Effect of timing of information about mitigating circumstances on emotional responses to provocation and retaliatory behavior

Effect of timing of information about mitigating circumstances on emotional responses to provocation and retaliatory behavior

JOURNAL OF EXPERIMENTAL SOCIAL PSYCHOLOGY 12, 38-55 (1976) Effect of Timing of Information about Mitigating Circumstances on Emotional Responses to ...

1MB Sizes 0 Downloads 22 Views

JOURNAL OF EXPERIMENTAL SOCIAL PSYCHOLOGY

12, 38-55 (1976)

Effect of Timing of Information about Mitigating Circumstances on Emotional Responses to Provocation and Retaliatory Behavior DOLF Z I L L M A N N AND JOANNE R. C A N T O R Indiana University Received March 3, 1975 To determine whether the reduction of retaliatory behavior by knowledge of mitigating circumstances is due to less motivation to retaliate or to an inhibition of motivated retaliation, subjects were provoked by a rude experimenter and informed of mitigating conditions (a) before provocation, (b) after provocation, or (c) not at all. Physiological data revealed that prior knowledge of mitigation prevented pronounced excitatory responses to provocation. In contrast, when mitigating conditions were not known, excitatory responses to provocation were intense. In addition, when mitigating information was supplied after provocation, excitatory responses decayed more rapidly than when no such information was supplied. Retaliatory behavior, as measured in complaints about the rude experimenter, was substantially lower in the condition in which mitigation preceded provocation than in the other two conditions. The retaliatory behavior of subjects who were informed of mitigation after being provoked did not differ significantly from that of subjects who were not informed of mitigation. The findings were interpreted as incompatible with the assumption that under mitigating conditions retaliation is motivated but inhibited and as generally supportive of the proposal that mitigation attenuates the response to provocation. In order to explain the failure of the reception of mitigating information after provocation to reduce retaliatory behavior in spite of the observed facilitation of excitatory decay, it was suggested that when subjects were experiencing high levels of anger, they formed a behavioral disposition to retaliate, which outlasted the state of elevated arousal. A number of investigations have demonstrated that when a frust r a t i n g a c t is p e r c e i v e d as nonarbitrary, justifiable, accidental, or simply as occurring under mitigating circumstances, less overt hostility and aggression toward the frustrating agent results than when the same a c t is p e r c e i v e d a s a r b i t r a r y , u n j u s t i f i e d , i n t e n t i o n a l , o r u n m i t i g a t e d

This research was supported by Grant GSOC-7205471~ from the National Science Foundation to Dolf Zillmann. Requests for reprints should be sent to Dolf Zillmann, Institute for Communication Research, Radio-TV Center, Indiana University, Bloomington, Indiana 47401. Joanne Cantor is now at the University of Wisconsin, Madison, Wisconsin. Jennings Bryant and Kenneth D. Day served as experimenters. Their assistance is gratefully acknowledged. Copyright © 1976 by Academic Press, lnc, All rights of reproduction in any form reserved.

38

TIMING OF MITIGATING INFORMATION

39

(Burnstein & Worchel, 1962; Cohen, 1955; Kregarman & Worchel, 1961; Mallick & McCandless, 1966; Pastore, 1952; Rothaus & Worchel, 1960; Zillmann, Bryant, Cantor, & Day, 1975). The mechanism underlying this effect has remained unclear, however. It is possible that when the individual is aware of mitigating circumstances, he develops less anger toward his frustrator and thus is less motivated to perform hostile or aggressive acts against him. It is also possible, however, that the perception of the nonarbitrariness of a frustrating act does not hinder the development of anger but causes the inhibition of motivated hostile and aggressive actions. Either of these two mechanisms may be solely responsible for the effect, or both may be operative, either independently or interdependently. Pastore (1952), who assessed the effect of arbitrariness in questionnaire responses to hypothetical situations, mentioned both mechanisms as possible underlying factors, but his study was not designed to implicate a particular rationale. Cohen (1955), who also used questionnaire measures, seemed to favor the mechanism of less instigation to aggress, although, again, his procedure did not permit a comparative evaluation of the two mechanisms. Worchel and his associates (Burnstein & Worchel, 1962; Kregarman & Worchel, 1961; Rothaus & Worchel, 1960), on the other hand, attempted to evaluate the two mechanisms comparatively and reported that their findings brought support for the inhibition mechanism. However, rather than assessing the critical distinction between the two rationales directly by measuring the level of aggressive motivation produced in arbitrary and nonarbitrary situations, they used highly indirect procedures. These procedures involved the assessment of aggression toward the frustrator in situations in which the social sanctions (i.e., external inhibitions) against behaving aggressively were varied and the assessment of displaced aggression toward innocent targets. The researchers thus interpolated constructs Whose validity remains as much in doubt as that of the phenomenon under investigation. Although they observed that aggression in response to nonarbitrary frustrations increased as the social sanctions against aggression decreased, the conclusion that the arbitrariness effect is due to an inhibition of motivated aggressiveness is not compelling. Similarly, although they reported that aggression toward the frustrator was higher under arbitrary than under nonarbitrary frustration, whereas subsequent aggression by the same subjects toward other targets was higher under nonarbitrary than under arbitrary frustration, it does not necessarily follow that subjects in the two conditions of arbitrariness were equally motivated to aggress against the frustrator. Thus, neither of these findings can be considered to demonstrate conclusively that aggressive motivation toward a frustrator is similarly high in response to arbitrary and nonarbitrary frustrations.

40

ZILLMANN AND CANTOR

In the studies described above, the arbitrariness manipulation was always introduced prior to frustration or concomitant with it. Thus, whether information about mitigating circumstances prevented anger from developing, rendering inhibition unnecessary, or whether anger developed fully and inhibition reduced its retaliatory consequences, the observed overt behavior would have been the same. Two studies conducted more recently have assessed the effect of mitigating information that is not acquired until after the provocation treatment. Mallick and McCandless (1966) showed that retaliatory behavior can be reduced when a reasonable explanation for the provoker's misbehavior is supplied subsequent to provocation. Similarly, Zillmann, Bryant, Cantor, and Day (1975) showed that informing a subject of mitigating circumstances after a provocation decreases the level of retaliation unless the subject is in a state of extremely high arousal when he receives the information. Neither of these studies can adequately resolve the controversy regarding the mechanism underlying the observed effect, however, because it still remains unclear whether the subjects who learned belatedly about the mitigating circumstances remained motivated to retaliate but inhibited retaliation or whether the knowledge of mitigating circumstances caused a reduction of hostile feelings and diminished subjects' desire to retaliate. Without some independent means of assessing emotional responses, it does not seem possible to establish conclusively the validity of one mechanism or the other. In order to test whether nonarbitrary frustrations produce less of a motivation to aggress than arbitrary frustrations, it must be determined whether the same frustrating or provoking act produces less emotional anger at the time of provocation when it is perceived as nonarbitrary than when it is perceived as arbitrary. Further, it must be determined whether an individual's recovery from his emotional response to a provoking incident i s facilitated by the subsequent acquisition of information about the nonarbitrariness of the provocation. To carry out these comparisons, the present experiment involves a constant provocation treatment and a manipulation of the time at which information about arbitrariness is introduced. An assessment of the emotional responses and retaliatory behavior of provoked individuals is made (a) when awareness of mitigating circumstances precedes the provocation, (b) when the same mitigating circumstances become known after provocation but before the opportunity to retaliate, and (c) when no mitigating information is supplied. Emotional responses following the provocation treatment are measured in physiological excitation, that is, in increases in sympathetic activation. In accord with the two-factor theory of emotion (cf. Schachter, 1964), it is assumed that the provoking circumstances induce in the subject

TIMING OF MITIGATING INFORMATION

41

a state of anger toward the provoker and that higher levels of physiological excitation reflect more intense feelings of anger. In line with the speculations of Cohen and contrary to the reasoning of Worchel and his associates, it is anticipated that prior knowledge of mitigating circumstances will interfere with the development of anger and hostility toward the provoker. Subjects who learn of mitigating circumstances before being provoked are expected to exhibit less intense physiological responses to the provocation than subjects who are unaware of mitigating circumstances when provoked. This prediction is based on the reasoning that knowledge of the nonarbitrariness of the act affects the subject's appraisal of the attack and helps the subject to control his emotional response (cf. Lazarus, 1966). Whereas subjects who are unaware of mitigating circumstances should attribute the attack to sheer malice on the part of the attacker, subjects with prior knowledge of mitigating circumstances should be able to attribute the assault, at least in part, to these circumstances. Consequently, they should perceive the attack as less malevolent. Moreover, subjects who remain unaware of mitigating circumstances are expected to perpetuate and possibly even increase their arousal state by brooding over their mistreatment, reiterating cognitions related to the upsetting experience (cf. Zillmann & Johnson, 1973), and justifying their anger to themselves. In contrast, subjects with knowledge of mitigating circumstances may be able to utilize this knowledge in reducing their emotional arousal. The process alluded to is related to Lazarus's notion of "secondary appraisal" or his hypothesized process of "short-circuiting threat" (cf. Lazarus, Opton, Nomikos, & Rankin, 1965). In sum, because they should develop less anger and recover more efficiently from their emotional disturbance, subjects who are made aware of mitigating circumstances before being provoked are expected to experience weaker motivations to retaliate and, as a consequence, to engage in less retaliatory behavior than subjects who are not made aware of mitigating circumstances. Regarding the comparison between subjects who receive the mitigating information subsequent to provocation and those who do not receive this information, it is expected that the belatedly acquired knowledge of mitigating circumstances will facilitate excitatory decay. This facilitation may be brought about by the same mechanism that was postulated for accelerated recovery in subjects with prior knowledge of mitigation. If the knowledge of mitigating circumstances should significantly lower the arousal associated with the experience of anger, retaliatory behavior should be reduced. Under these circumstances, the observed reduction in retaliatory behavior would

42

ZILLMANN AND CANTOR

be attributable to a reduced motivation to retaliate rather than to an inhibition of motivated retaliation. Finally, regarding the comparison of the behavior of subjects who are made aware of mitigating circumstances before vs. after provocation, predictions cannot be made until after inspection of the physiological data. If the belated awareness of mitigating circumstances should facilitate excitatory decay to such an extent that the levels of arousal in the conditions of prior and subsequent awareness of mitigation do not differ from each other at the time of the opportunity to retaliate, no difference in retaliatory behavior should be expected. On the other hand, recovery from anger may not be as complete in the condition of subsequent mitigation as in the condition of prior mitigation, since in the latter condition, anger should be aroused less intensely and decay should be initiated sooner. If, as expected, arousal levels remain higher in the condition of subsequent mitigation than in the condition of prior mitigation, retaliatory behavior is expected to be more intense in the former condition than in the latter. The above predictions contrast with predictions based on the reasoning of Worchel and his associates. According to their reasoning, aggression should be equally motivated in all three conditions. Thus, physiological responses to provocation and the decay of elevated excitation should remain unaffected by the manipulation of arbitrariness. As far as retaliation is concerned, subjects who hear about mitigating circumstances either before or after provocation should retaliate less than subjects who are never made aware of mitigating circumstances. Moreover, since the information known at the time of retaliation by subjects who hear about mitigation before and after provocation is identical, the social pressures to inhibit retaliation should also be identical. There should thus be no difference in the levels of retaliation performed by subjects in these two conditions.

METHOD

Subjects Forty-five male undergraduates at Indiana University served as subjects. They were recruited by means of announcements posted on campus and were paid for their participation. To minimize the potential effect of gossiping about the experiment, the data collection was completed in 3 consecutive days.

Design All subjects were provoked by a rude experimenter. In one condition, they were informed, before being provoked, of mitigating circumstances underlying the experimenter's behavior (prior mitigation). In a second condition, they received the same information after being provoked (subsequent mitigation). In a third condition, they were never informed of mitigating circumstances (no mitigation). Physiological re-

TIMING OF MITIGATING INFORMATION

43

sponses were assessed to determine the intensity of emotional responses to the provocation and to trace excitatory decay. Confidential evaluations of the rude experimenter, which purportedly were to have real consequences for his academic future, served to assess subjects' retaliatory behavior.

Procedure Subjects were tested one at a time and were randomly assigned to conditions. The procedure involved the subject's interaction with two experimenters, one who was polite, and one who was rude and insulting. During the course of the session, the subject saw the rude experimenter mistreat the polite experimenter on two occasions. These instances of misconduct, one occurring before provocation of the subject and one occurring after, 'gave the polite experimenter a plausible reason for telling the subject about the mitigating circumstances underlying the rude experimenter's behavior. All subjects were treated identically with the exception of the placement or absence of the mitigating-circumstances comment. The rude experimenter was not aware of the experimental condition of the subject. The subject was met in the experimental room by the polite experimenter, who played a tape recording of the instructions. The experiment was said to test some of McLuhan's hypotheses regarding the perception of visual stimuli. It was explained that McLuhan's speculations about "gut responses" to television were being put to a test by assessing subjects' physiological responses--heart rate, blood pressure, and skin temperature--while being exposed to visual stimuli. The experiment was said to involve two independent sections that were being run by two different experimenters. In the first section, subjects would be exposed to "high-information" stimuli, slides of magazine advertisements, and in the second part, they would be exposed to "low-information" stimuli, a video tape of the faces of people responding to a variety of stimuli. Subjects were told that they would not have to perform any task other than to look at the stimuli. After answering any questions the subject raised about the instructions, the experimenter hooked the subject up to the physiological measuring equipment, turned on the apparatus that took continuous readings of heart rate and skin temperature, and took a blood-pressure reading. He then turned on a slide projector that showed a series of slides of magazine advertisements, automatically changing slides at 8-sec intervals. After the slide-viewing period, which lasted 3 min, the experimenter again took the subject's blood pressure. After blood pressure had been recorded, and while the polite experimenter seemed to be making some numerical calculations related to the physiological measures, the rude experimenter came into the room and impatiently asked, "Haven't you finished yet?" The polite experimenter replied that it would take him just another minute. The rude experimenter then looked over at the subject and accused the polite experimenter of having hooked him up to the physiological equipment incorrectlY. The polite experimenter denied this, but the rude experimenter insisted. The polite experimenter then demonstrated that he was correct by pointing out the color-coding of the electrodes. Having been proved wrong, the rude experimenter gave a disgruntled look, but did not apologize in any way for his false accusation, and walked out of the room. It was at this point that subjects in the prior-mitigation condition were told by the polite experimenter, " H e ' s really uptight about a midterm that he has tomorrow." In the other two conditions, the polite experimenter made no comment at this time. Thirty seconds after the rude experimenter's exit, the polite experimenter, pretending to have finished his calculations, left the room and called in the rude experimenter. The rude experimenter entered the room and stated, with an air of impatience and apparent contempt for the subject, "All that you're required to do in this study

44

ZILLMANN AND CANTOR

is to watch the TV. All that I ask is that you sit very still. Any questions?" The subject never raised any questions at this point, and the experimenter turned on the television, which showed a video tape of a series of still pictures of undergraduates exhibiting a variety of facial expressions. The experimenter did not look at the subject at this time but kept his eyes on the physiological measures being recorded on the oscillograph. After the subject had viewed 2 min of the tape, the experimenter walked over to the video-cassette recorder and turned it off. With a disgusted look on his face, he said, "Apparently you don't listen to instructions very well. You were told that you had to sit still. This machine shows that you've been moving." At this point he paused, and the subject usually protested. No matter what the subject said, the experimenter continued, " I ' m going to start the tape over, and I'll have to take more base-level measures." The experimenter rewound the video tape and took a bloodpressure reading. He then started the tape again, making the comment, "Try and cooperate this time." At the end of the video tape, which lasted 3 min, the rude experimenter left, and the polite experimenter came back into the room and took a bloodpressure reading. Then, while the polite experimenter was making more computations, the rude experimenter reentered the room and asked the polite experimenter whether he had rewound the video tape yet. When the polite experimenter replied that he had not yet done it, the rude experimenter contemptuously walked over to the video-cassette recorder and rewound it himself, saying, "If you want your audiotape rewound, you had better start rewinding this." He then left, closing the door abruptly. At this point, subjects in the subsequent-mitigation condition heard the polite experimenter say " H e ' s really uptight about a midterm that he has tomorrow." The experimenter said nothing to the subjects in the other two conditions at this time. Thirty seconds after the rude experimenter left the room, the polite experimenter took a bloodpressure reading. He took a final reading 1.5 rain after that. He then disconnected the physiological equipment, told the subject the experiment was over, and sent him into another room to fill out a short questionnaire. In the next room, the subject was met by a third experimenter, who did not know which experimental condition the subject had been in. The subject was told that in accord with a new requirement on campus, he would have to fill out a questionnaire for the "Committee on Research Subjects" (a fictitious label). The experimenter explained that this committee had been set up for the protection of the subject and that they wanted to know if anyone had any complaints about what he was made to do in an experiment or how he was treated. She explained that the top half of the form related to the experiment itself and was to be used by the Committee as an administrative check on the experiment. The bottom half was said to deal with the experimenter who ran the study. The ratings made on this half of the form were to be forwarded to the experimenter's major department so that they could be used in evaluating experimenters and in deciding on the coming year's financial support for research assistants. She added that a separate form would be filled out for each experimenter and the part of the experiment that he conducted. She handed the subject two forms that showed the letterhead of the Committee on Research Subjects and had the experimenters' names written in at the appropriate places. She explained that the first form was for the experimenter who showed him slides and the second was for the one who showed him a video tape recording, The experimenter then said that the questionnaires were completely anonymous and told the subject to seal the completed forms in the attached intramural envelope. The envelope had been stamped with a rubber stamp reading "Committee on Research Subjects, 309 Bryan Hall" (this was an arbitrarily chosen room in the Research and Advanced Studies section of the Administration Building). When the subject had completed the forms and sealed the envelope, the experimenter slid the envelope through a slot in a ballot-type box stamped with the "Committee on Research Subjects" label. She then thanked the subject, debriefed him, and paid him.

TIMING OF MITIGATING INFORMATION

45

Apparatus Systolic and diastolic blood pressures were recorded by the cuff method on a Sears sphygmomanometergraph. Heart rate and skin temperature were recorded on a Hewlett-Packard oscillograph. Heart rate was measured from the electrocardiogram with a cardiotachometer, which computed an average over every 15. beats. Skin temperature was assessed on the distal pad of the middle finger of the subject's left hand.

Dependent Measures Physiological changes. The heart-rate and skin-temperature readings that occurred immediately before each blood-pressure measurement were used in the analyses. The initial measure was taken simply to allow the subject to become accustomed to the procedure of blood-pressure measurement. The second measure, taken after exposure to the slides of advertisements, constituted the base level (to). Response to provocation was assessed in the third measure, taken after the subject had been accused of not sitting still and had been told that he would have to start viewing the videotape again from the beginning (tO. Decay or perpetuation of the provocation-induced response was measured when the polite experimenter returned to the room after the subject had seen the complete video tape of the faces (t2). The effect of mitigation subsequent to provocation was measured 30 sec after the comment in the appropriate condition and at the same time in the procedure in the other two conditions (t3). Further excitatory decay was measured 1.5 min later (t4). Evaluations. The confidential questionnaire, which was filled out at the end of the session, gave the subject the opportunity to express his feelings about the way he had been treated in the experiment. The first three questions on the form allowed him to inform the Administration of any complaints he had. The last three questions permitted him to influence decisions to be made about the experimenter's future. Thus, negative ratings on these latter scales can be considered deliberate attempts to inflict negative consequences on the experimenter. The three questions on the top half of the form were answered by marking unipolar scales ranging from 0 to 100. The first question read, "Are you in any way dissatisfied with the way you were treated in this experiment?" Answers were given by intersecting a scale ranging from "not at all dissatisfied" to "extremely dissatisfied." The second question read, "Do you think the demands made upon you in this experiment were in any way excessive?" and the scale following it ranged from "not at all excessive" to "extremely excessive." The third question, "Were you in any way mistreated by the experimenter conducting this experiment?" was answered on a scale ranging from "not mistreated at all" to "mistreated to a great extent." A space for comments was provided below this scale. The last three questions were introduced with the statement, " T o aid t h e departments in determining stipends for research assistantships, we are including the following questions to be forwarded to the experimenter's major department." The questions were answered by marking bipolar scales ranging from -100 to +100. The fourth question, "How well did the above-named graduate student perform in his role as an experimenter?" was answered on a scale ranging from "poorly" to "excellently." The fifth question, "How would you rate his manner of interacting with others?" was followed by a scale ranging from "extremely unpleasant and discourteous" to "extremely pleasant and courteous." The final question read, "In your opinion, should this student be reappointed as a research assistant?" and the answer could range from "definitely not" to "definitely yes." This last question was considered to be the clearest behavioral measure of retaliation because the subject was led to believe he was making an explicit recommendation that would be considered in determining the experimenter's future.

46

ZILLMANN AND CANTOR

RESULTS

Physiological Responses To control for interindividual variation in basal levels of physiological activity, the measures taken at the various critical times during the experiment were converted to change scores, with a subject's baselevel score for a particular measure being subtracted from each of his subsequent scores. Mixed-design analyses of variance, with mitigation conditions as an independent-measure factor and times of measurement as a repeated-measures factor, were performed on all change scores. Systolic blood pressure. The analysis of variance revealed that mitigation conditions produced a significant main effect on changes in systolic blood pressure [F (2, 42) = 5.13, p < .025]. This effect was due to the fact that, overall, systolic pressure was less elevated in the condition in which subjects were aware of mitigating circumstances before being provoked than in the other two conditions. Time of measurement also exerted a significant main effect [F (3, 126) = 33.75, conservatively corrected by the Geisser-Greenhouse method, p < .001] due to a successive decline following the initial elevation from provocation. The interaction between mitigation conditions and time of measurement was nonsignificant [F (6, 126)= 1.27], apparently because there was a decline in this measure over time in all conditions. However, the means of the individual conditions and the subsequent tests of differences, displayed in Table 1, reveal that the course of decay was distinctly different in the three conditions and that awareness of mitigation was a crucial determinant of both the induction and the decay of arousal. As can be seen from the table, all three conditions showed significant systolic blood-pressure increases due to provocation (tl), but subjects who were already aware of mitigating circumstances exhibited significantly less intense responses. Assuming that increases in systolic blood pressure reflected the degree to which subjects responded with anger or irritation to the provocation, this finding is consistent with the notion that an awareness of mitigating circumstances produces a weaker motivation to retaliate. The measures taken after the 3-min exposure to the video tape of faces (tz) reveal that whereas subjects in the priormitigation condition had returned to their base levels during this time, the systolic pressure of subjects in the other two conditions remained significantly elevated. At t3, the newly acquired knowledge of mitigating circumstances in the subsequent-mitigation condition produced a steep and significant decline in systolic pressure--again consistent with the notion that mitigation reduces the motivation to retaliate. In contrast, decay continued to occur at a slower rate in

TIMING OF MITIGATING

47

INFORMATION

TABLE 1 MEAN CHANGES IN PHYSIOLOGICAL RESPONSES AT CRITICAL TIMES DURING THE EXPERIMENT Time of m e a s u r e m e n t

Measure

Immediately after instigation (tl)

3 min after instigation (L2)

30 s e c after second mitigation manipulation (t3)

2 rain after second mitigation manipulation (t4)

A Systolic blood pressure Prior mitigation S u b s e q u e n t mitigation N o mitigation

5.1 a* 10.9 b* 10.8 b*

0.9 a 9.5 b* 7.9 b*

-2.1 a 3.5 b* 5.7 b*

-2.9 a 0.9 ab 4.0 b*

A Heart rate Prior mitigation S u b s e q u e n t mitigation N o mitigation

-8.0 a 6.0 b* 7.8 b*

-6.9 a 7.5 b* 10.2 b*

-8.8 a 1.5 b 6.2 b*

-5.0 a 0.8 a 10.2 b*

Note. All scores are relative to a basal m e a s u r e . Systolic blood p r e s s u r e is m e a s u r e d in m m Hg, a n d heart rate in bpm. L e t t e r superscripts denote vertical comparisons. M e a n s having no letter in their superscripts in c o m m o n differ significantly at p < .05 by C o c h r a n ' s test. M e a n s denoted by asterisks are significantly greater than base level (0) by F test.

the no-mitigation condition. The final measure (t4) revealed that subjects in the subsequent-mitigation condition had returned to base level by the end of the measurement period and that their systolic pressure was no longer significantly above that of subjects in the priormitigation condition. In contrast, subjects who had still not heard about the mitigating circumstances remained significantly above their base levels throughout the measurement period and ended up experiencing significantly higher levels of systolic blood pressure than subjects in the other two conditions. Diastolic blood pressure. Due to high intersubject variability, the analysis of changes in diastolic blood pressure failed to yield any significant effects. Although diastolic pressure tended to remain lowest in the condition in which mitigating information was received prior to provocation, none of the observed differences can be considered reliable. H e a r t rate. Like systolic blood pressure, heart rate was highly responsive to the experimental manipulations. There was a highly significant main effect of mitigation conditions [F (2, 4 2 ) = 11.33, p < .001) due to the fact that, overall, heart rate was the least elevated in the condition in which subjects heard about mitigating circumstances before being provoked. The repeated-measures effect

48

ZILLMANN AND CANTOR

approached significance [F (3, 126)= 3.28, conservatively corrected, p < . 10], reflecting a variety of up-and-down movements in the different conditions. Again, although the interaction was not significant [F (6, 126)= 1.89, conservatively corrected, p > .10], the most revealing data came from the changes in excitation over time in the three conditions. As can be seen from Table 1, subjects who were aware of mitigation prior to being provoked actually showed reduced heart rate after instigation, and their heart rate remained at a low level throughout the rest of the assessment period without changing significantly. In both conditions in which subjects were unaware of mitigation at the time of provocation, heart rate increased significantly from instigation (tl) and tended to be even higher 3 min later (t~). The contrast between the latter two conditions and the condition of prior mitigation again supports the notion of less motivation to retaliate under conditions of nonarbitrary frustration. Moreover, the data suggest that subjects who were unaware of mitigating circumstances may have brooded over their mistreatment while being exposed to the minimally involving video tape, and may even have increased their anger. The measures taken at t3 reveal that heart rate was very sensitive to the mitigation manipulation and changed very rapidly in response to it. In the subsequent-mitigation condition, the comment regarding mitigating circumstances caused a significant drop in heart rate and rendered this measure no longer significantly different from base level, although it was still significantly higher than in the prior-mitigation condition. By the final measurement (t4), the mean heart-rate changes in these two conditions no longer significantly differed. In contrast, the heart rate of subjects in the no-mitigation condition remained elevated throughout the time of assessment. No significant decline in heart rate was observed at any time, and, at the end of the measurement period, heart rate was still significantly higher in this condition than in the other two conditions. Skin temperature. The effect of mitigation conditions on overall skin-temperature changes was nonsignificant [F (2, 42) = 1.80, p > . 10], but the means tended to reflect the least intense emotional responses in subjects who heard about mitigation prior to provocation and the most intense responses in those who did not hear about mitigation. A highly significant main effect of time of measurement was observed [F (3, 126)= 11.43, conservatively corrected, p < .001], due to a series of up-and-down movements of the measure over time. The interaction was nonsignificant. Cochran's test between means at the various time levels revealed a significant differentiation between conditions for the last measure taken. Subjects who had known about the mitigating circumstances from the beginning exhibited a final mean skin-temperature change indicative of the greatest amount of-relaxa-

TIMING OF MITIGATING

49

INFORMATION

tion (0.8°C), which differed significantly from that of subjects who never heard about the mitigating circumstances (-0.3). The mean for subjects who heard about the mitigating circumstances after provocation (0.2) held an intermediate position.

Evaluations Ratings of the nice experimenter. The evaluations given to both the nice and the rude experimenter were subjected to analyses of variance. In the ratings of the nice experimenter and his part of the experiment, the effects of the mitigation manipulation were trivial [for the second measure, F (2, 4 2 ) = 1.52; all other F ratios were approximately 1 or smaller]. The lack of significant differentiations was due to minimal differences between means rather than to high variability. In the evaluations for administrative purposes, few complaints were registered, the range of mean ratings being from 0 to 3 on 100-point scales. The nice experimenter himself received high ratings with great consistency. His mean ratings ranged from a low of 88 to a high of 94 on scales ranging from -100 to +100. These data show that randomization was adequate in that subjects in the different conditions did not differ in their use of the scales. They also show that the comment made by the nice experimenter about the rude experimenter either before or after provocation had no appreciable effect on subjects' evaluations of the nice experimenter or their ratings of the part of the experiment he conducted. Complaints about the experiment. In contrast, as can be seen from Table 2, evaluations of the rude experimenter and his part of the experiment were significantly affected by the mitigation manipulation. In the first three ratings, which the subject was told would be sent to the Administration so that it could correct any abuses being perpetrated on TABLE2 COMPLAINTS ABOUT THE EXPERIMENT Experimental condition

Evaluation

Prior mitigation

Subsequent mitigation

No mitigation

General dissatisfaction E x c e s s i v e n e s s of requirements Mistreatment

26 a 4a 21 a

20 a 8 ab 22 a

58 b 22 b 60 b

Note. T h e higher the score, the stronger the complaint. C o m p a r i s o n s are within m e a s u r e s only. M e a n s having no letter in their superscripts in c o m m o n differ significantly at p < .05 by N e w m a n - K e u l s ' test.

50

ZILLMANN AND CANTOR

experimental subjects, mitigation reduced the level of complaints about the experiment whether it occurred before or after provocation. In the ratings o f the degree to which subjects were dissatisfied with the treatment they received, the main effect of the mitigation manipulation was associated with F (2, 42) = 5.23, p < .01. T h e ratings of the excessiveness of the demands made upon the subject yielded a main effect associated with F (2, 42) = 3.62, p < .05. The table shows that in these ratings, the mean o f the subsequent-mitigation condition did not differ significantly from that of the no-mitigation condition; this difference approached significance, however (p < .10). T h e ratings of how much subjects felt they had been mistreated by the experimenter yielded F (2, 42) = 11.59, p < .001. Retaliatory complaints about the rude experimenter. Subjects' direct retaliatory behavior against their provoker, as reported in Table 3, was affected quite differently by the mitigation manipulations. In the ratings that subjects thought would affect the experimenter's future, the timing o f mitigation was the most influential factor. On all these measures, subjects who had heard about the mitigating circumstances before being provoked gave the p r o v o k e r significantly (p < .01) higher ratings than subjects who received the same information after being provoked or did not receive this information at all. The ratings of the experimenter's performance yielded F (2, 42) = 12.29, p < .001. As can be seen from the table, although the experimenter tended to receive higher performance evaluations in the subsequent-mitigation condition than in the no-mitigation condition, this difference was not significant (p > .10). In the evaluations of the experimenter's manner of interacting with others, the manipulation of arbitrariness was associated with F (2, 42) = 22.39, p < .001. In these ratings, the difference between the subsequent-mitigation condition and the no-mitigation condition reached the .10 probability level. Subjects' recommendaTABLE 3 RETALIATORY COMPLAINTS ABOUT THE RUDE EXPERIMENTER

Experimental condition Evaluation

Prior mitigation

Subsequentmitigation

No mitigation

Performance Courtesy Reappointment

59b 13b 28b

-8 a -58 a -47 a

-36 a -86 a -70 a

Note. The more negative the score, the more negative the evaluation. Comparisons are within measures only. Means having different superscripts differ significantly at p < .01 by Newman-Keuls' test. Means having the same superscript do not differ at p = .05.

TIMING OF MITIGATING INFORMATION

51

tions regarding whether or not the experimenter should be reappointed were associated with F (2, 42) = 18.85, p < .001. Again, the difference between the subsequent-mitigation condition and the no-mitigation condition was nonsignificant. These recommendations yielded the most pronounced difference between the effect of prior mitigation and that of subsequent mitigation. Only one of the 15 subjects who had prior knowledge of mitigation gave the rude experimenter a negative rating on this measure, thus recommending that he not be reappointed. In contrast, 12 of the 15 subjects who heard about the very same mitigating circumstances after being provoked gave negative ratings on this measure. It may be noted that when given the opportunity to retaliate, subjects in the condition of subsequent mitigation had been supplied with an excuse for their provoker's behavior more recently than subjects in the condition of prior mitigation. Because of this time discrepancy, it might be argued that the subsequently acquired mitigating information could have had a stronger effect if more time had elapsed before retaliation. However, it seems more likely that the greater recency of the information served to enhance rather than detract from its retaliation-reducing effect, making the observed difference between the two mitigation conditions conservative. Written c o m m e n t s . In the presence of the third experimenter, many subjects commented spontaneously on the rude experimenter's behavior. Since this had not been anticipated, no preparations had been made to record these potentially revealing comments. However, many subjects also exercised their option to write in comments on the evaluation form, and these comments were subjected to analysis. The written comments, all of which amounted to added complaints about the rude experimenter's behavior, were differentially frequent in the various conditions. As might be expected, the greatest number of comments, eight, occurred in the no-mitigation condition. In the condition of prior mitigation only four comments were recorded, and in the condition of subsequent mitigation only three. The comments seemed to differ substantially in tone, and it appeared that they were particularly severe in the conditions of no mitigation (e.g., "This jackal is sure to learn some manners soon," was one of the more moderate comments; most of these comments involved vulgarities) and subsequent mitigation (e.g., " I felt that I was horribly mistreated. I could not get my breath I got so mad. He should be helped"). In the condition of prior mitigation, the comments seemed more moderate (e.g., "Mr. Day was slightly rude-insinuating that I was attempting to ruin his experiment"). This intuitive assessment was corroborated by the following procedure. Thirteen judges, drawn from the same general population as the subjects, rated the various comments with respect

52

ZILLMANN AND CANTOR

to the degree of hostility expressed in them. The judges were informed about the instigational conditions under which the comments were made. However, no mention was made of the mitigation variation. Each judge was given a booklet presenting the 15 comments, one comment per page. In each booklet, the order of presentation of the comments was individually randomized. Below each comment was a scale ranging from "not hostile at all" (0) to "extremely hostile" (100). For each judge, a mean hostility score was computed for the comments deriving from each of the three experimental conditions. These scores were subjected to a Hotelling's T 2 analysis, which yielded T2o~= 29.95, p < .01. Subsequent analyses of differences between the means (Newman-Keuls' method) revealed that the mean score associated with the condition of prior mitigation (34) was significantly (p < .01) lower than the mean scores associated with both the condition of subsequent mitigation (49) and that of no mitigation (45). The difference between the latter two conditions proved trivial (p > . 10). Overall, then, the analysis of the written comments registering complaints about the rude experimenter produced results that were consistent and redundant with the basic findings concerning retaliatory complaints.

DISCUSSION The findings of the present investigation clearly do not support the notion of an inhibition mechanism, as proposed by Worchel and his associates. The physiological data provide strong evidence that emotional responses are less intense when mitigating circumstances are known than when they are not. Those subjects who knew about the mitigating circumstances before the provocation treatment apparently did not become as angry upon being treated in a provoking manner as those who did not know about them. In addition, among subjects who were provoked while remaining ignorant of mitigating circumstances, those who subsequently heard about the mitigating circumstances recovered more quickly from their emotional response than those who never heard about the mitigating circumstances. The physiological data also revealed that subjects who had received information about mitigating circumstances either before or after provocation had returned to their base levels of physiological arousal by the time of the opportunity to register complaints, whereas those who had not been told of mitigating circumstances had not. Thus, the finding that complaints to the Administration were similar in the two mitigation conditions and significantly higher in the condition of no mitigation can be accounted for by the rationale of differential instigation. This finding is also consistent with the inhibition rationale,

TIMING OF MITIGATING INFORMATION

53

but in light of the physiological data, the inhibition mechanism cannot be considered to have been supported. With respect to subjects' retaliatory actions against the rude experimenter, neither of the discussed rationales can account for all of the findings. The observation that the timing of mitigation played such a crucial role in determining subjects' tendencies to inflict negative consequences on their provoker is clearly inconsistent with expectations based on the inhibition mechanism. Since subjects in the two mitigation conditions had the same knowledge of mitigating circumstances at the time they gave their evaluations, inhibitions on retaliation should have been identical, and the two conditions should have shown equal levels of retaliatory behavior. In addition, both conditions of mitigation should have exhibited lower levels of retaliation than the condition of no mitigation. The physiological data show the findings regarding retaliatory behavior to be inconsistent with the notion of differential instigation, too, however. Since arousal in the two mitigation conditions did not differ significantly at the time evaluations were made, the levels of retaliation should not have differed. One way of partially reconciling these data would be to point out that although the arousal of subjects in the condition of subsequent mitigation did not differ significantly from that of subjects in the condition of prior mitigation at the end of the assessment period, it was still at a somewhat higher level. But even taking this nonsignificant difference into consideration, the mechanism of differential instigation alone cannot account for the fact that retaliation in the condition of subsequent mitigation failed to be significantly lower than that in the condition of no mitigation, when physiological arousal levels were significantly lower in the former condition than in the latter. It will be noted that the mean evaluations were consistently, although not significantly, less negative in the former than in the latter condition. (Regarding comments, the difference was in the opposite direction, yet trivial.) However, it still remains to be explained why the behavior of subjects who learned about mitigating circumstances after being provoked was so much more similar to that of subjects who never learned about mitigating circumstances than it was to that of subjects who were aware of mitigation at the time of provocation. In order to account for the ineffectiveness of the knowledge of mitigating circumstances when it came after provocation, it is proposed that subjects who had reacted truly angrily in the absence of the mitigating circumstances had committed themselves to a retaliatory course of action that could not readily be altered by new information. Previous research (Zillmann & Bryant, 1974) has shown that intensely aroused anger may produce a disposition to retaliate that

54

ZILLMANN AND CANTOR

has committal value and that lasts beyond the duration of the excitatory residues associated with the experience of intense anger. In the present experiment, the angry subject may have felt that the only way to rid himself of his anger would be to get back at the person who had made him angry (cf. Zillmann, 1972). Having made up his mind to retaliate if the opportunity should arise, he may have rejected the implications of the subsequently acquired knowledge of mitigating circumstances. He may have felt that since he had already been made to suffer, the damage had already been done, and the excuse had come too late. Thus, he may have ignored the mitigation in his retaliatory behavior in spite of the fact that the awareness of mitigating circumstances had caused his emotional anger to subside. It is of interest to note that complaints to the Administration about the experiment were affected in one way and direct retaliatory behavior was affected in another. Apparently, subjects were more "rational" in their complaints to the Administration. In these ratings, the knowledge that the experimenter was under pressure, and thus that what was witnessed may not have been a fair sample of his behavior, affected ratings equally, whether it had been acquired before or after provocation. However, when the opportunity to inflict noxious consequences on the experimenter was at stake, the ability or the desire to take mitigating circumstances into account was strongly reduced in subjects who had experienced high levels of anger.

REFERENCES Burnstein, E., & Worchel, P. Arbitrariness of frustration and its consequences for aggression in a social situation. Journal of Personality, 1962, 30, 528-540. Cohen, A. R. Social norms, arbitrariness of frustration, and status of the agent of frustration in the frustration-aggression hypothesis. Journal of Abnormal and Social Psychology, 1955, 51, 222-226. Kregarman, J. J., & Worchel, P. Arbitrariness of frustration and aggression. Journal of Abnormal and Social Psychology, 1961, 63, 183-187. Lazarus, R. S. Psychological stress and the coping process. New York: M c G r a w Hill, 1966. Lazarus, R. S., Opton, E. M., Nomikos, M. S., & Rankin, N. O. The principle of short-circuiting of threat: Further evidence. Journal of Personality, 1965, 33, 622-635. Mallick, S. K., & McCandless, B. R. A study of catharsis of aggression. Journal of Personality and Social Psychology, 1966, 4, 591-596. Pastore, N. The role of arbitrariness in the frustration-aggression hypothesis. Journal of Abnormal and Social Psychology, 1952, 47, 728-731. Rothaus, P., & Worchel, P. The inhibition of aggression under nonarbitrary frustration. Journal of Personality, 1960, 28, 108-117. Schachter, S. The interaction of cognitive and physiological determinants of emotional state. In P. H. Leiderman & D. Shapiro (Eds.), Psychobiological approaches to social behavior. Stanford: Stanford Univ. Press, 1964.

TIMING OF M I T I G A T I N G INFORMATION

55

Zillmann, D. The role of excitation in aggressive behavior. In Proceedings of the Seventeenth International Congress of Applied Psychology, 1971. Brussels: Editest, 1972. Zillmann, D., & Bryant, J. The effect of residual excitation on the emotional response to provocation and delayed aggressive behavior. Journal of Personality and Social Psychology, 1974, 30, 782-791. Zillmann, D., Bryant, J., Cantor, J. R., & Day, K. D. Irrelevance of mitigating circumstances in retaliatory behavior at high levels of excitation. Journal of Research in Personality, 1975, 9~ 282-293. Zillmann, D., & Johnson, R. C. Motivated aggressiveness perpetuated by exposure to aggressive films and reduced by exposure to nonaggressive films. Journal of Research in Personality, 1973, 7, 261-276.