Rage and reason: the psychology of the intuitive prosecutor

This study explores the conditions under which experimentally primed anger in ̄uences both attributions of responsibility and the processes by which people make such attributions. Drawing on social functional theory, it was hypothesized that people are best thought of as `intuitive prosecutors' who lower their thresholds for making attributions of harmful intent and recommending harsh punishment when they both witness a serious transgression of societal norms and believe that the transgressor escaped punishment. The data support the hypotheses. Anger primed by a serious crime `carried over' to in ̄uence judgments of unrelated acts of harm only when the perpetrator of the crime went unpunished, notwithstanding the arousal of equally intense anger in conditions in which the perpetrator was appropriately punished or his fate was unknown. Participants in the perpetrator-unpunished condition also relied on simpler and more punitive attributional heuristics for inferring responsibility for harm. Copyright # 1999 John Wiley & Sons, Ltd. A classic tension exists between rage and reason ( for discussions, see Aristotle, 325 BC/1962; Freud, 1924; Kant, 1781/1900; Lerner, Goldberg & Tetlock, 1998; Solomon, 1990, 1994; Tavris, 1989). Scholars have long suspected, and researchers have subsequently demonstrated, that there are numerous ways in which anger, once activated, degrades subsequent reasoning processes. Even when the object of subsequent judgments bears no relation to the source of one's anger, anger increases: (1) a desire to blame individuals, (2) tendencies to overlook mitigating details before CCC 0046±2772/99/050781±15$17.50 Received 9 March 1998 Copyright # 1999 John Wiley & Sons, Ltd. Accepted 12 October 1998 European Journal of Social Psychology Eur. J. Soc. Psychol. 29, 781±795 (1999) *Correspondence to: Julie Goldberg, 3210 Tolman Hall, Department of Psychology, University of California, Berkeley, Berkeley, CA 94720, USA. e-mail: [email protected] Contract/grant sponsor: US National Science Foundation. Contract/grant number: SRB-9505680. attributing blame, (3) tendencies to perceive ambiguous behavior as hostile, (4) tendencies to discount the role of uncontrollable factors when attributing causality and (5) punitiveness in response to witnessing mistakes made by others (Keltner, Ellsworth & Edwards, 1993; Lemerise & Dodge, 1993; Lerner et al., 1998; Quigley & Tedeschi, 1996). One might conclude, just as Kant and Freud did, that rage and reason are mutually exclusive. Anger only interferes with `rational' cognitive processes. It turns `intuitive scientists' (who seek to understand why events occur) into `intuitive prosecutors' (who have low thresholds for axing blame and imposing penalties). But is indiscriminate punitiveness the inevitable result of anger? Are there no stopping mechanisms for the tendency of anger to color unrelated judgments? In this article, we test hypotheses derived from a new social±functional analysis of emotion and judgment that challenge the view that rage and reason are always mutually exclusive. More speci®cally, we specify the conditions under which anger will and will not in ̄uence subsequent judgments in situations that are unrelated to the source of one's anger. This social±functional approach to emotion and judgment posits that the human emotion system serves both interpersonal and intrapersonal functions ( for related discussions, see Durkheim, 1843/1984; Frijda & Mesquita, 1994; Lerner & Keltner, in press; Parkinson, 1997). The approach assumes that emotions serve interpersonal functions by producing speci®c action tendencies which allow individuals to adaptively interact with their social environment, such as by forming attachments, resolving injustices, negotiating hierarchies, and adhering to social norms (Barrett & Campos, 1987; Ekman, 1992; Frank, 1988; Keltner & Kring, in press; Lutz & White, 1986; Nesse, 1990; Schwarz, 1990). Emotions also serve intrapersonal functions by interrupting ongoing cognitive processes to direct attention, memory, and judgment to deal quickly with encountered problems or opportunities in the environment (see Johnson-Laird & Oatley, 1992; Lazarus, 1991; Schwarz, 1990; Simon, 1967; Tooby & Cosmides, 1990). For example, cognitive appraisals of injustice lead to anger, whereas appraisals of uncertainty regarding future harm invoke fear (Smith & Ellsworth, 1985). These appraisals then trigger a set of cognitions and actions to address the goal of redressing injustice, in the case of anger, and avoiding harm in the case of fear (Frijda &Mesquita, 1994; Lerner et al., 1998; Lerner & Keltner, in press). Indeed, the tendency to perceive new information in ways that are consistent with the appraisal pattern underlying a salient emotion can be so strong that they persist beyond the event that elicited the emotion. For example, one study found that feelings of anger persisted to color participants' attributions of causality in other, unrelated situations, such that they were more likely to view individuals as causal agents rather than situational factors (Keltner et al., 1993). In summary, like other advocates of a social functional approach to emotion (e.g., Barrett & Campos, 1987; Frijda & Mesquita, 1994; Lazarus, 1991), we argue that emotions tie individuals to their social world by rapidly activating cognitive and behavioral response tendencies which allow individuals to adapt to their environment. More importantly, though, as Lerner and Keltner (in press) also suggest, we specify a stopping mechanism for these tendencies. To the extent that the emotionally arousing situation is resolved or the opportunity responded to, emotions should no longer in ̄uence subsequent judgments (cf. Frijda, 1988; Haidt, Silvia & Dias, 1993), even if the emotion persists experientially. When applied to anger, two testable hypotheses ̄ow from this social±functional approach. The ®rst hypothesis addresses whether anger activated in one situation will 782 J. H. Goldberg et al. Copyright # 1999 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 29, 781±795 (1999) in ̄uence judgments in other, unrelated situations. Speci®cally, we predict that if participants learn that justice was not served after an anger-eliciting event, then their anger will direct attention, memory, and judgment toward the goal of redressing injustice, creating `intuitive prosecutors'. As a consequence, participants' anger over injustice in one situation should predict participants' willingness to punish future, unrelated transgressors. By contrast, learning that justice was served should deactivate the goal of redressing past injustice. Under these circumstances, participants' anger should not predict their subsequent willingness to punish in unrelated situations, even if their anger persists experientially. The second hypothesis addresses whether or not facts in subsequent judgment tasks will be assessed in a selective and simplistic way or in an open-minded and complex way. If participants learn that justice was not served, their anger should activate an indiscriminate tendency to punish others in unrelated situations without regard for whether their actions were intentional or not. By contrast, learning that the perpetrator was punished should deactivate this tendency. As a consequence, participants in this condition should take into account whether others' actions were intentional before determining the severity of punishment (Figure 1 presents a schematic summary of these hypotheses).