Sunteți pe pagina 1din 6

Journal of Personality and Social Psychology Copyright 1985 by the American Psychological Association, Inc.

1985, Vol. 49, No. I, 281-286 0022-3514/85/S00.75

A Study of Salience and Motivational Theories of Humor


Thomas L. Kuhlman
Mendota Mental Health Institute
Madison, Wisconsin

Experiments by Goldstein, Suls, and Anthony (1972) discredited motivational


theories of humor in favor of a cognitive salience model, which was proposed as
more parsimonious. The present experiment also tested the salience model versus
motivational explanations. College students assigned humor ratings to three sets
of jokes containing salient themes (examination content), taboo themes (sex and
violence content), and neutral themes. The joke sets had been matched a priori
for incongruity structure. The salience of examination content was manipulated
by having the students rate jokes under one of three conditions: control (normal
classroom setting), before taking an examination, and 20 min after beginning an
examination. Results did not confirm hypotheses derived from the salience model;
motivational factors (taboo content, mastery theme) were associated with high
humor ratings. A conceptual integration of cognitive and psychoanalytic mecha-
nisms of humor is suggested.

Two general lines of humor theory that Dworkin & Efran, 1967; Goldstein, 1970;
have conceptual appeal and empirical support Lamb, 1968; Landy & Metee, 1969; Mueller
have emerged. Motivational content theories, & Donnerstein, 1983; Singer, 1968; Trice,
of which Freud's (1905/1960) is best known, 1982). This confusing empirical picture and
emphasize thematic properties of jokes, car- conceptual problems with the drive concept
toons, and the like. Freud viewed sexual and of psychoanalysis have prompted the devel-
aggressive humor as important release valves opment of microtheories linking humor and
in the psychic economy. Such "tendentious aggression to situation-specific variables (see
humor," as he termed it, functions as a Zillmann & Bryant, 1974, 1980; Zillmann &
defense mechanism, and the acute, fleeting Cantor, 1972).
pleasure of laughter stems from gratifying a The second line of humor theory empha-
drive that otherwise would have remained sizes process variables over content and is
pent up. The prominence of sexual and ag- embodied by the varieties of incongruity
gressive humor in adult interactions and en- theory. Effective humor stimuli are those that
tertainment cannot be dismissed without pose some incongruity, puzzle, or cognitive
some such explanation, but studies addressing dissonance to the individual. This must be in
Freud's model have yielded inconsistent re- a modest amount; too little incongruity will
lationships between experimentally induced not divert the normal direction and pace of
drive states and the enjoyment of sexual and information processing, and too much incon-
aggressive humor (Baron, 1978a, 1978b; gruity stresses the information processor and
leads to curiosity/concentration behaviors or
This study was initiated while the author was a member anxiety. When a modest incongruity is quickly
of the faculties of the University of Cincinnati and Xavier resolved (e.g., by viewing the stimulus from
University. A preliminary report was presented to the a second perspective, as in the pun), the
Third International Conference on Humor held in Wash- accompanying physiological arousal changes
ington, D.C., August 1982.
The author wishes to acknowledge the contributions are experienced as the amused sense of plea-
of Lee Callaway, Schifferau Gibran-Marion, Wilma Gil- sure. Readily observable support for incon-
bert, Jeff Goldstein, Frank Prerost, Susan Sorenson, Jane gruity theory lies in the importance of timing
Steif-Nute, and Bob Zechnich. in delivering a joke and the fact that jokes
Requests for reprints should be sent to Thomas L.
Kuhlman, Aggression Management Unit, Mendota Mental lose most of their amusement potential from
Health Institute, 301 Troy Drive, Madison, Wisconsin the first to the second presentation. Convinc-
53704. ing research evidence comes from studies of
281
282 THOMAS L. KUHLMAN

children's humor revealing that the sense of of the salience model of humor. There were
humor for particular stimuli is greatest when three significant deviations from the Goldstein
a child arrives at pivotal mastery points of et al. design. First, manipulation of the sa-
various Piagetian stages of cognitive devel- lience variable was less obtrusive. The salient
opment. McGhee (1979) has provided a com- humor content employed was the topic of
prehensive review of various incongruity school examinations. Groups of subjects rated
models and the evidence supporting them. examination jokes in one of three contexts:
The humor mechanisms posed by content (a) in a normal college classroom meeting,
and process models appear radically different. (b) just before taking an examination, and
Much would be gained by forging an integra- (c) 20 min after beginning an examination.
tion between them, including among other The salience model would predict that as
things, a closer bridge between psychoanalysis examination taking increases in salience, hu-
and cognitive psychology. There have been mor ratings of examination jokes should also
such beginnings (Ehrenstein & Ertel, 1978; increase. Second, subjects in all conditions
Godkewitsch, 1976; Suls, 1977), but a more rated not only salient jokes but also others
widely cited set of experiments sought to with neutral or socially taboo content. The
discredit motivational models as being super- joke content sets were matched for incongruity
fluous. techniques employed and other structural
Experiments by Goldstein, Suls, and An- properties. The salience model would predict
thony (1972) disputed research showing that that with structural properties constant, sa-
humor reduced experimentally induced drive lient jokes would be preferred to the other
states. Goldstein et al. argued that the exper- types of humor by subjects in the two salience
imental manipulations in such research af- conditions, Third, the salient jokes were di-
fected not a drive state but a cognitive set for vided into two subsets. In one subset the
processing sexual/aggressive stimuli more student succeeded in an examination situa-
easily, thus making such humor more enjoy- tion; in the other he or she failed or was put
able. This cognitive set interpretation of the down. The salience model would predict no
findings was termed salience. In one experi- difference in the humor ratings assigned by
ment, Goldstein et al. exposed undergraduate subjects to these two subsets of salient humor.
subjects to photographs of either automobiles
or aggressive scenes and then asked them to Method
rate the humorousness of cartoons containing Subjects
either automobile or aggressive themes. Con- Subjects were 88 female and 101 male undergraduate
sistent with the salience model, the authors students enrolled in three different sections of an intro-
found that prior exposure to automobiles ductory psychology course at a midwestern university.
yielded a better appreciation for automobile
cartoons, just as prior exposure to aggressive Humor Stimuli
scenes enhanced appreciation of aggressive Humor stimuli were 30 verbal jokes presented in a
cartoons. Manipulation checks on control written format. Ten jokes consisted of neutral content,
groups validated an effective manipulation of and 10 others had socially taboo themes (i.e., sex, pro-
fanity, untimely death, or sadistic violence). The third
content salience but not of subjects' hostility. set of 10 jokes dealt with humor around test taking and
A second experiment supported the salience examinations; this was designated as the salient content
model, with verbal jokes serving as humor category.
stimuli. The authors concluded in their dis- In joke selection the investigator drew stimuli from a
variety of popular magazines. Care was taken to ensure
cussion that the motivational model of humor that the various structural dimensions of humor (i.e.,
is superfluous: incongruity technique, number of words and dialogue
On the whole . . . the most parsimonious explanation segments) were closely matched among the three content
of the data would rule out motivation as an explanatory categories. Each joke in the neutral set corresponded to
concept, since the salience hypothesis can account for a similar joke in the taboo and salient content sets with
the appreciation of nonsense as well as aggressive and respect to these dimensions. Two examples of this match-
sexual humor, (p. 169) ing outcome are presented in Table I. 1 The 10 jokes in

The experiment to be described in the 1


A set of the humor stimuli used in this study is
present study was designed foremost as a test available from the author.
SALIENT HUMOR 283

the salient content set were selected so as to divide into control group (« = 62) rated the jokes during part of a
two subsets of 5: salient victor and salient victim. In normal lecture period. The before-exam group (n = 65)
victor jokes (see Table 1, Group 6), the student scored a rated the jokes just prior to taking an examination. The
real or symbolic victory over the professor and/or the during-exam group (n = 62) rated the jokes after working
examination; in victim jokes (see Table 1, Group 2), the on an examination for 20 min.
student experienced a real or symbolic defeat or failure After completing the joke ratings, subjects in the
at the hands of the examination and/or professor. before- and during-exam groups were asked to indicate
Finally, to investigate the validity of joke content how confident they were feeling about their performance
sampling, the 30 jokes were presented in random order on the examination in question by use of a 7-point scale
to judges who were asked to indicate whether the theme (7 = unusually high confidence about exam performance
of each joke related to examination taking, some socially compared to other exam situations; 4 = normal degree
taboo behavior, or some other theme. All salient jokes of confidence; 1 = much less confidence than usual).
and 9 of 10 taboo and neutral jokes were correctly
classified by all five judges; the remaining taboo and
neutral jokes (1 each) were correctly classified by four of Results
five judges. Judges had not been informed beforehand Hypotheses 1 and 2 were tested by means
that 10 jokes were intended for each category.
of a three-way analysis of variance (Sex X
Salience Condition X Joke Content Class)
Procedure performed on the summed ratings assigned
Subjects assigned humor ratings to each of the 30 to the jokes by subjects. The General Linear
jokes in the group setting of the classroom. Jokes were Model Procedure of the Statistical Analysis
presented on paper in a randomized fashion except that System was used. This model breaks down
two jokes from the same content set never appeared between-groups differences by the least
consecutively. Subjects provided no identifying information
except for sex. Extra credit points toward the course
squares means method. The overall F value
grade were offered to all subjects who participated (none for this analysis proved significant, F(17,
declined). Subjects were instructed to assign a humor 552) = 5.67, p < .0001. The relevant variance
rating to each joke by using a 7-point scale (1 = boring, source for testing Hypotheses 1 and 2 is the
7 = extremely amusing). They were permitted 15 min Salience Condition X Joke Content Class
for the task and were asked (a) to read all of the jokes
once before assigning ratings and (b) to assign at least interaction. This F value was not significant,
one 7 score and at least one 1 score. The goals of these F(4, 552) = 0.17, p = .95, and indicates that
instructions were to obtain relative rather than absolute Hypotheses 1 and 2 were not supported.
humor ratings; printed jokes, devoid of spontaneity and The analysis did yield significant main
a social context, seldom evoke the heights of amusement
and laughter.
effects for joke content class, F(2, 552) =
The salience variable was manipulated by testing the 36.03, p < .0001, and salience condition, F(2,
189 subjects under one of three group conditions. The 552) = 3.66, p < .05. With respect to joke

Table 1
Examples of Jokes From the Three Content Sets as Matched for Structural Properties
Content set Group 6 Group 2

Neutral A tern once graciously shared a fish Said the old man to the little boy who was
it had caught with a hungry eating an apple: "Be sure to look out for the
otter, and as a result of this worms, Sonny." "When I eat an apple the
gesture the two became quite worms have to look out for themselves,"
friendly. It only illustrates the old replied the boy.
saying that one good tern
deserves an otter.
Taboo There was a case recently in Texas The two rich old Englishmen were taking target
where a wealthy rancher sued his practice on the back lawn. "Watch your
wife for divorce. He had found shooting, old chap," one said, "you almost
his dear and an interloper at play. winged my servant." "Dreadfully sorry! Do
have a shot at mine over there."
Salient (Victor) Upperclassmen are well (Victim) The student approached his prof after
known for getting quite chummy class with his corrected exam. "I don't think
with their profs before exam I deserve an absolute zero," said the student.
time. They follow the old adage "Neither do I," responded the professor, "but
that familiarity breeds exempts. it is the lowest mark I'm allowed to give."
284 THOMAS L. KUHLMAN

content, least squares means analysis indicated would increase as examination taking in-
that taboo humor was significantly preferred creased in salience and (b) salient victor jokes
by subjects over the other two types (ps < would be preferred to the other types of jokes
.0001 in both comparisons) and that salient by subjects in the two salience conditions. A
humor was preferred over neutral humor three-way analysis of variance was performed
(p < .05). Probing the main effect for salience (comparable to the initial analysis), yielding
condition revealed that the during-exam a significant overall F, F(\7, 552) = 4.10,
group generally assigned higher joke ratings p < .0001, but the critical Salience Condi-
than did the control group (p < .005); other tion X Joke Content Class interaction was
between-groups comparisons were nonsignif- not significant, F(4, 552) = 0.81, p = .50).
icant. Predictions (a) and (b) thus did not hold, but
Finally, the analysis yielded a significant a significant main effect for joke content
Sex X Joke Content Class interaction, F(2, class, F(2, 552) = 21.95, p < .0001, revealed
552) = 5.19, p < .01. Least squares means that ratings assigned to salient victor and
analysis revealed that (a) females had assigned taboo humor were not significantly different,
significantly (p < .0001) higher ratings to sa- although both were preferred to corresponding
lient jokes than had males and (b) males neutral jokes (ps < .0001 in each case).
showed a clear preference for taboo humor Considering the salient victim cohort alone,
whereas females rated taboo and salient jokes it was predicted that (c) salient victim joke
equally. In light of this sex difference, Hy- ratings would decrease as examination taking
pothesis 3 was tested separately for males and increased in salience and (d) neutral and
females. The summed ratings assigned to taboo jokes would be preferred to salient
salient victor jokes were compared with the victim humor by subjects in the two salience
summed ratings assigned to salient victim conditions. Again an overall significant anal-
jokes via t tests performed on the six Sex X ysis yielded a nonsignificant Salience X Joke
Salience Condition combinations. The results Content Class interaction, F(4, 552) = 0.72,
revealed a significant preference for salient p < .57. Predictions (c) and (d) were not
victor humor over salient victim humor in validated; a significant main effect for joke
five of six group comparisons. For the sixth content class, F(2, 552) = 43.49, p < .0001,
group (female students-during exam), the yielded the finding that taboo humor was
results were consistent with Hypothesis 3, preferred over the other two types, but salient
revealing no difference between salient victor victim and neutral humor were not signifi-
and salient victim humor. Comparable t tests cantly different.
for all six groups were performed on the
neutral joke subsets, which had been matched Discussion
a priori with the salient victor and salient
victim subsets. None of these / tests yielded This experiment produced no support for
significant differences, thus ruling out a joke the salience concept advanced by Goldstein
structure explanation for the sample's pref- et al. (1972) as the most parsimonious expla-
erence for salient victor over salient victim nation for sources of humor in joke content.
jokes. Furthermore, it is submitted that the present
A reanalysis. Salient victor jokes and the design better approximated the manner in
subsets of neutral and taboo jokes that struc- which cognitive sets are generated outside the
turally corresponded to them were considered laboratory than did the original research. The
as a new and separate data base. Salient failure to generate any hypothesized salience
victim jokes and their neutral and taboo effects in the present study suggests that (a)
cohorts were also separated as a second new the presence of nonsalient humor has signif-
data base. The rationale for this separation icant impact on salience effects and/or (b)
and a reanalysis of the data was the possibility the salience variable is more situation specific
that salience effects operate differently on than Goldstein et al. concluded.
divergent content themes. For example, con- A third possibility is that the present design
sidering the salient victor cohort alone, it was confounded salience with emotional arousal,
predicted that (a) salient victor joke ratings specifically test or performance anxiety. No
SALIENT HUMOR 285

measures of state anxiety were made to ad- the jokes in a college classroom at the request
dress this variable directly; however, the self- of a professor, a mild salience effect for
confidence ratings made by subjects in the examinations is induced. An alternate control
two experimental groups have so ne relevance. group might consist of a comparable subject
The during-exam group displayed a signifi- sample rating the jokes outside the classroom
cantly greater degree of confidence about setting without the presence of the professor.
their collective exam performance than did However, a nonprofessored, nonclassroom
the before-exam group (t = 3.15, p < .01). setting introduces new confounding variables:
One may infer that lower self-confidence is The exact sociological and milieu aspects of
associated with increased test anxiety. The a college classroom would be extremely dif-
experimental groups then align in the follow- ficult to recreate in a student union, for
ing order of emotional arousal: control group, example. Furthermore, it is well-established
during-exam group (subjected to perfor- that the social context of humor has a decided
mance-for-grade stress), and before-exam impact on its quantity and quality. Other
group (subjected to anticipation-of-the-un- designs merit consideration in addressing the
known as well as performance-for-grade "Is it salient motivation or just motivation?"
stress). With respect to salience, the three question. One might simply include a fourth
groups align in the following order: control, content set of jokes wherein victor and victim
before-exam (subjective time and attention themes are played out with police officers,
spent in anticipating and preparing for the for example.
exam), and during-exam (subjective time and Perhaps the most striking finding overall
attention spent in anticipating, preparing, is the consistently greater mirth aroused by
and performing the exam). Thus it would taboo humor. The twin influences of content
appear that there is no confound of the salience and mastery theme are not sufficient
salience and arousal constructs. One unex- to merit salient victor jokes' higher ratings
pected finding was the emergence of a salience than comparably structured taboo jokes. This
effect on general humor responsiveness. Be- cannot be explained as content variety. Al-
cause it was not associated with any specific though the taboo jokes embraced only four
class of humor content, terming it a salience different themes and the salient jokes only
effect is a misnomer. The greater sense of two, the neutral jokes were the most varied
humor shown by the during-exam group is among the three content classes and yet were
better interpreted as evidence that a modest consistently rated as less funny than taboo
degree of state anxiety elevates humor ratings jokes. Because incongruity structure was con-
more than a high degree of anxiety (before- trolled, is the only recourse some variant of
exam group) or a low one (control group). Freud's drive reduction model?
Dividing the salient jokes into victor and When one looks at Freud's (1905/1960)
victim theme subsets revealed differential re- theory in its process aspects (the violation of
sponding by subjects. The question must be a social taboo) rather than the content aspects
raised as to what degree the salience of the that he emphasized (sex, aggression), there
content contributed to these effects. Does appears to be a productive shift in pespective
salience make any appreciable impact above (Kuhlman, 1984). The experience of trans-
and beyond a general motive to identify gressing or transcending social strictures in a
oneself with mastery and alienate oneself safe play setting such as humor can be seen
from failure? The relative humor content as inherently pleasurable. One advantage of
preferences of the control group, who rated shifting from a drive reduction to a taboo
the jokes without the specter of an examina- violation perspective is the strong parallel
tion, were identical to the relative preferences that immediately appears with respect to
of the experimental groups. This suggests that cognitive theories of humor. Incongruity
general mastery-failure motives are being structures violate principles of logic (thought);
tapped by the jokes and that the salience of taboo content violates the principles of con-
examination content is of little consequence. duct (behavior). Alford (1982) has drawn a
It may be argued that the control group is similar analogy. He describes incongruity hu-
not genuine; that is, that by virtue of rating mor as a violation of "phenomenal expecta-
286 THOMAS L. KUHLMAN

tions" and taboo humor as a violation of Ehrenstein, W. H., & Ertel, S. (1978). Zur genese des
"idealized expectations." The taboo violation lustigkeitseindrucks [On the genesis of the mirth re-
concept also explains the popularity of a wide sponse]. Psychologische Beitrage, 20, 360-374.
Freud, S. (1960). Jokes and their relation to the uncon-
range of common joke topics—drugs, alcohol, scious. New York: Norton. (Original work published
politicians, police officers—that touch on ob- 1905)
stacles to one's self-determination. Godkewitsch, M. (1976). Thematic and collative properties
Taboo violation is also superior to the of written jokes and their contribution to funniness.
Canadian Journal of Behavioral Science, 8, 88-97.
drive reduction concept in examining humor Goldstein, J. H. (1970). Repetition, motive arousal, and
within its interpersonal contexts. Societal in- humor appreciation. Journal of Experimental Research
junctions against sexual and aggressive be- in Personality, 4, 90-94.
havior may be respected but not fully accepted Goldstein, J. H., Suls, J. M., & Anthony, S. (1972).
Enjoyment of specific types of humor content: Moti-
(except by those who show no sense of humor vation or salience? In J. H. Goldstein & P. E. McGhee
whatsoever). In a "dirty joke" interaction, (Eds.), The psychology of humor. New York: Academic
the sender violates a social taboo and the Press.
receiver's laughter registers both surprise and Kuhlman, T. L. (1984), Humor and psychotherapy.
acceptance. A momentary intimacy bond is Homewood, IL: Dow Jones-Irwin.
Lamb, C. W. (1968). Personality correlates of humor
formed between the two, which sets the dyad enjoyment following motivational arousal. Journal of
apart from some larger group. On the other Personality and Social Psychology, 9, 237-241.
hand, if the receiver is not well known to the Landy, E., & Metee, S. (1969). Evaluation of an aggressor
sender or is, in fact, an adversary, the same as a function of exposure to cartoon humor. Journal
of Personality and Social Psychology, 12, 66-71.
dirty joke may provoke resentment or outrage. McGhee, P. E. (1979). Humor: Its origin and development.
In this case the apparent intimacy invitation San Francisco: W. H. Freeman.
is rejected because the taboo in question is Mueller, C. W., & Donnerstein, E. (1983). Film-induced
off-limits to the two parties before further arousal and aggressive behavior. Journal of Social
negotiation of the relationship. Freud (1905/ Psychology, 119, 61-67.
Singer, D. L. (1968). Aggression, arousal, hostile humor,
1960) was not ignorant of such interpersonal and catharsis. Journal of Personality and Social Psy-
aspects of humor. However, his greater interest chology Monograph Supplements, 5(1, Pt. 2).
in intrapersonal dynamics and his hydraulic Suls, J. M. (1977). Cognitive and disparagement theories
conception of drives contributed to the still- of humour: A theoretical and empirical synthesis. In
A. J. Chapman & H. C. Foot (Eds.), It's a funny thing,
prevalent disjunction of cognitive and moti- humour (pp. 41-47). Oxford, England: Pergamon Press.
vational approaches to humor. Trice, A. D. (1982). Ratings of humor following experience
with unsolvable tasks. Psychological Reports, 51, 1148.
References Zillmann, D., & Bryant, J. (1974). Retaliatory equity as
a factor in humor appreciation. Journal of Experimental
Alford, R. (1982, August). The evolutionary significance Social Psychology, 10, 480-488.
of the human humor response. Paper presented at the Zillmann, D., & Bryant, J. (1980). Misattribution theory
Third International Conference on Humor, Washington, of tendentious humor. Journal of Experimental Social
D.C. Psychology, 16, 146-160.
Baron, R. A. (1978a). Aggression-inhibiting influence of Zillmann, D., & Cantor, J. R. (1972). Directionality of
sexual humor. Journal of Personality and Social Psy- transitory dominance as a communication variable
chology, 36. 189-197. affecting humor appreciation. Journal of Personality
Baron, R. A. (1978b). The influence of hostile and and Social Psychology, 24, 191-198.
nonhostile humor upon physical aggression. Personality
and Social Psychology Bulletin, 4, 77-80.
Dworkin, E. S., & Efran, J. S. (1967). The angered: Their
susceptibility to varieties of humor. Journal of Person- Received February 1, 1983
ality and Social Psychology, 6, 233-236. Revision received March 5, 1984

S-ar putea să vă placă și