University of Southern California Dissertations and Theses

The Effects Of Justice, Balance, And Hostility On Mirth

(USC Thesis Other)

The Effects Of Justice, Balance, And Hostility On Mirth

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content This dissertation has bean
microfilmed exactly as received 67*17,676
GUTMAN, Jonathan, 1938-
THE EFFECTS OF JUSTICE, BALANCE, AND
HOSTILITY ON MIRTH,
U niversity of Southern C alifornia, Ph,D ., 1967
P sychology, general
University Microfilms, Inc., Ann Arbor. Michigan
THE EFFECTS OF JUSTICE, BALANCE, AND
HOSTILITY ON MIRTH
by
Jonathan Gutman
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Psychology)
September 1967
UNIVERSITY O F S O U T H E R N CA LIFO RN IA
THE GRADUATE SC H O O L
UNIVERSITY PARK
LOS ANGELES, CA LIFO RN IA 9 0 0 0 7
This dissertation, written by
mittee, and approved by all its members, has
been presented to and accepted by the Graduate
School, in partial fulfillment of requirements
for the degree of
D O C T O R OF P H I L O S O P H Y
JONATHAN GUTMAN
under the direction of hJk?....Dissertation Com-
D tan
Date September 1967
DISSERTATION COMMITTEE
Chairman
. . . . At KA.
TABLE OF CONTENTS
LIST OF TABLES .......................................
LIST OF FIGURES .....................................
Chapter
I. THE PROBLEM ................................
Introduction
Purpose of the Study
Definitions
Major Hypothesis
Secondary Hypotheses
II. BACKGROUND ...................................
Freud's Tension Reduction Theory
of Humor
Studies on the Tension Reducing
Aspects of Humor
Studies in which tension is experi
mentally induced in subjects
Studies in which subjects are selected
by pre-testing on relevant variables
Heider’s Balance Theory
Humor Studies Reinterpreted in Balance
Theory Terms
Experiments on Justice As It Relates to
Arbitration of Frustration
Moral Judgment
III. METHOD .......................................
Materials
Pre-testing
Subj ects
Presentation of Material to Subjects
Counterbalancing
Chapter
IV. RESULTS
Page
65
Manipulation Effectiveness
Major Hypothesis
Analysis Based Directly on Perceptions of
Agent and Recipient
Summary of Major Hypothesis
Secondary Hypothesis
Abstract Situations
Results Pertaining to Abstract Situations
V. DISCUSSION ...................................
Situational Effects on Mirth
Balance
Justice
VI. SUMMARY
REFERENCES
APPENDICES
Appendix A
Appendix B
Appendix C
Appendix D
Appendix E
Appendix F
Story-Version Combinations
Instructions to Subjects
Questionnaire
Abstract Situations
Counterbalancing Design
Free Response Data
ii i
97
106
110
115
ge
66
67
68
69
70
70
72
72
73
73
75
75
76
LIST OF TABLES
Mean Social Acceptability Ratings of the
Agent and Recipient by Versions ............
Percentage of Subjects Vho Rated the Social
Acceptability of the Agent and Recipient
As Intended ...................................
Summary of Analysis of Variance on Perceived
Justice Ratings for Stories and Versions
Mean Ratings of Perceived Justice by
Version ........................................
Summary of Analysis of Variance on Perceived
Hostility Ratings for Stories and
Versions............................. . . . .
Mean Ratings of Perceived Hostility by
Version ........................................
Summary of Analysis of Variance on Humor
Ratings for Story-Versions ................
Mean Humor Rankings by Version ...............
Summary of Analysis of Variance on Humor
Ratings for Stories and Versions .........
Mean Humor Ratings by Version ................
Summary of Analysis of Variance on Humor
Ratings for Perceived Justice Rating . . .
Mean Humor Ratings by Justice Rating
Categories ...................................
Summary of Analysis of Variance on Humor
Ratings for Perceived Hostility Rating . .
iv
Table Page
14. Mean Humor Ratings by Hostility Rating
Categories................................... 76
15. Summary of Analysis of Variance on Hostility
Ratings for Justice Rating........... 77
16. Mean Hostility Ratings for Justice Rating
Categories................................... 78
17. Mean Humor Ratings by Perceived Hostility of
Punch Line and Justice of Recipient's
Outcome..................................... 79
18. Mean Humor Ratings by Perception of the
Agent and Recipient....................... 81
19. Analysis of Variance of Humor Ratings by
Perception of the Agent and Recipient . . 81
20. Ratings of Perceived Justice by Perception
of the Agent and the Recipient........... 82
21. Ratings of Perceived Hostility by Perception
of the Agent and the Recipient............ 83
22. Summary of Analysis of Variance on Humor
Ratings for Story-Version and Sex of
Subject, Stories A and C .................. 85
23. Summary of Analysis of Variance on Humor
Ratings for Story-Version and Sex of
Subject, Stories B and D .................. 85
24. Summary of Analysis of Variance for Simple
Main Effects of Sex of Subject at
Version Level for Humor Ratings
for Story B ................................ 87
25. Summary of Analysis of Variance for Simple
Main Effects of Sex of Subject at
Version Level for Humor Ratings
for Story D ................................ 88
26. Summary of Analysis of Variance on Perceived
Justice for Version and Sex of Subject,
Stories A, B, C, D ......................... 90
v
I
Table Page
27. Summary of Analysis of Variance on Perceived
Hostility Ratings for Version and Sex of
Subject, Stories A, B, C, D .................. 91
28. Correlations of Adjective Checklist Ratings
for Agent and Recipient with Humor
Ratings....................................... 93
29. Correlations of Adjective Checklist Ratings
with "Just" Ratings for Agent and
Recipient..................................... 93
30. Summary of Analysis of Variance on Pleasant
ness Rankings for Abstract Descriptions
of Versions................................... 95
31. Mean Pleasantness Rankings by Abstract
Descriptions of Versions ................... 95
32. Free Response Data on Reasons for Justice
Ratings for Story A .......................... 134
33. Free Response Data on Reasons for Justice
Ratings for Story B ........................... 136
34. Free Response Data on Reasons for Justice
Ratings for Story C .......................... 138
35. Free Response Data on Reasons for Justice
Ratings for Story D .......................... 139
vi
LIST OF FIGURES
Figure Page
1 . Three-Person Interaction System Showing
Relations Involved with the Perception
of Justice...................................... 6
2. Schematic Representation of Version 1
Showing Necessary Relations for Justice
and Balance................................... 8
3. Schematic Representation of Version 2
Showing Necessary Relations for Justice
and No Balance................................. 8
4. Schematic Representation of Version 3
Showing Necessary Relations for No Justice
and No Balance................................. 8
5. Schematic Representation of Version 4
Showing Necessary Relations for No Justice
and Balance................................... 8
6. Schematic Representation of a General
Triadic System ................................. 29
7. Schematic Representation of Experimental
Materials Showing Stories, Versions, and
Sex of Story Characters..................... 55
8. Schematic Representation of Pre-Testing
Design and Actual Experimental Design . . . 58
9. Matrix of Stories and Versions of Each Story . 63
10. Simple Main Effects of Mean Humor Ratings
for Story B for Sex of Subject at Version . 87
11. Simple Main Effects of Mean Humor Ratings
for Story D for Sex of Subject at Version . 88
12. Counterbalancing Design Showing Stories,
Counterbalancing Group and Versions of
Each Story Shown to Groups.................... 132
vii
I
CHAPTER I
THE PROBLEM
Introduction
In most societies the open expression of hostility
is not socially acceptable. Hostility can be disguised in
humor so that persons can release hostile impulses in a
socially acceptable manner. There is, nevertheless, a
limitation to the expression of hostility in humor. The
limitation is that even though these hostile impulses are
cloaked in humor they still must be perceived as being
justified. It follows that this perceived degree of
justification will affect the appreciation of humor.
The aim of this study is to specify the conditions
(in terms of justice) under which people will be able to
release their hostility through the appreciation of tenden
tious humor. Tendentious humor expresses some degree of
covert hostility or aggression. And it is the subject’s
reaction to this hostility or aggression that we wish to
study. In the remainder of this dissertation when the word
"humor" is used, it will mean ’ ’tendentious humor.”
All humor expressing hostile impulses should not be
equally effective as vehicles for this release of tension.
1
2
Situational factors such as the justice in the social rela
tions between the characters used in the stories influence
the perceived humor.
"Justice in social relations" must be perceived by
the subject in order to have any effect on mirth (mirth
refers to the response a subject gives that demonstrates
the appreciation of humor). The social situation, as it is
perceived by the subject, is intimately bound up with the
attempt to create humor. Certain factors present in the
social situation (such as lack of justice in social rela
tions between story characters) tend to act against the
subject's readiness to respond in mirthful fashion (Freud,
1938).
Kenny (1955) has noted that the situational approach
has not been popular:
Owing to the impact of Freud's penetrating analysis of
tendency wit, investigators have paid scant attention to
how the stimulus properties of the joke set boundaries
or limits to the humor response. In any case, it will
be necessary for these two approaches to converge in
order to give an adequate theoretical account of humor.
However, Freud has also pointed up the importance of these
factors despite research on other aspects of his theorizing
on humor. The quote below clearly shows that he has con
sidered how situational factors can affect mirth.
He himself (the subject or third person) must be in a
jovial, or at least indifferent state of mind, in order
to become the third person of the jest. The same
hindrance is present in the case of both harmless and
tendency wit, but in the latter the antagonism to the
tendency which wishes to serve wit appears as a new
3
hindrance. The readiness to laugh about an excellent
smutty joke cannot manifest itself if the exposure con
cerns an honored kinsman of the third person .... In
the presence of my opponents' friends the wittiest
invectives with which I might assail him would not b-e
considered witticisms but invectives, and in the minds
of my hearers it would create not pleasure, but indigna
tion. A certain amount of willingness or a certain
indifference, the absence of all factors which might
evoke strong feelings in opposition to the tendency are
absolute conditions for the participation of the third
person in the completion of the wit process. (Freud,
1938:732)
Empirical studies on humor have mainly dealt with the
thematic content (aggression, hostility, etc.) and not the
formal content (social relations between story characters).
The emotional state of the subject and the type and strength
of impulse expressed in the humor have been emphasized.
This study focuses on the social relations existing in the
stimulus situation which influence the perceived justifica
tion for the hostility.
In dealing with thematic content, humorous stimuli
have been categorized as hostile, aggressive, socially
disparaging, anxious, etc. The stimuli are sorted by judges
into appropriate categories. In the typical study the
investigator seeks relationships between the appreciation of
jokes in each thematic category and personality variables or
induced tensions. Relationships are obtained by summing
ratings for all jokes in each category and comparing them.
It is the hypothesis of this study that the sub
ject's reactions to the formal content of the story will
affect mirth. That is to say, what is crucial here is not
4
the fact that hostility is expressed but whether the sub
ject's reaction to the social relations between story
characters leads him to believe the expression of hostility
was justified.
Purpose of the Study
The general purpose of this experiment is to deter
mine what effect justice has on the appreciation of
tendentious humor in stories.
It has been hypothesized that mirth in response to
humor is tension—reducing (Levine and Abelson, 1959;
O'Connell, 1960; Strickland, 1959; Villiams and Cole, 1964).
However, the studies that demonstrated this effect dealt
with the thematic content and not with such factors as
social relations between story characters. It is contended
that the subject's perception of the relationships among the
various story characters will affect the degree of humor the
subjects attribute to the study.
Definitions
Interpersonal relationships vary in the degree of
justice involved. According to Homans (1961: 243),
. . . justice is realized in the relationship between
two men when the investments, rewards, and costs of one
are all better than those of the other— or when, if the
investments of the two are equal, their rewards and
costs are equal too.
Heider (1958) defines justice to mean that a person’s out
comes or experiences are in accord with his perceived worth.
5
These two definitions of justice are roughly equiv
alent. Homan's "investments" can be equated to Heider's
"perceived worth." And rewards and costs can refer to the
quality of outcomes or experiences. Heider's definition is
more germane to the nature of this study. Therefore,
"justice" in this study will refer to a situation that is
characterized by the fact that the experiences or outcomes
of a person are in accord with his perceived worth.
If a person behaves badly or in a morally or
socially unacceptable manner we (socially responsible
people) feel that he should encounter some misfortune or
negative outcome. However, if a person behaves in a manner
we deem proper and good we feel that any outcome connected
with this behavior should be positive.
The concept of justice can be subsumed by balance
theory. Balance theory is a very broad systematic way of
looking at interpersonal behavior (Heider, 1958; Jordan,
1953; Newcomb, 1953; Osgood and Tannenbaum, 1955). It is
a general model. As Heider (1958: 180) puts it, "By a
balanced state is meant a harmonious state, one in which the
entities comprising the situations and the feelings about
them fit together without stress." The task of research is
to determine the limitations of this model by applying it
to different situations. In this study balance theory and
the related theory of justice will be applied to the
explanation of mirth. Balance applies to systems of three
cognitive objects, whereas justice is concerned with two
object systems.
In bringing these two theories together in this
study our purpose is to determine whether the perception of
justice causes balanced situations to be more humorous and
pleasant than balanced situations in which there is no
perceived justice. This problem can be^ studied by incor
porating, in the same situations (stories or jokes), balance
or no balance and justice or no justice.
The operational definition of justice used in this
study is taken from Heider (1958): justice occurs when a
person's outcomes are equal to his perceived worth. This
principle is demonstrated below.
0
Fig. 1.— Three—person interaction system showing
relations involved with the perception of justice.
(A) and (B) refer to the agent (aggressor) and the
recipient (victim) in the story. In all versions the agent
(A) provides a negative outcome for the recipient (R) by
making him the butt of the joke. (0) is the observer, or
in this case, the subject in the experiment. Let the plus
and minus signs indicate good and bad or favorable and
unfavorable evaluations— in general, positive or negative
7
affect. From the description of R in the story, 0 evaluates
R; 0 also evaluates A's remark to R. 0's evaluation of R
corresponds to R's perceived worth. And 0's evaluation of
A's remark to R (the A-R relation) corresponds to R's out
come. The stories used in this study were selected for
hostile content. Therefore, the A-R relation should be per
ceived as being negative.
In two of the stories R is described in a manner that
would lead one to give him a favorable evaluation and in two
stories his description should induce a negative evaluation.
When R is negatively evaluated, the 0—R relation is the same
algebraic sign as the A-R relation, or in other words, R is
evaluated negatively and he receives a negative outcome.
This condition represents justice— a situation where there
is justification for hostility.
The stories were not only varied in degree of
justice but also in degree of balance. In considering
balance we must take into account not only the two relations
depicted in Figure 1 but also the third relation (0-A).
To have a balanced system, the three relations (having a
plus or minus sign) must have a product of "plus" when
multiplied together algebraically. There are four ways of
combining justice or no justice and balance or no balance
in the stories. These four ways of writing each story in
the future will be referred to as "versions" and are
described below.
0
Fig. 2.— Schematic representation of Version 1
showing necessary relations for justice and balance.
0
Fig. 3.— Schematic representation of Version 2
showing necessary relations for justice and no balance.
0
Fig. 4.— Schematic representation of Version 3
showing necessary relations for no justice and no balance.
Fig. 5.— Schematic representation of Version 4
showing necessary relations for no justice and balance.
9
In Figure 2, R's behavior is described as socially
unacceptable (boring, rude, taking advantage of others,
etc.). Therefore the observer (0) should evaluate R's
behavior negatively (0 — R) and in agreement with the
treatment he gets from A. This should fulfill the require
ments for justice.
A, although he makes the hostile remark, is
described as a person who doesn't look for opportunities to
express hostility. His action toward R is made to defend
himself against R's onslaught.
Figure 3 shows a situation in which 0 will not have
much to choose between A and R; both have behaved in viola
tion of common standards of conduct. However, the require
ments for justice have been met because R does deserve to be
the butt of the joke. If forces toward balance are strong
the subject may adopt a positive opinion of A (0 + A), or of
R (0 + R). Or he may misinterpret A's remark to be positive
(A + R). Only one of these alternatives, if any, should
take place. Changing one and only one sign is the simplest
way to achieve balance (Rosenberg, 1960). „ - -
The version shown in Figure 4 should be the most
unstable. Not only is it unbalanced but a logical incon
sistency has been created. We have tried to describe A as
positive. But A then attacks R who hasn't given him any
justifiable reason to do so. 0 could induce balance in the
system by changing his evaluation of R's behavior (0 — R) or
10
by changing his opinion about (0 - A). The latter is the
more likely, for it is A's description and his behavior that
create the inconsistency. The subject might also misinter
pret A's remark to R, in which case all relations would be
positive. However, it is felt that the subject's tendency
will be toward changing his opinion of A to conform with A's
actual behavior. In this case the model would approach the
one shown on page 8.
In Figure 5, 0 should think it unfair or unjust that
R be the butt of the joke. A's behavior is not socially
acceptable and furthermore he attacks R who has behaved in
a proper fashion. The reader (observer) should feel that A
is totally in the wrong and that there is no justification
for his remark. However, according to the formal definition
of balance, this represents a balanced situation.
Major Hypothesis
It is contended that balanced situations which are
justified are more pleasant than balanced situations which
are not justified. It is therefore the specific purpose of
this study to examine justice theory as a subset of balance
theory to see if it is a better predictor of mirth response
to the two-person interactions presented in the stories.
The following hypothesis is proposed as a basis for
this research: Stories in which the relations between story
characters are just will have higher humor ratings than
11
stories in which the relations between story characters are
not just. Also, stories in which the conditions for justice
are met, and which also represent balanced situations will
have higher humor ratings than such stories which are not
balanced. Conversely, those stories in which the conditions
for justice are not met, and which also represent balanced
situations will have lower humor ratings than such stories
which are not balanced.
Thus, within a single story the conditions for
justice and balance can be met; or the story can be devised
so that neither the conditions for justice nor balance will
be met. A story can also meet the requirements for justice
but not for balance, or vice versa. These four types of
stories are listed below in the order of their humor ratings
(from high to low) as hypothesized above:
1. justice and balance
2. justice and no balance
3. no justice and no balance
4. no justice and balance
However, if the results are in accord with balance
theory and not justice theory the humor ratings for the
justice and no justice conditions, 1 and 4, should be equal.
This would mean that one balanced situation is equivalent in
humor to another. And the humor ratings for both these con
ditions should be greater than for the two intermediate con
ditions (which are not balanced).
12
Secondary Hypotheses
A secondary hypothesis for this study is: When the
subject and the story characters are of opposite sex,
stories which depict the social relations as being just will
be rated as more humorous than when the subject and story
characters are of the same sex.
In the literature on humor (Hammes and Wiggins,
1962; Malpass and Fitzpatrick, 1959; O'Connell, 1958, 1962),
differences in humor appreciation for males and females have
been found. However, there is no evidence relating these
sex differences to the sex of the story characters used in
the stimulus material. A subsidiary purpose of this experi
ment is to see if these sex differences can be explained in
terms of severity of moral judgments and their effects on
perceived justice.
Studies on moral judgment (Kinder, 1925; Kitay,
1940; Klinger et al., 1964) have indicated that females
judge males more harshly than they judge females and vice
versa. Therefore, the effect of justice should be exagger
ated when the subject and story characters are of opposite
sex. The harsher the moral judgment the more the person is
deserving of a negative outcome. When he (or she) receives
this negative outcome (by being the butt of the joke) the
more satisfying it should be to the subject. And the better
the subject will be able to release hostility through mirth.
Interest of this study in humor has led to the
1 3
offering of a somewhat indirect test of the efficacy of
balance and justice theory. The desired ordering with
respect to justice could occur but not make itself evident
in the humor ratings. Therefore this indirectness will be
supplemented with a more direct comparison of justice and
balance theory. Using the general model employed by Jordan
(1953), simple statements were made up characterizing the
four conditions presented in conjunction with the first
hypothesis. These statements were presented to the subjects
who were asked to rank them on the basis of "pleasantness."
It was hypothesized that the preference for the situations
would be in accord with justice theory and not balance
theory.
CHAPTER II
BACKGROUND
Freud's Tension Reduction Theory of Humor
Most of the recent studies on humor and especially
those that were examined in formulating the hypotheses for
this study were based on Freud's tension reduction theory
of humor (Williams and Cole, 1964). Therefore, before
these studies are reviewed, Freud's thoughts on the tension
reducing aspects of humor will be discussed.
Freud dichotomized humor into harmless and tendency
humor (Freud, 1938). Harmless humor is humor for its own
sake. It serves no particular purpose other than to pro
duce a feeling of satisfaction and a certain amount of
laughter in the hearer. On the other hand, humor may serve
some particular tendency such as expressing hostility.
Where one once had recourse to physical violence,
now one relies on verbal invective. The aim of this invec
tive is to enlist the support of a third party by whose
laughter (at our enemy) we obtain our pleasure. Thus, as
Freud says:
We are now prepared for the role that wit plays in
hostile aggression. Wit permits us to make our enemy
ridiculous through that which we could not utter loudly
14
1 5
or consciously on account of existing hindrances; in
other words, wit affords us the means of surmounting
restrictions and of opening up otherwise inaccessible
pleasure sources. (Freud, 1938: 698)
Thus, humor, in serving a hostile tendency, makes
possible the gratification of a hostile desire despite a
hindrance which stands in the way. The hindrance is eluded
and pleasure is derived from a source that has been
inaccessible on account of the hindrance. The hindrance
often consists of nothing more than cultural restrictions on
our behavior.
In tendentious humor three persons are involved: the
one who makes the humor, the person who is the object of the
hostility expressed in the humor, and the listener, whom
Freud called the third person.
Thus far we have considered the first person— the
one who produces the humor. A rationale for his behavior
has been provided; and some benefits resulting from his
resort to humor have been mentioned. However, in this
study our interest is in the third person— the casual
observer of the situation. Let us first discuss how he can
derive pleasure from response to humor and then spell out
in detail some necessary factors in the relationship between
the first person (producer of the humor) and the third
person (casual observer).
Pleasure arises from response to humor because
wishes or motivations are gratified that otherwise would
16
have been repressed or suppressed. Freud assumes that in
both repressing and suppressing motivations an expenditure
of psychic energy is required. However, if through humor
we avoid this, we have saved that amount of psychic energy.
The pleasure arising from response to humor occurs when
this "saved amount of psychic energy" is released suddenly
in response to humor.
This process of sudden release of psychic energy has
come to be known as the tension reduction theory of humor.
As a response to humor a person may express unacceptable
impulses in a manner acceptable both to himself and to
others. In so doing he may reduce tension produced by some
blocked or ungratified need. This breakthrough will not
produce the anxiety that would normally occur because the
humorous stimuli act as a disguise. Thus potentially pain
ful situation can be resolved through pleasurable response
to humorous stimuli.
All of the above presupposes that the producer of
the humor happens to strike some cordant note in the third
party or casual observer. Freud (1938) discusses some of
the necessary ingredients of the relationship between the
third and first persons that enable the third party to
release tension through response to humor.
Firstly, the third person (hearer of the humor) must
be in psychic harmony with the first person. That is to say,
he should possess the same inner inhibitions that the first
17
person has overcome by resorting to humor. It was mentioned
earlier that a person assailed his enemy with verbal invec
tive rather than physical violence. In so doing he hoped to
gain the support of any spectators or observers (i.e., third
persons). The first person eluded his inhibitions against
expressing this hostility by doing so through humor. The
third person must possess similar inhibitions in order to
appreciate the humor.
Secondly, the humorous stimuli must awaken in the
third person
compulsively and automatically a readiness for this
inhibition. This readiness for the inhibition, which
I ["I" refers to Freud] must conceive as a true expen
diture analogous to the mobilization of an army, is
simultaneously recognized as superfluous or as belated,
and is thus immediately discharged . . . through the
channel of laughter. (Freud, 1938: 737)
The above quotation indicates that the humorous
stimuli must first strengthen and heighten the inhibitions
in the third person before providing a sudden release for
the pent-up energy. Thus, in releasing tension through
response to humor the hearer or third person must have
similar inhibitions to those eluded by the person producing
the humor. And the humorous stimuli must itself first
enhance these inhibitions and provide a sudden release for
them in the "punch line" or the technique employed in
creating the humor.
18
Studies on the Tension Reducing Aspects
of Humor
Studies demonstrating the tension reducing aspects
of humor are of two basic types: (l) studies in which ten
sion is experimentally induced in subjects; (2) correla
tional studies in which subjects are selected by pre—testing
on relevant variables in order to get groups high and low
on those variables. The studies will be reviewed in that
order.
Studies in which tension is
experimentally induced
in subjects
Bryne (1958) studied the effect of general . ’rive
level on subsequent responses to humorous stimuli. He used
four experimental conditions designed to create four degrees
of tension or drive level. His hypotheses were that sub
jects under higher drive levels would rate cartoons as more
amusing than subjects in a more neutral state; and that car
toons would be rated less amusing under conditions of low
drive level as compared to neutral conditions.
His hypotheses did not prove to be correct. There
were no differences in ratings among his experimental condi
tions. Therefore, Bryne concluded that general drive level
was not related to responses to humorous material.
There is one aspect of Bryne's study that is rele
vant to all other studies which follow. The dependent
variable is the subject’s rating of the cartoons he has been
19
exposed to. This is supposed to indicate whether or not
tension was reduced by response to humor. Tension reduc
tion, per se, is never measured. The reasoning is that the
more tension released through response to humor the more
pleasurable the humor will be to the subject. Therefore,
the more amusing the subject will find the humor and the
higher he will rate it.
Strickland (1959) did a study that demonstrated that
specific drive level is related to humor response. He
hypothesized that a person who is annoyed and frustrated but
is prevented from expressing his hostility should respond
more favorably to humorous material of a hostile nature.
Similarly, a subject who is aroused sexually but has no
means of expressing this motivation should prefer humor with
sexual content. Only relevant humor stimuli reduce drives
in specific need areas. Strickland found a significant
interaction between type of arousal and type of cartoon
preferred as predicted. Induced tension was released
through humor when the humorous stimuli mirrored some aspect
of the tension inducing situation.
Williams and Cole (1964) investigated the relevance
of a tension—inducing social situation upon an individual's
appreciation of humor. They created a sense of inadequacy
in subjects which resulted in an increased appreciation for
jokes in which the main character in the joke behaved
inadequately. Their results support the tension reduction
20
theory of humor. Unfortunately, they did not include any
non—relevant humor. Therefore, one cannot be sure that the
results were due to the need relevance of the humor and not
just general tension reducing aspects of humor.
In the above studies tension was experimentally
induced in subjects. They sought a direct test of the
tension reduction theory of humor. The results of these
studies are somewhat equivocal but tend to point up the
importance of the relevance of the humor stimuli to the
specific drive or tension induced.
Studies in which subjects are
selected by pre-testing on
relevant variables
In the following studies the approach is less direct.
They are based on the assumption that people with certain
personality traits or social backgrounds will experience
more tension than other people. Subjects are separated into
groups based on these traits and differences in humor appre
ciation are sought. These studies confirm the tension
reduction theory of humor and give additional support to the
importance of the relevance of the humor stimuli to the
tension that the subjects experience. Also, an interesting
finding in the following studies is that if persons experi
ence too much tension, tension reduction through response to
humorous stimuli will not occur.
Levine and Abelson (1959) were interested in reac
tions to humor as a function of how well the subject could
21
deal with anxiety. They made the hypothesis that for per
sons poor at dealing with anxiety, humorous material relevant
to the sources of tension they have may well create tension
instead of reducing it. In support of this assumption they
cited Redlich, Levine, and Sohler (1951) who made a similar
assumption connecting low humor ratings with an increase in
anxiety:
When a stimulus, ostensibly humorous, evokes anxiety,
disgust, shame, guilt, or horror it is assumed that the
release of some primary suppressed or repressed need
produced a threat with the resulting affect of dis
pleasure. (Redlich, Levine and Sohler, 1951: 718)
Also, Levine and Redlich (1955) did a study using
psychiatric patients. When these patients were shown car
toons, some of them evoked not laughter, but rather anxiety
which often resulted in a failure to "get” the joke. The
events depicted in these cartoons were traced to incidents
in the patients' histories that were the cause of much
anxiety for them.
If cartoons caused anxiety instead of reducing it
the natural assumption would be that they would not be con
sidered amusing. Furthermore, persons who are unable to
manage their anxiety will respond less favorably to anxiety-
producing cartoons and only get pleasure from milder ones.
Normal persons should prefer humor that is at least mildly
anxiety—arousing (i.e. , strengthening and heightening the
inhibition before providing release).
Levine and Abelson used psychiatric patients in the
22
experimental condition and men accepted for submarine duty
(regarded as being very stable) in the control condition.
The cartoons initially were rated by psychiatrists for the
"disturbingness" associated with each cartoon. When these
cartoons were rated by the patients only the more innocuous
ones were appreciated. In the control group the disturbing—
ness of the cartoon had no effect on their responses.
O'Connell (1960) pre-tested his subjects using the
Taylor Manifest Anxiety Scale. He selected subjects with
either high or low scores. He then used insult techniques
to induce tension in half of the subjects in each group.
No overall differences were found in humor ratings between
insulted subjects and those who were not insulted.
O'Connell also had hypothesized that persons who
scored low on the anxiety scale would employ humor to
reduce tension when they were subjected to these insulting
experiences. But they would not do so when not subjected
to this tension. On the other hand, the high scorers on the
anxiety scale would act in an opposite fashion. Under
induced tension conditions the repressions of these
"maladjusted" persons would become so exaggerated that dis
guised release would not be possible. The data supported
his hypotheses. There was a significant interaction effect
between level of anxiety and tension.
The results of an experiment by Doris and Fierman
(1956) can be explained by the fact that subjects with too
23
high a level of tension cannot reduce tension through
response to humor. They were trying to demonstrate that
subjects with high anxiety responded differently to humorous
material than subjects with a low level of anxiety. The
results did not support their data. Later it was found that
their results were confounded by the fact that they used two
experimenters— one male and one female. When the data were
re—analyzed, a significant interaction effect was found
between level of anxiety and same or opposite sex of the
subject and the experimenter. There were significant dif
ferences in preference for hostile cartoons, but only when
an opposite-sexed experimenter was present. In these cases
the high-anxious subjects showed less appreciation for the
hostile humor. Supposedly the opposite-sexed experimenter
raised the tension level of the high-anxious subjects so
that they were unable to reduce tension through humor.
Hammes and Wiggins (1962) used college students who
were at the upper and lower extremes of the Taylor Manifest
Anxiety scale. They hypothesized that highly anxious people
would not enjoy situations in which emotions such as worry
and tension are treated in humorous fashion as much as
people who had low anxiety levels. Their hypothesis is
based on the assumption that individuals possessing a high
degree of anxiety have more problems in a greater number of
behavioral areas than do those individuals with less
anxiety. Also, highly anxious individuals tend to project
24
greater danger or threat into common environmental stimuli
than do low-anxious individuals.
Their predictions were partially supported by the
data. The main effect of anxiety was not significant. But
the interaction between anxiety and sex of subjects was
significant. The prediction that highly anxious subjects
would give lower humor ratings than low-anxious subjects was
true only for males. There was a non-significant reversal
for female subjects. The explanation they provide is based
on identification. The humor stimuli were "Peanuts" cartoon
strips. Most of the characters in them were male. Hammes
and Viggins say that the male subjects evidently identified
with the male story characters more than did the female
subjects. Hence the cartoons were more threatening to them.
Hetherington and Vray (1964) did a study that demon
strates the importance of relevance of humor stimuli to the
source of tension in tension reduction. And it also shows
how the need for social approval affects mirth. They
investigated the effects of aggressive drive and need for
social approval on humor preferences. They selected sub
jects who were high or low on their need for social approval
and the need to express aggressive impulses. Half their
subjects rated aggressive and nonsense cartoons after having
consumed some alcohol; the other half did not receive any
alcohol.
They found that subjects with a high need for social
25
approval (high NSA) rated nonsense cartoons higher than did
low NSA subjects. In their interpretation Hetherington and
Wray assume that having a sense of humor is a positive
social trait and hence will be correlated with a high need
for social approval. In the nonsense cartoon as opposed to
the aggressive cartoons no social prohibitions were being
violated. Therefore, high NSA subjects were not inhibited
in responding to it and their better sense of humor resulted
in higher humor ratings.
Subjects with a high need to express aggression
(high NA) liked aggressive cartoons more than did low NA
subjects. Presumably high NA subjects had more aggressive
tension to release. And the release of this tension through
response to aggressive cartoons accounted for their increased
ratings. There was no such difference between high and low
NA subjects for nonsense cartoons. Nonsense cartoons
evidently did not provide an adequate outlet for persons
with a high need to express aggression.
Both aggressive and nonsense cartoons were rated
higher under the alcohol condition than they were under the
nonalcohol condition. It seems that alcohol lowers inhibi
tions and enables subjects to respond more freely,
especially in the case of aggressive cartoons. Further,
with respect to the above difference between alcohol and
nonalcohol conditions for aggressive humor, high NA sub
jects but not low NA subjects were the ones responsible for
26
this difference. Presumably the low NA subjects had so
little aggression to release it did not matter whether their
inhibitions against expressing aggression were lowered or
not.
As part of the above findings, Hetherington and Wray
discovered an interaction between need aggression and the
need for social approval. Subjects high in both NA and NSA
gave higher ratings under the alcohol condition than under
the nonalcohol condition. This was not true for subjects
high in need aggression but low in need for social approval.
Thus it seems that the need for social approval has some
limiting or inhibiting effect on the release of tension.
The aggressive cartoons depicted situations in which
social prohibitions were being violated. If one assumes a
mirthful response to this type of humor constitutes some
sort of sanction of this behavior it would explain why the
lowering of inhibitions from the influence of alcohol makes
a difference in the case of high NSA subjects but not for
low NSA subjects (Hetherington and Wray, 1964). When high
NSA subjects1 inhibitions against violations of social pro
hibitions are lowered through the influence of alcohol,
their appreciation of this type of humor goes up— especially
if they have a need to express aggression. The alcohol
makes no difference for low NSA subjects because they do not
have these inhibitions to begin with.
The last study in this review takes us far afield in
demonstrating the tension reducing aspects of humor. More
27
and Roberts (1957) did a study which tried to relate mirth
to the subject's position in the social structure of our
society. They assert that social learning determines humor
preferences. Humor arises out of social interaction, and
hence, each individual is conditioned by his particular set
of past experiences.
Two social groups differing markedly in developmental
experiences, but having fairly high homogeneity of
experiences within groups should show significant dif
ferences in types of humor enjoyed and/or intensity of
humor response. (More and Roberts, 1957: 235)
They used subjects from three different social
groups: public welfare trainees, upper middle class adults,
and upper middle class college students. The humorous
material was divided into categories based on their
dominant theme, such as hostility toward authority, hostil
ity toward peers, escape from unpleasant situations, anxiety
over money, sexual impulses, etc. The subjects rated the
cartoons by ranking them in sets of four according to how
funny they thought they were.
Differences in the themes thought most amusing were
found. The upper middle class adults thought cartoons with
the theme of hostility toward peers the funniest, whereas
the welfare group thought the hostility toward authority
themes the funniest. The college students found cartoons
with narcissistic and supernatural themes among the least
funny. Males, in general, found the cartoons funnier than
did females. They especially found the themes of hostility
28
toward peers and authority funnier than did females. Where
sociological theories predicted that expression of a partic
ular theme was blocked for one of the groups, it was found
that this group tended to rate cartoons with this theme as
being most amusing.
The preceeding studies demonstrate that response to
humor is tension reducing under appropriate conditions.
When the humor represents a blocked need that creates ten
sion, but not too much tension, the subsequent release of
tension causes the humor to be more appreciated than humor
that doesn't meet these conditions.
In this dissertation interest is in situational
factors that affect this release of tension. The situa
tional factor of interest is the justice in social relations
between story characters. To the writer’s knowledge, there
have been no studies on humor dealing with this particular
variable. However, there have been studies that have
systematically investigated the effects of certain other
situational factors that can affect the tension reducing
ability of humor. We wish to consider these other situa
tional factors in terms of balance theory. Balance theory
is a general theory of interpersonal behavior; justice
theory, as defined in this study, is based in part on
balance theory but is more specific in nature.
Heider's Balance Theory
In its simplest form, balance theory involves
29
sentiments and unit formations. Sentiments refer to how a
person, p, feels about or evaluates something. The "some
thing" may be another person, p, or an impersonal entity or
event, x* Sentiments, roughly speaking, may be positive or
negative; they may refer to a liking relation or a dis
liking relation. Unit formation refers to the fact that
separate entities comprise a unit when they are perceived
as belonging together. For example, a person and his deed
belong together.
As Heider (1958) put it,
The concept of balanced state designates a situation in
which the perceived units and the experienced senti
ments co-exist without stress; there is thus no pres
sure toward change, either in the cognitive organiza
tion or in the sentiment.
Heider views the relationship between p, p, and x
as constituting a system. They are interdependent; if one
relationship changes others may be shifted in order to
maintain balance (Heider, 1958).
+ or '+ or
Fig. 6.— Schematic representation of a general
triadic system.
The relationships between p, p, and x can either be
sentiments (liking or disliking) or unit relations
(belonging or not belonging together). By giving these
30
relations algebraic signs (+1 or -1), and then combining
the signs for all the relationships in the system, a posi
tive or negative total will result. If this product is
positive the system is balanced; if it is negative the
system is imbalanced. There have been various refinements
to Heider's method for determining balance: considering the
lack of any unit relation as having no sign (instead of -1)
(Cartwright and Harary, 1956; Harary, 1959); giving the
signs of the relations more degrees of strength (more than
+1 and -1)(Morrisette, 1956). But the essential nature of
the system is as expressed above.
This is a deceptively simple system and very broad
in scope. The two studies below demonstrate one of the
most basic tenets of balance theory— that people show a
preference for balanced states.
In a study done by Esch (in Heider, 1958), subjects
were asked to predict the probable outcomes of various
imbalanced situations. A typical example is:
Bob thinks Jim very stupid and a first-class bore. One
day Bob reads some poetry he likes so well that he
takes the trouble to track down the author in order to
shake his hand. He finds that Jim wrote the poems.
In response to this dilemma (j> dislikes o, likes x» and
discovers x belongs with o_) and other similar ones, the
outcomes predicted by the subjects involved changing one of
the three above relations (between o, and x)• Practi
cally all such changes resulted in balanced situations.
31
In a study by Jordan (1953) subjects were asked to
indicate their preference for various triadic situations
stated in abstract form. Some of these situations were
balanced, others were not. The balanced situations were
preferred over the imbalanced ones.
Humor Studies Reinterpreted in
Balance Theory Terms
To the writer's knowledge there have been no studies
published on humor using Heider's balance theory as a
theoretical basis. Below are some studies in which the
stimulus properties of the humorous material are important.
These studies will be explained in balance theory terms.
A study by Kenny (1955) demonstrates that the
better the elements in a humorous situation fit together
the greater the mirth resulting. He sought to relate humor
appreciation to the stimulus-confirmation of joke-ending
expectations. He too, as are we, was interested in the
stimulus qualities of humor-evoking situations. His study
dealt with one aspect of the stimulus qualities of jokes—
the stimulus confirmation of joke-ending expectancies. He
defines three categories or degrees of what he chooses to
call "stimulus-expectancy incongruity."
1. low stimulus expectancy incongruity— jokes classi
fied as containing stimulus endings which largely
confirm one's expectations.
2. moderate stimulus expectancy incongruity— jokes
characterized as containing moderately expected
stimulus endings.
32
3. high stimulus-expectancy incongruity— jokes con
taining stimulus endings which are highly
unexpected.
His results indicate:
The critical finding in the present study is that the
lowest stimulus-expectancy incongruity jokes are rated
the most humorous. This result argues strongly against
both the incongruity and the discrepancy hypotheses.
(Kenny, 1955; italics mine)
However, these results are in accord with predictions made
from balance theory. In a balanced state the entities com
prising the situation and the feelings about them fit
together without stress. When the stimulus-expectancy
inconguity is low— the stimulus endings confirm one's
expectations— balance is high. And Kenny's results indi
cate that humor appreciation is also high under these
conditions.
Another study that demonstrates that high forces
toward balance are associated positively with humor appre
ciation is by LaFave (1961). His study concerns reference
group effects and humor appreciation. He used jokes
having to do with religion. Subjects were members of
different religious groups; a group of agnostics was also
included. His design was such that for each joke which
esteemed one group and disparaged another there was another
which did the opposite. His results support the hypothesis
that jokes will be judged funny by subjects whose reference
group is esteemed and whose outgroup is disparaged.
These results are certainly in accord with balance
33
theory. The sentiments and unit relations existing because
of the subject's acceptance of the reference group are
balanced for a joke that esteems his group and disparages
an outgroup. And conversely the opposite condition would
be imbalanced.
Lastly, a study by Elbert (1962) relating authori
tarianism to humor preference is of interest here. She
used cartoons, some of which directed hostility at high
status targets and others at low status targets. One of
her main findings was that cartoons with upper status
figures as targets of hostility were preferred signifi
cantly less by the high-F (high in authoritarianism) group
than by the not-high-F group.
It is well known that persons high on authoritar
ianism are more status conscious and value high status
figures more than do persons not high in authoritarianism
(Adorno et al.. 1950). Therefore, it should be more
stressful (less harmonious) for them to encounter stimulus
material which is hostile toward high status persons. On
this basis balance theory is also in accord with the
results.
The above post-hoc analyses demonstrate that balance
theory can be used to explain mirth. However, in the main
hypothesis of this study it was put forth that some balanced
situations (namely those that are just) are more satisfying
than other balanced situations (those that are not just).
34
And that mirth will result in response to just-balanced
situations. In formulating this hypothesis the thinking
was that balance theory is so broad that other factors in
the situation might also have an effect in determining
preference or desirability. And that these factors might
render some balanced situations more desirable or satis
fying than other balanced situations. The "other" factor
in this dissertation is justice of social relations.
A study done by Rosenberg and Abelson (i960)
demonstrates how "other" factors operate. In this case the
other factors are "the fulfillment or frustration of
motives other than those driving the individual toward
formal cognitive balance." (Rosenberg and Abelson, 1960:
144) These and other factors caused subjects in some cases
to prefer imbalanced situations to balanced ones.
Subjects were divided into three groups and assigned
to one of three roles. Each role contained a source of
imbalance among its concepts and beliefs. After the sub
ject had assimilated the role he was given three counter
communications to read. These communications were designed
to resolve the imbalance originally existing in the roles.
The results indicate that after reading these counter
communications only one of the three groups achieved
balance within their roles. All groups showed some change
within the role structure as shown by before and after
ratings. But two of the three groups changed in accordance
35
with some principle other than balance.
For all three groups the role demanded that they,
as owner of a department store, keep sales as high as
possible. In Group 1 the subjects could adhere to this
goal and at the same time achieve balance in their role
structures. This was not true for groups II and III. In
achieving balance Group II could at best hope for a situa
tion that would keep sales at their present level. For
Group III to achieve balance they would have to agree to
a situation that would promise lower sales. Both groups
II and III preferred imbalanced role structures and high
sales to balanced role structures and the same or lower
sales.
Rosenberg and Abelson summarize the results of this
part of their experiment in the following way:
In resolving cognitive discrepancies of the sort repre
sented by our materials, subjects seek not only the
attainment of cognitive balance and consistency but
they seek also to alter their beliefs and evaluations
in ways that will maximize expected gain and minimize
expected loss; when both forces converge so that they
may be gratified through the same change or changes a
formally "balanced" outcome will be achieved; when
these forces diverge the typical outcome will not meet
the requirements of a simple formal definition of cog
nitive balance.
In the situations in this dissertation it is hypo
thesized that justice or lack of it will be of more
importance in determining preference than will balance.
Justice
The concept of justice, as defined in this study,
36
has to do with the balancing of a person's outcomes or
experiences with his perceived worth (Heider, 1958; Homans,
1961). The definition we have chosen was deliberately
formulated in balance theory terms so as to facilitate the
comparison between justice theory and balance theory. The
idea of "perceived worth" involves some one perceiving and
evaluating "others" behavior.
For example, assume a third person observing the
interaction between two other persons. He observes the
first person and evaluates him (determines his "perceived
worth")— and also as a result of this what the "rights" of
that person are. Then he observes the behavior of the
second person toward the first person. At this point he
can decide if the second person has behaved justly or
unjustly. This decision is based on whether the second
person has respected the rights of the first person. In
terms of the above definition: justice results when the
behavior of the second person is in accord with the
rights accruing to the first person as a result of his
behavior.
A man's rights depend to a great extent upon his
actions. According to Westermarck (1932: 132):
Other things being equal, the criminal has not the same
rights to inviolability as regards reputation, freedom,
property, or life, as the innocent man; the miser and
egoist have not the same rights as the benefactor and
philanthropist. On these differences in rights due to
differences in conduct the terms "just" and "unjust"
lay stress.
37
Concerning injustice, Westermarck (1932: 130)
wrote;
Injustice is a kind of wrongness. To be unjust is
always to be unjust to somebody, a violation of some
one's rights. Justice is a kind of rightness. It
involves the notion that a duty to somebody, a duty
corresponding to a right in him, is fulfilled; . . .
It will be noticed that both justice and injustice
involve a person's rights. If a person has a right to
behave in a certain way it follows that he ought not to be
prevented; and this character of inviolability is basic to
the concept of justice.
People make judgments about the behavior of others.
These judgments are a basis for determining the rights of
the person executing the behavior, and the justice in the
situation. These judgments are of a particular nature.
They are referred to as moral judgments. In making a
moral judgment a person determines, for himself, what is
right and wrong or good and bad. Insofar as determining
justice is concerned, these judgments should be impersonal.
If I pronounce an act done to a friend or to an enemy
good or bad, that implies that I assume that act to be
so independently of the fact that the person to whom it
is done is my friend or my enemy. Conversely, if I
pronounce an act done by a friend or my enemy good or
bad, that implies that I assume it to be so indepen
dently of my friendly or hostile feelings towards the
agent. (Westermarck, 1932)
Personal wishes should not influence the perception
of what is good and what is bad. This is not to say that
they do not— but rather that they "should not." In
38
principle moral judgments and hence justice should be
established by the objective requirements of the situation
(Westermarck, 1932; Heider, 1958).
The objective requirements of a situation are in
part a function of the social relations existing between the
persons interacting in the situation. Moral judgments may
be altered because of these social relations. But as long
as these changes are due to objective requirements and not
because of changes in personal wishes, they still belong
on the just-unjust dimension.
Albert (in Heider, 1958) has demonstrated that
perception of justice changes with the situation because
the social relations between the agent and the recipient
of the action vary. These changes cause the act to be
interpreted differently.
It was shown that certain provocative acts, such as
playing a practical joke, criticising, kidding, or
disagreeing vocally, were considered to be less
acceptable morally when directed against a stranger
than a friend. In the former case the acts tended to
be interpreted as harms with overtones of hostility or
aggression, whereas in the latter case the act was not
interpreted as real harm. The same general difference
in interpretation and moral judgment depended on
whether the social relationship involved disliking
the recipient or liking him.
Before we go on to look at other studies using the
concept of justice, the question of why the effects of
justice should affect the response to humor will be dealt
with. Why will mirth be greater when hostile actions are
justified and less when their justification is low? The
39
answer comes from the fact that people are able to release
hostile impulses (i.e., tension) through humor. When we, or
some friend we like, are attacked or treated unjustty our
hostility is aroused. We then get great satisfaction when
justice is done and the offending person "gets what's coming
to him."
The progression is from perceiving an unjust action,
to the arousal of our hostility, to the releasing of this
hostility in the perception of a negative outcome for the
unjust person. This process when duplicated in humor can
cause the same reaction to take place. The satisfaction
allows us to see the humor in the situation.
All situations in which justice prevails are not
funny. The production of humor is dependent on more than
creating justified situations. However, if the inter
personal relations are not justified we will not have the
satisfied feeling associated with justice being done. And
we will not release our hostile impulses in mirth because
the situation is inappropriate (Freud, 1938: 698).
Experiments on Justice As It Relates
to Arbitrariness of Frustration
The experiments that follow use the concept of
justice in a somewhat different way than it is employed in
this study. These studies are interested in the elicita
tion of aggression by frustration. The frustration is
varied so that in some cases it is arbitrary or unjustified.
40
The specific interest is in the difference in aggressive
responses given- by subjects in response to justified and
unjustified frustration.
They focus on whether the frustration itself was
justified or not. Our focus is on whether the aggression
or hostility expressed in the stories is justified or not.
In the following studies frustrating situations are pre
sented to subjects. The subjects themselves supply the
aggression or hostility in reaction to them. In our study
a complete situation is presented to subjects; the ante
cedent conditions and the aggressive or hostile responses
to them. The perception of justice is based on the matching
of the hostility to the behavior occurring before it.
In the studies that follow there is no aggression
or hostility in the situation as presented— that is supplied
by the subject. Therefore there is nothing to match. The
perception of justice is based solely on internal norms
built up by experience which are used to evaluate the
frustration. In our study this same type of judgment
occurs but it is then used as a basis for evaluating the
"justness" or "unjustness" of the following behavior.
The concept of justice is highly similar, both in
this study and in those reviewed below. Justice is applied
to different parts of the sequence of events— in the studies
below, the antecedent conditions; in this study, the
hostility resulting from the antecedent conditions.
41
However, the variables of importance in the studies below
should be relevant to the study presented in this disserta
tion.
In presenting these studies we wish to demonstrate
that people feel justified in making aggressive responses
when confronted with an unjustified frustrating agent. And
therefore they should also find situations more satisfying
when the aggression in a situation appears justified, and
not satisfying when the aggression is not justified.
Pastore (1952), Cohen (1955), and Rothaus and
Vorchel (i960) took exception to the generality of the
frustration-aggression hypothesis. Their thinking was that
aggression would follow frustration when the frustrating
agent was seen as arbitrary or unjustified but not when the
action of the frustrating agent was nonarbitrary or
j ustified.
Subjects were presented with short stories which
depicted the frustrating agent as behaving in either a
justified or unjustified manner. Subjects were then asked
for their reactions to these situations. The unjustified
situations elicited more aggressive responses than the
justified situations.
Allison and Hunt (1959) did a study whose results
were similar to those above. They were interested in the
effect of the social desirability scale factor upon the
response of expressing aggression under varying hypothetical
42
conditions of frustration. The subjects were given the
Edwards Social Desirability Scale and then presented with
12 hypothetical frustrating situations.
In filling out questionnaires subjects were asked
to express their reactions to these situations. There was
no difference in aggressive responses between subjects high
and low on the need for social desirability for either
justified or unjustified frustration. However, there was,
overall, more aggression expressed in response to the
unjustified frustrating situations than to the justified
frustrating situations.
Lee (in Buss, 1961) investigated the justifiability
of frustration in a classroom situation. She studied the
effect of a teacher's acting in opposition to the desires
of a small class. In one condition the teacher gave as a
rationale for her behavior the good of the students
(justifiable frustration). In the other condition the
teacher gave as the reason for her behavior her own desires
(unjustified frustration). The unjustified frustration led
to significantly more aggressive verbalizations by the
students than did the justified frustration.
Cohen (1955) (cited above) incorporated some other
variables in his study to see what their effect on the
justice of frustration was. He had subjects respond by
telling what they would "actually" do in response to the
frustrating situation; and they also told "ideally" what
43
their response would be. "By ideally is meant the socially
acceptable or socially circumscribed response as opposed to
what one would actually do regardless of social harms."
(Cohen, 1955) For both justified and unjustified frustra
tion aggressive responses were significantly lower when the
subjects indicated their ideal responses as opposed to their
actual responses to the situations.
In expressing his opinion on the impact of this
finding, Cohen says:
These norms (social norms of American middle class
society) discourage open aggression and stress certain
ideal standards of behavior. The pressure to conform
to circumscribed behavior patterns exerts itself with
enough strength to bring a great deal of communality
among the 60 subjects as to how people should ideally
respond when frustrated.
Evidently people are definitely aware of what
behavior is socially acceptable and how they "ought" to
behave— even though they might not "actually" behave in
that manner. And although the ideal-actual distinction was
significant for both justified and unjustified frustration
it made a bigger difference in the unjustified condition.
Rothaus and Worchel (1960), in addition to the
general finding that aggression occurs more frequently in
response to unjustified frustration than justified frustra
tion, also demonstrated that other variables affect this
relationship. One variable they investigated was the
"maladjustment" of the subject. The maladjustment variable
was defined as the discrepancy between the self concept and
44
the ideal self concept. The reasoning behind using this
variable was:
Under unreasonable frustration, one would assume that
the well-adjusted person would make some constructive
attempt to alter the situation whereas he would more
readily accept or submit to frustration when it is per
ceived as reasonable, necessary, or nonarbitrary. The
"maladjusted" person, however, would resort to more
destructive, immature, disorganized, and regressive
modes of behavior regardless of the nature of the
frustration. In a sense he does not discriminate
arbitrary from nonarbitrary frustration but rather
responds to frustration per se. (Rothaus and Worschel,
1960)
After subjects were divided into high and low groups
on self-ideal discrepancy they were shown the stories with
the justified and unjustified frustrating situations. Sub
jects with high self-ideal discrepancies responded with
more hostility than did subjects who were more well
adjusted (low S-I discrepancy).
Rothaus and Worschel (1960) also investigated the
effect of sex of subject on aggressive responses to
justified and unjustified frustration. The important
finding here is that females showed as much hostile feeling
as males but were more inhibited in expressing this
hostility in overt behavior.
The authors attribute the above difference to the
effect of culture:
Culture undoubtedly influences the aggressive responses
to frustration. Aggression identifies the masculine
role, while passivity is characteristic of the feminine
role. Experimental evidence seems to support the
obvious; namely that males usually display more aggres
sion than females. (Rothaus and Worschel, 1960)
45
But as the results indicate, females seem to develop as
much hostile feeling as males although they don't express
it in their behavior.
An experiment by Berkowitz and Rawlings (in Berko-
witz, 1962) demonstrates the differences in reactions to
justified and unjustified aggression. Their subjects, both
males and females, were to give their reactions to filmed
violence. They were shown a movie about a prize-fighter.
Preceeding the movie, the subjects were given a synopsis
of the movie. For half the subjects the protagonist was
depicted so as to make the beating he received seem justi
fied. For the other half of the subjects the protagonist
was described in a sympathetic manner so as to make his
beating seem unjustified.
The findings indicated that media violence is most
likely to strengthen the audience's own hostile tendencies
when the make-believe aggression seems justified. The sub
jects expressed the greatest hostility toward the test
examiner after they had seen an "evil" person receive a
deserved beating. This was interpreted to mean that when
fantasy violence can be justified or sanctioned by the
audience it evidently heightens the probability of overtly
hostile actions by the audience.
A likely explanation offered by the authors is that
the sight of the ethically unwarranted aggression may
have aroused aggression anxiety, or perhaps made the
audience aware of moral standards of conduct. Now
46
conscious of "right" and "wrong" behavior, the social
ized observer presumably felt the need to behave in a
socially proper fashion. In other words, scenes of
unjustified hostility decreased the likelihood of
hostile actions by the audience. (Berkowitz and
Rawlings; in Berkowitz, 1962)
Unjustified aggression apparently makes persons aware of
their moral obligations concerning right and wrong. And it
seems that the unjustified aggression is r e.j ected as a
model for behavior whereas the justified aggression is
incorporated and used as a model for behavior. Vhen
encountering unjustified aggression, subjects are aware
that it is wrong because its perception activates their
moral standards, and consistent with this moral judgement,
it is rejected or deemed improper.
Moral Judgment
In the previous section on justice, moral judgments
were discussed largely in philosophical fashion. The
emphasis was on their contribution to the establishment of
the justness or unjustness of a situation. In this section
studies on the making of moral judgments about others'
behavior are reviewed. Factors affecting moral judgments
lie within the general areas of: the relationships within
the situation in which the behavior occurs; the character
istics of the person judging the situations; and the
characteristics of the person being evaluated in the
situation.
The largest amount of research has been done
47
involving the relationship between the subject or person
making the moral judgment and the object-person of that
moral judgment. The foremost factor in the above relation
ship is the sex of the subject versus the sex of the object-
person. Kinder (1925) demonstrated that females ascribe
more superior qualities to other women than they do to men.
Kinder's study involved women rating the "average college
women" and the "average college man" on various social
traits describing everyday behavior. The women consistently
rated the "average college woman" as possessing superior
qualities as compared to their ratings of the "average
college man."
Kitay (1940) used an attitude test consisting of 90
items dealing with the role and value of women in society.
All the items compared women to men; some items were
favorable toward women, others were unfavorable toward
women. The results indicated that more males than females
accepted negative statements about women; and more females
than males accepted positive statements about women.
A study by Sherriffs and Jarrett (1953) indicated,
contrary to the above studies, that males and females show
surprising agreement with respect to the behaviors and
characteristics which they attribute to males and females
and to the values they place on these qualities. Subjects
were given a list of 58 traits, 29 believed to be positive
and 29 believed to be negative. They were asked to indicate
48
to which group, male or female, the trait was most appro
priate. Then, on another occasion, the same subjects indi
cated the importance that they attached to these traits.
The results show that males and females both have highly
similar stereotypical attitudes about both men and women.
Skaggs (1940) did a study in which male and female
subjects made moral judgments about various social
behaviors. Twenty social acts or situations were used;
they were not specific as to the sex of the person involved
in the act. These acts were rated on an 11-point scale.
When ratings were compared, there was a high degree of
agreement between the male and female subjects.
A study by Crissman (1942) suggests that women make
harsher moral judgments than men. The study consisted of a
survey in which respondents were asked to evaluate 50 acts
or situations in terms of rightness or wrongness. The sur
vey was done in 1929 and again in 1939— both times with
college students. Of particular interest is the analysis of
the 1939 data by sex of respondent. In all, 40 items were
rated more harshly by women than by men. Of these, 24 were
statistically significant.
DiVesta and Bossart (1958) investigated sex differ
ences in attitude toward an ethical situation and the
extent to which such attitudes could be mofied by the mani
pulation of the labels used in the descriptions of the
situations. Each subject was presented with a short
49
paragraph about a family living in a low rent housing
development. Their income had to be below a certain level
for them to be eligible to live there. The family in ques
tion had to start making more than this amount to save for
their son’s college education. But they had to keep this
extra income secret, since if they reported the extra income
they would have to live somewhere else and the increase in
rent would absorb their extra earnings.
In one condition this problem was presented as an
ethical dilemma. In another condition it was described as
a case in which social judgments are dominant; and in the
last condition, a situation involving the management of
one’s financial affairs. Subjects were asked to indicate
their approval or disapproval of the father’s actions.
The results indicate that females were more negative
in their judgments than were males. With an ethical set,
the attitude responses were more negative than those induced
by either of the other two conditions for both males and
f emales.
Klinger, Albaum, and Hetherington (1964) designed a
study to make more clear some of the factors operating in
DiVesta and Bossart’s study. Klinger et al.'s position was
that the generality of sex differences cannot be considered
established unless both the sex of the subjects and the sex
of the story character are specified and varied.
With this in mind they generated four forms of
50
DiVesta and Bossart's story. In half the stories the
characters were a mother and her son and in the other half
they were a father and his son. The labeling condition was
also replicated— either the "ethical" or "economic" labels
were used.
An additional condition was the administering of the
A1lport—Vernon—Lindzey Scale of Values to all subjects. The
reasoning for this was that a person's judgments of a situa
tion will depend in part on his or her interpretation of the
situation. Subjects who have different dominant value
systems should not be expected to perceive and judge a
situation in the same way.
The results indicate that females are not more
severe in their judgment than males. And that such findings
in past studies might have been due to the sameness or
oppositeness of sex of subject and the judged object-person.
Males were more lenient in judging males than females. And
females were more lenient in judging females than males.
Also, subjects' values influenced their moral judg
ments of the story characters. Subjects with high religious
values judged the action of the story character most criti
cally. Subjects who scored high on political values judged
the story character's actions most favorably. The labeling
of the situation as either ethical or economic had no sig
nificant effect on the judgments.
Thus it seems that there are basic similarities in
51
moral judgments made by both males and females. But differ
ences between males and females occur depending on whether
the evaluated person is of the same or opposite sex as the
person doing the evaluating. Also, the significance of
values in determining moral judgments suggests that where
people have common values their moral judgments are more
similar than they would be if their basic values were
different.
CHAPTER III
METHOD
Materials
The materials used in this study were four stories,
each written four different ways (four versions). The
stories and versions of each story were especially con
structed to meet the requirements for this study. It was
necessary to create "personalities" for the characters in
the stories. It is the perception of these characters that
act as the independent variables.
Each story has two characters, an agent (aggressor—
makes the final remark or punch line) and a recipient
(victim— the butt of the joke). The perception of the
story characters should lead to more complex perceptions,
such as justice, hostility, and balance that are crucial in
explaining the resulting humor ratings. Because of the
demands on the materials great care was taken in writing
and presenting them, as indicated below.
Initially, jokes in written form were examined—
including those used in other studies (Williams and Cole,
1964) and those found in popular magazines such as Post,
Look. and Reader's Digest. The primary criteria for
52
53
selection were that the humor or intent of the joke be some
what hostile in nature; that each joke contain two main
characters; that both characters be human and of the same
sex; that the basis of the humor not be sexual in nature
(off color); and that the character descriptions of the two
story characters be interchangeable and capable of being
switched one with another without changing the sense or
essential meaning of the joke. After sifting through hun
dreds of likely jokes a dozen good prospects were selected.
The task now was to expand these short jokes into "situa
tions" or stories by describing each of the two main charac
ters more fully and presenting a small bit of interaction
between the two characters personifying their basic charac
ters along the dimensions of interest in the study.
There was a particular method selected for making
the story characters, A and/or R, appear favorable or
unfavorable. This was accomplished by describing the story
characters' behavior as being socially and morally
acceptable or unacceptable. The rules of etiquette (Hadida,
1952; Post, 1955; Roosevelt, 1962; Vanderbilt, 1963) were
used as a standard. It was felt that most persons brought
up in the mass American middle class culture should be aware
of what constitutes good and bad social behavior even though
they might not practice it (Cohen, 1955; Klinger, et al..
1964), and further, that they would evaluate persons as
good or bad on this basis if other information about the
54
story characters was lacking. The characters were varied
on this basis and during the pre-testing a check was made
to see if these initial assumptions were correct.
In actually writing the versions of any one story
the descriptions used for the agent were interchanged with
the descriptions used for the recipient. For versions 1
(A+ R-) and 4 (A- R+) the same working was used in both
stories but the descriptions were reversed. The story line
remained the same but all actions used to create a favorable
impression of the agent in version 1 were then used to
create a favorable impression of the recipient in version 4.
The same was true for the description of the recipient in
version 1 and the agent in version 4. The only thing that
does not change about the two rolee is that the recipient
in each case gives the "straight line" and the agent in each
case provides the "punch line" regardless of the preceeding
behavior.
In versions 2 (A- R-) and 3 (A+ R+) the roles were
rewritten to resemble the original roles as closely as
possible. This was done to avoid the oddity of having two
story characters whose actions were described with exactly
the same words.
In two of the stories (all 4 versions of each) both
the story characters are males and in the other two they
are females. However, none of the stories makes any illu
sion to sexual-type humor. Following is a figure depicting
55
the arrangement of the materials used in this experiment:
Stories
A
(Poor Bridge
Player)
B
(Party)
C
(Restaurant)
D
(Law
School)
1
Justice-
balance
2
Justiee-
no balance
female story mal e story
3
No justice-
no balance
charac-^ers characters
No justice-
balance
Fig. 7.— Schematic representation of experimental
materials showing stories, versions, and sex of story-
character s .
There is another set of materials associated with
this study. This material consists of abstract statements
describing the essential characteristics of the four ver
sions. This material was modeled after Jordan's (1953)
study. He used abstract statements to describe triadic
situations. His method was followed in the present study
in order to avoid putting any unnecessary content into the
situations. The stories and their versions provide com
plete illustrations of the principles of balance and
justice in action; however, there is the possibility of
56
these principles "getting lost" in the story-version com
binations. Subjects could respond to other aspects of the
stories, i.e.. the technique involved in creating the
humor— which might supersede what is of experimental
interest. Therefore, in order to provide a clear test of
justice and balance the basic relationships have been
described by which justice and balance are determined.
In Jordan's study, interest was in the balance of
the relationships between three persons. For example, "I
dislike 0; I have a sort of bond or relationship with X;
0 likes X" (Heider, 1958: 204). Jordan's subjects were
asked to rate the degree of pleasantness associated with a
number of situations described in this manner.
In this study the essential characteristics of each
of the four versions include: the agent's actions toward
the recipient in terms of social acceptability; the recip
ient's actions toward the agent in terms of social accepta
bility; and then the fact that the agent expresses hostility
toward the recipient. Each subject, after seeing and
responding to the story-version combinations, was shown
these materials in counterbalanced order and asked to rank
order them on the degree of perceived pleasantness of each.
The materials are presented in the Appendix.
Fre—testing
There were a number of things concerning the
materials that had to be determined before actual testing
57
began. The materials had to fulfill the conditions neces
sary for testing the hypotheses. The four versions had to
create variations in the amount of justice perceived in the
stories; the stories had to be equally humorous within
reasonable limits; the story characters had to be perceived
in specified ways as was indicated in the previous descrip
tions of the versions. To this end the pre-testing pro
ceeded along a schedule, starting with the most basic
requirements and proceeding to more refined criteria.
The first concern was whether or not the stories
were humorous at all— and if so, could stories that were
"nearly equally" humorous be found? To determine this,
version 1 (A+ R-) was written for twelve stories. These
stories were then shown to a small sample of male and
female college students.
At this point the eight best stories were selected—
four with male story characters and four with female story
characters. All four versions were written for each of
these stories. These story-version combinations were then
presented to subjects. Each subject saw the same version
(1, 2, 3, or 4) of all stories. All pre-testing from this
point on was done in this manner because the actual testing
would be done so that each subject saw four different ver
sions and four different stories (see Figure 8).
Next, working with the two stories with male, and
two stories with female story characters that most closely
58
Pre-Testing Design
Stories
A B C D
^ Group 1 y
2 Group 2 y
Vers ions
3 Group 3__^
4 Group 4__^
Actual
Experimental Design
Stories
A B C D
1
2
3
4
Fig. 8.— Schematic representation of pre-testing
design and actual experimental design.
59
fit the requirements, two final pre-tests were given. The
first of these involved presenting subjects with incomplete
stories. In these incomplete stories the endings were
omitted— that is, the final remark or punch line. By so
doing the point of the story was left undetermined and the
necessary similarities between versions could be studied.
The goal of this phase of pre-testing was to find out if
the role-reversal between the agent and recipient in the
stories produced the desired similarities. For example,
in versions 1 (A+ R-) and 4 (A- R+), the character descrip
tions given the agent in one version is the same description
given the recipient in the other version and vice versa.
Therefore, the characters should be the same in both
versions: agent in version 1 = recipient in version 4;
recipient in version 1 = agent in version 4. These charac
ters were perceived as being higher similar (adjective
checklist ratings of each story character) in spite of the
fact that in one case the character was an agent or
aggressor and in the other case a recipient or victim.
The second pre-test was concerned more directly
with whether the story-version combinations met the require
ments for the study. Results indicated that the stories
were moderately humorous. Also, there were differences in
the amount of perceived justice for the versions of the
stories. It was felt at this point that the stories ful
filled the basic requirements for the hypotheses.
60
Subjects
One hundred and ninety-two subjects participated in
the experiment. Of these, one half (96) were males and
\
one-half were females. The subjects were college students
at the University of Southern California and the California
State College at Fullerton. There were no special require
ments for participation in the study. All subjects except
a small percentage at Fullerton were volunteers who
received credit for an introductory psychology course for
their participation.
Presentation of Material
to Subjects
The stories and accompanying questionnaires were
presented to the subjects in booklet form. There were four
separate sections, each stapled together separately and
then all clipped together. Each story was stapled together
with its questionnaire— with the story on top. The order
of the stories presented to the subjects was varied
statistically.
The subjects were instructed to read all four
stories before answering any questions. Before reading the
stories the subjects were told something about the experi
ment. The instructions (reproduced in full in the Appen
dix) indicated that there were four stories and that in each
story there were two main story characters. Subjects were
61
asked to pay close attention to the behavior of the persons
in the stories. They were cautioned not to be too analyt
ical and to take a straightforward approach in reading the
stories. It was impressed on them that this was not a
"tricky" experiment and that all would be explained to them
upon completion of the experiment.
These cautions were deemed necessary from the prob
lems encountered in pre-testing. Subjects would consis
tently try to outguess the experimenter and read "between
the lines" to get at what they thought was the "real"
meaning of the stories. The subjects were told that the
experiment was not a test of their "sense of humor" but
rather that their opinions about the materials were being
sought. It was emphasized that there were no "right" or
"wrong" answers.
After being read the instructions aloud, question
naires were passed out to the subjects. They were told to
read all the stories through before beginning. The first
thing they were to do was to rank the stories according to
how humorous they thought they were. Each story was to be
given a rank of 1, 2, 3, or 4. In ranking the stories they
were advised to think of the stories as situations and base
their rankings on the entire story. This was done to avoid
the tendency of subjects to treat them as jokes and just
pay attention to the last line and ignore the rest of the
material.
62
After ranking the stories they were instructed to
proceed, one story at a time, and fill out the accompanying
questionnaire. The questionnaires were the same for each
story with the exception of the names of the story charac
ters referred to in the questions.
Aside from ranking the stories for humor, subjects
were asked to rate each story for humor. The other ques
tions concerned the subject’s perception of the story
characters and his perception and evaluation of the justi
fication and hostility of the final remark in the story.
They were given a free response question concerning the
basis for their rating of the justification in the story.
They were asked to respond in terms of the events in the
story.
After the subject had seen all four stories he was
given a single sheet on which were written the four
abstract situations. He was instructed to rank these
situations according to how pleasant he thought he would be.
Most subjects were given the experiment in small
groups. Two groups of about 25 were included— the rest of
the groups varied from three to nine subjects. Sometimes
the groups were all females and sometimes all males but
most groups were mixed. Each subject sat at an individual
desk in a classroom. The running time for the experiment
varied between 20 minutes and 50 minutes with most subjects
taking about 35 minutes.
63
Counterbalancing
There are sixteen story-version combinations (see
Figure 9 below). Each subject sees only four story-version
combinations. However, certain constraints were followed
in the selection of story-version combinations for any one
subj ect.
1
2
Vers ions
3
4
Figure 9.— Matrix of stories and versions of each
story.
It was necessary that each subject be exposed to
four different stories and four different versions. If a
subject is shown a story from the first row (version 1),
there is a choice of four stories. Moving to the second
row (version 2), the choice is reduced to three stories—
that is, excluding the story chosen in the first row.
Similarly, in the third row (version 3) there is a choice
of two stories, and in the bottom row (version 4), a choice
of one story. In this manner, one can choose four stories
and four versions with no duplications. In the first row,
four choices; in the second row, three choices; in the
Stories
A B C D
1A 1 B 1C 1D
2A 2B 2C 2D
3A 3B 3C 3D
4A 4B 4C 4D
64
in the third row, two choices; and in the fourth row, only
one choice.
4 x 3 x 2 x 1 =24 choices
In other words, there are 24 ways available to
choose four stories and versions to show each subject.
Because the stories, as vehicles, play such an important
part in the study and because the basic data involve each
subject rating four story-version combinations it was
thought best to use all possible combinations of stories
and versions.
By proceeding in this manner, we have 24 groups
of subjects; each group seeing the same set of stimuli.
With each of these 24 groups the order of presentation of
the stimuli was varied systematically. Each of the 24
groups contained eight subjects; four males and four
females. (24 x 4 = 96 males; 24 x 4 = 96 females; and
96 + 96 = 192— total number of subjects.) The counter
balancing scheme is presented in Appendix E.
CHAPTER IV
RESULTS
The purpose of this study was to determine what
effect justification for hostility has on humor apprecia
tion. The independent variable was the way in which the
stories were manipulated. These manipulations, described
previously, resulted in four versions for each story.
Before discussing the results pertaining directly to the
hypotheses, the way in which the versions were perceived
will be examined. Any analysis based on versions will
depend for its relevance on the degree to which the per
ception of the versions fits the necessary conditions for
justice and balance.
Manipulation Effectiveness
The stories each had two characters, an agent and a
recipient. The subjects, as part of their task, rated the
story characters on the basis of whether or not their
behavior was socially acceptable or not. The results are
shown below. Any rating above zero (+) can be considered
as indicating socially acceptable behavior; any number
below zero (-) indicates socially unacceptable behavior.
65
66
TABLE 1
MEAN SOCIAL ACCEPTABILITY RATINGS OF THE AGENT
AND RECIPIENT BY VERSIONS
Versions
Mean
Agent
Rating
Recipient
1 +0.86 -1 . 59
2 -1 .00 -1 .05
3 -0.3 5 +1 .40
4 -1 . 56 + 1 .46
All except the agent in version 3 conform to the
desired results (see page 8). Version 3 is the most
unstable version and the possibility of its being perceived
in other ways than desired was discussed earlier. In ver
sion 3 an agent who has behaved in a socially desirable way
directs a "hostile" remark to a recipient who has also
behaved incorrectly. It seems that the nature of the final
remark influenced the perception of the agent. This mean
rating conceals the fact that there were some respondents
who did perceive the version 3 agent as behaving in a
socially desirable manner.
If the data from Table 1 are organized as shown
below (Table 2) one can see the extent to which the versions
were perceived as intended. The versions are a function of
the manipulations of the agent and recipient. The manipu
lations consisted of varying the behavior of both story
characters so that it was either good (socially approved
67
behavior) or bad (socially unapproved behavior). As can
be seen in the table, the lowest percentage of correct
perceptions occurred when both characters were supposed to
be good. From Table 1 we see that this was due to the
agent being perceived as a bad person.
TABLE 2
PERCENTAGE OF SUBJECTS WHO RATED THE SOCIAL
ACCEPTABILITY OF THE AGENT AND
RECIPIENT AS INTENDED
Agent Intended to Be
Seen As:
Good
Bad
Recipient Intended to Be
Seen As:
Good Bad
34# (V3) 74# (V1)
81# (V4) 69# (V2)
The basic hypothesis of this study was that humor
appreciation would be a function of justice and not related
to balance. In order for this to be a possibility there
must be differences between versions with respect to
j ustice.
The type of analysis used on this data deserves
some explanation before it is presented. A model had to be
found that would take into account the four stories, four
versions, and twenty-four counterbalancing groups. To do
this, six latin squares were used. Each square consisted
of four counterbalancing groups, four stories, and four
68
versions. It is possible to sum the variances from all
these latin squares to get the total sums of squares
(Conrad, 1958; Winer, 1962).
TABLE 3
SUMMARY OF ANALYSIS OF VARIANCE ON PERCEIVED JUSTICE
RATINGS FOR STORIES AND VERSIONS
Source df
Sum of
Squares
Mean
Square
F
Between subjects 485.3
Squares 5 4.6 .92 .375
Groups within squares 18 69.6 3.87 1 .579
Subjects within groups 168 411 .1 2.45
Within subjects 1931.2
Stories 3 127.7 42.57 22.76 ;
Versions 665.9 221.97 118.71 '
Stories x squares 1 5 26.6
1 .77 .946
Versions x squares 1 5 43.6 2.91 1 .556
Residual 36 122.9 3.41 1 .823
Error (within) 504 944. 5 1 .87
aP < .01
As can be seen, the main effects of stories and
versions are significant. It was not possible to obtain
stories that do not differ or that subjects would react to
in the same way. However, these story differences do not
affect the results (any story effects were balanced out by
the design).
A Duncan multiple range test was done on the mean
justice ratings for each version. The means and signifi
cant groupings are presented in Table 4.
69
TABLE 4
MEAN RATINGS OF PERCEIVED JUSTICE BY VERSION
Version Mean Justice Rating
1 +1 .11a
2
1
o
•
o
00
3 -0.61
4 -1 .45
All means differed from each other at the ,05
level.
The justice ratings are all significantly different
from one another. They are in the predicted direction, but
only V1 is rated as being just. Version 2 is at the neutral
point between justice and no justice. Evidently justice is
more than a function of the A - R, 0 - R relation (see
page 6). This is a necessary but not a sufficient condition
for justice. A person’s definition of justice seems to
include all relations in the system. Justice occurred in a
particular subset of balanced situations in which the per
son who had behaved in a socially acceptable manner
directed a negative remark to a person who had behaved in a
socially unacceptable manner. If the agent is ’ ’bad" the
situation as a whole is not justified— neither is it unjust.
The same design was used in analyzing the ratings
of hostility. Hostility ratings relate in particular to
the final remark made by the agent (the punchline). The
70
main effects of versions and stories were both significant
at the .05 level. The analysis is presented below.
TABLE 5
SUMMARY OF ANALYSIS OF VARIANCE ON PERCEIVED HOSTILITY
RATINGS FOR STORIES AND VERSIONS
Source df
Sums of
Squares
Mean
Square
F
Between sub.iects 207.3
Squares 5 13.7 2.74 2.61
Groups within squares 18 16.3 .91 .867
Subjects within groups 168 177.3 1 .05
Within subjects 491 .8
Stories 3 28.8 9.60 13.521a
Versions 3 32. 5 10.83 15.253
Stories x squares 1 5 11.1 .74 1.042
Versions x squares 1 5 14.4 .96 1 .352
Residual 36 47.1 1 .31 1 .845
Error (within) 504 3 57.9 .71
Significant at .01 level
TABLE 6
MEAN RATINGS OF PERCEIVED HOSTILITY BY VERSION
Version Mean Hostility Rating
3 2.49a (low hostility)
1 2.261
2 2.14J
4 1.93 (high hostility)
^eans within brackets do not differ at the .
05
1evel.
71
Of all possible differences between versions only
the difference between versions 1 and 2 was not signifi
cant. Judgments of hostility seem to depend on the combined
perceptions of the agent and recipient. This is not the
opposite of justice, however. It seems that when the agent
and recipient both behave in a socially acceptable manner
the remark will be judged least hostile. None of the ver
sions quite fit this pattern but version 3 comes closest
and is the least hostile. It will be noted that in version
4 the recipient is equally favorably perceived to the ver
sion 3 recipient, but the agent's behavior has not been
socially acceptable. This version is the most hostile.
And in version 1 the agent is good but the recipient is
bad. This version is significantly more hostile than
version 3 and not any less hostile than version 2, in which
both A and R are negatively perceived.
Major Hypothesis
Now that some initial findings about justice and
hostility have been presented, data basic to the main
hypothesis can be shown. The first task of the subjects
was to rank the story-version combinations on the basis of
how humorous they thought they were. Because the data are
in the form of ranks, the chi-square statistic is used
instead of the F-test. This type of test is referred to as
the Friedman test (Siegel, 1958; Winer, 1962).
72
TABLE 7
SUMMARY OF ANALYSIS OF VARIANCE ON HUMOR RATINGS FOR
STORY-VERSIONS
Source df
Sum of v 2 D .
c Ranks
Squares
Versions 3 30.2 18.8a
Res idual 573 893.88
Within people 576 924.0
£L
Significant at the .01 level.
As can be seen, the main effects of versions are
significant. Multiple comparisons between mean humor
rankings are shown below in Table 8.
TABLE 8
MEAN HUMOR RANKINGS BY VERSION
Version Mean Humor Ranking
1 2.23~|a
3 [2.43J
2 |_2. 57
4 2.77
eans within brackets do not differ at the .05
1evel.
Version 1 is more humorous (lower ranking) than
versions 2 and 4. Also, version 3 is more humorous than
version 4. However, version 3 is not less humorous than
73
version 2 or version 1.
Analysis of humor ratings indicate the same result.
Humor ratings were obtained after the initial rankings were
made.
TABLE 9
SUMMARY OF ANALYSIS OF VARIANCE ON HUMOR
FOR STORIES AND VERSIONS
RATINGS
Source
Sum of
Squares
Mean p
Square
Between subjects 976.7
Squares 5 43.1 8.64 1 .789
Groups within squares 18 121.6 6.75 1 .397
Subjects within groups 168 811 .9 4.83
Within subjects 2630.0
Stories 3
267.2
89.07 22.16 a
Versions 3 100.8 33.60 8.36 a
Stories x squares 1 5
30.8
2.05 . 510
Versions x squares 1 5 54.0 3.60 .895
Residual 36 1 52.8 4.24 1 .054
Error within 504 1024.4 4.02
0/
Significant at .01 level
TABLE 10
MEAN HUMOR RATINGS BY VERSION
Version Mean Humor Ratings
1 4.31T
3 “3.92J
2 _3.67l
4 3.32J
*M eans within brackets do not differ at the .05 level
74
That version 1 is more humorous than version 4 is
basic to the main hypothesis. The hypothesis was that
humor ratings would not be a function of balance but rather
of justification for hostility. With respect to balance,
the hypothesis is established. Version 1, which is
balanced, is more humorous than version 4, which is also
balanced. However, tying the results to the effects of
justice has been confounded by the fact that versions 1
and 4 also differ in the hostility attributed to the final
remark. At this point we cannot unconfound them, but later
results will point to a joint effect of the two as deter
mining the results. This will also explain why version 3
is more humorous than was hypothesized.
Let us now compare humor ratings directly with the
subject's own ratings of the justice of the story. Analysis
by version was based on the justice differences that exist
between versions— this, however, is a direct comparison.
¥e will use humor ratings instead of the rankings for the
remaining analyses because they are more amenable to
statistical manipulation. The analysis is presented in
Table 11. The analysis is cast in the form of a one-way
analysis of variance with unequal cell frequencies (Winer,
1962).
As shown in Table 11, humor ratings do vary sig
nificantly with the perception of justice. However, when
we look at the comparison between mean humor ratings for
75
TABLE 11
SUMMARY OF ANALYSIS OF VARIANCE ON HUMOR RATINGS
FOR PERCEIVED JUSTICE RATING
Source df
Sum of
Squares
Mean p
Square
Justice rating
^category
5 284.9 56.98 I3.07a
Error 762 3321.8 4.36
£ L
Significant at .01 level.
TABLE 12
MEAN HUMOR RATINGS BY JUSTICE
RATING CATEGORIES
Justice Rating Number of Mean Humor
Category Responses Ratings
-3 20 5 1 .83
-2
49 “3.87
-1 102
4.05
+1 134 4.09
+2
125 _4.24
+3 103 4. 53
eans within brackets do not differ at .05 level.
76
different degrees of justice (Table 12), we see that it is
only at the lower extremes that humor ratings vary with
justice. This finding supports our initial hypothesis that
justice is a pre-condition for mirth. Its absence will
work against mirth but the presence of justice will not
cause mirth but does help at the extremes.
Next, because of the hostility differences between
versions we will look at humor ratings in direct relation to
hostility ratings. The analysis is presented below.
TABLE 13
SUMMARY OF ANALYSIS OF VARIANCE ON HUMOR RATINGS
FOR PERCEIVED HOSTILITY RATING
df
Sum of
Squares
Mean j,
Square
Hostility category 3 80.29 26.76 5•779a
Error 762
3526.41 4.63
Significant at the .01 level.
TABLE 14
MEAN HUMOR RATINGS BY HOSTILITY
RATING CATEGORIES
Hostility Rating Number of Mean Humor
Category Responses Ratings
1 226 3.31a
2 260 3.99“
3 186 4.05
4 96 4.00
^eans within brackets do not differ at .05 level.
77
As with justice we see that humor ratings are only
affected at the extremes. When hostility is high, humor
ratings are low. Otherwise the degree of hostility has no
effect on humor ratings. It is interesting to note that
"no hostility" (category 4) is neither more nor less
humorous than "some" but not a great deal of hostility
(categories 2 and 3). Perhaps some stories in this study
were humorous because the subjects vicariously released
hostility in response to the hostility humorously expressed
in the stories. Or, as the above data indicate, some of
the stories were humorous even though there was no
(perceived) hostility expressed in them. This might be due
to technical aspects of humor— the play on words.
To look further into the relationship between
justice and hostility the next analysis concerns the direct
relationship between hostility and justice. It seems that
hostility is low when justice is high and conversely.
However, in the midranges there is no particular relation
ship. The analysis is presented below.
TABLE 15
SUMMARY OF ANALYSIS OF VARIANCE ON HOSTILITY
RATINGS FOR JUSTICE RATING
df
Sum of
Squares
Mean
Square
F
Justice category 5 60.3 5 12.07 13.12a
Error 762 705.57
.92
aSignif i cant at the .01 level.
78
TABLE 16
MEAN HOSTILITY RATINGS FOR JUSTICE
RATING CATEGORIES
Justice Rating Number of Hostility
Category Responses Ratings
-3 205 1.79a
-2
99
I
c r v
o
•
CM
-1 102 “2.32
+1 134 2.33_
+2 126 _2.37 |
+3 103 2.61_I
% eans within brackets do not differ at .05 level.
Because of this relationship (and a correlation
between perceived justice and perceived hostility of .268)
between justice and hostility it was thought that any
independence of justice from hostility ought to be
examined. To this end for each level of hostility, the
justice ratings were combined into two groups. One group,
justice, included those ratings that indicate any justice,
whether "slightly," "somewhat," or "very" (+1, +2, +3). The
other group includes all degrees of lack of justice (-3,
-2, -1). The humor ratings for each of these two groups at
each level of hostility were then obtained. The results
are presented in Table 17.
It can be clearly seen that when a story outcome is
perceived as just it is rated funnier than a story at the
same level of hostility that is not just. Also any story
79
that is just, no matter what the level of hostility is, is
rated funnier than any story that is not just.
TABLE 17
MEAN HUMOR RATINGS BY PERCEIVED HOSTILITY OF
PUNCH LINE AND JUSTICE OF
RECIPIENT'S OUTCOME
Hostility
Justice
High Low
1 High 3.96 3.14
2 4. 50 3.62
3 4.35 3.67
4 Low 4.20 3.33
When the process is reversed and humor ratings for
hostility at given levels of justice are examined, then
results are less clear. However, they still support pre
vious data indicating some independent effects of hostility
on mirth.
The lowest humor ratings are associated with the
highest and lowest hostility conditions. This throws some
additional light on a point raised earlier in this study
about the "no hostility" condition being as humorous as the
"some" hostility condition. When the effects of justice
are partialled out, the "no hostility" condition is less
humorous relative to the other condition than it was pre
viously thought to be. Therefore, when there is no
hostility, mirth seems to be less than when there is some
80
hostility, but equal in mirth to a situation where there is ;
a great deal of hostility.
The multiple correlation of justice and hostility
with humor ratings was obtained. The multiple correlation
is rather low: .263. Justice contributes .258 and hostility
the remainder.
Analysis Based Directly on Perceptions of
Agent and Recipient
To this point we have tried to demonstrate the
relationship between justice, balance, and mirth, dealing
mainly with the versions. The versions were somewhat less
than complete representations of the conditions we were
trying to create in this study. Nevertheless, within the
limitations of the variability of perception it is thought
that the versions, with the exception of version 3, were
quite representative of the desired conditions. The results
based on these data supported in part the hypothesis and in
part left the picture somewhat clouded. In an effort to
make more clear the dynamics involved in this study, a
2 x 2 contingency table was formed. In so doing, the
variables used were the ratings of social acceptability or
unacceptability of the agent and recipient. It is the
ratings of these variables that are the basis for our
operational definitions of justice and balance. To restrict
the cells of the table to combinations of justice/no
justice and balance/no balance, we have omitted all sets of
81
ratings in which the subject rated hostility as a 4 (no
hostility). By so doing, all agent-recipient relations are
given a minus sign (see page 6).
TABLE 18
MEAN HUMOR RATINGS BY PERCEPTION OF THE
AGENT AND RECIPIENT
Good (+)
Agent Perceived As:
Bad (-)
Recipient Perceived As:
Good {+) Bad (-)
4.36 4.31
N=69 N=1 51
3.31 3.78
N=262 N=190
As Table 18 shows, an A+ R+ version is as funny as
an A+ R- version.
TABLE 19
ANALYSIS OF VARIANCE OF HUMOR RATINGS BY PERCEPTION
OF THE AGENT AND RECIPIENT
df
Sum of
Squares
Mean j,
Square
Recipient 1 11.88 11.88 2.44
Agent 1 85.80 85.88 19.28a
Agent x recipient 1 3.96 3.96
Error 668 2977.00 4.45
Or
Significant at the .01 level.
82
The analysis of variance of these data shows a
significant agent effect. Humor ratings are higher when
the agent is good than when he is bad. There is no sig
nificant difference in humor ratings for the perception of
the recipient. This is an important finding for this study.
It was thought that mirth would vary only with the percep
tion of the recipient because of his key role in the deter
mination of justice. This finding can be explained, in
part, by the justice and hostility ratings for the corre
sponding cells of Table 20.
TABLE 20
RATINGS OF PERCEIVED JUSTICE BY PERCEPTION OF
THE AGENT AND THE RECIPIENT
Good (+)
Agent Perceived As:
Bad (-)
In Table 20, the justice ratings for the two cells
in which the recipient is good are of immediate interest.
The cell in which the agent is also good, while not being
just, is much less unjust than when the agent is bad. Thus
it seems that the perception of the agent does affect
perceived justice. The same phenomenon is demonstrated in
the pair of cells in which the recipient is bad.
Recipient Perceived As:
Good (+) Bad (-)
-0. 52 + 1 .27
-1 . 68 -0.05
83
Also, it can be seen that only the A+ R- cell (V1)
is rated as just. In terms of hostility (Table 21), A+ R+
is the least hostile of the four conditions— but it is
slightly unjustified. So in the two cells, A+ R— and
A+ R+, we have respectively high justice and low hostility.
The other two cells have neither of these two attributes.
And it is the former two cells that have the highest
humor ratings.
TABLE 21
RATINGS OF PERCEIVED HOSTILITY BY PERCEPTION OF
THE AGENT AND THE RECIPIENT
Good (+)
Ag ent Perceived As:
Bad (-)
Summary of Major Hypo thesis
Summarizing the data relevant to the main hypothe
sis, humor ratings seem to be a function of justice and, to
a lesser extent, hostility. The perception of justice is
not only affected by the perception of the recipient but
also the perception of the agent.
Hostility seems to affect humor ratings only when
hostility is high, in which case humor ratings are low.
Humor ratings are slightly lower for "no hostility" ratings
Recipient Perceived As:
Good ( + ) Bad (-)
2.45 2.07
1 .79 1 .84
than for the midranges of hostility. But this is only when i
the effects of justice are controlled.
At each level of hostility rating, humor ratings !
are higher for justice than for no justice. And the humor
ratings are higher for perceived justice at any level of
hostility than when there is no perceived justice. The
basic finding is that version 1 (A+ R-) is more humorous
than version 4 (A- R+). It is in these conditions that
justice is highest and lowest. There are also hostility
differences but not as great as for justice.
Secondary Hypothesis
Now let us examine some data relevant to the
secondary hypothesis. It concerns sex differences in mirth
as a function of harshness of judgment of the actions of an
opposite-sexed story character. Vhen the subject and the
story characters are of opposite sex, stories which depict
the social relations as being just will be rated as more
humorous than when the subject and story characters are of
the same sex.
In testing this hypothesis, separate analyses of
variance were done on each story. The data for stories A
and C are presented in Table 22. Story A has female story
characters and story C has male story characters.
As can be seen, the main effects of versions are
significant but neither the main effects for sex nor the
sex x version interaction is significant.
TABLE 22
85
SUMMARY OF ANALYSIS OF VARIANCE ON HUMOR RATINGS FOR
STORY-VERSION AND SEX OF SUBJECT, STORIES A AND C
Source df
Sum of
Squares
Mean
Square
F
Story A
Vers ion 3 30.9 10.3 3.244a
Sex 1 .3 .3 .095
Vers ion-Sex 3 15.6 5.2 1 .638
Error (within cell) 1 84
Story C
584.2
3.175
Version 3 53.9 17.97 3.54 a
Sex 1 1 .7 1 .7 .334
Version-Sex 3 6.4 2.13 .419
Error (within cell) 184 934.8 5.08
E L
Significant at the .05 level.
TABLE 23
SUMMARY OF ANALYSIS OF VARIANCE ON HUMOR RATINGS FOR
STORY-VERSION AND SEX OF SUBJECT, STORIES B ANEI D
Source df
Sum of
Squares
Mean
Square
F
Story B
Version 3 22.73 7.58 1 .87
Sex 1 1 .73 1 .73 .428
Version-Sex 3 37.23 12.41 3.07 a
Error (within cell) 184
Story D
743.83 4.04
Version 3 37.5 12.5 3.12 a
Sex 1 .6 . 6 .149
Vers i on—Sex 3 46.0 1 5.33 3.82 a
Error (within cell) 184 738.0 4.01
E L
Significant at the .05 level.
The following two stories (Table 23, page 85) both |
have significant differences in humor ratings between sex
of subject within versions.
To support the hypothesis, the simple main effect
of sex at version 1 must be significant. And the male sub
jects should have higher humor ratings for story B and
female subjects have higher ratings for story D. The
analysis for simple main effects and accompanying figure
for story B is presented fri Table 24.
The simple main effect of sex at version 1 is not
significant and does not support the hypothesis. Males
find this version more humorous but not at a level that
reaches statistical significance. However, when we look at
version 4, we see a significant difference. However, no
logical reason can be found for such a difference within
the context of our hypothesis.
Vhen the simple main effects for story D are
examined, we find no significant differences at version 1
(just above the .05 level). At version 2 there is a sig
nificant difference but it does not fit into any logical
pattern and is in the opposite direction from sex differ
ences at version 1.
Given that the main effects of sex are not signifi
cant for any story and that the necessary sex-version simple
main effects are not significant, we have to conclude that
the secondary hypothesis is not supported by the data.
87
TABLE 24
SUMMARY OF ANALYSIS OF VARIANCE FOR SIMPLE MAIN EFFECTS OF
SEX OF SUBJECT AT VERSION LEVEL FOR HUMOR RATINGS
FOR STORY B
Source df
Sum of
Squares
Mean
Square
F
Simple effects of sex:
For version 1 1 6.8 6.8 1 .683
For version 2 1 4.2 4.2 1 .039
For version 3 1 6.7 6.7 1 .658
For version 4 1 21 .3 21 .3 5.272a
Error (within cell) 184 743.8 4.04
Of
Significant at .05 level
males
Mean Humor Rating
5.00
4.75
4. 50
4.25
4.00
3.75
3. 50
3.25
3.00
fenale
2 1 3 4
Version
Fig. 10.— Simple main effects of mean humor ratings
for story B for sex of subject at version.
88
TABLE 2 5
SUMMARY OF ANALYSIS OF VARIANCE FOR SIMPLE MAIN EFFECTS OF
SEX OF SUBJECT AT VERSION LEVEL FOR HUMOR RATINGS
FOR STORY D
Source df
Sum of
Squares
Mean
Square
F
Simple effects of sex;
For version 1 1 15.2 15.2 3.790
For version 2 1 18.7 18.7 4.629a
For version 3 1 3.5 3.5 .873
For version 4 1 9.2 9.2 2.294
Error (within cell) 184 738.4 4.01
Significant at .05 level.
females
Mean Humor Rating
5.75
5. 50
5.25
5.00
4.75
4. 50
4.25
4.00
3.75
3. 50
3.25
3 .00
ma.es
2 1 3 4
Fig. 11.— Simple main effects of mean humor ratings
for story D for sex of subject at version.
89
Furthermore, any differences in humor ratings would have to
be based on differences in perceived justice. In Table 26
the analyses of variance for the justice ratings by story
are presented. They are in the same form as the previous
tables (sex by version).
There are two significant differences within Table
26— one main effect in story C and one interaction effect
in story A. Of these the interaction effect is in agreement
with the hypothesis and the main effect is in the opposite
direction from that hypothesized.
Table 27 presents the corresponding hostility
ratings. Of all the stories there is only one significant
result (other than versions). The interaction effect is
significant for story B. It is in agreement with the
hypothesis in that males judged the female story character's
remark more hostile than the female subjects did.
There may be specific situational factors respon
sible for the few sex differences that exist. But overall
it has to be concluded that no significant pattern of sex
differences was found, either in humor ratings or in per
ceived justice or hostility.
There are two other sources of data that are
related to the subjects1 perception of the stories. One of
these is the free response question in which subjects were
asked to explain their justice ratings for each story in
terms of the events in the story.
90
TABLE 26
SUMMARY OF ANALYSIS OF VARIANCE ON PERCEIVED JUSTICE
FOR VERSION AND SEX OF SUBJECT,
STORIES A, B, C, D
Source df
Sum of Mean
F
Squares Square
X
Story A
Vers ion 3 283.1 94.37 54.236a
Sex 1 2.3 2.3 1.322
Version-Sex 3 14.2 4.73 2.71 b
Error (within cell) 184
Story
320.0
B
1 .74
Version 3 165.3 55.1 28.9 a
Sex 1 1 .3 1 .3 .684
Version-Sex 3 10.1 3.4 1 .789
Error (within cell) 1 84
Story
3 50.0
C
1 .9
Vers ion 3 210.8 70.2 35.6 a
Sex 1 5.7 5.7 2.89
Version-Sex 3 1 .1 .367 .186
Error (within cell) 184
Story
361 .6
D
1 .97
Version 3 108.7 36.12 18.3 a
Sex
1 .2 .2
.091
Version-Sex 3 12.9 4.3 1 .96
Error (within cell ) 184 405.0 2.20
Significant at .01 level.
^Significant at .05 level.
TABLE 27
91 ,
I
SUMMARY OF ANALYSIS OF VARIANCE ON PERCEIVED HOSTILITY
RATINGS FOR VERSION AND SEX OF SUBJECT
STORIES A, B, C, D
Source df
Sum of
Squares
Mean
Square
F
Story A
Version 3 30.0 10.0 11.94 a
Sex 1 .4 .4 .43 5
Vers i on—Sex 3 1 .8 .6 .652
Error (within cell) 184
Story B
169.3 .92
Vers i on 3 4.0 1 .33 1 .839
Sex 1 1 .3 1 .3 1 .798
Vers i on-Sex 3 12.1 4.03 5.7 a
Error (within cell) 184
Story C
133.0 .723
Vers i on 3 4.7 1 . 57 1 .42 5
Sex 1 .2 .2 .181
Vers ion—Sex 3 3.0 1 .00 .907
Error (within cell) 1 84
Story D
202.8 1 .102
Versi on 3 17.8 5.93 3.68 b
Sex 1 0 0 0
Version-Sex 3 .7 .23 .261
Error (within cell) 184 161.5 .88
Significant at the .01 level.
^Significant at the .05 level.
There are few instances which show females and males
differing in the ways in which they perceived the stories.
In stories A and B, both having female story characters, 1
females showed a tendency to disparage the agent for
answering unpleasantness with further unpleasantness. Males!
showed no such tendency. This is demonstrated in the
responses associated with versions 1 and 2 for each story.
There were no noteworthy differences for stories C and D.
All in all, the responses reflect the incidents in the
stories that were purposely built into them to create the
differences between versions. The tabulation of the free
response data is presented in Appendix F.
Abstract Situations
A final set of data pertaining to the stories is
the adjective checklist data. The correlations of the
adjective data (all stories and versions combined) for the
agent and recipient with humor ratings are shown in Table 28.
For the agent, high humor ratings are associated
with him or her being: witty, just, likable, friendly,
polite, and warm. Lower humor ratings more frequently occur
when the agent is cold, hostile, vain, and insincere. For
the recipient the highest correlations with humor ratings
are for: aggressive, critical, and witty— evidently good
qualities for being the butt of a joke.
The multiple correlation of the agent and adjective
checklist ratings with humor ratings is .483. However, most
93
TABLE 28
CORRELATIONS OF ADJECTIVE CHECKLIST RATINGS FOR AGENT
AND RECIPIENT WITH HUMOR RATINGS
Adj ective Agent Recipient
Friendly- .235 -.034
Just (good or moral) .239 -.025
Hostile -. 1 34 .059
Vain -.108 .103
Likable .230 -.037
Critical -.096 . 114
Polite .196 -.075
Warm .192 -.052
Aggres sive -.064 .165
Witty .450 .135
Insincere -.107 -.057
Cold -.137 -.089
TABLE 29
CORRELATIONS OF ADJECTIVE CHECKLIST RATINGS WITH "JUST"
RATINGS FOR AGENT AND RECIPIENT
Adj ective Agent Recipient
Friendly- .750 .607
Hostile -.300 -.328
Vain -.264 -.242
Likable
.745 .718
Critical
-.197 -.261
Polite .701 .711
Warm .699 .635
Aggressive
-.147 -.232
Witty
.425 .1 51
Insincere
-.247 -.255
Cold
-.329 -.347
94
all of this is due to vitty with a correlation coefficient
of .450. For the recipient ratings the multiple correla
tion with humor ratings is .223.
It is also interesting to see what adjectives
correlate highly with just. For both agent and recipient
they are: friendly, likable, polite, and warm. These
correlations are consistent with the manipulations used to
vary the perception of justice by way of the socially
approved or disapproved behavior of the story characters.
Results Pertaining to Abstract Situations
The final hypothesis for this study has to do with
the abstract descriptions of the versions. Because of the
possibility that the humor ratings of the story-versions
might not reflect the differences in justice and balance
incorporated in them, these four versions were described
in abstract terms. That is, just the bare outline of the
critical events were presented as stimuli (see Appendix D
for actual stimuli). Subjects were asked to rank the four
descriptions in terms of how pleasant they would find each
situation if they were present as an observer. It was
hypothesized that the rankings would vary, in the same way
as the story-version combinations, according to justice and
not balance. The analysis of variance summary is presented
in Table 30, along with the mean rankings for the descrip
tions, as shown in Table 31.
95
TABLE 30
SUMMARY OF ANALYSIS OF VARIANCE ON PLEASANTNESS RANKINGS
FOR ABSTRACT DESCRIPTIONS OF VERSIONS
Source df
Sum of -v 2 t, .
„ A Ranks
Squares
Versions 3 231.2 69.36a
Residual 185 729.8
Within people 288 960.0
Significant at the .01 level •
TABLE 31
MEAN PLEASANTNESS RANKINGS BY ABSTRACT DESCRIPTIONS
OF VERSIONS
Abstract Situation
Corresponding to
Versi on
Mean Pleasantness
Rankings
1
a
1 .71“]
3 r 2 .27_|
2 |_2.88"|
4 3.12_|
^eans within brackets do not differ at .05 level.
The results are the same as for the rankings of the
complete story-version combinations (tables 7 and 8).
Version 1 was found to be more pleasant than version 4.
Both versions are balanced but only version 1 is just.
And again, as in the previous results, version 3 is ranked
higher (more pleasant) than version 2. Thus it seems that
96 |
the pleasantness rankings demonstrated the same dependence
on justice and hostility as did the humor rankings and
ratings for the actual stories.
Perhaps "pleasantness" as a function of justice and
hostility is the determining factor that causes some ver
sions to be more humorous than others. A situation is most
pleasant when the hostility is justified. And if hostility
is justified it is not perceived as being "too hostile."
A situation is also pleasant if both characters behave in
a socially acceptable manner and if any hostility expressed
is "not very hostile." However, situations are unpleasant
if there is no justice and there is a medium to high degree
of perceived hostility. This description of the joint
effects of justice and hostility is consistent with all
the results in this study.
CHAPTER V
DISCUSSION
The results showed that a balanced situation in
which the social relations between story characters were
justified (version 1) was more humorous and pleasant than
a balanced situation in which there was no justice
(version 4). Justice, however, was related not only to
the "badness" of the recipient or victim but also to the
"goodness" of the agent or aggressor. Humor ratings were
higher when the agent was good than when he was bad.
Version 3, in which both the agent and recipient were good,
was as humorous and pleasant as version 1. Version 3 was
the least hostile of the four versions. And version 1 was
the only version that was justified.
Therefore, the above rankings seem to be due, at
least in part, to both justice and hostility. And the
combined effects of both should be related to "pleasantness"
because of the similarity of the two sets of rankings
(humor and pleasantness). Other researchers (Allison and
Hunt, 1959; Cohen, 1955; Lee, 1961; Pastori, 1952; Rothaus
and Vorchel, 1960) have found that unjustified acts lead
97
98 j
more frequently to aggression than do the same acts when
they seems justified. If this is true, presumably situa
tions in which acts are justified should be more pleasant
than when these acts do not appear justified. Applying
the same line of reasoning to the expression of hostility,
very hostile remarks should be less pleasant than less
hostile remarks.
This reasoning fits the data obtained in this
study. Versions 1 and 3 are respectively high in justice
and low in hostility. Versions 2 and 4 have neither of
these attributes. And the data indicate that versions 1
and 3 are ranked higher than 2 and 4 for both pleasantness
and humor.
Situational Effects on Mirth
The results of this study are consistent with
Freud's theorizing about situational effects on humor and
with other studies on humor (Elbert, 1962; LaFave, 1961)
that dealt with situational aspects of humor.
Freud (1938) has indicated that humor appreciation
should be greatest when the agent and the observer (third
person) are in agreement (psychic harmony) and possess the
same inhibitions. When the agent is perceived as a good
person by the observer and the recipient is judged to
deserve receiving a hostile remark, the observer should
agree with the agent's actions and, if not being in psychic
99
harmony, at least in agreement with the agent. Given these
conditions in this study, version 1 was the most humorous.
If the hostility is low and, therefore, the offense not too
serious, the observer should still approve more of the
agent's actions than if the agent were bad. The data are
in agreement with this position. Also, the two most
humorous versions were those in which the agent was liked.
This "liking" seems to affect the perception of justice
and hostility. The difference between versions 1 and 2 is
that in version 1 the agent is liked and in version 2 he is
not liked. Similarly, in version 3 the agent is good and
in version 4 he is bad.
These results are also consistent with the results
obtained by LaFave (1961) and Elbert (1962). In both of
these studies humor was appreciated more when either the
agent was liked and/or the recipient was disliked.
Balance
Insofar as balance is concerned, the results indi
cate that the principles of balance seem to be effective
in determining the perception of the versions and story
characters. But balance does not determine preference or
humor ratings (version 1 is more humorous than version 4).
This finding is consistent with Rosenberg and Abelson's
(1960) conclusion that when other factors operate beside
balance the most satisfying state may not be one that is
100 :
balanced but one in which all factors are maximized as much
as possible. This may or may not result in a balanced
state.
However, one aspect of balance theory is certainly
demonstrated by the results. Heider (1967) reiterates that ;
the special significance of work on balance is ". . . that
it shows how often the total set of propositions rather
than any single item is involved in determining the pre
ference for a triad . . , ." Our results cannot be
adequately explained using any one or any two of the three
relations in our triads. This is why our initial position
that justice is a simple function of the A - R, 0 - R
relations proved incorrect. The situations were only seen
as justified when the algebraic signs of the A - R, 0 - R
relations were the same and the agent was rated positively.
In Heider's (1967) article he discusses a recent
series of experiments, the results of which are of interest
relative to the results of this study. Subjects were asked
to rate persons described briefly in two-person interaction
situations. When person A benefited from person B, person
B's being good or bad had little influence on person A's
rating. But when person A harmed person B it made a great
deal of difference whether person B was good or bad. The
findings were that it was worse to harm a good person than
a bad person. Our results are the same. The agent's
behavior in version 3 (A+ R+) is rated less socially
101
acceptable than in version 1 (A+ R—). However, the
pleasantness rankings based on these perceptions do not
agree with Heider's results. Both the pleasantness and
humor rankings of these two versions were not significantly
different.
This could have been due to the difference in the
extent of harm (hostility) done to the recipient. Less
hostility was attributed to the harm done to the good per
son than the bad one. This could have been true for the
abstract statements also, and if so, this factor was not
operating in Heider's experiment. It is not likely that
the subjects in our experiment behaved differently in
rating the abstract situations because of the preceding
stories they rated. Most respondents were questioned about
the nature and purpose of the experiment and the relation
between the two parts. Very few saw the abstract state
ments as the fundamental outlines of the stories. And
practically no one was aware of the four separate conditions
created by the versions. The fact that the stories were so
different in content probably helped in this respect.
The pleasantness rankings of the abstract situa
tions do not conform to results obtained in other balance
theory experimentation. Esch (in Heider, 1958) and Jordan
(1953) both found that balanced situations were preferred
to imbalanced ones. Our results indicate that some
balanced situations are preferred to others and some not
102
balanced states are preferred "to balanced ones. Balance
theory is too broad to provide specific predictions across
a wide variety of situations. Hdwever, the principles and
the system approach are useful in formulating hypotheses if
predictions can be based in part on more specific but per
haps less powerful theories.
Justice
Justice is not a direct function of the matching of
the perception of the agent's remark to the recipient and
his perception of the recipient. The observer's perception
of the agent also enters into the determination of justice.
This is why only version 1 was rated as being just. In
version 2 the agent behaved improperly and so lost his
"right" to direct a final hostile remark to the recipient,
who also behaved badly. Versions 3 and 4 were not just
because the recipient behaved in a proper social manner.
This is further support of the balance theory system
approach but not its predictions. All relations in the
triad are interdependent.
For this reason it is believed the perception of
justice determines, to some extent, the degree of hostility
attributed to the remark. Albert (in Heider, 1958) found
that acts are interpreted differently as relations between
the agent and recipient change. Evidently the perception
of harm or hostility changes also. In a sense, then, one
103 j
could say that justice and hostility are co-determined by i
the agent—recipient relations. !
It is felt that justice or justness is basic to the !
meaningfulness of the story. The observer is aware of the
antecedent conditions (relations between A and R) that lead
up to the final remark. The remark itself would be meaning
less without the story context. And putting the remark in
context means becoming aware of the justness or unjustness
of the entire situation. How is one to decide how hostile
the remark is without considering the remark in light of
preceding conditions? Why, in version 1, should the remark
be thought less hostile than in version 4? The remark is
exactly the same— the events in the two versions are exactly
the same but ascribed to different story characters. It is
less hostile, apparently, because of the relationship
between A and R. And knowing the agent and recipient and
how they fit together in the story, one can decide what
rights accrue to each. This in itself is basic to the con
cept of justice.
Justice and hostility do, however, show a certain
independence from one another. In version 3, where the
agent and recipient both behave properly, at least up to
the point where the final remark is made, hostility is low
but there is a lack of justice. This is a peculiar situa
tion. It was included in the experiment for theoretical
completeness, not because it represented a logical-type
104 |
I
:
situation that would be likely to occur. There is not j
justice because the recipient is not bad— and this is a
necessary but not sufficient condition for justice. How
ever, why should A, unprovoked, direct hostility toward R?
There is no evident reason supplied in the story. So if
hostility is perceived, it is likely to be of slight degree.
Version 3 is the only version that is out of its
hypothesized order. It is believed that it was found more
humorous and more pleasant because it was low in hostility.
This brings it to the point where it is equal to version 1
on both respects. Version 1 has its position at the top
because it is high in justice, while at the same time being
more hostile than version 3.
Of all the humor studies dealing with hostility
(Bryne, 1958; Doris and Fierman, 1956; Hammes and Viggins,
1962; Levine and Abelson, 1959; O'Connell, 1960; Strickland,
1959; Williams and Cole, 1964), none have taken into account
possible justification for hostility or other aspects of
agent-recipient relations. It would seem from the results
of this study that situational factors have a more important
place in the study of humor than the literature indicates
they have been given. Even if the main purpose of a study
is involved strictly with the tension-reduction aspects of
mirth or of personality differences related to tension—
reduction, situational factors are important. The percep
tion of hostility is mediated by the perception of justice.
105
And justice, and liking or approval of story characters,
have important effects independent of hostility. It seems
from our results that these factors also affect mirth.
I
I
CHAPTER VI
SUMMARY
The present study investigated the effects of
justification for hostility on the appreciation of tenden
tious humor. It was thought that many studies on humor do
not take into account the importance of situational factors
in studying response to humor. These situational factors
can set limits on mirth, by affecting the ability of persons
to release tension in responding to humorous stimuli.
Describing the social relations between story
characters in terms of justice is not the only approach
that can be taken. Justice theory, as outlined in this
dissertation, is subsumed by balance theory. And, balanced
situations which are justified should be more satisfying
than balanced situations which are not justified. It is
therefore the specific purpose of this study to examine
justice theory as a subset of balance theory to see if it
is a better predictor of mirth response to the two-person
interactions presented in the stories.
The following hypothesis was proposed as a basis
for this research: Stories in which the relations between
106
107
story characters are just will have higher humor ratings
than stories in which the relations between story characters
are not just. Also, stories in which the conditions for
justice are met, and which also represent balanced situa
tions will have higher humor ratings than such stories
which are not balanced. Conversely, those stories in
which the conditions for justice are not met, and which also
represent balanced situations, will have lower humor ratings
than such stories which are not balanced.
In addition to the major hypothesis, two other
hypotheses were made. One posited sex differences in
responsing to the stories because of sex-related differences
in making moral judgments of the story characters. The
other hypothesis was the same as the major hypothesis but
the materials were a set of situations which were abstract
descriptions of the stories used in this study.
One hundred and ninety-two subjects comprised the
total sample in the study. One half of them were males;
one half were females. All were college students, most
of whom were introductory psychology students. Most sub
jects were tested in small groups.
The procedure consisted of having subjects read,
and subsequently rate, four stories on the basis of how
humorous the subjects found them. The four stories were
four different jokes expanded into short paragraph form.
In each story there were two principal characters. The
108
descriptions of these story characters were systematically
varied as to whether each behaved in a socially acceptable
or unacceptable manner. These variations resulted in four
versions for each story. Each subject read four different
story—version combinations.
The basic data were the rankings and ratings of how
humorous the subjects found each story. Other data were
checks on the perception of various aspects of the stories
and the story characters.
The results showed that a balanced situation in
which the social relations between story characters were
justified was more humorous and pleasant than a balanced
situation in which there was no justice. However, justice,
as it was operationally defined in the beginning of the
study, was not the same "justice" that the subjects per
ceived during the course of the study. The perception of
justice was not only associated with perceived "badness"
of the recipient but also with the perceived goodness of
the agent. Humor ratings and pleasantness ratings were
higher when the agent was good than when he was bad. This
relationship is more important in determining humor ratings
than justice, balance, or hostility.
There is a relationship between perceived justice
and the perceived amount of hostility expressed in a
humorous remark. Hostility is lower when there is justice
in the social relations between story characters. However,
109
hostility is lowest for one particular version for which
there is no justice in social relations. This occurs when
both the agent and recipient are made to look good.
Justice was more important in determining humor
ratings than was hostility. At each level of hostility
rating, humor ratings were higher for justice than for no
justice. And the humor ra(tings are higher for perceived
justice regardless of the level of hostility than when
there is no perceived justice.
There was no particular trend indicated by the ways
in which male and female subjects perceived the stories.
There were a few isolated sex differences but in general
male and female subjects reacted to the stories in the same
way.
The pleasantness rankings of the abstract situations
were in the same order as the humor rankings of the complete
versions. This correspondence between overall pleasantness
and mirth seems to indicate that some of the same factors
were operating in making both sets of rankings. The most
humorous and most pleasant situations were those in which
there was either high justice or low hostility. One of
the most important determinants of these factors is the
perception of the agent or aggressor in the stories.
R E F E R E N C E S
110
REFERENCES
Adorno, T. Frenkel-Brunswik, E., Levinson, D. J. , and
Sanford, R. N. The authoritarian personality.
New York: Harper and Brothers, 1950.
Allison, J., and Hunt, D. P. Social desirability and
expression of aggression under varying conditions
of frustration. Journal of Consulting Psychology.
1959, 23, 528-32.
Berkowitz, L. Aggression: a social psychological analysis.
New York: McGraw-Hill, 1962.
Buss, A. The psychology of aggression. New York: Wiley and
Sons, 1961.
Bryne, D. Drive level, response to humor, and the cartoon
sequence effect. Psychological Reports. 1958, 4,
439-42.
Cartwright, D., and Harary, F. Structural balance: A
generalization from Heider's theory. Psychological
Review. 1956, 63, 277-93.
Cohen, A. Social norms, arbitrariness of frustration, and
status of the agent of frustration in the frustra-
tion-aggression hypothesis. Journal of Abnormal and
Social Psychology, 1955, 51, 222-26.
Conrad, R. Accuracy of recall using keyset and telephone
dial, and the effect of a prefix digit. Journal
of Applied Psychology. 1958, 42, 285-88.
Crissman, P. Temperal change and sexual difference in
moral judgments. Journal of Social Psychology.
1942, 16, 29-38.
DiVesta, F., and Bossart, P. The effects of set induced by
labeling on the modification of attitudes. Journal
of Personality, 1958, 26, 379—87.
111
11 2
Doris, J., and Fierman, E. Humor and anxiety. Journal of
Abnormal and Social Psychology. 1956, 53., 59-62.
Elbert, S. Humor references related to authoritarianism.
Dissertation Abstracts. 1962, 23(2), 701.
Freud, S. Wit and its relation to the unconscious, In:
The basic writings of Sigmund Freud. New York:
Modern Library, 1938.
Hadidia, S. Manners for millions. New York: Barnes and
Noble, 1952.
Hammes, E., and Wiggins, A. Manifest anxiety and apprecia
tion of humor involving emotional content. Per
ceptual and Motor Skills. 1962, J_4, 291-94.
Harary, F. On the measurement of structural balance.
Behavioral Science. 1959, 4, 316-23.
Heider, F. The psychology of interpersonal relations.
New York: Wiley and Sons, 1958.
Heider, F. On social cognition. American Psychologist.
1967, 22_j_ 25-32.
Hetherington, E., and Wray, N. Aggression, need for social
approval, and humor preferences. Journal of Abnor
mal and Social Psychology. 1”964, 68, 685-89.
Homans, G. Social behavior; Its elementary forms. New
York: Harcourt, Brace and World, Inc., 1961.
Jordan, N. Behavioral forces that are a function of
attitudes and of cognitive organization. Human
Relati ons. 1953, 6, 273-87.
Kenny, D. The contingency of humor appreciation on the
stimulus—confirmation of joke-ending expectations.
Journal of Abnormal and Social Psychology. 1955,
51_, 644-48.
Kinder, J. Through our own looking glass. School and
Society. 1925, 22, 533-36.
Kitay, P. A comparison of the sexes in their attitudes and
beliefs about women: A study of prestige groups.
Soci ometry. 1940, 3., 399-407.
Klinger, E., Albaum, A., and Hetherington, M. Factors
influencing the severity of moral judgments.
Journal of Social Psychology. 1964, 63., 319-26.
113
LaFave, L. Humor judgments as a function of reference
groups: An experimental study. Dissertation
Abstracts. 1961, 22(4), 1284.
Levine, J. and Abelson, R. Humor as a disturbing stimulus.
Journal of General Psychology. 1959, 60, 191-200.
Levine, J. and Redlich, F. Failure to understand humor.
Psychoanalytic Quarterly, 1959, .23, 528-32.
Malpass, L. and Fitzpatrick, E. Social facilitation as a
factor in relation to humor. Journal of Social
Psychology. 1959, .50, 295-303.
More, D. and Roberts, J. Societal variations in humor
response. Journal of Social Psychology. 1957, 45.
233-43.
Morrissette, J. 0. An experimental study of the theory of
structural balance. Human Relations. 1956, 11
239-54.
Newcomb, T. An approach to the study of communicative acts.
Psychological Review. 1953, 60, 393-404.
O'Connell, ¥. A study of the adaptive functions of wit
and humor. Dissertation Abstracts. 1958, 1_2, 1126.
O'Connell, ¥. The adaptive functions of wit and humor.
Journal of Abnormal and Social Psychology. 1960, 61.
263-70.
Osgood, C. and Tannenbaum, P. The principle of congruity
in the prediction of attitude change. Psychological
Review. 1955, 62^, 42-55.
Pastore, N. The role of arbitrariness in the frustration-
aggression hypothesis. Journal of Abnormal and
Social Psychology. 1952, 47, 728-31.
Post, E. Etiquette: The blue book of social usage. New
York: Funk and ¥agnalls, 1955.
Redlich, F., Levine, J., and Sohler, F. Mirth response
test, preliminary report on a psychodiagnostic
technique utilizing dynamics of humor. American
Journal of Orthopsychiatry. 1951, 2J_, 717-34.
Roosevelt, E. Book of common sense etiquette. New York:
Macmillan, 1962.
114
Rosenberg, M. An analysis of affective-cognitive consis
tency. In C. I. Hovland and M. J. Rosenberg (Eds.),
Attitude organization and change. New Haven, Conn.:
Yale University Press, 1960.
Rothaus, P. and Worchel, P. The inhibition of aggression
under non-arbitrary frustration. Journal of
Personality. 1960, 28, 108-117.
Sherriffs, A. and Jarrett, R. Sex differences in attitudes
about sex differences. Journal of Psychology.
1953, 25, 161-68.
Siegel, S. Nonparametric statistics for the behavioral
sciences. New York: McGraw-Hill, 1956.
Skaggs, E. Sex differences in moral attitudes. Journal
of Social Psychology, 1940, 1 _j_, 3-10.
Strickland, J. The effect of motivation arousal on humor
preferences. Journal of Abnormal and Social
Psychology, 1959, .59, 278-81 .
Vanderbilt, A. New complete book of etiquette: the guide
to gracious living. Garden City, New York:
Doubleday, 1963.
Vestermarck, E. Ethical relativity. New York: Harcourt,
Brace and Company, 1932.
Williams, C. and Cole D. The influence of experimentally
induced inadequacy feelings upon the appreciation
of humor. Journal of Social Psychology, 1964, 64
113-17.
Winer, B. Statistical principles in experimental design.
New York: McGraw-Hill, 1962.
A P P E N D I C E S
11 5
APPENDIX A
STORY-VERSION COMBINATIONS
116
117 ;
STORY A— VERSION 1
One afternoon in a bridge game two girls, Ellen and
Jane, were partners. Jane spent as much time complaining as
she did playing. She pointed out all Ellen’s mistakes and
ridiculed the opposition. Ellen didn't want to get into an
argument so she just kept quiet and tried to enjoy the game
as much as possible under the circumstances. During one hand
Jane played particularly badly. She didn't do anything ,
right. However, she thought she played splendidly. She
moaned about how everything was against her and that there
was no way to do better. After registering all her com
plaints she asked Ellen with a note of scorn, "If you were
me how would you have played that hand?" This was too much
for Ellen. Her prompt reply was, "Under an assumed name."
STORY A— VERSION 2
One afternoon in a bridge game two girls, Ellen and
Jane, were partners, Ellen and Jane both spent as much time
complaining as playing. They pointed out each other's mis
takes and ridiculed the opposition. Whenever either made a
mistake they moaned and groaned about how everything was
against them and that there was no way to do better. Each
thought herself a splendid player. During one hand Jane
played particularly badly. She didn't do anything right.
After registering all her complaints she asked Ellen with a
note of scorn. "If you were me how would you have played
that hand?" Ellen replied equally scornfully, "Under an
assumed name."
STORY A— VERSION 3
One afternoon in a bridge game two girls, Ellen and
Jane, were partners. They both were enjoying the game. But,
during one hand Jane didn't play too well. She couldn't
figure out where she went astray and asked Ellen, "If you
were me how would you have played that hand?" Ellen replied
scornfully. "Under an assumed name."
STORY A— VERSION 4
One afternoon in a bridge game two girls, Ellen and
Jane, were partners. Ellen spent as much time complaining as
she did playing. She pointed out all Jane's mistakes and
ridiculed the opposition. Whenever Ellen herself made a mis^
take she moaned and groaned about how everything was against
her and that there was no way to do better. She always
thought she played splendidly. Jane, not wanting to get in
to any arguments, just kept quiet and tried to enjoy the game
as much as possible under the circumstances. During one hand
Jane didn't play too well. She couldn't figure out where she
went astray and asked Ellen, "If you were me, how would you
have played that hand?" Ellen replied with a note of scorn,
"Under an assumed name."
STORY B— Version 1
118
A few girls were standing in a group talking at a
large reception. One of the girls, Mary, was saying what a
nice party it was. She had started to say something else
when Betty walked up to them and broke into the conversa
tion. Betty was just returning from her third trip to the
buffet and had a plateful of rich canapes which she was
hungrily devouring. "Not a bad party, huh?" she asked Mary.
"I even saw somebody wearing a dress just like yours." Mary
wanted to tell Betty a thing or two but she just smiled
bravely and blushed a little. Changing the subject Betty
said, "The food is really good, but I do have to watch my
waistline." Mary purred, "And how lucky you are to have it
right out there where you can."
STORY B— Version 2
A few girls were standing in a group talking at a
large reception. Mary had just walked to the group. She
was returning from her third trip to the buffet and had a
plateful of rich canapes which she was hungrily devouring.
Betty was also doing a lot of eating and was attacking a new
plateful with gusto. Betty said to Mary, "Not a bad party,
huh? I even saw somebody wearing a dress just like yours."
Mary snapped back, "At least mine is new. Yours looks
familiar.' Didn’t you wear it to that reception last month?"
Changing the subject Betty said, "The food is really good
but I do have to watch my waistline." Mary purred, "And how
lucky you are to have it right out there where you can."
STORY B— VERSION 3
A few girls were standing in a group talking at a
large reception. One of the girls, Mary, was saying what a
nice party it was. As she was talking, Betty came up and
joined the group. She agreed with Mary that the party was
really great. Changing the subject Betty continued, "The
food is really good, but I do have to watch my waistline."
Mary purred, "And how lucky you are to have it right out
there where you can."
STORY B— VERSION 4
A few girls were standing in a group talking at a
large reception. Mary had just walked up to the group. She
was returning from her third trip to the buffet and had a
plateful of rich canapes which she was hungrily devouring.
Betty was saying what a nice party it was. She had started
to say something else when Mary broke into the conversation.
"Not a bad party, huh?" she asked Betty. "I even saw some
body wearing a dress just like yours." Betty wanted to tell
Mary a thing or two but she just smiled bravely and blushed
a little. Changing the subject Betty said, "The food is
really good but I do have to watch my waistline." Mary
purred, "And how lucky you are to have it right out there
where you can."
STORY C— VERSION 1 j
|
This little exchange was overheard at one of the j
better restaurants in town recently. Jim and his friend were
having a quiet business discussion. At the next table Earl :
and a friend were reminiscing about old times. They were
quite noisy. Earl was talking loudly and laughing even more >
loudly. They were so loud Jim couldn’t even hear his friend
over the din. Jim wanted to change tables and was trying to
spot the headwaiter. Earl, looking around for a waiter, said!
quite loudly, "What do I have to do to get a glass of water !
around here?" Jim thoroughly disgusted by this time suggest
ed quietly, "Why don't you try setting yourself on fire?"
STORY C— VERSION 2 j
This little exchange was overheard at one of the
better restaurants in town recently. Jim and a friend were
talking and laughing about some experience they had been
through together. They were quite noisy. They had some
rivals at the next table who were making even more noise than
they were. Earl and his friend were reminiscing about old
times. They were talking loudly and laughing even more
loudly. They were so loud Jim couldn't even hear his friend
over the din. At one point Earl, looking around for a
waiter, said quite loudly, "What do I have to do to get a
glass of water around here?" Jim couldn't help hearing Earl
and suggested, "Why don't you try setting yourself on fire?"
STORY C— VERSION 3
This little exchange was overheard at one of the
better restaurants in town recently. Jim and his friend were
having a quiet business discussion. At the next table Earl
and a friend were also talking over some business affairs.
At one point Earl, looking around for a waiter, said quietly
to his friend, "What do I have to do to get a glass of water
around here?" Jim, with a keen ear, overheard him and sug
gested nastily, "Why don't you try setting yourself on fire?"
STORY C— VERSION 4
This little exchange was overheard at one of the
better restaurants in town recently. Jim and a friend were
reminiscing about old times. They were quite noisy. Jim was
talking loudly and laughing even more loudly. At the next
table Earl and a good friend of his were having a quiet busi
ness discussion. The din from the next table was so loud
that Earl couldn't even hear his friend. Earl thoroughly
disgusted by this time said quietly to his friend while
looking around for a waiter, "What do I have to do to get a
glass of water around here?" Jim, with a keen ear, over
heard him and suggested nastily, "Why don't you try setting
yourself on fire?"
120
STORY D— VERSION 1 j
During a mock trial in law school Bill was defense 1
attorney. He was a good student and well liked by his felloW
students. He was giving his closing argument and as he
talked he gestured at times with his glasses for emphasis.
In the reflection from his glasses he saw his opponent, Fred,;
seated behind him, shaking his head in disagreement. Fred
was known around school as a sore loser. He had been dis
rupting the proceedings with constant irrelevant objections
and making life miserable for all concerned. Bill had some
how managed to keep from losing his temper in spite of Fred's
behavior. At one point Bill said, "Notwithstanding the fact
that Fred is shaking his head in disagreement with every
thing I have said ..." He then continued with his state
ment. Fred interrupted: "Objection. I suppose defense
attorney would have us believe he has eyes in the back of his
head? How does he know I was shaking my head?" "I could
hear it rattle," said Bill without turning around.
STORY D— VERSION 2
During a mock trial in law school Bill was defense
attorney. Bill was known around school as a sore loser. He
had been disrupting the proceedings with constant irrelevant
objections and making life miserable for all concerned. He
was giving his closing arguments and as he talked he gestured
at times with his glasses for emphasis. In the reflection
from his glasses he saw Fred shaking his head in disagree
ment. Fred was also a hot head. He had already started
several heated exchanges with Bill. And, the short fuse on
his temper was getting even shorter. At one point Bill said,
"Notwithstanding the fact that Fred is shaking his head in
disagreement with everything I have said ..." He then con
tinued with his statement. Fred interrupted: "Objection. I
suppose defense attorney would have us believe he has eyes
in the back of his head? How does he know I was shaking my
head?"
"I could hear it rattle," said Bill without turning
around.
STORY D— VERSION 3
During a mock trial in law school Bill was defense
attorney. He was a good student and well liked by his fellow
students. He was giving his closing arguments and as he
talked he gestured at times with his glasses for emphasis.
In the reflection from his glasses he saw his opponent, Fred,
seated behind him, shaking his head in disagreement. Fred
was also very bright and got along well with almost everyone.
At one point Bill said, "Notwithstanding the fact that Fred
is shaking his head in disagreement with everything I have
said ..." He then continued with his statement. Fred
interruped, "Objection." And then he gave his reason for
objecting and as he was sitting down again he added, mostly
STORY D— VERSION 3— Continued
121
to himself, "And I suppose defense attorney would also have
us believe he has eyes in the back of his head? How does he
know I was shaking my head?" \
"I could hear it rattle," snapped Bill without
turning around.
STORY D— VERSION 4
During a mock trial in law school Bill was defense
attorney. Bill was known around school as a sore loser. He
had been disrupting the proceedings with constant irrelevant
objections and making life miserable for all concerned. He
was giving his closing argument and as he talked he gestured
at times with his glasses for emphasis. In the reflection
from his glasses he saw Fred shaking his head in disagree
ment. Fred was a good student and well liked by his fellow
students. He had somehow managed to keep from losing his
temper in spite of Bill's behavior. At one point Bill said,
"Not withstanding the fact that Fred is shaking his head in
disagreement with everything I had said ..." He then
continued with his statement. Fred interrupted: "Objection."
He then gave his reason for objecting and as he was sitting
down again he added, mostly to himself, "And I suppose
defense attorney would also have us believe he has eyes in
the back of his head? How does he know I was shaking my
head?"
"I could hear it rattle," snapped Bill without
turning around.
APPENDIX B
INSTRUCTIONS TO SUBJECTS
122
INSTRUCTIONS TO STUDENTS
Before you read, these stories I would like to give
you some idea of what you are to do. I would like you to
read each of these four stories when you are requested to do
so. In each story there are two people who interact with
one another. As you are reading each story keep in mind how ;
the people in the stories behave toward one another and how J
this initial behavior relates to the ending of the story.
Do not worry about tricks or subtleties— there are
not any deliberate misleading statements in the stories.
Just take what the characters say at face value. If some
thing seems to be left unsaid it was not meant to be said or
implied— don't look for hidden meanings. I know the ten
dency in psychological experiments is to be very cautious and
to analyze everything very carefully— sometimes to think too
much. Just relax and take a straightforward approach. The
nature and purpose of the experiment will be explained to you
when you have finished.
Now I would like you to read each of these four
stories. You are to rank them according to how humorous you
think they are. Do not think of the stories as jokes but as
situations that might or might not be humorous. Imagine
yourself present at the situation— observing the actions of
the story characters. Take into account the whole sequence
of events in the story— then rank them for humor. Give rank
1 to the story you think is most humorous; give rank 2 to
the story you think is next-most humorous; give rank 3 to
the story you think is next-to-least humorous; give rank 4
to the least humorous story. Just write a 1, 2, 3, or 4 on
top of the story and draw a circle around it.
APPENDIX C
QUESTIONNAIRE
124
QUESTIONNAIRE
Now that you have read the above story we would like you
to tell us how humorous it is. Please do this by
circling a number on the scale below. Some of the num
bers have phrases beside them; others do not. The
phrases are guides to help you in using the scale.
Please feel free to circle any one of the numbers below
that best indicates your feeling about the story.
8 very humorous
7
6 humorous
5
4 somewhat humorous
3
2 slightly humorous
1
0 not at all humorous
Now we would like you to think about the final remark in
the story. In answering this question do not consider
whether the remark should or should not have been made.
Just considering that the remark was made by one person
to another: Do you think the remark at the end of the
story is:
very hostile ___
somewhat hostile ___
slightly hostile
How would you judge (name of agent)'s overall behavior
in this story?
circle one number
_3 -2 -1 OR +1 +2 +3 j
very____ somewhat____slightly slightly somewhat very;
UNACCEPTABLE SOCIAL BEHAVIOR ACCEPTABLE SOCIAL BEHAVIOR
How would you judge (name of recipient)'s overall
behavior in this story?
circle one number
-3 -2 -1 OR +1 +2 +3
very____ somewhat____slightly slightly somewhat very
UNACCEPTABLE SOCIAL BEHAVIOR ACCEPTABLE SOCIAL BEHAVIOR
Do you think (name of agent) was JUSTIFIED OR UNJUSTI
FIED in making the remark (he) (she) did at the end of
the story?
circle one number
-3 -2 -1 OR +1 +2 +3
very____ somewhat____slightly slightly somewhat very
UNJUSTIFIED JUSTIFIED
Why did you circle the number you did? Please explain
in terms of what happened in the story.
1 27 (
We would like you to describe (name of agent) by placing an i
X in the column that indicates your feeling about (him)
(her) for each of the characteristics listed below.
(1) (2) (3) (4)*
NOT AT ALL SLIGHTLY SOMEWHAT VERY
1 1 1 ' ' — —
FRIENDLY______________ ___________ _________ _________ ____
JUST (GOOD OR MORAL) ___________ ■
HOSTILE ;
VAIN 1- ;
LIKABLE___________________________ _________ _________ ____
CRITICAL . _________ _________ ____
POLITE ___________ _________ _________ ____
WARM :
AGGRESSIVE ___________ _________ _________ ____
WITTY ___________ _________ _________ ____
INSINCERE_____________ ___________ _________ _________ ____
COLD ___________ _________ _________ ____
We would like you to describe (name of recipient) by placing
an X in the column that indicates your feeling about (him)
(her) for each of the characteristics listed below.
NOT AT ALL SLIGHTLY SOMEWHAT VERY
FRIENDLY ___________ _________ _________ ____
JUST (GOOD OR MORAL) ___________ _________ _________ ____
HOSTILE ___________ _________ _________ ____
VAIN ___________ _________ _________ ____
LIKABLE ___________ _________ _________ ____
CRITICAL__________________________ _________ _________ ____
POLITE ___________ _________ _________ ____
WARM ___________ _________ _________ ____
AGGRESSIVE ___________ _________ _________ ____
WITTY ___________ _________ _________ ____
INSINCERE_____________ ___________ _________ _________ ____
COLD ___________ _______ _______
*These numbers were used in the analysis but were
not part of the Questionnaire.
APPENDIX D
ABSTRACT SITUATIONS
128
129
Now we would like you to read
it are brief descriptions of
situation the behavior of two
described. Read each of the
Your task is to rank each of
following basis: Imagine you
following situations and then
pleasant you would find each
situation you would find most
situation you would find next
the situation you would find
rank 4 to the situation you w
the following material. In
four situations. In each
persons toward each other is
four descriptions carefully,
these situations on the
have observed each of the
rank them according to how
one. Give rank 1 to the
pleasant; give rank 2 to the
-most pleasant; give rank 3 to
next-to-least pleasant; give
ould find least pleasant.
Rank
(Vers ion 1)* A behaves in a socially acceptable manner
toward B.
B behaves in a socially unacceptable manner
----------- toward A.
A expresses hostility toward B.
(Version 2) Both A and B behave in a socially unacceptable
manner toward each other.
A expresses hostility toward B.
(Version 3) Both A and B behave in a socially acceptable
manner toward each other.
A expresses hostility toward B.
(Version 4) A behaves in a socially unacceptable manner
toward B.
B behaves in a socially acceptable manner
----------- toward A.
A expresses hostility toward B.
*The order of presentation was varied.
Filmed as received
Without page(s)___
130
UNIVERSITY MICROFILMS
APPENDIX E
COUNTERBALANCING DESIGN
131
132
Story
B C D
ft
3
O
U
c3
GO
fl
•H
O
fl
(8
r— I
(8
JO
U
a>
-p
C
3
O
O
1
2
3
4
5
6
7
8
9
10
1 1
12
13
14
1 5
16
17
18
19
20
21
22
23
24
V*1 V2 V3 V4
V1 V2
V4 V3
V1 V3 V2 V4
V1 V3 V4 V2
V1 V4 V2 V3
V1 V4 V3 V2
V2 V1 V3 V4
V2 V1 V4 V3
V3 V1 V2 V4
V4 V1 V2 V3
V3 V1 V4 V2
V4 V1 V3 V2
V2 V3 V1 V4
V2 V4 V1 V3
V3 V2 V1 V4
V4 V2 V1 V3
V3 V4- V1 V2
V4 V3 V1 V2
V2 V3 V4 V1
V2 V4 V3 V1
V3 V2 V4 V1
Y4 V2
V3 V1
Y3 V4 V2 V1
V4 V3 V2 V1
^Version
Fig. 12.— Counterbalancing design showing stories,
counterbalancing group and versions of each story shown to
groups. Each group has four male and four female subjects.
APPENDIX F
FREE RESPONSE DATA
133
134
TABLE 32 '
FREE RESPONSE DATA ON REASONS FOR JUSTICE
RATINGS FOR STORY A
Story Characters
Per
Subj ect
Cent
Response
(Female; Bridge Players)
Males
(N=24)
Females
(N=24)
Version 1
Jane aggravated Ellen— a person can
only take so much
87 54
Ellen was justified— (non-specific
answer
62 37
Ellen was polite— she took all she
c ould
46 25
Ellen was not justified— could have
shown more tact
12 12
Version 2
(Justice)
Ellen's remark appropriate— justified 46 17
Both behaved badly
29 38
Jane complained too much
17 8
Humor eased the tension in the
situation
17 17
Jane and Ellen were friends— no
hostility intended
12 —
Ellen was unjustified— (non-specific
answer
8 29
Jane behaved badly
8 8
TABLE 32— Continued
135
Story Characters
Per
Subj ect
Cent
Response
(Female; Bridge Players)
Males
(N=24)
Females
(N=24)
Version 3
Not justified— non specific 54 37
Jane just asked for advice— no call
for such a remark
25 17
Remark made to humor her partner 21 17
Inconsistent with Ellen’s previous
behavior
17 4
Justified because of Jane's poor
playing
17 8
Between friends— no hostile intent 4 25
Version 4
Ellen unjustified— her remark totally
out of place
83 71
Jane only asked for advice
29 33
Anybody that plays that bad deserves
such a remark
12 —
Justified because Ellen is a better
player
1 2 4
136
TABLE 33
FREE RESPONSE DATA ON REASONS FOR JUSTICE
RATINGS FOR STORY B
Per Cent
Story Characters Subject Response
(Female; Party)
Males Females
(N=24) (N=24)
Version 1
Betty made unkind remark to Mary
about her dress
71 46
Justified— (non-specific) 46 46
Mary wasn't nice to Betty 8 12
Mary showed restraint 8 4
Not justified— no reason to be
nasty just because someone
else is
4 25
Version 2
Justified— (non-specific) 38 21
Betty started it
33 29
Both behaved badly
25 17
Mary over-reacted, Betty tried to
change the subject
21
29
Mary was rude to Betty
17 29
Not justified (non-specific) 12 42
TABLE 33— Continued
Story Characters
Per Cent
Subject Response
(Female; Party) Males Females
(N=24) (N=24)
Version 3
No apparent reason for Mary's
remark— didn't fit in with her
previous behavior
59 42
Betty was polite— didn't deserve
to receive remark
42 21
Unjustified— (non specific) 38 50
Justified— Betty interrupted 8 8
Version 4
Mary behaved badly, she had no
right to make such a remark
54 75
Betty was polite, she didn't
deserve the remark
50 17
Betty made nasty remark about
Betty's dress
17 21
Both behaved badly
8 4
138
TABLE 34
FREE RESPONSE DATA ON REASONS FOR JUSTICE
RATINGS FOR STORY C
Story Characters
Per
Subj ect
Cent
Response
(Male; Restaurant) Males
(N=24)
Females
(N^24)
Version 1
Earl was rude and noisy, deserved
what he got
84 63
Jim was justified— (no specific answer) 50 42
Not justified— (no specific answer) 21 12
Jim showed restraint 21 8
Version 2
Both bad but Earl louder 42
17
Jim was loud— he wasn’t justified 38 54
Jim was justified (no specific answer)
29 21
Both were bad, loud
29
21
Version 3
Remark was out of place— shouldn't
have said it
46 50
Unjustified— (no specific reason) 33 25
Jim was listening otherwise he wouldn't
have heard
33 21
Justified— (no specific reason)
17 —
Remark was made just for the humor 4 12
Version 4
No call for such a remark from a loud mouth 46 50
Unjustified— (no specific answer) 33 33
Earl caused no trouble 17 29
Justified— said to be funny 12 4
TABLE 3 5
FREE RESPONSE DATA ON REASONS FOR JUSTICE
RATINGS FOR STORY D
Story Characters
(Male; Law School)
Per Cent
Sub.ject Response
Males Females
(N=24) (N=24)
Version 1
Justified because of Fred's behavior
Bill's behavior was all right
Irrelevant objection Fred made— he
asked for it
Bill was just being funny
Version 2
Irrelevant objection Fred made— he
asked for it
Justified because of Fred's behavior
Improper courtroom behavior on the part
of both parties
Both behaved badly
A humorous reply to Fred
Bill was just being nasty
Version 3
Justified as a witty reply to Fred
Unjustified— too hostile, personal
Behavior all right— anything goes in a
courtroom
Bill started it with the remark about
Fred
Justified because of Fred's behavior
79
17
17
8
50
21
21
12
12
12
46
38
12
8
4
84
8
8
29
29
17
17
8
8
42
21
29
TABLE 3 5— Continued.
140
Per Cent
Story Characters Sub.i ec t Response
(Male; Law School) Mai es Females
(N=24) (N=24)
Version 4
Bill was a nuisance 29 17
Unjustified— too hostile, personal 25 29
Inappropriate remark for a courtroom 8 25
Justified with respect to Fred's
remark
25 21
Bill's remark was made just for fun 17 8

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The Effects Of Sex, Assigned Therapist Or Peer Role, Topic Intimacy, And Expectations Of Partner Compatibility On Dyadic Communication Patterns

PDF

The Effects Of Anxiety And Threat On Self-Disclosure

PDF

The Effects Of Nonreward Dissonance And Secondary Reward On Extinction And Attractiveness

PDF

Transfer Of The Partial Reinforcement Extinction Effect Across Tasks In Normal And Retarded Boys

PDF

The Enhancement Of Eeg - Alpha Production And Its Effects On Hypnotic Susceptibility

PDF

Effects Of Group Behavior Therapy Imagery On Basketball Performance

PDF

Verbal Reports Of Emotional States And Onsets And Offsets Of Conditioned Stimuli

PDF

The Effect Of Dissonance In Self-Esteem On Susceptibility To Social Influence

PDF

Prognostic Expectancy Effects In The Desensitization Of Anxiety Over Invasion Of Body Buffer Zones

PDF

The Effects Of Feedback On The Communication Of Medical Prescription To Diabetic Patients

PDF

A Multiple Investigation Of Child-Rearing Attitudes

PDF

Non-Specific Treatment Factors And Deconditioning In Fear Reduction

PDF

The Effects Of A Self Shock Procedure On Hallucinatory Activity In Hospitalized Schizophrenics

PDF

Effects Of Video Tape Feedback Versus Discussion Session Feedback On Group Interaction, Self Awareness And Behavioral Change Among Group Psychotherapy Participants

PDF

An Application Of A Two-Stage 'Attention' Model To Concept Formation In The Mentally Retarded

PDF

Imagery And Response Styles In Desensitization

PDF

Human Operant Eye Blink Conditioning, Awareness, And The Extraversion-Introversion Dimension Of Personality

PDF

Differences Between Cues In Effectiveness As Retrieval Aids

PDF

Recognition Accuracy In The Method Of Single Stimuli: A Test Of An Operational Definition Of The Distinctiveness Of Stimuli

PDF

The Influence Of Communality And N On The Sampling Distributions Of Factor Loadings

Asset Metadata

Creator Gutman, Jonathan (author)

Core Title The Effects Of Justice, Balance, And Hostility On Mirth

Degree Doctor of Philosophy

Degree Program Psychology

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag OAI-PMH Harvest,psychology, general

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Priest, Robert F. (committee chair), De Nike, L. Douglas (committee member), Myers, James H. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-588138

Unique identifier UC11360891

Identifier 6717676.pdf (filename),usctheses-c18-588138 (legacy record id)

Legacy Identifier 6717676.pdf

Dmrecord 588138

Document Type Dissertation

Rights Gutman, Jonathan

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA