Đang tải... (xem toàn văn)
ABSTRACT The purpose of this paper is to highlight the important role of individual difference factors in the experience of emotion We begin by describing several commonalties across two major approaches to the study of emotion, namely, the neuropsychological and cognitive perspectives Both approaches provide some degree of support for the role of individual differences and cognitive factors in the experience of emotion This paper builds on these commonalities by reviewing personality and psycho.
ABSTRACT The purpose of this paper is to highlight the important role of individual difference factors in the experience of emotion We begin by describing several commonalties across two major approaches to the study of emotion, namely, the neuropsychological and cognitive perspectives Both approaches provide some degree of support for the role of individual differences and cognitive factors in the experience of emotion This paper builds on these commonalities by reviewing personality and psychopathology findings that indicate the contribution of both positive and negative personality characteristics (e.g., extraversion, optimism vs neuroticism, trait anxiety) to the types of cognitive appraisals and emotional responses exhibited by different individuals A self-schema model of emotion is presented as a means of integrating more fully this individual differences perspective with a theory of emotion In this model, selfschema content provides the basis for individual differences in underlying core themes and self-evaluative beliefs The model describes how self-schema content distinctions across individuals may have a differential impact on the initial processing of an event, evaluation of this event with respect to the self, and emotional and behavioral output Several examples are then presented to illustrate the increased predictability afforded by this individual differences-based self-schema model of emotion The application of this model to treatment and prevention issues in clinical and health psychology is also briefly considered Finally, the model is integrated with other theoretical perspectives on emotion by describing a number of additional research and theoretical implications Emphasis is placed on the need for further clarification of both cognitive and emotional components of an individual differences perspective on the study of emotions © 1997 Elsevier Science Ltd THE PAST decade has witnessed a resurgence of interest in the study of emotions A wide variety of theoretical perspectives have been advanced, ranging from neuropsychological approaches to cognitive-based theories of emotion (Izard, 1993; Parkinson, 1997) At the neuropsychological level, researchers have typically attempted to identify various brain systems that may mediate the experience of emotions As one illustration, it has been proposed by several investigators that the two hemispheres of the Authorship is equal Correspondence should be addressed to Kathy A Winter, Department of Psychology, University of Western Ontario, London, Ontario, Canada N6A 5C2 791 792 K A Winter and ;\: A Kuiper brain are differentially related to emotion Activation of the left frontal region is thought to be associated with positive affect or approach-related emotions, whereas activation of the right frontal region of the brain is associated with negative affect or withdrawal related emotions (Davidson, 1993; Zajonc & McIntosh, 1992) Congruent with this proposal, it has been found that righthemisphere arousal and vigilance mechanisms are particularly sensitive to increased levels of negative affect (such as depression), resulting in impaired functioning (Liotti & Tucker, 1992) In a somewhat different vein, other neuropsychological investigators have focused on mapping out the biologically based motivational systems that may underlie different behaviors and emotions A prime illustration is provided by Gray's (1990) neuropsychological model for the control of emotional behavior In his model, Gray has identified three fundamental neurologically based emotion systems, namely, the behavioral approach system (BAS), the behavioral inhibition system (BIS), and the fight/flight system (F/FL S) The behavioral approach system regulates the individual's behavior in the presence of signals of reward, producing positive emotions and filrther approach behaviors The behavioral inhibition system regulates behaviors in the presence of signals of punishment, leading to increases in anxiety, negative arousal, and the further inhibition of behavioral responses Finally, engagement of the fight/flight system may lead to escape behaviors and/or defensive aggression, coupled with the emotions of terror and rage Along with a neux'opsychological perspective, another approach to the study of emotions that has become increasingly prominent over the past decade involves a cognitive orientation This approach is based on the premise that emotional experience is contingent upon a person's evaluation and interpretation of the situation for his or her well-being (Frijda, 1993; Izard, 1993; Lazarus, 1993; Ortony, Clore, & Collins, 1988; Parkinson, 1997; Roseman, 1991; Scherer, 1993) This personal evaluation, known as cognitive appraisal, is conceptualized slightly differently by various cognitive emotion theorists Some of these theorists describe the emotional experience without attempting to detail the particular appraisals involved (Arnold, 1960; Epstein, 1983; Leventhal, 1984) Other theorists contend, however, that each emotion is characterized by a specific and unique set of cognitive appraisals (Frijda, 1987; Lazarus, 1991a, 1993; Ortony et al., 1988; Scherer, 1988; Smith & Lazarus, 1993) Commonalties Across Different Theoretical Perspectives on Emotion Despite the apparent diversity across various theoretical perspectives on emotion, several commonalties exist To begin, both the neuropsychological and cognitive approaches generally distinguish between emotional and cognitive components when describing their theoretical models (Izard, 1993; Parkinson, 1997) Considering the emotional components first, one relatively common theme across both perspectives is that a basic set of emotions can be identified and classified For example, the four basic emotions proposed by Gray (1990) are hypothesized to correspond to each of his three neurologically based motivational systems That is, the behavioral approach system is hypothesized to correspond to reward, the behavioral inhibition system is hypothesized to correspond to anxiety, and the fight/flight system is hypothesized to correspond to rage and terror (Ortony & Turner, 1990) Other emotion researchers have advanced somewhat different sets, depending upon their theoretical perspectives Plutchik (1980) has proposed that eight primary emotions exist These are fear, surprise, sadness, disgust, anger, anticipation, joy, and acceptance Oatley and Individual Differences and Emotion 793 Johnson-Laird (1987), in comparison, have proposed that the number of basic emotions totals five, and includes sadness, disgust, anger, anxiety, and happiness Although the exact number and identity of basic emotions may vary from one theoretical perspective to the next, a high degree of convergence is still evident in terms of positing a fundamental distinction between positive and negative emotions Past models in both domains have typically focused on negative emotions in greater detail This imbalance has been recognized in more recent theoretical approaches, which have also provided a more fine-grained analysis of positive emotions As one illustration, a recent cognitive theory of emotion by Lazarus (1993) describes six positively valenced emotions (happiness, pride, relief, love, hope, and compassion), in addition to nine negatively valenced emotions (anger, anxiety, fright, guilt, shame, sadness, envy, jealousy, and disgust) Both the neuropsychologicai and cognitive perspectives on the study of emotion have also acknowledged, to varying degrees, the important role that cognitions may play in the experience of emotion (Izard, 1993) In Gray's (1990) neuropsychological system, for example, information processing or computational functions of the brain are suggested to underlie all psychological processes, including emotion Within current cognitive theories of emotion, cognitions are posited to play a central role (Lazarus, 1993; Parkinson, 1997) As suggested earlier, these theories are governed by the proposition that the onset of emotional responses or affective states is dictated by the manner in which an individual ascribes meaning to the environment That is, emotions are elicited and differentiated via the perception of the situation or event as appraised by the individual (Scherer, 1993) There is considerable debate among current cognitive theorists, however, regarding the number and type of cognitive appraisals that are thought to differentiate the emotions (Parkinson, 1997) Lazarus (1993) has advanced a cognitive-motivationalrelational theory of emotion in which both positive and negative emotions are thought to be shaped by two cognitive mediators primary and secondary appraisals According to Lazarus, primary appraisal is defined as an evaluation of the personal relevance of an encounter for one's well-being, and consists of three well-defined cognitive dimensions: goal relevance, goal congruence, and type of ego involvement (Lazarus, 1991b, 1991c, 1993) Secondary appraisal, on the other hand, is defined as the evaluation of the person's options and resources for coping with the situation and future prospects, and includes three specific and detailed components: blame/credit, coping potential, and future expectancy In comparison to the set of appraisals specified by Lazarus, Roseman (1979) suggested that emotion is dependent upon a different set of cognitive attributes, including motivational state, situational state (presence/absence of reward/punishment), probability (certainty of outcome), legitimacy, and agency (responsibility) Adding to this list, Smith and Ellsworth (1987) have found evidence for three further appraisal dimensions, namely, attentional activity, anticipated effort, and perceived obstacles Finally, Frijda and his colleagues (Frijda, 1987; Frijda, Kuipers, & ter Schure, 1989) conducted a meta-analysis of several proposed cognitive appraisals and concluded that support existed for the presence of five dimensions: valence, certainty, agency, interestingness, and globality While much of current cognitive emotion theory focuses on appraisals as the primary means of eliciting emotions, attributions and construals have also been identified as antecedents to emotions (Lazarus & Smith, 1988) Even more recently, however, cognitive theories of emotion have begun to incorporate discussions of 794 K A Winter and N A Kuiper relational meaning, dynamic systems, and cognitive schemata (e.g., Frijda, 1993; Lazarus, 1993; Lewis, 1996) Thus, in addition to specifying the type of cognitive content that may be important in emotions (e.g., cognitive appraisals and attributions), more recent cognitive models of emotion have increasingly focused on the cognitive processes that may pertain to the dynamic, unfolding nature of emotions As one example, Lazarus (1993) has proposed that each emotion can be defined by a particular pattern of thought, known as a core relational theme These core themes are higher order cognitive constructs that are derived from particular combinations of primary and secondary appraisal components Fifteen core relational themes have been advanced by Lazarus (1993), with each representing a specific relationship between the person and environment With respect to anger, for example, the underlying core theme relates to a demeaning offense against the individual and her possessions In contrast, the emotion of happiness relates to a core theme of making reasonable progress toward a goal Sadness relates to the experience of an irrevocable loss, whereas pride is associated with a core theme of enhancing one's ego-identity by taking credit for an achievement or valued object A second example of a recent cognitive process orientation toward the study of emotion is provided by Lewis (1996) Building upon a dynamic systems approach, his model of emotion recognizes the importance of the ongoing interplay between cognitions and emotions In accord with current cognitive theories of emotion, Lewis (1996) incorporates the notion of a reciprocal relation between cognitive appraisals and emotions His model further proposes, however, that the feedback between cognition and emotion is the basis for self-organizing appraisals that are recurrent for an individual These appraisals assemble over time, and constitute global interpretations of situations that correspond with particular emotions Thus, for each individual, self-organizing appraisals help guide the direction and magnitude of an emotional response by shifting that person's attention to certain features of an event or situation, and also highlighting particular memories, associations, and scripts These cognitive processes elicit emotions In turn, these emotional changes provide further feedback that helps guide that individual's subsequent cognitive processing of other salient features of the situation and/or directs attention to other relevant memories, associations, and scripts Overall, Lewis' (1996) model clearly acknowledges the complex and ongoing interplay between appraisals, emotions, and cognitive processes, such as selective attention and memory A final common theme evident across both the neuropsychological and cognitive perspectives on emotion relates to the role of individual differences At a neuropsychological level, investigators have described various individual differences in neurological systems underlying the experience of emotion Davidson (1992), for example, has reported that individual differences in measures of brain activation asymmetry are associated with substantial variations in dispositional mood, affective reactivity, and temperament In a similar fashion, Gray (1990) has acknowledged the existence of prominent individual difference factors relating to the degree of sensitivity associated with each of his neurologically based emotion systems As corroborated by a number of other investigators (Eysenck & Eysenck, 1985; Larsen & Ketelaar, 1991; Watson, Clark, & Harkness, 1994; Watson & Tellegen, 1985), individuals high on the general personality dimension of extraversion exhibit heightened sensitivity to signals of reward (BAS), producing increased levels of positive affect In contrast, individuals high on the general personality dimension of neuroticism display heightened sensitivity to punishment (BIS), with subsequently enhanced levels of negative affect When Individual Differences and Emotion 7~5 taken together, this pattern suggests that individual difference factors mapping onto underlying biologically based systems may play a prominent role in the differential experience of situations and emotions Current cognitive theories of emotion have also begun to consider the potential role of individual differences in the experience of emotion Roseman (1991) has hinted at the need for future research to investigate individual differences in cognitive appraisals, and their potential relation to emotional predispositions In turn, Lazarus' (1991c, 1993) model has introduced several person-mediated variables that readily lend themselves to further consideration from an individual difference perspective In discussing cognitive appraisals, for example, Lazarus has proposed that, beyond sharing certain common beliefs and goals, each person also has their own specific set of well-established and stable goal hierarchies, beliefs, and cognitive styles that influence how situations are appraised Thus, due to unique learning and developmental experiences, individuals may vary significantly in the types of cognitive appraisals that are made for similar situations Paralleling his views relating to cognitive appraisals, Lazarus (1991c) has suggested that great variations in emotions are noted among individuals as well Although he proposes that general laws can be formulated about the emotion process, and that people share emotional experiences, he believes that "an emotion happens to an individual with a distinctive history who wants, thinks, and confronts specific environments, evaluates their significance, and acts as adaptively as possible" (Lazarus, 1991c, p 7) Moreover, according to Lazarus, when faced with a recurrent person-environment relationship, stable emotional patterns will result in the individual Summary In addition to distinguishing between cognitive and emotional components, both the neuropsychological and cognitive perspectives have commented on the number of basic emotions, with a fundamental distinction being drawn between positive and negative emotions Furthermore, although current cognitive theories of emotion are primarily dedicated to uncovering the role of cognitions within the experience of emotion, neuropsychological perspectives have also acknowledged the contribution of cognitive elements as well For cognitive-based theories of emotion, this emphasis has recently resulted in more refined theoretical models that incorporate both cognitive content and processes In addition to articulating various types of cognitive content (e.g., appraisals, core themes, and attributions), cognitive-based emotion researchers have also begun to specify more fully the cognitive processes (e.g., selective attention and information processing effects) that may relate to emotion A final commonality, and one that forms the principal concern of the present article, pertains to the role of individual differences in emotion Across both perspectives there is now reasonable support for the proposal that various individual difference factors may bear importantly on the experience of emotion THE ROLE OF INDIVIDUAL DIFFERENCES IN EMOTION In this article, we propose that a detailed consideration of individual difference factors is integral to a more complete theory of emotion This proposal is quite congruent with a substantial body of literature that supports an integrative hierarchical model of personality and emotionality advanced by Watson and Clark (1984, 1992a, 1992b, 1994) This hierarchical model consists of both a general and specific level The general level refers to broad distinctions across personality attributes, namely extra- 796 K A Winter and N A Kuiper version versus neuroticism, that also map onto broad distinctions across emotions, namely, positive versus negative affect Within each general level of this model there is also a lower order specific level As one example, the general level of neuroticism and negative affect can be further subdivided into different content domains at the specific level, such as depression or anxiety (Watson et al., 1994) The hierarchical model was briefly alluded to in our previous discussion of individual difference factors relating to Gray's (1990) emotion systems In particular, the general dimension of extraversion is characterized by a temperamental sensitivity to pleasurable stimuli (BAS); and is associated with the more frequent and intense expression of positive affect, including the specific content emotions of joyfulness, enthusiasm, boldness, pride, and selfconfidence Neuroticism, on the other hand, is thought to be characterized by a temperamental sensitivity to painful or negative stimuli (BIS) Individuals scoring high on this general negative dimension of personality more frequently and intensely experience negative affect, including such specific content emotions as fear anxiety, sadness, guilt, hostility, and self-dissatisfaction Considerable research findings now converge on the proposal that individuals scoring high on neuroticism generally perceive their world in a negative manner, viewing their life experiences as threatening, problematic, and distressing In contrast, individuals with high levels of extraversion generally approach their environment in a positive and energetic manner, displaying a strong willingness and enthusiasm to seek out and actively engage in various life events (Watson & Clark, 1984, 1992a, 1992b, 1994) The Watson and Clark hierarchical model is relevant to our individual differences approach to emotions in several ways First, the numerous findings associated with this model provide strong empirical support for the proposal that individual differences in personality are of fundamental importance to the differential experience of emotions Second, the most general level of this hierarchical model (i.e., positive vs negative valence) provides a useful framework for organizing our presentation of additional findings that pertain directly to the cognitive and emotional components of theories of emotion Third, and as discussed in subsequent sections of the paper, the specific content level of the hierarchical model (e.g., depression vs anxiety; pride vs enthusiasm) is of particular importance to the development and presentation of our individual differences-based self-schema model of emotion Finally, the last section of the paper indicates how both the specific and general levels of this hierarchical model might be integrated with other theoretical and research perspectives to further advance our understanding of individual difference factors pertaining to emotions Individual Differences in Personality and Psychopathology Congruent with the general level of the Watson and Clark hierarchical framework described above, this section of the paper reviews research findings from both the personality and psychopathology domains that are of direct relevance to an individual difference perspective on theories of emotion In particular, the impact of both positive and negative personality characteristics on cognition and emotion, the two central aspects of theories of emotion, is described When considering cognition, this review focuses on a construct of primary interest to emotion theorists, namely, cognitive appraisals of events Following this, the review of emotion describes additional personality and psychopathology findings that highlight the important role of individual difference factors in emotional responses to events Individual Differences and Emotion 797 Cognition A review of the literature indicates that individual differences in personality mediate cognitive appraisals of both positive and negative life event.s To begin with negative personality characteristics, Gallagher (1990) found that individuals high in neuroticism were more likely to appraise academic events as more threatening and less challenging than those low in neuroticism In a similar fashion,Jerusalem (1990) discovered that individuals high in trait anxiety appraised a series of difficult performance tasks as threatening and involving loss, whereas low trait anxiety individuals buffered the experience of stress by appraising these situations as less threatening In addition to the specific cognitive appraisals of loss, threat, and challenge described above, negative personality characteristics have also been found to be related to higher general stress appraisals for life events Olinger, Kuiper, and Shaw (1987), for example, found that individuals who endorsed a large number of rigid and dysfunctional self-evaluative standards were found to think about stressful events more often, and rate them as more displeasing, than individuals who displayed few dysfunctional selfevaluative standards This pattern of findings was evident even when these individuals were currently nondepressed, suggesting a persistent depressive personality style related to negative cognitive attitudes or beliefs about self-evaluation Similarly, Westra and Kuiper (1992) found that Type A dispositions were significantly related to increased perceived stress, with higher Type A individuals appraising an arithmetic task as more stressful than those lower in Type A attributes These heightened stress appraisals, however, were only found to be correlated with increased Type A characteristics when the arithmetic task was presented in a manner that maximized self-worth evaluations While negative personality characteristics tend to be associated with increased appraisals of threat and loss, positive personality attributes have been shown to be related to more positive appraisals of challenge In terms of extraversion, for example, Gallagher (1990) found that extraverted individuals were more likely to appraise academic stressors as challenging than were individuals low in extraversion A similar appraisal pattern has been demonstrated for sense of humor, with Kuiper, Martin, and Olinger (1993) reporting that individuals with a higher sense of humor viewed stressful events as more challenging than individuals with a lower sense of humor Other research has shown that positive personality characteristics are also associated with reduced negative appraisals of threat and loss regarding stressful life events In a sample of German migrants, for example,Jerusalem (1993) found that those individuals high in optimism and self-efficacy were less likely to appraise their current adaptational demands as threatening, and were less likely to consider that they had suffered a loss, when compared with those who were low in optimism Furthermore, the relationship between an individual's level of optimism and both challenge and threat appraisals has been examined in a sample of 73 dental patients presenting for wisdom-tooth extraction under general anesthesia (Siim, 1994) In this study, higher levels of optimism were positively related to appraisals of challenge and negatively related to appraisals of threat That is, those people who had a positive outlook on life felt more confident and less afraid when it came to surgery than those people who have a negative outlook on life As well, optimists were more likely to appraise pain following surgery as more challenging, less threatening, and more controllable than pessimists A similar appraisal pattern has been demonstrated for individual differences in sense of humor In particular, when faced with performing a novel task in a laboratory setting, Kuiper, McKenzie, and Belanger (1995) found that individuals with 798 K A Winter and N A K'uiper an increased level of humor provided higher challenge and lower threat appraisals, compared with individuals displaying a lower sense of humor Emotion A review of the literature also indicates that individual differences in personality mediate affective responses to life events There is some agreement in the literature that a depressive personality style is associated with a variety of negative emotional features, including increased levels of dysphoric affect, sadness, anxiety, and guilt (Beck, Rush, Shaw, & Emery, 1979; Kuiper, Olinger, & Martin, 1990; Segal & Swallow, 1993) As one example, Krantz and Hammen (1979) found that when faced with ambiguous information, depressed individuals favored emotionally negative interpretations As a second illustration, Olinger et al (1987) found that individuals with a large number of dysfunctional self-evaluative standards rated negative life events as having a greater emotional impact, ruminated more often about these events, and displayed higher perceived stress levels than individuals who were less cognitively vulnerable to depression Consistent with a depressive personality style, these effects were also evident even when these individuals were nondepressed Congruent with the above findings, individuals high in Type A characteristics also perceive their lives as involving experiences The specific ramifications of this depressive personality self-schema can be highlighted by considering the following scenario, adapted from Smith and Lazarus (1993), in which person A, a close friend of person B, betrays the trust of B by revealing to a third party that B has cheated on his lover In terms of initial processing of this situation, our model would predict that, if person B has a depressive personality selfschema, he would be more likely to report negative cognitive appraisals that center around construals of threat, loss, and increased personal responsibility At the evaluation stage, the self-schema model would propose that person B would be more likely to employ core relational themes that are again negatively valenced Given their depressive personality self-schema, these themes may include the perception that they are hopeless with regard to their ability to handle or improve this situation, and that they are personally to blame tor everything that has happened Finally, in terms of emotion, our model would predict that, consistent with a depressive personality self-schema, person B would be more likely to exhibit negative feelings, such as sadness and/or guilt, at the output stage For this same scenario, the self-schema model of emotion would make quite different predictions for individuals displaying positive personality characteristics, such as trait optimism Our earlier review indicated that individuals with higher levels of optimism have generally positive self-schema content, consisting of favorable attitudes and beliefs (e.g., a generalized expectancy for good rather than bad outcomes), Individual Differences and Emotion 809 as well as adaptive cognitive strategies (e.g., social reasoning that emphasizes positive personal attributes) Compared to a depressive personality, then, the unique selfschema content associated with an optimistic individual would have quite different effects on each phase of selfschema functioning That is, when the self-schema content of an optimist is mapped onto schema-related functioning for person B in the above scenario, different initial processing, evaluations, and outcomes would be expected In terms of initial processing, our model would predict that an optimistic individual B would be likely to appraise this situation as generally nonthreatening and, perhaps, even somewhat challenging In terms of evaluation, the generalized expectancy for good rather than bad outcomes may translate into a core relational theme pertaining to increased hope (i.e., fearing the worst, but wanting better) Finally, in terms of outcome, the model would predict that optimism would be associated with more favorable emotions, such as hope and/or compassion Overall, the self-schema model's inclusion of specific individual differences results in differential predictions at each phase of the experience of emotion As such, this suggests that a more fine-grained analysis of individual difference factors may serve to greatly increase the predictive utility and accuracy of a model of emotion Finally, it should be noted that the self-schema model proposes that individuals may sometimes be quite similar with respect to one or two stages of self-schema functioning, but then differ critically on a third stage, resulting in the expression of quite distinct emotions This can be illustrated with a final example Suppose that two different individuals similarly construe an event as being important for well-being, inconsistent with desired goals, and something for which they are not personally responsible In turn, this subsequently activates a core theme of "a demeaning offense against me and mine." Lazarus' model of emotion would suggest that anger is the resulting emotion for both persons The self-schema model, however, would predict that only Certain individuals, such as those high on Type A characteristics, for example, may show this anger In contrast, our model would suggest that an individual with a depressive personality selfschema would be more likely to experience guilt in this situation than anger As such, knowledge of individual difference factors would again lead to increased predictability of resulting emotions Applications of the Self.Schema Model This section further illustrates the potential utility of the proposed self-schema model of emotion by mapping out applications of this model to clinical psychology treatment and prevention issues Following this, several implications of the model for health psychology issues will be briefly summarized Treatment and prevention issues We would argue that the selfschema model is highly relevant to clinical treatment issues In particular, a basic tenent of this model is that both self-schema content and functioning are integral aspects of how each individual construes reality, and then acts upon that information Thus, a model of emotion that maps individual differences in self-schema content onto selfschema functioning would enable therapists to more clearly identify specific areas that could be targeted for the most appropriate treatment strategies Following the basic principles that guide cognitive therapy (Clark & Beck, 1989), one specific goal of treatment might be to bring aspects of self-schema content and functioning into consciousness, so that the dysfunctional appraisals, maladaptive core themes, and cognitive strategies related to 810 K A Winter and N A Kuiper unfavorable emotions and ineffectual coping strategies can be readily identified and targeted for treatment Using the example of a depressive personality style, these individuals may benefit from intervention at several phases of serf-schema functioning In particular, strategies aimed at recognizing and redefining dysfunctional self-evaluative self-schema content, as well as learning more adaptive cognitive processes, may help to alter negative cognitive appraisals, evaluations, and emotional output A second treatment-related advantage of the proposed model is that this framework enables therapists to tailor interventions to the current stage of self-schema development Research from a developmental psychology perspective has shown that the child's self-schema content shifts over time, from one that is concrete and based on observable attributes, to one that is more abstract and based on generalizations relating to one's beliefs, wishes, and motives (Harter, 1990) Commensurate with this shift, the strategies used to make judgments about the self become more effective As the individual progresses through developmental stages, cognitive limitations are lessened, making it possible to draw more fully on information obtained via social comparisons and perspective-taking when making selfevaluations According to Harter (1990), knowledge of these normative shifts and changes may not only help the therapist assess whether or not an individual is at the appropriate stage, but may also facilitate utilization of treatments that are developmentally appropriate Along these lines, the self-schema model of emotion could provide one framework for guiding individuals through the developmental stages that they may have missed, or that they may have experienced in a less than adaptive manner In terms of depression, for example, by tracing cognitive development stages, primary caregiver relationships, and any unusual life experiences, it may be possible for the therapist to more fully articulate both the content and fimctioning aspects of a depressive personality self-schema Perhaps individuals displaying a depressive personality may not have progressed smoothly through the stages of self-schema development Alternatively, these individuals may have had certain damaging experiences that hindered developmental progression As a case in point, Guidano and Liotti (1982) argue that schemata formed as a result of distressful experiences and maladaptive learning during childhood underlie mnch of adult psychopathology In addition to treatment, the self'schema model of emotion also has implications for prevention In the case of child abuse, for example, knowledge that different individuals perceive similar situations differently, depending on both self-schema content and functioning, may help in predicting which caregivers are at increased risk fbr engaging in child abuse In this regard, Hillson and Kuiper (1994) have reviewed a number of studies that support their proposal that abusive caregivers more often perceive misbehavior on the part of their child as intentional, and specifically aimed at compromising their own identity as competent caregivers For the abusive caregiver, this perceived violation of an important core relational theme then results in the expression of anger and child abuse In response, Hillson and Kuiper (1994) have proposed that one aspect of a treatment program for such individuals would include cognitive interventions that help identify and appropriately modify these maladaptive cognitive appraisals and underlying core themes Moreover, these interventions could be readily adapted to educational and preventative programs that provide early identification of the patterns of maladaptive appraisals and evaluations that result in increased child abuse and neglect Individual Differences and Emotion 811 Health psychology issues, The self-schema model of emotion has application to health psychology issues as well Within the domain of health psychology, recent work has attempted to ascertain the psychological characteristics and underlying mechanisms of individuals who are prone to disease Studies have examined factors within the social environment (e.g., stress, social support), as well as personal variables (e.g., dispositional tendencies, physiological mechanisms) that influence disease susceptibility (Adler & Matthews, 1994) However, in order to adequately identify who becomes sick and for what reasons, an approach needs to be taken in which the differences underlying personal variables can be examined more closely The proposed self-schema model of emotion is one example of a theoretical framework that could be readily adapted to a health psychology approach In addition to providing specialized information regarding self-schema content, and shedding light on aspects of self-schema functioning that may be implicated in disease proneness, the sell: schema model of emotion may provide insights into preventing and controlling various diseases As one specific example of this type of application, knowledge of the impact of individual differences in cognitive appraisals, core relational themes, and resulting emotions may help health care providers better prepare patients for surgery Targeting the areas within self-schema content and functioning that are particularly problematic may help to reduce recovery time, postoperative symptom reporting, and medication usage In accord with this proposal, Siim (1994) found that optimists, who had more positive cognitive appraisals relating to surgical procedures and postoperative pain, reported significantly less pain days following surgery for the removal of wisdom teeth than pessimists, who conversely had more negative cognitive appraisals relating to the two phases of dental surgery FURTHER THEORETICAL IMPLICATIONS AND RESEARCH DIRECTIONS Drawing from a broad range of literature pertaining to emotions, this section integrates a number of different theoretical issues and suggests additional research avenues for the self-schema model of emotion This section begins with a further examination of the role of cognitive factors in theories of emotion Following this, several theoretical and research recommendations regarding the emotional components of the self-schema model are advanced The paper then concludes with a discussion of how both the specific and general levels of analysis associated with the Watson and Clark (1992, 1994) hierarchical model might also be employed with our selfschema approach to facilitate the continuing refinement of theories of emotions Clarifying the Role of Cognitions in Emotions In presenting our self-schema model of emotion, we have emphasized cognitive elements pertaining to cognitive appraisals and core themes It is clear, however, that both of these cognitive components require more detailed theoretical elaboration and clarification Beginning with core themes, we would suggest that considerable research is still required to more fully articulate and describe the sets of core themes that are of particular relevance to the experience of emotions In particular, our model would suggest at least two further research directions First, we would encourage an increased research emphasis on determining the positive belief statements and attitudes that guide an individual's evaluations of life events and experiences As indicated throughout this paper, positive personality factors, such as extraversion, sense of 812 K A Winter and N A Kuiper humor, and optimism, have each been identified as important individual difference characteristics that bear strongly on how a person evaluates and responds to life situations Accordingly, we would propose that the identification of positive core themes related to each of these dimensions is a worthwhile avenue for further investigation Second, and in accord with Frijda (1994), our model would also propose that core themes may vary tremendously in terms of their degree of specificity, with some themes reflecting broader and more general areas of self-evaluative concern, and others being relatively specific Future research could then determine if the general self-evaluative themes are more closely tied to individual difference variables, such as trait anxiety, for example When considering cognitive appraisals, we concur with a recent suggestion by Parkinson (1997) that further research is required to refine this construct Cognitive emotion theorists have long debated the number and type of cognitive appraisals that are thought to differentiate the emotions (Frijda et al., 1989; Lazarus, 1991b; Roseman, 1979; Smith & Ellsworth, 1987), with several different sets being advanced (Frijda, 1987; Roseman, Spindel, & Jose, 1990) As such, it is clear that further research is required to identi~, the most parsimonious set of cognitive appraisals that fully captures the richness and complexity inherent in any assessment of life experiences From our theoretical perspective, it would be particularly interesting to determine the degree of association between these proposed appraisals and the types of core themes identified in the self-schema model In other words, to what extent, and under what circumstances, might cognitive appraisals and core themes be distinct or overlap? Future research could determine the degree to which individual differences in self-schema content may play an important role in the amount of convergence or divergence displayed for various cognitive constructs As one illustration, findings reported by Smith, Haynes, Lazarus, and Pope (1993) support the existence of specific core themes relating to threat and danger In turn, these core themes may predominate in individuals with anxious self:schemata, enhancing their threat appraisals for a wide variety of events A more detailed focus on individual difference t~tctors relating to core themes and appraisals may also assist in explaining discrepant findings reported in the literature Earlier in the paper it was indicated that extraverted individuals were more likely to appraise academic stressors as challenging In that study by Gallagher (1990), extraverted individuals did not provide lower threat appraisals for these same academic events From the current perspective, this pattern suggests that various core themes (and their associated appraisals), may, under certain circumstances, be relatively independent of one another Thus, future work could directly address this issue by mapping out individual difference factors associated with the potential links between core themes and appraisals across a wide variety of situations Mthough our model has emphasized cognitive appraisals and core themes, we have briefly noted that other cognitive variables, such as causal attributions, may also play a role in emotion In the depression domain, for example, research by Metalsky and his colleagues (Metalsky, Abramson, Seligman, Semmel, & Peterson, 1982; Metalsky, Halberstadt, & Abramson, 1987) has found that individuals with a depressive attributional style, which is characterized by internal, stable, and global attributions, were more likely to experience depression following exposure to stressful life events than individuals who made external, unstable, and specific attributions regarding these negative events Individual Differences and Emotion 813 b Although this depressive attributional pattern has been demonstrated several times, other research findings provide a somewhat different perspective In particular, Brewin (1985) reviewed evidence suggesting that depressive attributions may change following recovery from depression, leading to the conclusion that a dysfunctional attributional style may be a symptom or correlate of depression, rather than a causal variable As such, more recent work in this domain has attempted to determine the precise conditions under which dysfunctional attributions may lead to depression Of particular interest to our self-schema model of emotion are findings by Metalsky, Joiner, Hardin, and Abramson (1993) that indicate that individual differences in self-esteem levels and cognitive expectations relating to hopelessness are both critical variables In accord with the emphasis placed on a self-evaluative component in our model, Metalsky et al (1993) found that only those individuals with poor self-esteem showed the expected depression effects for a dysfunctional attributional style Furthermore, these effects were mediated by a cognitive construct of hopelessness, which consists of the expectation that highly desired outcomes will not occur or that highly undesired outcomes will occur It should be noted that this construct is highly similar to the types of core themes described in the self-schema model Smith et al (1993), for example, have identified core themes relating to both loss/helplessness ("Nothing can ever been done to fix this bad situation"), and the converse, hope/effortful optimism ("In the end, there is a chance that everything will be OK") More generally, the above parallels suggest that future research might investigate directly the associations between attributional approaches to emotion and components of the self-schema model, such as core themes, cognitive appraisals, and selfevaluative processes In addition to examining currently depressed individuals, it is also important that this work determines the factors that may influence the use of these dysfunctional cognitions and self-evaluative processes when an individual is currently nondepressed Such an approach would help elucidate more clearly the degree to which these various components of the model may have a causal versus concomitant status The self-schema model might also benefit from a fllrther integration with other cognitive-based theories of emotion that emphasize the self A prime example is self-concept discrepancy theory, which postulates that incompatibilities between various self-representations, including actual, ideal, and ought-self, produce distinct emotional responses (Strauman & Higgins, 1988) Research findings converge on a pattern in which discrepancies between actual and ideal self are associated with increased depression, whereas discrepancies between actual and ought self result in heightened anxiety (Strauman, Lemieux, & Coe, 1993) This pattern of findings suggests that the self-schema model, in addition to incorporating actual selfdescriptors, might also consider how the inclusion of self-representations relating to ideal and ought views would impact on the experience of emotions One research possibility would be to more fully elaborate upon the cognitive processes that culminate in a self-discrepancy evaluation producing a specified emotion Of particular interest would be an assessment of the extent to which these self-discrepancy judgments also incorporate core themes pertaining to either anxiety or depression Finally, we would like to reinforce the notion that further research is certainly necessary to determine the limits and boundary conditions that may be associated with the selfschema model Our model was derived largely from social cognition and emotion theories that have placed a great deal of emphasis on the role of cognitions in both selfrepresentations and emotions It should be clearly acknowledged, how- 814 K A Winter and N A Kuiper ever, that not all models of emotion insist on this cognitive component As one illustration, Zajonc (1980) has proposed that positive and negative affective reactions can be evoked under conditions that allow for virtually no cognitive processing The empirical support garnered for this affective primacy hypotheses (e.g., Murphy & Zajonc, 1993) suggests that any theory of emotion must also acknowledge instances in which noncognitive factors may lead to emotion activation In terms of a self-schema model, this may require consideration of additional emotion systems that are independent of cognition In this regard, Izard (1993) has specified three emotionactivating systems, namely, neural, sensory, and motivational The neural system, for example, may generate emotions as a result of activities of certain neurotransmitters and brain structures Although cognitions not play a role in these three systems, Izard (1993) has also specified a fourth system, which places cognitive constructs at the top of her hierarchy of emotion activators A.s reflected in our selfschema model, this cognitive level is seen as the most complex and versatile of the four systems Thus, the challenge will be to incorporate emotional effects that stem from the lower levels, while simuhaneously preserving the unique contributions associated with a cognitivebased self-schema model of emotion Clarifying the Emotional Components of the Model In this section, we highlight several ambiguities and issues in the area of emotion research that have both theoretical and research implications for our selfschema model We begin with a re-examination of the notion of basic emotions, followed by consideration of more complex emotions The section concludes with a discussion of the role of emotions in determining both behaviors and cognitions With respect to basic emotions, our earlier review revealed that there is only modest agreement between theorists regarding the exact set of fundamental emotions (Ekman, 1992; Johnson-Laird & Oatley, 1992; Lazarus, 1991c; Plutchik, 1980) These distinctions between various emotion theorists point to a need for future research in this domain to clarifv more precisely the emotional components of any proposed theoretical framework With respect to the self-schema model, two avenues of research may be of particular benefit First, in keeping with the inclusion of both positive and negative individual differences in our model, we would suggest a greater emphasis on determining and clarifying the positive emotions In addition, this work could also provide a more detailed output component for various combinations of positive appraisals and positive core relational themes in our model In pursuing this line of research, a second related avenue of investigation might involve determining the tactors that relate to the more complex expression of emotions Several emotion theorists have suggested that the existence of pure emotions is rare, compared to the simultaneous experience of two or more emotions As one empirical demonstration, Smith and Ellsworth (1987) found that over 60% of a sample of college students who reported feeling guilt, either before or after an exam, simultaneously reported feeling angry and/or apathetic In terms of our model, it is interesting to note that several types of personality characteristics may often co-occur, such as a depressive personality style coupled with higher levels of trait anxiety (Kuiper et al., 1990; Westra & Kuiper, 1996) Future research could thus determine if these personality combinations also give rise to multiple core themes and multiple appraisals that may be employed in a single situation Here, a real-time process examination Individual Differences and Emotion 815 may be of particular benefit, as appraisals and relevant themes may shift as the situation unfolds across time (Kuiper et al., 1995; Lewis, 1996; Parkinson, 1997) Although the self-schema model of emotion clearly acknowledges the bidirectional, reciprocal nature of the relationship between cognitions and emotion, our presentation of the model has thus far focused primarily on the effects of personality attributes and cognitive factors on emotions Clearly, however, the further development and refinement of this model must also consider the obverse relationships in more detail In our literature review we have pointed out instances where personality traits may result in certain emotions, but it is also the case that emotions may produce behaviors indicative of certain underlying traits, Malatesta and Wilson (1988), for example, reviewed the literature in this domain and found that a happy mood is significantly related to increased levels of optimism Similarly, Myers and Diener (1995) report that happy individuals engage in more outgoing behaviors that are indicative of the personality characteristic of extraversion This latter effect has also been demonstrated for positive affect, with individuals high on this dimension engaging in a greater number of social interactions, which are also judged of higher quality by independent raters (Berry & Hansen, 1996) In addition to these important behavioral effects, emotions may also have a pervasive impact on the cognitive processing aspects of the self-schema model As described earlier in the paper, individuals that are currently depressed or anxious often show cognitive information processing biases that favor negatively valenced stimuli (Mathews & MacLeod, 1994) This type of pattern was also evident in Malatesta and Wilson's (1988) review, with sadness being related to impairments in cognitive performance, and the induction of anger leading to facilitative encoding for negative information about others When taken together, these findings reinforce the need for future research studies to obtain multiple assessments across time for the different components of the self-schema model This dynamic approach may then provide data that elucidates more clearly the reciprocal effects of these components as they unfold to generate further expressions of behaviors, emotions, and cognitions (Kuiper et al., 1995; Lewis, 1996; Parkinson, 1997) Specific and General Levels of Analyses Throughout this paper we have employed Watson and Clark's (1992, 1994) hierarchical model as an organizational framework to help facilitate our review of past research findings relevant to an individual differences, social-cognitive perspective on emotion In addition, this framework has proven of further value in our presentation of the self-schema model itself Recall that the hierarchical model consists of both a general and specific level The general level of personality, namely extraversion versus neuroticism, maps onto the general level of emotions, namely, positive versus negative affect Each general level can then be further subdivided into specific content domains (e.g., depression versus anxiety within the general negative domain) In this final section of the paper, we argue that the continued application of both the specific and general levels of analyses embedded in the Watson and Clark framework may provide benefits for future research directions and the integration and refinement of theories of emotion Specific level In presenting our self-schema model of emotion, we have shown how the specific content level of the Watson and Clark's framework is of principal relevance to 816 K A Winter and N A Kuiper our individual differences perspective This level of analysis allows for a ready integration of various components of the self-schema model with other research approaches that also focus on specific content distinctions As one illustration, benefits may accrue from a further integration of our model with research in experimental psychopathology on the content-specificity hypothesis (Clark et al., 1990) This hypothesis proposes that each form of psychopathology may be identified by a unique and specific set of dysfunctional cognitions, with Westra and Kuiper (1996) further suggesting that these cognitions are primarily self-descriptive and self-evaluative in nature Depressed individuals, for example, typically display specific and unique dysfunctional cognitions that relate to negative self-evaluative themes of personal loss and failure (Clark et al., 1990; Mathews & MacLeod, 1994; Westra & Kuiper, 1996) These content-specificity distinctions in experimental psychopathology research are fully compatible with the specific content level of the self-schema model of emotion As such, they suggest that our model may provide a useful theoretical and research perspective for shedding further light on the proposed self-schema content and processes underlying these differences Specifically, the self-schema model of emotion could be used in such research programs to begin to document individual differences in psychopathology or maladjustment for initial processing, evaluation, and outcome Illustrative is a recent study by Westra and Kuiper (1997), which explored initial processing distinctions relating to selective attention for four different types of negative self-schemata, namely, depression, anxiety, Type A, and bulimia Using a modified probe detection task that assessed accuracy in reporting various probe positions, it was found that participants showed consistent selective attention effects for personal adjectives that were highly relevant to their own selfschema As such, this study extends both the content-specificity hypothesis and the selfschema model by documenting selective attention processing effects across a much broader range of individual difference characteristics than has typically been the case in past research Several of the design features associated with the Westra and Kuiper (1997) study may also prove useful in future investigations of cognitive-based theories of emotion at the specific level of analysis In this regard, Parkinson (1997) has pointed out that contemporary research on cognitive appraisals and emotion has relied too extensively on self-report measures This overreliance can result in substantial conceptual and measurement overlap among appraisal and emotion constructs In turn, this may lead to serious interpretative problems, including unwarranted causal explanations that are circular or tautological in nature To address these types of concerns, Westra and Kuiper (1997) have moved beyond a strict reliance on self-report measures to also include an on-line processing measure of self-schema selective attention (i.e., probe position accuracy) In addition, great care was taken in this study to minimize any content overlap between measures of psychopathology and measures of self-schema content (i.e., self-referent personal adjectives) for each of the four domains Finally, participants in the Westra and Kuiper study were carefully selected to ensure that they did not exhibit elevated levels on more than one of the four measures of psychopathology Taken together, these design features were used to minimize the type of conceptual and measurement overlap problems described by Parkinson (1997), thus minimizing subsequent interpretative difficulties As such, the inclusion of these types of features may prove quite worthwhile in future research studies on cognition and emotion Individual Differences and Emotion 817 General level In addition to the specific level, further benefits may also accrue from continued research and theoretical integration at the general level of the Watson and Clark framework As one illustration, Parkinson (1997) has recommended that research on the connection between cognitive appraisals and emotions needs to examine more closely the interpersonal communicative functions of emotions Parkinson also suggests a movement away from the overreliance on self-report data to include more real-time measures and social interaction paradigms that are sensitive to the dynamic unfolding nature of appraisals and emotions (see also Kuiper et al., 1995 and Lewis, 1996) As a starting point, this type of work might build upon research at the general level of the Watson and Clark framework indicating that positive affect is much more strongly related to various social activity measures than is negative affect This pattern of findings includes such measures as the overall quality of social interactions, time spent with friends, and the number of social activities engaged in (Berry & Hansen, 1996; Watson, 1988; Watson, Clark, McIntyre, & Hamaker, 1992) Accordingly, one possibility for future research would be to extend the real-time social interaction videotaping paradigm used by Berry and Hansen (1996) to incorporate measures of cognitive appraisals, core themes, and individual difference variables, as well as assessments of positive and negative affect This extension might then provide one means of determining the ongoing impact of selfschema distinctions on emotional experiences in a social context In addition, work at this level might also elaborate on the underlying neuropsychological mechanisms involved in social interactions and emotions, employing Gray's (1990) emotions systems model as one possible theoretical framework CONCLUDING COMMENTS In this paper we have discussed the central role of individual differences in theories of emotion, culminating in the presentation of a self-schema model of emotion Building upon the strengths of prior theoretical models in the social cognition and emotion domains, our individual differences model consists of both self-schema content and functioning In this model we have emphasized that unique core themes and specific self-evaluative beliefs are important content components of the self-schema; and that these themes and beliefs differentially influence the manner in which life events are processed and experienced by different individuals In particular, the self-schema model of emotion contends that a certain set of cognitive appraisals, core themes, and emotions is more accessible to any one individual; and that the composition of each of these sets varies markedly as a function of individual difference factors In describing and discussing this self-schema model of emotion, we have integrated a wide range of theory and research findings from the personality, psychopathology, health psychology, and clinical domains of psychology This integration has then lead to a number of implications of the proposed self-schema model of emotion for further research, theory, and clinical practice Acknowledgment We wish to thank Jim Olson for his thoughtful comments on an earlier version of this paper 818 K A Winter and N A Kuiper REFERENCES Adler, N., & Matthews, K (1994) Health psychology: Why some people get sick and some stay well? Annual Review of Psychology, 45, 229-259 Arnold, M B (1960) Emotion and personality: Volume P~ychological a6pects New York: Columbia University Press Banaji, M R., & Prentice, D A (1994) The self in social contexts Annual Review of Psychology, 45, 297-332 Beck, A T., Rush, A.J., Shaw, B E, & Emery, E (1979) Cognitive therapy of depression New York: Guilford Press Berenbaum, H., Fujita, E, & Pfennig, J (1995) Consistency, specificity; and correlates of negative emotions Journal of Personality and Social Psychology, 68, 342-352 Berry, D S., & Hansen,J S (1996) Positive affect, negative affect, and social interaction.Journal ¢?fPersonality and ,Social Psychology, 71, 796-809 Breck, B E., & Smith, S H (1983) Selective recall of selfdescriptive traits by socially anxious and nonanxious females Social Behavior and Personality, I1, 71-76 Brewin, C (1985) Depression and causal atu'ibutions: What is their relation? Psycholog-ical Bulletin, 98, 297-309 Cane, D B., Olinger, L.J., Gotlib, I H., & Kuiper, N A (1986) Factor structure of the dysfunctional attitude scale in a student population Journal ofClinieal Prycholog3; 42, 30%309 Carvei, C S., & Gaines,.] G (1987) Optimism, pessimism, and postpartum depression Cognitive Therapy and Re.~earch, II, 449-462 Clark, D A., & Beck, A T (1989) Cognitive theory and therapy of anxiety and depression In E C Kendall & D Watson (Eds.), Anxiety and depression: Distinctive and overlapping features (pp 379-411) San Diego, CA: Academic Press Clark, D A., Beck, A T., & Stewart, B (1990) Cognitive specificity and positive-negative affectivity: Complementary or contradictory views on anxiety, and depression?Journal of Abnormal Psychology, 99, 148-155 Davidson, R.J (1992) Anterior cerebral asymmetry and the nature of emotion [Special Issue] The role of frontal lobe maturation in cognitive and social development Brain and Cognition, 20, 12.5 151 Davidson, R.J (1993) Parsing affective space: Perspectives from neuropsychology and psychophysiology [Special Section] Neuropsychological perspectives on components of emotional processing Neuropsychology, 7, 464-475 Derry, E A., & Kuiper, N A (1981) Schematic processing and self-reference in clinical depression Journal of Abnormal Psychology, 90, 286-297 Ekrnan, E (1992) An argument ibr basic emotions Cognition and Enu~tion, 6, 169-200 Epstein, S (1983) A research paradigm for the study of personality and emotions In M M Page (Ed.), Personality: Current theory and research Nebraska symposium on motivation (pp 91-154) Lincoln: University of Nebraska Press Eysenck, H.J., & Eysenck, M W (1985) Personality and individual differences: A natural science approach New York: Plenum Press Frijda, N H (1987) Emotion, cognitive structure, and action tendency Cogaition and Emotion, I, 115-143 Frijda, N H (1993) The place of appraisal in emotion Cognition and Emotion, 7, 357-387 Frijda, N H (1994) Lazarus' labour of love Cognition and Emotion, 8, 473-482 Frijda, N H., Kuipers, E, & ter Schure, E (1989) Relations among emotion, appraisal, and emotional action readiness Journal of Personality and S~'ial Psychology, 57, 212-228 Gallaghei, D.J (1990) Extraversion, neuroticism, and appraisal of stressful academic events Pe~aonality and Individual Differences, 11, 1053-1057 Gray, J A (1990) Brain systems that mediate both emotion and cognition [Special Issue] Development of relationships between emotion and cognition Cognition and Emotion, 4, 269-288 Guidano, V E, & Liotti, G (1982) Cognitive processes and emotional disorderx: A structural approatz to psychotherapy New York: Guilford Press Hatter, S (1990) Developmental diitierences in the nature of self-representations: Implications f ... personality characteristics Across several samples these analyses consistently revealed two underlying dimensions of positive personality, namely, Positive Individualism (PI) and Positive Relations... differentially associated with negative emotions Finally, it was demonstrated that cognitive schemata of anxious and depressed individuals are differentially related to the encoding of emotionally negative... that trait anxious individuals would most likely experience an entirely different set of core themes, namely, those centering around feeling uncertain, threatened, and overwhelmed by some immediate