate, independent instinctual disposition in
man" (1950) that he called the death wish. Its object is the death‘and
destruction of life. It forms the basis for masochism and, when turned
outward, sadism. Freud's "new" instinct theory has met with consider—
able criticism (see'Monroe, 1955). VBut many psychoanalysts have

1

followed Freud's theoretical scheme. ~Melanie Klein (1937) writes:

The idea of an infant from six to 12 months trying to destroy

its mother by every method at the disposal of its body, trans-

formed in imagination into all kinds of dangerous weapons-- 1

presents a horrifying, not to say unbelievable picture to'our

minds. . . . But the abundance, force and the multiplicity

of the-imaginary cruelties which accompany these cravings

are displayed before our eyes in early analysis so clearly and

forcibly that they leave no room for doubt (pp. 187-188).

Other analysts feel that emphasis on instinctive features of hostility
and aggression with their negative emotional experiences is one-sided.
Among those who criticize the idea of an aggressive instinct are Harry
Stack Sullivan, Sandor Ferenczi, Thomas M.. French,- Karen Horney,
..Clara Thompson, and Paul Schilder. In general, they take a more positive
approach and stress the adaptive aspects of aggression. Schilder (1949),

i for example, talks about an instinct to action and mastery and holds that
the tendency tocenstruct and reconstruct cannot be separated from
action and aggression. Schilder and some other theorists seem-to dis-
cuss aggression as embracing all positive outgoing actions. Another group
of theorists stress the reactive aspects of hostile aggression, i. e. ,
behavior whose-ultimate goal is injurious to another. Where Freud thought
that aggression was an instinctual drive, the ultimate goal of which was a
static state free from all tensions, i. e. , a return to the womb, these
theorists see the aim of aggression as a safeguard to the personality
against the outer world.

- Ackerman (1958) offers a convincing explanation for the striving
for static equilibrium that Freud called the death instinct, writing:

The striving for a static equilibrium can be understood as

secondary process, a defensive avoidance of shock and frus-

tration, or as an escape from psychic pain; that is, such

behavior can be looked upon as an aspect of failure, defeat, or

derailment of the life drive, rather than as a primary death
instinct. In any case, the motivational constellations that Freud

identified with his death instinct point dramatically to‘the con-
nection between a striving for static equilibrium and retreat
from life. The homeostatic process actuates and protects life.
It does not foster fear and withdrawal from life (p. 72).

In discussing the homeostatic process, Ackerman departs from the

usual concept of stability and states that it is:

. . . closer to the truth to conceive that the exact purpose of
homeostasisis to protect, not stability in any static sense, but
rather a creative but controlled "instability"* of the organism in
consonance with the necessary conditions for maturation and for
expanding relations with the outerworld. In other words, in the
interpretation of this concept I am shifting the emphasis to the
exact Opposite of the connotation of "staying the same. " I am sug-
gesting that "homeostasis, " or the principle of dynamic equi-
librium, signifies the capacity for creative, fluid adaptability to
change, which at the same time assures that measure of coordi-
nated control that prevents the organism from being overwhelmed
by a barrage of stimuli in excess of the organism's capacity to
accommodate. The principle thus reinterpreted is conceived as
creative, controlled "instability"; it regulates response to experi-
ence not in order tomaintain sameness, but rather to preserve a
resilient capacity for change while preventing change from becom-
ing too rapid--so rapid as to disintegrate resources for adaptation
and growth. . . . I do not conceive aggression and destructiveness
as the expression of a death instinct but rather as a derailment of
healthy adaptation, a defense, a means of controlling environment,
of counteracting frustration and anxiety, and of asserting the
identity of the self in interpersonal situations (p. 70, p. 50).

The Frustration-Aggression Hypothesis

 

That aggression is a counteraction of frustration and anxiety has
long been known. To name but a few, William James, McDougall and
Freud, in their earlier writings, all noted that aggression was aroused
by frustration. Before developing his theory of instincts, Freud regarded

the tendency to seek pleasure and avoid pain as the basis mechanism in

 

:9:
"Instability" here is used to emphasize readiness for change.

~ all mental functioning. .Whenever pleasure seeking or pain avoiding
behavior was blocked, aggression wasthe "primordial reaction" and
was directed toward persons or objects in the external world which
were perceived as causing the frustration.

In 1939, Dollard, Doob, Miller, Mowrer and Sears presented
their classical work on frustration and aggression as viewed in a learn-
ing theory framework. Their efforts as systematization and objectivity
havelmade the study of aggression more accessible to experimental
study. Since their early work, one writer (McNeil, 1959) has stated
that studies in this area number well over a thousand of which at least
400 are experimental.

The learning theory orientation such as that used by Dollard et al
gives greater attention to the origins of individual differences in the
amount and quality of aggressive behavior. -According to Sears (1961)
who points out that the learning model:

. . . makes no assumption about a quantum of internal insti-

gation that must be uninhibitedly tapped. On the contrary, it

presumes that whole classes of aggressive acts are learned as
responses tovmany different stimulus situations. By stimulus
generalization, these acts come to be evoked by great ranges of
stimuli, both internal and external, including ones that are
symbolic or have imaginal qualities only. By response generali-
zations, acts can shift from one action category to another

(e. g. , from physical to verbal) or from one psychological func-

tion to another (e. g. , from overt action to imagination or- fantasy).

’ Indeed, the lability of expression is so great that even entirely
novel stimulational settings and manipulanda seem able to elicit
and provide a means of expression of aggressive acts. Each such
setting has its own instigative value, deriving from direct rein-
forcements in the past, as well as from primary or secondary

stimulus generalizations. This effective instigative value is a

cumulation of both excitatory and inhibitory values (p. 469).

Dollard et al take as their point of departure the older notion that

aggression is always a consequence of frustration. From this‘idea they

develOpa series of psychological principles relating to the strength of

instigation, the degree of inhibition, and the direction of aggression.
Because the direct expression of aggression in our society isifrequently
prevented from occurring by a strong anticipation of punishment,
aggressive impulses must be compressed, delayed, disguised, dis-
placed, or otherwise deflected. The energies behind aggressive acts

are not lost by being deflected away from the frustrating object.

Displac ement of Aggro s sion

 

Miller has developed a model for predicting behavior in approach-
avoidance conflict situations, including those where display and/or
control of aggression are involved. ' Miller's (1948) assumptions; {which
have been borne out in research, are as follows:

(1) that the direct response to the original stimulus generalizes

to other similar stimuli, with the amount of generalization becom-

ing smaller the less similar the stimuli; (2) that the response

which conflicts with the occurrence of the direct response to the
original stimulus also generalizes to other similar stimuli, be-
coming weaker the less similar the stimuli; (3) that the gradient
of generalization of the conflicting response falls off more steeply
with dissimilarity than does that of the original response which

it inhibits; (4) that when two or more incompatible responses

are simultaneously excited, the one with the greatest response

will be its strength minus that of any response incompatible with

it which is excited at the same time (pp- 167-168).

In situations that involve aggression, the conflicting response
tendencies are aggression and its inhibition fortfear of punishment:

It has been noted that observations and interview estimates of aggression
do-not correlate highly with fantasy aggression expressed in play (Kroner,
1951; Sears, 1950), correlations between these variables ranging from
about .13 to . 21. Returning toMiller's assumptions, one would expect
aggressive responses and fear of punishment in the home to generalize

to school and play situations. - Situations will differ in their similarity to

home conditions where the child is punished most directly for aggression.

A school situation, because-teachers act as parental surrogates, will
be closer to home on a stimulus generalization gradient than will a
doll play situation. If the gradient of generalization for inhibiting
aggressionfalls off more steeply, then the inhibition brought about
by severe punishment would affect behavior toward the teacher; but
not the doll play where it would increase aggression because of in-
creased frustration.

Sears (1951) divided preschool boys into three groups based on
severity of punishment in the home. The group receiving moderate
punishment in the home showed the most aggression in school. The
High and Low Punishment groups both showed about the same frequency
of aggression in school. In a doll play situation, however, the most
severely punished group showed more aggression than either of the
others. The fear of punishment did not generalize to the doll play
situation, so displaced aggression was permitted to appear.

A study by Hollenberg and Sperry (1951) lends support tothe
frustration- aggression hypothesis and Miller's assumptions about the
displacement of aggression. From interviews Hollenberg and Sperry
derived a measure of home frustration based on number and kinds of
restrictive rules, such as forcing of the child, and the like. ‘Mothers
were divided into high and low frustration categories. Thirty children
from these homes were observed in freedoll play and their aggressive
responses measured. Children who were highly frustrated in the home
showed more aggressive behavior in doll play. The differences between
the high and low frustration groups, however, were not statistically
significant. In general, while frustrationmay lead tofantasy aggression,
these writers felt that other factors must alsobe considered. More
conclusive results were obtained when another factor, the. severity of
punishment for aggression the mothers inflicted on their children, was

taken intoaconsideration. A measure of punishment for aggression based

on an'interview-rating of frequency, intensity, and duration of spanking,
threatening, isolation, denying privileges, and derogating the child

was also available. Hollenberg and Sperry then subdivided each of the
two original groups into a low and a high punishment group. Thus a
given child might be, to use initials, Low P-Low F, Low P-High‘F,
High‘P-Low F, or High P-High F. When this distinction is made, the
difference among the groups, not significant when frustration alone was
considered, became more clear-cut. High frustration and high punish-
ment led to the greatest doll play aggression, low frustration and low

punishment to the least.

Imitative Aggre s sion

 

In their book on social learning and imitation, Miller and Dollard
(1941) developed the notion that changes in the child's social behavior
can be explained on the basis of social imitation properly related to
general principles of learning. Their reinforcement hypothesis is as
follows: (1) the child is motivated to action by an internal drive; (2) he
sees someone else satisfying a comparable drive in certain ways-~this
pattern of behavior is a cue for an imitation response; (3) the child
attempts to c0py this behavior and the response may be rewarded and
thus reinforced. On subsequent occasions when motivated by this drive,
he has a tendency to repeat the response which he previously imitated
from another person's behavior. Miller and Dollard have presented a
number of controlled observations and experiments which appear to
support this hypothetical paradigm of the socialization process.

While imitation or modeling can play a positive role in social
development, it can also have deleterious effects depending on the type
of behavior that is modeled. Newspaper accounts, for example, often
mention children or teenagers learning deliquent behavior from tele-

vision, comic books, or movies. In the home, parents serveas models

of undesirable as well as desirable behavior. -A part of the reason for
this‘would be incidental learning, i. e. , learning that takes place with-
out intent or set to learn. Sears, Maccoby and Levin (1957) make the
point that parents often automatically react aggressively to aggression

in their children, who are more likely to be punished for aggression than
for any other undesirable behavior. The children's aggressive behaviors
can act as a frustrating stimulus for parents who respond with counter-
aggression.

Punishment, which serves to connect anxiety and aggression and
eventually have a localized inhibiting effect, also provide the child with
an aggressive model after which he can pattern himself. Parents can
act as aggressive models just as can aggressive characters in films,
comic books, and television. In this way punishment both inhibits and
instigates aggressive behaviors. Bandura and Huston (1961) summarize
this relationship as follows:

. . . when a parent punishes a child physically for having

aggressed toward peers, the intended outcome of the training is

that the child should refrain from hitting others. Concurrent

with the intentional learning, however, a certain amount of

incidental learning may be expected to occur through imitation,

since the child is provided, in the form of the parent's behavior,
with an example of how to aggress toward others and this
incidental learning may guide the child's behavior in later social

interactions (p. 311).

Bandura has investigated imitative learning from a variety of
angles, including aggression. One experiment (1961) involved preschool
children who performed a two-choice discrimination problem and a model
who exhibited unusual, irrelevant, non-functional divertive behaviors
in the solution process in addition to the orienting tasks typical of
incidental learning studies. Half of the children observed a model who

aggressed against dolls as a part of the diverting behavior while the other

half observed a non-aggressive model. Both groups were further divided

so that half of each group had a warm rewarding interaction with the
model while the other half experienced a cold and non-nurturant relation-
ship. (Previous experiments have shown that nurturance promotes
identification. Mowrer (1950) suggests that affectional rewards increase
secondary reinforcing properties of the model for the satisfaction these
cues provide.) It was predicted that children who experience a warm,
rewarding interaction with the experimenter would imitate more of the
behaviors performed by the model than those children who experienced a
colder, more distant relationship. The findings indicated that, in the
presence of the model, the children imitated the behaviors of the model.
Warmth in the relationship facilitated modeling except for aggressive
behavior, which was readily imitated by all subjects regardless of the
warmth in the relationship. Children who experienced a warm nurturant
relationship also exhibited more pre-solution conflict behavior.

Bandura, Ross, and Ross (1961), observing the facility with which
aggressive behaviors were imitated, investigated the generalization of
imitative response patterns to new settings in which the model was not
present. Pre-schoolers were divided into three groups. One group ob-
served an aggressive model, another observed inhibited non-aggressive
models and the third, a control group, was not exposed to models.
Children exposed to aggressive models exhibited more imitative and
non-imitative aggression in a different environmental situation. Another
study by these investigators (1963) investigated the extent to which film-
mediated aggressive models serve as sources of imitative behavior.
Aggressive models can be ordered on a reality-fictional stimulus dimen-
sion, thus pro-school children were divided into three experimental
groups and exposed to three models, each occupying a different place on
the continuum. One group observed a live aggressive model, a second
group observed a film of the same model, while a third group viewed a

film depicting an aggressive cartoon character. Experimental subjects

10

exhibited nearly twice as much aggression as control subjects who were

.not exposed to models. Exposure to aggressive humans on film was

the most influential in terms of eliciting imitative aggressive behavior.
Mussen and Rutherford (1961) also investigated the effects of

aggressive cartoons on childrens' aggressive play, finding that exposure

to an aggressive animated cartoon served to stimulate the intensity of

first graders' impulses to aggression.

Hypotheses

 

The present study concentrates upon the effectiveness of verbal
controls on aggression shaped by exposure to an aggressive model.

- Since parents and others responsible for learning experiences in this
area expect their efforts to generalize to other situations when they are
not present, a situation is provided so that generalization of aggression
can be studied. First, it is predicted that boys exposed to an aggressive
model will, following a frustrating experience, reproduce the aggressive
behaviors of the model and will differ in this respect from boys not
exposed to the model. This hypothesis has been successfully tested by
Bandura, Ross, and Ross (1961) but replication of their findings is
necessary to check on the present experimental procedures.

The second hypothesis is that boys who are verbally prohibited
from performing the aggressive actions of the model will inhibit these
acts inra subsequent situation following frustration, and that they will
differ in this respect from subjects exposed to the model without receiv-
ing instructions and controls not exposed to the aggressive model.

The third hypothesis concerns the generalization of effects of
exposure to the aggressive model and the generalization of prohibiting
instructions. Following frustration, subjects not receiving inhibiting

instructions should display more imitative and non-imitative aggression

11

in the presence of the frustrater than in the presence of an experimenter
of the opposite sex.

Finally, it is predicted that the effects of verbal prohibition on
imitation of the model's aggressive behavior will be greater inthe
presence of the person invoking the prohibition than in the presence-of a

neutral experimenter.

METHOD

Subjects

The experimental design replicates many of the features of
Bandura's studies. Bandura's findings indicate that aggression is
differentially influenced by the sex of the model. Boys exhibited more
aggressive behavior than girls when exposed to a male model. To maxi-
.mize imitative aggression, only boys were used in this study. The
subjects were 60 boys attending the Spartan Cooperative Nursery School
at Michigan State University. They ranged in age from ‘36 to 66 months
with a mean age of 51. 6 months.

The author and four female assistants conducted the study.

Experimental De Sign

 

All of the boys in the experimental groups were exposed to a
5-minute film of an aggressive male model. One experimental group
was specifically prohibited from behaving like the model. As a test of
generalization, half of the boys receiving such instructions and half of
those not receiving instructions were exposed to'a new experimenter
in the test situation. The subjects were assigned randomly tothe experi-
mental groups as shown in Table 1.

The-number within each cell refers to the number of subjects
exposed to the experimental variables listed to the'left and above it.
The control group was not exposed to the film and received no instructions.
Half of the group was exposed to a male tester and half to-a female tester.
All subjects in this group will be used as a base line for examining the

effects of exposure to an aggressive model on subsequent aggression.

12

13

Table 1. Allocation of Subjectsto Experimental Groups

   

Pretest Treatment Nmber..'of..Su1jects‘in Doll Play

 

 

 

Groups Model Instructions ' E Same as ‘E Different from
Prete st ' Pr ete st
Expe rimental Y e s Pr ohibitive 10 10
Expe rimental Y e 5 None 10 10
Control No None 10 10

 

Experimental Conditions

 

In the first step of the procedure, the author brought individual
subjects to the experimental building located across the street from the
nursery school. The author had spent at least two nursery school
sessionswith each class group'in order to increase rapport. When they
were later asked to come to the experimental building, subjects under-
stood that there were toys there that they were going over to play with.
Experimental subjects were told on the way over that they would see a
movie and attempts were made to secure their interest.

The experimental subject entered the building from a door that put
him immediately into the room where he would see the film. He could
not see thevother rooms used later in the experiment. The child was
seated at a small table where there were a number of Colorforms Toys,
shown how these toys worked, and told a story connected with‘the
Colorforms. When adequate rapport was established, thesubject was
told that he was now going to see the movie. The: movie and sound
track was started by the author who acted as experimenter for all sub-
jects in the pretest situation. The film with sound track used in the
present study was obtained from Bandura and is representative of films

used in his studies on imitative aggression. The film is in color and lasts

14

a little more than five minutes. There are four specific behaviors
which-the model performs in the film and these four behaviors‘are
repeated twice in the five-minute period. In the first sequence, the
adult male model says, "Out of my way, you funny clown. I said out
of my way. 'I'll fix you. I'll sit on you and punch you inthe nose. "

He proceeds to sit on an inflated plastic clown (Bobodoll) which is
about four feet high and punches him in the nose. This sequence lasts
about 15 seconds. - The model then says, "Still here for more. I'll
pick up that mallet and hit you in the nose. " While saying, "Bang,
bang . . . Sock you down" etc. , the model strikes the Bobo doll with
the mallet for 35 seconds. The model then states, "I don't know what
to dowith you. I know what I'll do. Kick you! Right across the room.
Fly away . . . fly away‘you funny clown. " etc. After 15 seconds of
kicking the doll, he says, "I'll pick up those balls and hit you in the
head . . . throw them at you . . . Bang. . . Boom!" He throws the
balls and verbalizes for 45 seconds. The model then duplicates these
four aggressive acts again in the same sequence. He sits and punches
the Bobo for 30 seconds; then strikes him in the nose with the mallet for
30 seconds; kicks him for another‘40 seconds; and finally spends a full

minute throwing the balls at him.

' . Prohibitive Instructions: During and following the movie, the

 

prohibitive instructions were given‘to the‘Prohibitive Group. These
instructions, which were essentially, the same for each subject, were
given in aconversational way to the child as follows: "Look what he's
doing! You wouldn't act like that, would you? . . . He's hitting him
with a mallet and he (Bobo) didn't do anything. I couldn't allow you to
act that way . . . if you did, I'd have to take you back and you couldn't
come again. He shouldn't do that. . . . Spartan Nursery School
children wouldn't act that way. Mrs. (S‘s teacher) wouldn't like

it if you acted like that. He shouldn't act that way--hitting toys and

15

saying things like that. You wouldn't do that, would you? I would have
to send you back tonursery school if you did. "

' The instructions were as uniform. as possible from subject to sub-
ject. The verbalization followed the action of the film and were given in
a way similar to that used by parents using verbal prohibitions. Care
was taken so that the instructions did not overlap with the model's
verbalizations. Most of the subjects were quite involved with the film
and when instructions were given, they would respond by agreeing with
the prohibitions outlined by the experimenter. Following the film, the
experimenter briefly repeated that he knew that the subject would not
act like the man in the film and that that was a very bad way to act. No
further instructions or comments were made to the subject regarding the
film or their behavior for the remainder of the experiment.

Subjects not receiving prohibitive instructions were treated the
same as prohibitive subjects up until the time the film began. The author
maintained rapport with the subject by looking at the film and describing
the action briefly at intervals in as neutral a way as possible. Carefully
avoided were any indications of approval, such as signs of enjoyment
of the model's actions in the film, giving permission to act like the
model, or other expressions that might affect subsequent behavior. If
the subject's attention wandered from the film, he was encouraged to

watch what was happening on the screen.

~ Aggression Arousal: Following exposure to the film, the subject

 

was told that the author had other toys and asked if he wanted to. see
them. All of the children readily agreed and were taken to a separate
room for aggression arousal. This room, . decorated with balloons and
cut-outs on thewall, contained a variety of toys that would be especially
appealing to boys in this age group (see Figures 1,. 2, 3, and 4). There

was an electric train that the boys were shown and encouraged to Operate;

 

 

 

FIGURE 2

l7

 

 

FIGURE 3

 

 

FIGURE 4

18

"a moving ferriswheel; an electric submarine; a battery operated
rocket launcher; a miniature castle withsoldiers and horses; various
large constructiontoys including a motor grader, road scraper,
cement mixer, and trucks. There was also twolarge fire engines, a
large rocket launcher, and a variety of stuffed toys.

Most of the subjects became involved in operating the train,
the‘rocket launchers, or the large trucks. The author left the room
after introducing the toys and showing how they operated. The subjects
usually became quite engrossed in play activities and after about five
minutes, the author returned and told the subject that, "These are -my
very best toys and I cannot let just anyone play with them. I'm saving
these for someone else who is coming later. I have to keep them nice
for 'Mike' and I just can't let you play with them. " Most of the subjects
were very reluctant to leave the room and many asked if they could
complete play activities that they had begun. However, the author
made it a rule that the subjects’left the room immediately. The sub-
jects were told that there were other ‘toys in another part of the building
and that they could play with these toys. The subjects were thentaken

to the other end of the building and introduced to the test situation.

Doll Play

Half of the subjects in each of the two experimental groups were
introduced to a new experimenter for the doll play test situation. These
subjects were chosen at random. In order to maximize differences in
experimenters, the new experimenter was of the opposite sex. When
the subject was introduced to the new experimenter, he was told that

"This is Miss . She will stay withyou while I go and fix-the pro-

 

jector. There are a lot of toys here and you can play with any of them

you like. " The author then left the room.

19

The room used for doll play contained a variety of toys and ° -
included among them were those seen in the film (see Figures 5,. 6,1 7,
and 8.)". These included a large Bobo doll, balls, a tether ball suspended
from the ceiling, and mallets. Other toys available for the subject in-
cluded a small house with a family of rubber dolls; an Etch-A-Sketch toy;
a mechanical tank; a dart gun; a mechanical ray gun; Colorforms; a
small wooden train and track; miniature cars and trucks with a garage;
airplanes and hangar; a take-apart car; and stuffed animals. The toys
were arranged in a standard fashion for all subjects.

The play room was equipped with a transparent mirror and micro-
phones. The behavior of the subjects was rated during the 20-minute
test period in terms of predetermined response categories. The test
session was divided into five—second intervals by an electronic interval
timer which emitted a beep sound heard only by the raters, giving 240
response units for each subject. The raters remained in the observation
room during the experiment and had no knowledge of pretest experimental
conditions. The author acted as a second rater for some of the subjects
who were exposed to the new experimenter in doll play. In no instance
did he make ratings alone. In those cases where a new experimenter
was introduced for doll play, the author would reappear after the 20-
minute rating period to return the subject to the nursery school.

During doll play, the experimenters avoided initiating any inter-
action between the subject and themselves and minimized their presence
by busying themselves at a table on one side of the room. Efforts were
made to standardize the behavior of the experimenter in this test situ-
ation as much as possible. Most of the subjects busied themselves with
toys immediately upon entering the room and many of them conversed
freely withthe experimentar who was present. One subject refused to
stay for the whole twenty minute test period and another subject in the
control group would not leave his mother who helped bring him. Both

subjects were eliminated from the study.

 

 

20

 

FIGURE 5

 

FIGURE 6

 

 

 

FIGURE 7

 

FIGURE 8

22

Behavior Measures

 

The response measures used in the present study fall into four
categories. These cover imitative aggression, partially imitative
aggression, non-imitative aggression, and non-aggression. Attempts
were made to secure measures of verbal aggression, both imitative
and non-imitative, but a number of factors led to the discarding of these
variables. Often raters were not able to hear clearly because'of defects
in the sound system, extraneous noises, and/or the quality of the subject's
speech. The behavior categories which were checked when present are
as follows (adapted from Bandura):

No Play

Child does not handle any of the play material, e. g. ,
stands or sits quietly, looks about the room, etc.

Ball

 

Rolls or plays catch with the ball, or uses it in some
other non-aggressive fashion.

Bobo

 

The following responses directed toward the Bobo doll.
constitute imitative aggressive responses:
Mallet--child strikes, taps, or shoves Bobo
with the mallet.
Balk-child rolls a ball toward, or tosses it
at Bobo.
~_lii£_l_<_-—child kicks or shoves Bobo with his foot.

Sit and Punch--child punches, taps, or slaps Bobo

 

while sitting or lying on it.
Sometimes children combine two imitative responses in a
single act, e. g. , child sits on Bobo and pounds it with
the mallet. . . , child strikes Bobo with the mallet and

kicks it simultaneously. These acts are double» scored--

23

in the first example, the'Mallet and the Sit and Punch

 

categories are checked; in the second example, the

‘Mallet and the Kick categories are checked.

 

The following are non-imitative classes of aggressive
responses toward Bobo:
M--child punches, taps, slaps, shoves or
wrestles Bobo.
Strikes -- child strikes» tape, or shoves Bobo with
any object other than the mallet, e. g. , jabs Bobo
with darts, pokes it or strikes it with a gun, doll,
etc.
Non-aggressive responses toward Bobo doll:
Sit--child sits on Bobo, bounces up and down with

it, lies or rolls on it.

Non-aggressive--child embraces Bobo carries it

 

around, dances with it, stands alongside Bobo

with his arm around it, etc.

Tether Ball

 

Guns

 

Mallet--child strikes the tether ball with the mallet.
Bobo-~child strikes Bobo with the tether ball.

 

911331--chi1d punches or slaps the tether ball, slams
it against the wall; strikes it with objects other than
the mallet.

Non-aggressive--child swings on the ball, spins it,

moves in an arc, examines it, etc.

Explores or loads--child examines the gun, loads it,

 

carries it in his hand.

24

Shoots--child fires darts, or aims the gun and shoots

imaginary bullets. Check the Egbgcategory when the

doll is the target and the 9.9.13.5 category when any

other object is selected as the target (e. g. , shoots at

the wall, the animals, the cars, the tether ball, etc.).
Non-Ajgrressive 'Play

This category includes all non-aggressive play with the

doll house, the dolls, the bears, the cars, and the farm

animals.
Peg Board

Aggressive--child pounds the pegs with the mallet.
Non-wssive—examines the peg board, attempts to
pull out or punch in the peg with his hand.

Other Non-Imitative‘Aggressive Responses
This category included physically aggressive acts directed
toward the farm animals, the dolls, or the cars, e. g. ,
reenacts fights between the animals or the dolls, spanks
or aggresses in other ways toward the dolls, crashes the
cars, or runs them into the animals, etc.

Other Imitative Resgonses
Child places his hands on his hips; shakes finger at
the Bobo doll.

The time units were very brief so that during any interval the child

usually exhibited behavior that fell within a single response category.

If a child displayed behavior that fell within two or more categories, the

raters checked only the response that consumed most of the time during

the scoring interval. There were several exceptions to this scoring rule:

1. When a child combined two imitative responses in a
single act, raters checked both of the imitative response

categories.

25

When a child performed an imitative aggressive response
and a non-imitative aggressive response, they checked

both categories.

7 When a child exhibited non-aggressive behavior during

most of the time interval but performed some imitative
or non-imitative aggressive act, they scored only the

appropriate aggressive response category.

RESULTS
Before giving the results as they relate to the hypotheses dis-
cussed earlier, a section will be devoted to reliability of observed

ratings. Following this, results will be discussed.

Rater Reliabilitj

 

Bandura (1961) reports interscorer reliabilities in the .908 for
the highly specific concrete classes of behavior used as response
measures in his experimental test situation. These same response
measures were used in the present study. High interscorer reliabilities
similar to those obtained by Bandura were found between raters in the
current study. In order to provide an estimate of rater reliability, over
half of the subjects were observed and rated by two trained raters.

No one rater rated one experimental condition exclusively. The raters
alternated with one another in a random way and all of the raters served
as experimenters at one time or another in the experimental test situ-
ation. The author never rated subjects alone.

The analysis of the data reported in the next section is concerned
with three scores for each subject. These scores represent total
imitative aggression, total non-imitative aggression, and total non-
aggression. The imitative aggression scores combines both'aggressive
behavior exhibited by the subjects which exactly imitates the aggressive
behavior of the model and partially imitative aggression, such as,
striking the tether ball with a mallet rather than Bobo. Non-imitative
aggression includes aggressive behaviors such as shooting or punching
the Bobo doll. The total imitative aggression score is the sum of the

subject's scores on the four response categories of imitative aggression

26

2.7

and the fourcategories of partially imitative aggression. The scores

on the response categories represent the number'of times the rater
checked the behavior in question as being present during the doll play.
Ratings were made of the child's behavior every five seconds and there
was a total of 240 five-second periods in the doll play session. Scores
on any response category could vary from 0,. if the child never per-
formed the act in question, to 240, t which indicates that the-child per-
formed the behavior in question during the entire doll play situation.

The total non-imitative aggression score is the sum of the six response
categories involving aggression which does not resemble that of the model.
The total non-aggression score is the total number of five-second periods
checked by the rater in which the child played in a non-aggressive way.
There are six categories of non-aggressive behavior.

Three of the raters, A, B, and C, accounted for nearly all of the
doll play sessions where a single rater was involved. Because the
accuracy of outcomes depends greatly upon their ability as raters, rank
correlation coefficients were computed for each rater versus their
co-raters D and E. (Appendices D, E, and F. show total scoresand
ranks for Raters A, B, and C and their co-raters on the three scoring
categories.) In Table 2 below, rank correlation coefficients- are reported
for these three raters and their co-raters on the three major response

categories.

Table 2. Rank Correlation Coefficients for Raters A, B, and C on-the
Three Response Categories

M

Imitative Aggres sion ‘Non- Imitative Non-Agg res sion

 

Rater A: 1.00 1.00 .99
’Rater B: .99 .99 1.00
Rater C: .99 1.00 1.00

 

28

‘Another rank correlation coefficient was computed between Raters
A, B, and C combined and Raters D and E (see Appendix G). There
were 24 subjects who were rated by Raters A or B or C and co-rated by
D or E. The rank correlation coefficients are again quite high. - For
imitative aggression, the rank correlation, coefficient is . 91; for non-
imitative aggression, it is . 99; and for non-aggression, it is also . 99.

While there'was some variation between raters within the finer
response categories, i. e. , within the‘20 classes of behavior that were
scored, even here the inter-rater agreement is remarkably. high.
< Ranking the scores tends to eliminate minor variations and indicates that
the raters are in high agreement as to the level and amount of the be-

havior the subjects exhibited in the three‘major scoring categories.

Imitation of the Model's Behavior

 

Subjects who were exposed tothe film of the aggressive -model
and received no instructions displayed a great deal more imitative
aggression than either the prohibitive groupwor the controls. Nineteen
of the 20 boys seeing the film without instructions displayed some imi-
tative aggressionwhile only 12 of the controls and six of theprohibitive
group did so.

Using a similarfilm, Bandura (1961) found that very few children
who had not been exposed to the film exhibited behavior similar to
that of the model. Approximately 70 per cent .of his subjectsenot seeing
the film had zero imitative aggression scores. In the present study,
mean scores for boys seeing the film without prohibitive instructions
aremore than three times as'large as those of boys in the control groups
' (Table 3). Differences between, the film group, prohibitive group and
controls on total imitative aggression is highly significant using a two-

tailed test (Table 4). The differences are alsoapparent when film groups

29

Table 3. - Mean‘Scores for the Experimental and Control Subjects on the
Three Major Response Categories

 

 

Experimental Group

 

 

No ‘ Pr ohibitive Control
Response Categories Instructions Instructions Group
Imitative Aggression
(Same) Experimenter A 15 . 2 . 6
(Different) Experimenter 9. 4 . 2 . 8
Combined 12. 2 . 2 . 7
Partially Imitative Aggression
(Same) Experimenter 24. 2 4. 6 6. 3
(Different) Experimenter 8. 2 2. 0 10. 7
Combined 16.2 3.3 8.5
Total Imitative Aggression
‘(Sarne) Experimenter 39. 2 4. 8 6. 9
(Different) Experimenter 17. 6 2. 2 ll . 5
Combined 28.4 3.5 9.2
Non-Imitative Aggression
(Same) Experimenter 155. 8 98. 1 154. 1
(Different) Experimenter 62.1 123.6 118.8
Combined 108.9 110.8 131.4
Non-Aggression
(Same) Experimenter 67. 8 147. 3 94. 3
(Different) Experimenter 168. 1 119. 9 108. 5
Combined 117.9 133.6 101.4

 

Table 4. Significance of Differences (H-Test) Between Experimental
Groups and Control Groups in the Expression of knitative and
Non-Imitative Aggression. .

o

 

 

 

a I 1
Test Situation
Same Different
Response Category Experimenter Experimenter Combined
‘ ‘P ' P ‘P

IInitative Aggression . 001* . 01* . 01*
Non-Imitative Aggression . 05* n. s. . 0.5
Non-Aggression n. s. . 02 . 05*

 

at:
Corrected for tied ranks.

30

-who had a new experimenter in doll play and those having the same
experimenter throughout are compared with subjects receiving prohibi-

tive instructions and controls (Table 5).

Table 5. Significance of Differences Between Groups in the Expres sion. of
Aggression (Mann-Whitney Test)

 

Response Category

 

Experimental Groups Imitative Aggression Non-Imitative Aggression

 

U p 'U P
Film vs. Prohibitive
Same EXperimenter ‘11 . 02 30. 5 n. 3.
Different Experimenter 8 . 002 23 . 05
Combined 42 . 002 199 n. 5.
Film vs. Control
Same Experimenter l3 . 02 48. 5 n. s.
Different Experimenter 30. 5 n. s. 19. 5 . 02
Combined 92.5 .02 141.5 n. s.
Prohibitive vs. Control
Same Experimenter 38. 5 n. s. 31 n. 5.
Different Experimenter 31. 5 n. s. 49. 5 n. 8.
Combined 137.5 .10 159.5 n. s.

 

Effects of Verbal Instructions

 

The most significant differences on imitative aggression occurred
between the groups seeing the film without prohibitive instructions and
the groups who received such instructions (Table 5). - These groups differ
significantly both when the same experimenter remained in the test situ-
ation and when a different non-frustrating experimenter is substituted.

When film groups are compared with controls, similar significant
differences in favor of the film groups again appear except for groups

having a different experimenter in the test situation. The value of U in

31

this instance is 30. 5 and it would need to be 23 to reach the . 05 level of
significance.

Examination of individual scores shows that all 10 subjects in the
film group whohad a different experimenter in doll play showed some
imitative aggression. Their mean score was 17. 6. 1 Only six of the
controls displayed aggression similar to that displayed by the model.
Their mean score was 11. 5. - One of the subjects in the control group
had a score of 53 in the partially imitative aggression category for sitting
on Bobo. This was the only imitative aggression he displayed and he was
the second lowest on total aggression; yet his score on total imitative
aggression was the highest in the group. His sitting on Bobo artificially
inflated the mean of his group and though he exhibited partially imitative
aggression, he apparently was not as aggressive as most of those seeing
the film.

It seems clear that while exposure to an aggressive. model has a
definite effect on the shaping of aggression following frustration and this
effect generalizes to new situations, it can be. controlled by prohibitive
instructions directed at specific behaviors of the model.

Subjects receiving prohibitive instructions during exposure to the
film consistently showed less total imitative aggression than those
seeing the film without instructions and those in the control group (Table 3).
The controls showed nearly three times as much total imitative aggression
as the prohibitive groups and the film groups without instructions showed
more than three times as much as the control groups. Again the most
outstanding differences occur between groups seeing the film without
instructions and theprohibitive groups. The prohibitive and control
groups did not differ significantly when a U Test was applied to ranks.
The effects of instructions on the prohibitive group were quite dramatic.
While only one subject of the twenty not receiving instructions failed to

imitate the model, thirteen of the prohibitive group were inhibited from

32

doing so. .Specific verbal instructions have a very significant inhibitory
effect on subsequent imitative aggressive behavior following frustration.
Though there were differences in non-imitative aggression related to
new experimenters in the test situation, both the film groups and the
prohibitive groups showed a very similar amount of non-imitative
aggression. The prohibitive instructions had effects which were quite

specific to aggression against the Bobo doll.

Generalization of Imitative Aggression

 

Subjects who viewed the film without prohibitive instructions dis-
played more aggression, both imitative and non-imitative, in the
presence of the experimenter who frustrated them (see Table 3). The
differences are highly significant when non-imitative aggression is con-
sidered, while differences on imitative aggression only approach signifi-

cance (see Table 6).

Table 6. Significance of Differences in Expression of Aggression and
Experimenters in Doll Play

Response Category

 

Experimental Groups Imitative Aggression ~ Non-Imitative Aggression
U p U P

 

Film - No Instruction
Same vs. Different
Experimenter 25 . 05-. 10 2 . 002
Film - Prohibitive Instruction

Same vs". Different
Experimenter 42. 5 n. s. 38. 5 n. 8.

Controls

Same vs. Different
Experimenter 50 n. s. 32 n. s.

 

33

Generalization of Prohibitive Instructions

V It was hypothesized that those subjects given prohibitive instruc-
tions would display greater inhibition of prohibitive acts in the presence
of the experimenter who gave the prohibition. While itis clearthat
these instructions had a pronounced inhibitive effect on both imitative
and non-imitative aggression (Table 3), there is only a slight trend,
contrary to hypothesis, for subjects to display more partially imitative
aggression in the presence of the experimenter who prohibited them
and later frustrated them. The differences between the groups having
the same experimenter in doll play and those having a new experimenter
is not significant on either imitative aggression or on non-imitative '
aggression. Both» of the groups who were given prohibitive instructions
displayed very little imitative aggression of any kind. <However, the
prohibitive group having a different experimenter showed significantly
- more non-imitative aggression than the experimental group not receiving
instructions who had a new experimenter in doll play. Comparison of
mean scores on non-imitative aggression between prohibitive and control
groups having a different experimenter suggest that the effect in question
is taking place in the film subjects not receiving instructions whohad a

new experimenter indoll play.

DISCUSSION

The major focus of this study has been on the effects of prohibitive
instructions on imitative aggression in nursery school boys. ‘ Bandura
(1961) has shown that preschool children exposed to an aggressive model
will, following instructions, reproduce a good deal of aggression re-
sembling that of the model. The present study incorporates many
features of Bandura's study and provides additional evidence regarding
the effectiveness and importance of imitation in social learning.

Subjects in his study who observed a model performing specific
aggressive acts later reproduced them when the need for aggression was
present. Bandura suggests that mere observation of aggression regard-
less of the quality of the model- subject relationship is a sufficient con?
dition for learning of imitative aggression in children. He has demon-
strated that children will imitate the aggressive behavior of a number of
types of models including nurturant, non-nurturant, and neutral. Other
studies have shown that cartoon figures may also act as models for A
subsequent aggression. With the wide variety of opportunities for ob-
serving aggressive behavior performed by potential models in the home,
neighborhood, and on television, in films, and in comic books, etc. ,
the control of imitative aggression is critical. ‘

Current thinking among those psychologists concerned 'with psycho-
logically healthy patterns of child rearing suggest that an effective way
of handling aggressive behavior in children is to make it clear that such
behavior is undesirable. ‘This approach was taken in the present study
where prohibitive verbal instructions were given to the subject relating.
to specific acts of aggression being performed by the model. These
instructions were contiguous with the acts being performed by the model.

In doll play following a frustrating experience, children who received

34

35

these instructions while seeing‘the film inhibited the aggressive behavior
specified-in the instructions. With-prohibitive instructions, imitative
aggression scores of subjects in, the experimental group dropped to a
level even lower than those of controls who occasionally by chance
imitated some of the behaviors of the model. Many of the prohibited
subjects carefully avoided the Bobo doll altogether, so that no imitative
aggressive responses occurred.

7 The following comments by the raters illustrate the behavior of
some of the children who had received prohibitive instructions. The
comments dramatically reflect the inhibiting effects of the instructions.

"Fascinated by Bobo. Would come over to the clown and,
particularly in the beginning, some imitative aggression would
leak out. Showed visible restraint however. Said, 'It's not nice
to. punch clown, ' and 'What's this?‘ giving the clown a half-
restrained punch. Showed lots of curiosity about Bobo. Asked if
mallets were to hit the clown. Would say 'You. . . . ' and not
complete the phrase to the clown. "

"When S came in, he said, 'You're not supposed to punch
him in the nose hard. ' Generally avoided the side of the room
with Bobo except for shooting at him. In the middle of the
session, 5 said, 'Is that the one the guy was hitting?”

"'I'm not going to do like that guy in the picture. I‘m
just going to shoot him. "'

"S said he wanted to kill the clown and then went over and
hugged it. He shot at the clown most of the time but when he
went near it, he would pat it. Kissed the clown. On leaving,
said, :1. love those toys, ' indicating those used during the frus-
tration sequence in the other room. "

"S said, 'Could I hit that awhile?’ indicating Bobo; but when
told that he could, he didn't touch it. S avoided the side of the

room where Bobo was standing. "

36

‘ "Child at very first half punched Bobo and then hugged the
clown. - Later he really let go by picking up and heavily dropping
the clown. Used Bobo like a bat to hit the tether ball. "

"S looked at the clown and then-looked quickly away. For

the first half of the test session, he didn't go near the half of the

room where the clown was. "

While there was wide variation in behavior among the subjects
receiving prohibitive instructions, it seems clear that an inhibitory
effect was operating. This effect seemed to operate in various ways,
including displays of affection for Bobo strongly reminiscent of reaction
formation, obvious introjection of the experimenter's prohibition which
some of the subjects repeated aloud to themselves, and channeling of
aggression along lines other than those specifically prohibited by the
experimenter. 1 The presence of the experimenter who gave the prohibitive
instructions tended to increase inhibition of all forms of aggression, but
effects of instructions were also highly significant in the absence of the
prohibitor.

Rosenblith (1959) in a study of imitative learning found that male
experimenters were more effective than females in influencing children's
behavior. - She suggests that this may be due to increased reward value
of male figures resulting from their relative absence in most of the child's
day to day experiences. In the present study, subjects tended to display
more aggression, both imitative and non-imitative, in the presence of
the male experimenter except for the prohibitive group, who showed
greater inhibition of non-imitative aggression in his presence. Only five
prohibitive subjects showed non-imitative aggression directed at the
Bobodoll, while nine of the controls did so. The trends toward inhibition
of aggression in the presence of a female experimenter in nearly all of
the groups may reflect generalization of controls from both the home and

the nursery school where female figures are generally responsible for

37

discipline and control of aggression. Male figures on the other hand
may serve to activate aggression in boys because of identification
with cultural stereotypes of masculinity.

The mere fact that the female experimenters were adults may
account for the generalization of inhibition. It seems clear that the pro-
hibitive instructions were very effective in suppressing imitative
aggression. Less firm instructions may show less of an effect. The
instructions in the present study were so strong that they seemed to
suppress all aggression and easing of the prohibitions may show a dif-
ferential effect and allow more non-imitative aggression to be shown.

It would be expected that this non-imitative aggression would show up
-more in the presence of a different experimenter. Perhaps if the test
for generalization were more removed both temporally and specially,

the prohibitions might not have had such a strong effect and differences
would appear in the expression of imitative and non-imitative aggression.

Most explanations of imitative aggression refer to Freud's (1946)
notion of "imitation with the aggressor" or Mowrer's (1950) "defensive
identification. " These theories assume that subjects when threatened
with aggression take on the qualities of the aggressor and become agents
of aggression rather than objects of aggression in order to allay their
anxiety. Bandura's studies, however, indicate that his subjects imitated
aggressive behaviors of models who were non-threatening and even
nurturant. He prefers a learning theory orientation as underlying
imitative aggression and postulates that object displacement accounts
for aggression found in children of punitive parents. He found, for
example, that parents of aggressively antisocial adolescents (Bandura
and Walters, 1959) and of hyper-aggressive boys (Bandura, 1960) were
punitive of aggression directed toward themselves while actively rein-
forcing aggression toward othersoutside the home. aThiss differential
reinforcement tends to inhibit aggression. towards the .origina1:i2n'sti-

gators while fostering displacement of aggression towards other outside

38

the home. Bandura (1961) claims that Miller and Dollard's (1941)
demonstrations of imitative learning in which observed models were
rewarded for certain acts and subjects rewarded for duplicating the
behaviors of the model are actually cases of discrimination learning
in which the behavior of others act as a discriminatory stimulus for
responses already part of the subjects' response repertoire. He
stressed that:

An adequate theory of the mechanisms underlying imitative learn-
ing is lacking. The explanations that have been offered (Logan,
Olmsted, Rosner, Schwartz, and Stevens, 1955; Maccoby, 1959)
assume that the imitator performs the model's responses covertly.
If it can be assumed additionally that rewards and punishments
are self-administered in conjunction with the covert responses, the
process of imitative learning could be accounted for in terms of
the same principles that govern instrumental trial-and-error
learning. In the early stages of the developmental process, how-
ever, the range of component responses in the organism's
repetoire is probably increased through a process of classical
conditioning (p. 580).

It is interesting to note that one child who was being told by the
experimenter in the prohibitive instructions, "You wouldn't hit him with
a mallet, would you?" became quite excited and exclaimed, "N02". . . .
I'd hit him with a hatchet!" It appears that more observation, without
apparent rewards, shaped the aggressive behaviors aroused by mild
frustration. There was no opportunity for performance of the model's
behavior except in the test situation and there were no external rewards.
The experimenter abstained from making any evaluative comments about
the model and his only verbalizations were intended to keep the subject's
attention on the film.

The frustration experience in the present study seemed to provide
the desired effect. Most of the children became deeply involved with
the various mechanical and electrical toys that were provided in a
pleasant party-like atmosphere. It was necessary to gently but firmly

take some of the subjects from this room to the doll play test situation.

39

One of the experimental subjects who had not received prohibitive
instructions told the teacher on his return to nursery school that,

. "There was a pounding thing in there and I punched and punched and
punched until he was dead: There was a real man in there but I

couldn't punch him. " When asked by the teacher if he wanted to, he
replied, "Yah . . . but he was too big. " His displacement of aggression
to the Bobo doll was obvious.

The most extreme aggressive responses occurred in the experi-
mental group who had not received prohibitive instructions and these
aggressive responses were nearly always directed at the Bobo doll.

7 Several of the children attacked the doll violently with a wooden mallet
in each hand. The following are comments made by the raters while
observing non-prohibitive experimental subjects:

"Before S picked up the mallet to hit Bobo, he looked at
E and then at Bobo several times. Waited until E looked out of
the window before really letting go. Child commented on his
disappointmentin having to leave 'the best toys' and said that he
liked the big train better than the small wooden one. "

"Child was very aggressive verbally and physically
toward Bobo. At one point he said he would 'knife‘ Bobo.

S struck the father family doll twice and said he would
beat up the E (who in this case was a female). S said the mother
and father family dolls got killed and had to go to the hospital. "

Occasionally subjects would shoot the dart gun at the experimenter but
generally their aggression was directed toward the Bobodoll, the
family dolls and one child even shot the teddy bear.

This study has several implications for further research. One
area which may yield interesting findings relates to the effects of
instructions when no adult figure is present. During one of the doll play

sessions, the experimenter was called into the observation room to

40

make a minor adjustment on the interval timer. No sooner had the
experimenter left the room when the subject, who had not received
prohibitive instructions, exploded in a violent bur st of aggression
directed against the Bobo doll. Several other subjects were observed
following the twenty minute test session and in some, the absence of
an adult figure seemed to lessen previous inhibitions.

Another area which might be investigated is the relative effective-
ness of different types of instructions on imitative aggression. For
example, certain subjects might be given channeling instructions which
would have as their purpose the channeling of aggression along more
socially acceptable lines such as verbal aggression at the experimenter.

It would be interesting to observe these same children after a
period of weeks or months to see how well prohibitive instructions hold
up with the passage of time, and to observe the duration of effects of
exposure to an aggressive model.

Bandura has suggested a comparative study of the imitation of
aggressive models who are feared, who are liked and esteemed, or who
are essentially neutral figures in order to throw light on whether or not
a more parsimonious explanation other than that involved inii'identifi-
cation with the aggressor" can be used to explain imitative aggression.
Currently, Bandura is trying to evaluate the effects when the model is
rewarded, is punished, or experiences no reinforcement after an
aggressive sequence. All groups of children in this research, as a final
step, will be given permission to reproduce the aggressive actions of
the model for an attractive reward. The expectation is that subjects
‘who may be prohibited or otherwise inhibited from performing the
behavior of the model would imitate this behavior for a reward; thus

clearly demonstrating the inhibitory effects of the experimental variables.

SUMMARY

Sixty nursery school boys were randomly assigned to three
conditions. Subjects in one experimental group individually observed
a five-minute film sequence of an aggressive model; a second experi-
mental group received, during the film presentation, prohibitive verbal
instructions regarding specific aggressive acts which the film model
directed at an inflated clown. A control group did not see the film and
received no instructions of any kind. Half of the subjects in each of
the groups were randomly assigned to a neutral female experimenter
in a doll play test situation which followed mild frustration designed
to elicit aggressive behavior. Subjects were observed in doll play
for a 20-minute period during which ratings were made of their behavior
at five-second intervals. Over half of the subjects were rated by two
independent raters in order to determine rater reliability. Ratings
were made on 27 different behaviors which were broadly classified as
Imitative Aggression, Partially Imitative Aggression, Non—Imitative
Aggression, Non-Aggression, and Imitative Verbal Aggression. The
latter category was eliminated because of poor sound transmission and
the lack of intelligibility in the speech of these subjects.

It was predicted that boys exposed to an aggressive model would,
following a frustrating experience, reproduce the aggressive behaviors
of the model and would differ in this respect both from boys prohibited
from performing such acts and boys not exposed to the model. A second
hypothesis was that boys who were verbally prohibited from performing
the aggressive actions of the model would inhibit these acts in subsequent

doll play. The third hypothesis concerned the generalization of effects

41

42

of exposure to an aggressive model and the generalization of prohibit-
ing instructions. It was expected that following frustration, subjects
not receiving inhibiting instructions would display more aggression in
the presence of the frustrator than in the presence of a more neutral
figure of a different sex. Finally, it was expected that the effects of
verbal prohibition on imitating of the model's aggressive behavior would
be greater in the presence of the person invoking the prohibition than
in the presence of a neutral experimenter.

The responses scored involved highly specific concrete classes
of behavior and yielded high inter-scorer reliabilities, the rank corre-
lation coefficients being in the . 908.

Subjects seeing the film without instructions clearly imitated the
aggressive behaviors of the model following the mild frustration
experience. These differences were generally highly significant.

Those subjects who were prohibited during the presentation of the film
were inhibited in imitating aggressive behaviors displayed earlier by
the model.

As a test of generalization of the effects of exposure to the film
and as a test of the generalization of prohibitive instruction, half of
the subjects in each group were exposed to a "neutral" experimenter
in doll play. The highly significant differences between the experimental
groups on imitative aggression remains, indicating that both effects
readily generalize. Subjects who viewed the film without prohibitive
instructions displayed more aggression, both'imitative and non-imitative,
in the presence of the experimenter who frustrated them. The difference
is highly significant for non-imitative aggression, but only approaches
significance for imitative aggression.

It was expected that the effects of verbal prohibitions would be

greater in the presence of the person invoking the prohibition.

43

However, the differences here were not significant, both groups receiv-
ingprohibitive instructions showing little imitative behavior.

The results were discussed in terms of a learning theory of dis-
placement rather than other theories utilizing such concepts as

"identification with the aggressor" or' "defensive identification. "

REFERENCES

44

REFER ENC ES

Ackerman, N. W. The psychodynamics of family life. New York:
Basic Books, Inc., 1958.

Bandura, A. , and Walters, R. H. Adolescent Eggggssion. New York:
Ronald Press, 1959.

 

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. , and Huston, Aletha C. Identification as a process of
incidental learning. J. abnorm. soc. 'Psychol., 1961, _6_3_ (2),
311-318.

 

Bandura, A., Ross, Dorothea, and Ross,-Sheila A. Transmission of
aggression through imitation of aggressive models. 7 J. abnorm.
soc. Psychol., 1961, 63(3), 575-582.

Bandura, A., Ross, Dorothea, and Ross, Sheila A. Imitation of

 

 

film-mediated aggressive models. J. abnorm. soc.lPSychol.,
1963, Q (1), 3-11.

Dollard, J., Doob, L. W., Miller, N. E., Mowrer, 0, 1:1,, and Sears,
R. R. ~ Frustration and agggession. New Haven: Yale University
Press, 1939.

Freud,*Anna. The ego and the mechanisms of defense. 4 New York:

 

International University Press, 1946.
Freud, S. Beyond the pleasure principle. - In Jones, E. , (ed.),
. Collected Papers, Vol. 2. London: Hogarth‘Press and Institute
of Psycho—Analysis, 1950.

 

Hollenberg, Eleanor, and: Sperry, Margaret. Some antecedents of
aggression and effects of frustration in doll play. - Personality,

1951, _1_, 32-43.

 

45

46

Klein, Melanie. The psycho-analysis of children. London: Hogarth

 

Press and the Institute of Psycho-Analysis, 1937.
- Korner, Anneliese F. Relationship between overt and covert hostility-

economy and dynamics. Personality, 1951, _1_, 20-31.

 

Logan, F., Olmsted, O. L.,*Rosner,oB. S., Schwartz,»R. S., and

Stevens, C. M. Behavior theory and social science. New Haven:

 

Yale University Press, 1955.
Maccoby, Eleanor E. Role-taking in childhood and its consequences

for social learning. Child Develpm., 1959, 3_0_, 239-252.

 

McNeil, E. B. Psychology and aggression. ‘ J. Conflict and Resolution.
1950, 3, 1-75.

 

Miller, N. E. Theory and experiment relating psycho-analytic dis-

placement to stimulus-response generalization. J. abnorm. soc.

 

Psychol., 1948, g, 155-178

Mowrer, O. H. Identification: a link between learning theory and
psychotherapy. In, Learninitheory and personality dynamics,
New York: 'Ronald Press, 1950, pp. 573-616.

Munroe, Ruth‘ L. Schools of psychoanalytic thougfl. New York:
The Dryden‘Press, 1955.

 

Mussen, P. , and Rutherford, E. Effects of aggressive cartoons on
children's aggressive play. J. abnorm. soc..Psychol., 1961,

95, 461-464.

 

Rosenblith, Judy F. Learning by imitation inkindergarten children.
Child Develpm. , 1959, 32, 69-80.

 

Schilder, P. Goals and desires of men. New York: Columbia University

Press, 1949.

 

Sears, R. R. Relation of fantasy aggression to interpersonal

aggression. Child Develpm., 1950, _2_1_, 5-6.

 

Sears, R. R. A theoretical framework for personality and social

behavior. -Amer. Psychologist, 1951, _6, 476-483.

 

47

Sears, R. R., Maccoby, Eleanor E., and Levin, H. Patterns of child

 

rearing. Evanston: Row, Peterson, 1957.

Sears, R. R. Relation of early socialization experiences in‘aggression
in middle childhood. J. abnorm. soc. Psychol., 1961, 6343),
466-492.

 

APPENDICES

48

 

m .Hm ow< smog

 

Hwoa Ame ova Nm #0

co OOH 0m 0 H¢ mm mm
00." Nm N. N O 0* WH
NON NM w w 0 mm 4N
Hwa hm: A .H 0 mm Hh
com 5N m n. N mm NH
NF.” om mN O 0H 00 OD
«VNH mo mm NN OH me 1n”
mHN mm N O N 5* mm
0mg Mn. mm m; a; me NH
med we m o M 0m mm

 

coflmsfim umoh GM Hoodocfiuomxm unouomflfl

m .wm om< smog.

 

49

whxw wmma mom New. om:

no omd m» we. mm mm mm
mm «LN on N; MN ow hU
uvm com «um mm o mm M4
Ho m: vm Na Na om m4
«.0 owa OH OH 0 om m4
mo ooa on NM was he M4
.mm 1: He or NH mm M4
co mod or am ma do 02
om mod 0 o o mm G¢4
mp A om.” om mg m cm 00

 

:oflmgfim umoh 5 soucogfiuoaxnm 05mm

 

onoom ouoom onoom ouoom ouoom 93:02, 330
cofimmosmwdw :ofimmosmmww cofimmoummd. . :ofimmosww.‘ gammouwmlw om<
#3 Z o>$mfic§ earn—mung o>flofiEH 7. o>flofiEH
$52 #308 333st

 

 

 

 

mcofiosfimfi eZfiQEOum 305E? 53h on» wfioom maoofindw mo mouoom 3mm :4. NHQanmnm<

50

 

G oQV 0W< Gmmz

 

 

 

 

 

moHH «mug . NM 0N N
Ma; Q: o o 0 me Ah
om: mm o o 0 mm Nd.
#mm b c o 0 Sq mm
:1 mo 0 o o a. 4m
NM NAN N o N Nm Hm
:. end 0 o 0 me E...
Nwa we 0 o o oo <0
w: NNH H H 0 mm MM
mm 00H m; «L 0 we 43
co om: o o o me 40
cofimsfim “ooh. E soucoafisomxm ucouofiwfl
o .8 he. emmE
m5; :3 we we N
omm w o o 0 mm 40
Nmm p o o o oo AHA
#0 3; o o o Nm UG
NNN mm o o 0 iv a: 7
mo N“; “v w. 0 mo m4
mm 0mg 0 o M 1v 2..
MAN mm o o 0 Ho EU
H: wma o o o Ev DU
«ma mm cm om o no. 04
Am mmm o o 0 do MO
cofimsfim pooh. cw soucogfisomxm oEmm
ouoom ouoom ouoom osoom onoom 93:02 @330
:ofimmopmwaw cofimmosmwaw cofimmosmwmw Goflmmonww/w :ofimmoswaN om<
unoZ o>3mfiEH o>flmfifld 038.939: ozumfig

uGoZ 130R. >Zmflsmm

 

 

mcoflosfimcH 9,355on £35 53...». 05 mcfloom whoeflhsm mo mosoom 3mm .m NHDZHnngw

51

 

m .3 ha. 522

 

 

 

 

 

mon mm: m: hog w
HwH wo h h o no NJ
no 3: m o m hm .mH
wm mom 0 o o wm no
em mmH o o o mw Mb
mama. on o o o on hU
on NMH o o 0 mm OH
.92 nod o o o ow >4
mo mod H a o ww 00
9: mo ow hm m ww 0m
omH mm mm mm o ww mm

coflmsfim pooh. 5 poucocfiuomxm usesofififl

c .mm om< Geog

mwm mea me mo 0
mm: pm 0 o o ww MA
MNH m: ma ma 0 mm MG
2: mmH om em 0 Ho OM
wh 9: o o 0 on Maw
ww ENN N N 0 pm and
nod Nma o o 0 mm NO
h mmm o o o we hN
co 9: ma w m ow >N
o: NNH OH o H wm >H
a: w]: m m 0 we 4.?

cofiodfim “ooh. a“ sousogwuomxm 05mm
ouoom osoom osoom ouoom onoom mficoz UHEO
cofimmosmwwt Cofimmosmmws coflmmouwqu sofimmoummw. :ofiwmosmwxw omr.
acoZ met/Sofia»: o>3mfiEH «53.32»: 0.2331»:
IcoZ H.308 >23»qu

ill) l
L

95ch 33:00 «.0 mouoom 3mm .0 UHQZMQQ<

 

32

 

 

 

 

 

 

 

m.HH mm oH mm m hmH m me m m w m NA

wH wm wH wm N NON N. wON o o B
seasoEHuomxm onth u mHoh—GOU
m S: m omH 3 cm S cm 0 o MA.
w mNH w- NNH w mm: m wHH m mH m mH mm
seasoEHnomxfl 3.32 n mHonucoO
w ooH w ooH NH Nm NH Hm o o N<
HounoEHsomme aconoHHHQ .. mcoHuosuumGH o>HanH~Honm .. EHHM
N OMN N OMN mH w mH w a. h n o 40
H NMN H NmN wH N. wH N. o 0 m4
me Q: o HHH w hmH w omH o o 00
o HmH o mmH HH mw HH om m on m mm 04
mH oN mH Hm H NNN H wwN o o VHO
noucoEHsomxH 08mm u mcoHuoshmsH, 8,3330%“ u EHHh
m nON m nON mH mN mH wN o w o w <N
w omH h owH o Nu m NH. w mm w mm NH
poucoEHuomxm Bohemia .. 9832535 02 u EHHh

m.HH no m.HH co m wNH a. omH H NH. H we mm

m.HH mo m.HH oo o me @ NmH N ow N ow 03H
.mH Hw mH mu m S: m moH ...o o O<

nousoaHuomxH 05mm .. mGoHuoduumcH oZ .. EHHh

vHsmm osoom . vHsmm osoom ”Mcmm ouoom . VHGom enoum anmm choom Hcmm onoom Hoofiasm

schemiou 4 sound nonMuoU .4 sound noummnou .4. soomm

Gonmonw<nGoZ

Goa. m m on mw< o>HumuH§H Inez

L!

Con mouwmws 039.32g:

sound-OD mHHH Hose .04 sou—mm

>3 poGmem< moHsowoqu mGHuoom .Hoflmw/H cough. o5 Go 318% paw mouoom Hmuoh .Q XHQZHnmnHHw

53

 

 

 

 

 

 

 

w HwH w HwH m.N. mo w wo m.m N. w N. NA
HoacogHsomxH onth .. Houuaou
N. me N. NwH o NoH. o moH o o AM
o owH o wwH m .N no N. mm o o 4m
mH wN mH 0N N mHN N oHN m6 N m.N. H Hm
H H HN. HH or w AwwH w wwH o o 2.
nousoEHuome «souomfifl u mcoHuoshmGH ownHanHnHonnH .. EHHM
H OMN H OMN mH w mH w m c m o 40
m mmH m HmH m6 om HH mm H mm H em 04
NH Hm NH oN. H wwN H wNN o G M0
nouaoaHuome 08mm: mcoHaosbmcH 9,3330?“ u EHHh
w me w 0: m5 om o om w H mN H awn»...
m A5N m mON NH N.N NH 0N m .m N. m 9 NH
N NbN N CNN HH Nm oH Hm m.o N e N mm
noHGQEHuonH “domain u mcoHHoduumsHoz .. EHHMV
o 2. o Hm. m moH m th o o O<
OH 2. 0H me m me m wcH, N mN N mN GO
poucoEHuomNH oEom u mcofloduumfi oZ FEHHM
VHS...“ onoom Mood ouoom demm ouoom Hand, ouoom vHsmm onoom Mam“ osoom Hoonflsm
seHMMuoU . m sound soundsoo m soumm uonMuoO . m non—mm
:onmouwwwtucoZ conmoumw< o>HH3H§HucoZ conmoumw< 9,332»:

 

1
1‘

nous/muoO mHeH was m soumm

>9, pocmHmm< moHsomoumO wcHsoom soflmSH cough. 9H”— co mchmm Hess mouoom Hooch. .mH XHQZHnwnm¢

54

 

 

 

 

 

 

m _ owH m N.wH m em m cm o 0 MA
HopcofiHuomxm onHZ .. HonuGoO
N NHN N NHN N. N.N N. .N o 0 $0
HoHGoEHuomxfi oEmm n 9833533 o>HH3H~Honmi EHHnH
m on m moH w mm w ow m w m m 9
w MNH w NN.H o ww 0 mm w 0N w NN 00
o NNH o mNH m No m was m.N on N on Am
souGoEHuolonM Hcouowfifl u mcoflosflmfi 02 n EHHh
w HN w HN H oON H wNN m.N on m mm NU
N. oo N. ma N HoH N NoH H ow H New 02
noncoEHuome 08mm .. moon—053mg 02 u EHHM
Mama ouoom Mama onoom chom osoom chmm onoom chmm osoom uHcmm ouoom Hoonndm

soumHHuoO U sebum
son mosmmwfiucoz

sonMuoO 0 Hanna
con m oummwt 033357552

nonMHuoU U Heumd

:03 m e ummd. o>HumHHEH

 

noummuoO m3 Hose 0 son—mm

>nH HoocmHmm< moHuoonmO mcHsoum sofimHZ 00.23... 05. Go mvHGmm Hons mosoom H.308 3m XHQZHQnqu

posfiusoo

 

55

 

 

 

MH MwH MH NwH OH NOH OH MOH O O Ah
N. ooH h ooH N.H OM OH HM O O N4
H wMN H MMN wN O MN N. O O mm
NH nwwH NH me M .HH Mo HH mo O O 4m
MN ON MN ON H N.HN N OHN M .NH N M .wH H Hm
HN HN. HN o0 w OOH w wwH O O MM
noucoaHHomxH unouomfifl u muoHuoSfi—ms o>HanH~Houm u EHHh
N NMN ;N NMN MN N. MN N. O O Q..—
M NHN w NHN M.ON N.N M.HN ON O O HHO
N.H OOH N.H HHH M N.MH O nNMH O O DU
HoocoEHnomNH 05mm u mnoHuonnumGHoNrHuHQEounH .. EHHh
OH OOH OH MOH wH Mm wH om HH w M.w o a
M .w NON M N.ON NN MN ON MN m O OH O 4N
w oN.H m MOH OH OM OH OM MH H M.wH H 2.
O OON O MON M.ON N.N M.HN ON M.o N. M.m o NHN
o MN.H o NN.H_ wH ww N.H MM M ON O NN 00
OH NNH MH .MNH MH No M.NH wo M.N oM N OM 1mm
M .w N.ON M ONN oH NM oH HM M .NH N MH N mm
HH OMH HH owH MH MN. MH Nb w NM w MM NH
noHGoEHHomxm acouowwHQ n mGoHHofiuHmcH OZ .. EHHh
wH N.o oH co m wNH w OMH H NN. H wh mm
wN HN wN HN M MON H N.NN M .N oM M MM ._...0
ON MN. ON. MN. O MwH M wOH O MN M MN 00

 

peacoEHuomxH 05mm .. mcoHuodnude oZ .. EHHh

 

demm ouoom deoM onoom Manx ouoom xnmm onoom uHcmm ouoom Mood ouoom pooBdm
uonMuoO noumm sonMIOO Hound HoHMMnoO Hound
Gonmonww<acoZ :onmouwwdw o>HumHHEanoZ aonmoumm< 9.53qu

 

m one Q mueHmMuoU :93. was 0 one .m .< msoumm poGHnEoU
.3 eoamHmm< moHsomoumO msHsoom .3“me cough. or: so muHcmm Hose mouoom H.308 :U XHQZMnHm<

 

56

wH HwH wH HwH M.HH Mo M.NH wo
oH Mo OH mo N. N.MH N. wMH
NN wM NN «M M MON M wON

 

soucoEHsome onth u HohcoO

MH MNH OH NNH o MHH o wHH

 

noosoEHuomxH eHmSH .. HoSsoO

M.o N. HH
OH M NH

O
N. MH N.

N.H

Ln

O .NH

MH ma

 

xcmm ouoom vHsmm ouoom xsem ouoom demm ouoom
soundnoo sound noanuoO noumm
conmouww<ucoZ :onmouww< o>HumޤHuGoZ

Hood 980% Hand ouoom uooflndm

uoHNMuoU scum
:on m ouwmw. 92383::

m

 

 

 

poSGHHGoO u

.0 NHQZManqu