Abstract
In this introduction to the special section on the attributional approach to emotion and motivation, the character of Weiner’s attributional theory as an appraisal theory is discussed. I argue that the theory, although focusing on appraisal dimensions related to causal attribution, is actually a fairly general appraisal theory of emotion. Distinctive features of the attributional approach are its pioneering role in emotion research, its emphasis on the functional role of emotions, particularly for the motivation of action, and the existence of a large supportive data base. Given the solid evidence compiled by attributional researchers for emotion-specific effects of emotion on motivation, I propose that future research should focus on the details of the mental mechanisms through which emotions affect motivation, and formulate five questions that could guide this inquiry.
The attributional approach to emotion and motivation has its roots in Fritz Heider’s (1958) book The Psychology of Interpersonal Relations, which played a pivotal role in the origination of one of social psychology’s most productive research programs: the study of attribution. The core assumption behind this research program is that people are folk psychologists, and that explaining their actions, thoughts, and feelings requires firm acknowledgment of this fact (Reisenzein & Rudolph, 2008). Accordingly, the attributional approach to emotion and motivation focuses on people’s common-sense explanations of instigating events (including their own and other’s actions), that is, their beliefs about the causes of these events, as determinants of their emotional and behavioral reactions to them.
Although several variants of the attributional approach exist, the most elaborated, best known, and most influential attributional theory is that developed by Bernard Weiner (e.g., Weiner, 1986, 1995, 2006). A summary of the current state of this theory is provided in the first article of the special section (Weiner, 2014). Initially formulated for the achievement domain, Weiner’s theory of emotion and motivation (henceforth “theory of emotion” for short) has been extended to the explanation of helping behavior, aggression, and many other important psychological phenomena (e.g., Weiner, 1986, 1995, 2006, 2014). To this day, the theory continues to be elaborated further and remains a source of inspiration for theoretical analyses and empirical research. Two examples are presented by the other contributors to this special section: Rudolph and Tscharaktschiew (2014) present an analysis of moral emotions from the attributional perspective, and Hareli (2014) summarizes research on the use of common-sense knowledge about the links between cognitions and emotions to infer other’s situation appraisals from their emotions, as well as to influence their emotions.
The Attributional Theory of Emotion as an Appraisal Theory
As Weiner (2014) suggests, the attributional theory of emotion can be regarded as a variant of cognitive appraisal theory (e.g., Arnold, 1960; Frijda, 1986; Lazarus, 1966; Ortony, Clore, & Collins, 1988; Roseman, 2001; Scherer, 2001). What are the distinctive features of the attributional theory of emotion within the class of appraisal theories? Reisenzein, Meyer, and Schützwohl (2003, pp. 126–127) propose the following:
Pioneering role. Weiner (e.g., Weiner, 1985) has been one of the first postbehaviorist psychologists after Arnold (1960) who tried to enumerate the cognitions underlying specific emotions. In so doing, Weiner has focused on “cognitively complex” emotions such as anger, pity, pride, guilt, and gratitude; that is, emotions that presuppose elaborated event construals such as, specifically, beliefs about the causes of events and attributions of responsibility. Most of these emotions had been ignored by academic emotion psychology before. The fact that all subsequently formulated appraisal theories contain at least one appraisal dimension referring to causal attribution (e.g., “agency”; see Reisenzein & Spielhofer, 1994, for a systematic comparison) is undoubtedly owing to the influence of Weiner’s theory, even if this influence is not always acknowledged.
The attributional theory of emotion and the research it inspired have also been pioneering in other respects. Weiner and coworkers were the first to empirically demonstrate emotion-specific action tendencies (Weiner, 1980; see Rudolph, Roesch, Greitemeyer, & Weiner, 2004, for a summary). They (Weiner, Graham, Stern, & Lawson, 1982) were also the first to document that a person’s expressed emotions are used by social perceivers to infer that person’s appraisal of the eliciting event (see Hareli, 2014). In this context, Weiner and coworkers also made pioneering contributions to the study of the development of emotion concepts in children (e.g., Weiner, Kun, & Benesh-Weiner, 1980). Finally, Weiner’s theory of emotion seems to have been the first cognitive emotion theory for which a computational model was sketched (Bower & Cohen, 1982).
Emphasis on the functions of emotions. Different from some other cognitive appraisal theories, the attributional theory of emotion provides not only an analysis of the cognitive causes of emotions, but also speaks to the effects of emotions, in particular their functional effects. Two effects of emotions are emphasized: a motivational effect (emotions evoke action tendencies tailored to the eliciting situation-as-appraised), and a communicative effect (emotions provide information about the experiencer’s situation appraisals and action tendencies to social perceivers; Hareli, 2014; Weiner, 2014).
Empirical support. Weiner and his coworkers and students, together with numerous other contributors to the attributional research program, have amassed an impressive amount of evidence supporting the proposed relations between cognitions, emotions, and action tendencies. Indeed, there is probably no other cognitive emotion theory that has been as extensively tested as Weiner’s. As an example, consider Weiner’s (1980, 1986) attribution–emotion–action model of help-giving, according to which beliefs about the controllability of the cause of another’s plight evoke feelings of pity (uncontrollable cause) or anger (controllable cause), which in turn influence willingness to help (with pity increasing and anger decreasing it). As Weiner (2014) notes, by 2004 this model had been tested, and was mostly found to be supported, in 39 studies involving nearly 8,000 participants (Rudolph et al., 2004). Of particular note, in most of these (and many other) studies conducted by attributional researchers, the postulated cognition–emotion–motivation links were tested using causal modeling techniques, frequently in combination with experimental manipulations (see Rudolph et al., 2004). Studies of this kind are rare outside the attributional research program. Given that the motivational function of emotion postulated in the attributional theory is in similar form also assumed in other appraisal theories (e.g., Arnold, 1960; Frijda, 1986; Lazarus, 1991; Roseman, 2001), it is puzzling that the research on the attributional model of helping is not more widely cited by appraisal researchers. In addition, this research should also be of interest to appraisal researchers because it speaks to the much-discussed issue of the merits of hypothetical scenario studies (e.g., Parkinson & Manstead, 1993): The findings were very similar for hypothetical and real helping situations (Rudolph et al., 2004).
Yet another distinguishing feature of the attributional theory of emotion is its focus on appraisal dimensions related to causal attributions. Sometimes this focus has been so much emphasized by attributional theorists that readers could get the impression that other appraisal dimensions are ignored altogether. However, this impression is misleading: Although causal attributions certainly figure centrally in the attributional theory, closer examination reveals that they are by no means the only cognitions considered important for emotions (see Rudolph & Tscharaktschiew, 2014; Weiner, 2006, 2014). Emotion-relevant beliefs considered in the attributional theory also include beliefs about the certainty or probability of eliciting events (have they already occurred, or are they merely probable), beliefs about the focus of eliciting events (whom they primarily concern, oneself or another person), and of course, beliefs about whether the events are good or bad for oneself, as well as for the other (i.e., appraisals in the core sense of the term; Arnold, 1960). In addition, the emotion-relevant beliefs considered important by Weiner for some emotions include attributions of responsibility, as well as moral evaluations: According to Weiner (1995, pp. 8, 16) holding another responsible for a negative event implies the belief that the other person has violated, by her action, an accepted social or moral norm; but this belief is (or implies) an evaluation of the other’s action as socially or morally wrong. In sum, even though Weiner foregrounds causal beliefs in his appraisal theory of emotion—which indeed are the pivot of his thinking—the resulting theory is implicitly a fairly general appraisal theory of emotion. For a more detailed comparison of Weiner’s emotion theory with other cognitive emotion theories—specifically those of Meinong (1894) and Ortony et al. (1988)—see Reisenzein et al. (2003, Chapter 4).
Five Questions for the Attributional Theory of Emotion
Although there is good evidence for many assumptions of the attributional theory of emotion, a number of interesting questions are currently left unanswered by the theory. I restrict myself here to open questions related to the postulated motivational effect of emotions, that is, the assumption that (at least some) emotions directly elicit “fitting” action tendencies—pity the tendency to help, anger to aggress, guilt to repair the damage done, and so on. Note that the following questions are also largely unanswered by other appraisal theories; therefore, they are not only relevant for the attributional theory of emotion, but of appraisal theory at large.
Attributional research has provided solid evidence for the assumption that pity and anger influence helping and aggression (Rudolph et al., 2004). However, taken by itself this evidence is not diagnostic about the exact way in which these emotions affect motivation. Weiner (e.g., 1980, 1986) assumes that the effect of pity and anger (and other emotions) on motivation is direct and nonhedonistic—for example, pity generates, in a direct fashion, a desire to help the other, and anger the desire to aggress or withhold help. The alternative is that the effects of pity and anger on action are mediated by hedonistic concerns, that is, the desire to get rid of these feelings (note that both pity and anger are hedonically negative, unpleasant emotions). Intuitively, Weiner’s assumption seems to be correct (see also Reisenzein, 1996): If one helps out of pity or aggresses in anger, one does indeed seem to be motivated by these emotions—one helps because one feels pity, and aggresses because one feels angry; but one does not at all have the impression that one helps in order to get rid of one’s feelings of pity or anger! Experimental research supports this intuition (for pity, see Batson et al., 1989; for anger, see Gollwitzer & Bushman, 2012).
However, it seems to me that if one considers the same question for guilt—how does guilt generate the desire to repair the harm one has caused—intuitions are less clear. Here, it seems equally plausible to assume (as has in fact often been done) that the mediating mechanism is hedonistic: One seeks to repair the damage done to reduce the unpleasant feeling of guilt. Should we conclude from these considerations that some emotions (e.g., pity, anger) affect motivation via a nonhedonistic path, whereas others such as guilt influence motivation through the hedonistic mechanism? Or should we conclude that all emotions have both nonhedonistic and hedonistic effects, with different weights attached to them in different cases? Furthermore, how do hedonistic and nonhedonistic motivational effects of emotions combine?
Assuming that some emotions have a nonhedonistic, direct effect on motivation, the question still remains how these emotional action desires are translated into concrete actions. Are we to assume the existence of a separate, emotional mechanism of action generation? Or can we assume that the emotional action tendencies are translated into actions via the well-known decision-making and action selection mechanism that is also responsible for nonemotional actions (standardly some version of a belief-desire, or expectancy-value mechanism)? Attributional theorists, as well as other appraisal theorists who assume that emotions create specific action tendencies (e.g., Lazarus, 1991) sometimes convey the impression that they want to completely replace classical motivation and decision theory. However, in my view it is more reasonable to assume that emotional action impulses are translated into concrete actions by the same mechanisms as other desires (see also, Reisenzein, 1996; Reisenzein et al., 2013). To clarify this issue, an explicit discussion of the relation of appraisal theories of emotion to standard theories of decision making and action selection is called for.
Beyond that, the desire-generating function of emotions needs further explication. Are emotion-generated desires different from other action tendencies, and if so, in which sense? For example, is it just their particular urgency or priority that matters (Frijda, 1986)? Furthermore, how do emotions manage to generate particular goals or action tendencies in the first place? Is this a hardwired, evolutionary effect of emotions? Or is it, at least in part, a learned consequence of emotions? Relatedly, is this effect mediated by further cognitive processes, and as a consequence potentially malleable by additional information? For example, would my pity for another still evoke a tendency to help him even if I am convinced that I cannot help him in any way?
Finally, although emotions are the proximate motivators of action in the attributional theory of emotion and similar appraisal theories, they are presumably not the ultimate sources of motivation. On the contrary, according to standard appraisal theory, emotions themselves depend on appraisals of events as motive-congruent or motive-incongruent, and hence on the prior existence of motives (desires). A complete theory of emotion-based motivation will therefore also have to specify the motives underlying emotions—ultimately by describing a set of basic motives of humans. Clarification of this issue could also be important for refining the analysis of the preconditions of specific emotions. For example, pursuing this path suggested to me that moral emotions such as pity and guilt are not only based on specific kinds of thoughts, but also on specific (namely, nonegoistic) kinds of desires (Reisenzein, in press).