For a generation, the cognitive perspective dominated the social sciences. Emotions were relegated to the fringes of scientific work. But, with the inevitable cresting of a virtually exclusive interest in cognition and its problems, emotion has reemerged as a legitimate and interesting topic of inquiry. 1975 was the watershed year, with Arlie Russell Hochschild publishing an article on emotions in a feminist collection; Thomas Scheff organizing the first session on emotions at the American Sociological Association; and Randall Collins theorizing a central place for emotion in "the microdynamics of stratification." By the end of the 1970's, several books specifically devoted to emotions by sociologists had appeared, particularly those by Thomas Scheff and Theodore Kemper. At the brink of the 1980's, the sociology of emotions was poised for developmental take-off - and it has. Many new participants and publications emerged throughout the 1980's, culminating in the successful organization of a section on emotions in the American Sociological Association (ASA).
In this article, Kemper attempts to outline the current state of theoretical research based on the ASA's first two thematic sessions in 1987. The idea in these sessions was for some of those who had developed theories to present both an overview of their position and a program of research that would help to elaborate and test the theory they espoused. What Kemper does is to take each of these contributions and discusses them in relation to particular theoretical choices, options, or alternatives that are available in social scientific work today (e.g. micro vs. macro; positivism vs. anti-positivism, etc). These, he argues, "are organizing matrices for thought and research. They aid implicitly (and often unconsciously) in assembling certain ideas and, as certainly, in precluding others. Each of these will now be discussed in turn.
Micro vs. Macro: Today, more self consciously than before, sociologists are examining the micro-macro mix and the nexus between them. In the sociology of emotion, Randall Collins stands out in this regard, viewing some of the central processes of macro-sociology such as order, conflict, and stratification as resting on the long unappreciated micro-level foundation of emotion. For Collins, social order is Durkheimian solidarity and moral commitment. These emerge in the course of "interaction rituals" at the micro-level, when two or more actors focus on a common activity. In the proper circumstances, a common mood is experienced, and this leads to a sense of unity among the actors. Conflict, too, rests on a foundation of emotion, as much as it involves a mobilization of sentiments of anger towards carriers of opposing interests. Both perspectives, moreover, are joined in Collins analysis of stratification.
Collins views stratification in terms of two dimensions, namely, power and status. Power rituals are interactions structured by the division of roles into order givers and order takers, mainly in large scale organizations. Although these actors participate with different goals and interests, crucially a common mood of shared emotions arises nonetheless. Because of their dominance, order givers derive "emotional energy" from the interaction. On the other hand, order takers frequently experience loss of emotional energy. Similarly, Collins talks of status rituals which involve, for example, membership inclusion or exclusion which respectively increase or diminish emotional energy. Broadly speaking, these interaction patterns provide a foundation of emotional resources for the participation in further interactions - which Collins expresses in his idea of "interaction ritual chains," which cumulate across time and space to constitute the macro-structure of stratification.
A second theorist who focuses on the relationship between micro and macro is Michael Hammond. Taking an evolutionary view, he contends that, as a species, we have an inbuilt need for dependable, long-term emotional gratification, and that we rank choices among possible sources of "affective maximization" (e.g. persons, objects, experiences) according to a scale of preference established by the cumulated prior experiences of others. Now, early in history, Hammond argues, this hierarchically differentiated preference system would have led to very little hardship, insofar as there was scant economic surplus to distribute disproportionally among the stratified population. However, with the growth in scale and the introduction of technological complexity, increasingly large surpluses were appropriated in such a manner that those at the low end of the hierarchy experienced considerable deprivation. In these circumstances, a new form of differentiation emerged, particularly for the benefit of those who were ranked low in the existing preference order. This involved the hierarchical differentiation of time: the future was designated as the repository of a superior level of affective maximization, either in this life or in a postulated hereafter. Belief systems are then invoked to alleviate the burden of inadequate affective resources in the present. In essence, then, the human emotional matrix shapes social organization in order to facilitate the pursuit of affective goals, while cognitive management is simultaneously applied to the mending of emotional distress for those at the low end of the scale.
A third theorist who weighs in on the micro-macro issue is Thomas Scheff. He approaches this through shame and pride, which he views as the linchpins of social control. Scheff proposes that individuals engage in more or less perpetual emotional monitoring of the sentiments of approval or disapproval of self that are presented by others. Individuals not only cognize others' reactions, but also react emotionally with either pride (for approval) or shame (for disapproval). These emotions operate in somewhat gyroscopic fashion to guide the individual along a socially prescribed path. Cumulating this effect across the whole society provides a micro-basis for a macro-effect, namely, the general pattern of conformity that prevails in society. Social order is thus constructed by aggregating the total of individual cases of the experience of pride and shame.
Of course, the inevitable question arises, if pride and shame are so crucial to the maintenance of social order and so pervasive, then why don't we see these emotions manifested more often? Scheff proposes that both shame and pride are indeed present, only they are masked - sometimes even from those who are experiencing these emotions. Shame, specifically, can be of two types: "overt, undifferentiated shame," which includes painful feelings, blushing, lowered gaze, and low-volume stammering speech among the major signs; and "bypassed shame," which is marked by more covert symptoms such as obsessive, repetitive thought. Scheff argues that both of these, particularly the latter are so prevalent that they may contaminate many situations in which actors experience critical evaluation from others.
A fourth theorist who takes a macro view in relation to emotion is Norman Denzin, who focuses on movies and television. He offers a semiotic reading of dramas, discovering an underlying emotional logic in these productions in which cultural and societal definitions of class and gender provide the subtext for the manifest emotions of the characters. As vehicles of culture transmission, film and TV dramas define proper and legitimate emotions and serve to shape individuals' desires and self-definitions so that they can experience the emotions that society deems proper for them.
Finally in this regard, Steven Gordon explicitly links micro and macro through his employment of the classical social structure and personality paradigm. Gordon proposes that social structure and culture may differ in how they affect emotions: the former more influential on behavioral and motivational aspects of emotion, while the latter may have greater effect on the quality, intensity, object, and setting of emotion. Gordon poses questions about what specific elements affect emotion: (a) whether it is better to look at structural effects on emotion content, or at abstract elements of emotion; (b) what intermediate elements operate to translate macro structures into micro effects; (c) how structural change in society ultimately leads to change in the emotions that are socially relevant; and (d) about the cultural relativity of emotion paradigms.
Overall, Gordon distinguishes between emotion, which he proposes is a response based on inborn, undifferentiated bodily arousal, and sentiment, which he defines as a socially learned composite of bodily sensations, gestures, and cultural meaning connected with a social relationship and provided with a cultural label. This definition fosters a perspective on emotions (sentiments) that stresses their social, as opposed to biological, origin. Entailed are such matters as their (a) long term character (in contrast to the relatively short time span of emotions more closely linked to the biological; (b) the social constraints that structure emotional situations even in the absence of underlying physiological response; and (c) how social fiat overrides the physiological in dictating change in emotion. Gordon proposes three processes by which emotions are transformed into sentiments: Differentiation, which elaborates the raw emotional materials into highly nuanced and complex patterns that are coordinated with social variability; socialization, which entails the social processes, including rewards and punishments, and modelling, by which culture members learn about emotional experience; and management, which allows social determination through normative interventions that bring emotions into line with social prescriptions. Conclusively, in regarding the micro-macro link, Gordon focuses attention on the analytic possibilities in both directions - not only how societal structure affects emotions, but also how emotions affect societal structure.
Quantitative vs. Qualitative Methods: The second dimension that Kemper considers concerns whether various theorists feel that emotions can be measured and quantified on the one hand, or are entirely qualitative by their very nature.
Lynn Smith-Lovin and David Heise fudge the issue by answering yes to both questions. Their "affect control theory" asserts that emotions have both quantitative and qualitative aspects, and formulates the emotions question in a way that embraces both approaches (in their view). What they have done is calculate the numerical values of certain underlying meanings of common language terms for behaviors, identities, objects, and emotions. Given that a certain situation reflects a certain meaning, they assert that they are able to predict the emotions that are likely to be felt in the situation. They base their meaning analysis on three commonly found dimensions: (a) evaluation (how good or pleasant something is); (b) potency (how powerful or strong something is); and (c) activity (how aroused or active something is). All the items of a culture - including objects, actions, roles, and emotions - can be rated according to the three dimensions. In essence, this produces a profile of a culture in terms of evaluation (E), potency (P), and activity (A). Consensual ratings are obtained from culture members who provided numerical scores for each item and the average ratings of an item on the three dimensions comprise the meaning of the term.
Smith-Lovin and Heise's affect control theory holds broadly that actors behave so as to maintain their fundamental identities as reflected in EPA profiles. In cases of discrepancy between behavior and identity, and emotion will be experienced - which emotion depends on the degree of discrepancy. Emotion provides a signal that there is a discrepancy to be mended, or an identity to be reformulated. According to the theory, the actor will seek to return to a state of harmony with the fundamental identity at issue, and will tend to choose a subsequent act that will attain that harmony. Failing this, there can be a shift in identity.
In addition to Smith-Lovin and Heise, the other sociologists of emotion who have collected empirical data (and discuss it in Kemper's book) are Arlie Russell Hochschild, Peggy Thoits, and Candace Clark. Their approach to emotions is mainly qualitative.
Positivism vs. Anti-Positivism: One of the major intellectual debates in the social sciences today is between positivists and anti-positivists. In general, the positivist view is that emotions can be examined as more or less objective phenomena, determined by certain social structural and interactional conditions. Anti-positivists, on the other hand, argue against the possibility of treating emotions as objective, measurable phenomena. Emotions are considered cognitive constructions that have no reality aside from the mental processes that allow the individual to perceive situations that normatively demand certain emotional responses.
In addition to Smith-Lovin and Heise referred to above, Kemper argues that the positivist view is represented mainly in his own work. Kemper proposes that social structures give rise to specific emotions - at least modally - and that a sociology of emotions must also accommodate the physiological underpinnings of emotions. His idea that social structures determine specific emotions is based on the notion that we are phylogenetic inheritors of (i.e. have biologically evolved such that we exhibit) a set of primary emotions - fear, anger, joy, and depression. These, according to Kemper, serve evolutionary adaptive needs, and are responsive to certain environmental contingencies. For humans, the major environmental contingencies are social. Therefore, for Kemper, the vicissitudes of social relations determine emotions. In particular, he argues, in typical positivist fashion, that the emotional outcomes of interaction can be predicted from a model that centres on the social dimensions of power and status. Power and status interactions directly produce emotions. Fundamental to his model, however, is the idea that actors necessarily interpret their and others' power and status positions subjectively. Regardless of phenomenal differences from culture to culture in what constitutes power and status, and in what one judges to be a "good" or a "bad" outcome, however, once the judgement is made, the phylogenetically adaptive emotions will ensue when certain power and status outcomes occur in interaction.
In the sharpest contrast to this positivist view, Norman Denzin proposes a sociology of emotions that ignores the conventional science approach. He feels that it is inappropriate to treat emotion as a variable. Indeed, he opposes efforts to build a theory of emotions, whether grand or middle range. Instead, Denzin prefers to phenomenologically study what he calls "emotionality." His procedure includes examining how emotion is structured as a lived experience, the forms of emotional feeling and intersubjectivity, violent emotions, temporality and emotions, epiphanies and shattering emotions (which have the effect of bringing about decisive changes in lives). Nevertheless, despite Denzin's phenomenological orientation, Kemper asserts that some of the questions that he proposes may share some common ground with positivist theorists, for example, his interest in cultural constraints on emotionality, the relation of emotions to self and biography, and differences in emotionality according to gender.
Political Economy of Emotions: Another axial difference in contemporary sociology is whether an issue is examined apart from its historical rootedness, or whether the historical context is analytically central to the problem. At one end of this dimension is Norman Denzin, who treats the problem of emotions as rooted fundamentally in historical issues of political economy, along with gender and class. These structures, in our "post-modern" culture, have "cultural" effects: they provide a set of meanings embodied in cultural productions - such as movies and TV - that set forth models of action and emotionality. These reflect the political economy of postwar, late capitalist America, whose elements include: (a) bureaucratization, which organizes individuals into compartmentalized roles and interaction opportunities; (b) commodification, which translates all human interests, including desire, sexuality, and eroticism, into marketable goods; (c) mass-mediated reality, which removes individuals from direct encounters with the world, but overloads the senses with pseudoreality; and (d) the deconstruction of (or concentration of extreme skepticism on) major sustaining "myths" such as the value of science and/or religion, the prevalence of freedom, and the efficacy of democracy in favor of a focus on scandal, hypocrisy, conspiracy, and obfuscation.
Denzin seeks no "theory" of emotion, but rather to understand how emotionality is "lived" in such a world. This approach allows him to examine the cultural and media productions of our time not in terms of plot, but as strictures about the limits of emotion and reason in the very society that gives them life. Viewers' needs and emotional selves are shaped by these films. They learn how to be, and what it means to be, emotional in a particular historical moment of the productive process, in particular sites where particular emotions are endorsed as suitable and desirable. Denzin argues that the social relations of emotionality in the media are ideological, not the real relations that prevail, and must be examined.
Arlie Russell Hochschild also works in the area of political economy. One of her main concepts, as we will discuss in more detail later, is emotional labor. She defines this as the emotion work that one must do as a part of one's job. Emotion work is brought into play when what one feels is discrepant from the "feeling rules" proposed by society or self. Emotions must then be managed to bring them into line with feeling rules. Hochschild proposes that this can be done by such methods as surface acting, or by deep acting. In the former, one puts on the expressive visage or body stance of the emotion, in the hope of stimulating the authentic feeling. In deep acting, one resorts to more profound strategies, such as modifying bodily or mental states, designed to evoke an emotion more in line with feeling rules. Hochschild's analysis of the emotional labour of airline flight attendants provides insight into the way that emotions have become a commodity, sold along with one's labour power in capitalist society. Workers are constrained to adjust their emotions according to the feeling rules set down by their employers, rather than feeling the emotions that they would normally feel in the circumstances.
Gender Analysis: Another factor cutting across recent work in the sociology of emotions is gender analysis, which must address both biological differences in male and female hormonal processes, as well as social and cultural shaping and repression of emotion.
Hochschild addresses gender in her work on gender ideologies, which are justifications for the maintenance or change of social relations between the sexes. Focussing on two-job families that also have a young child, Hochschild applies her notion of feeling rules and emotion work to the problems the couples face in the management of the complex fit between work roles and child care, which she imaginatively refers to as the "second shift." According to Hochschild, gender ideologies among these couples divide into three types: (1) "traditional," where the husband works and the wife mainly stays at home engaged in child care; (2) "egalitarian," where both husband and wife should work and share responsibilities equally for child care; and (3) "transitional" (something between the first two). Hochschild asserts that when partners have discrepant ideologies, the stage is set for powerful emotions. She examines the emotional pathways that are used by women in developing a gender strategy for change, and by men in resisting their partner's demands for a more egalitarian arrangement.
Another theorist who employs gender analysis to emotion management is Peggy Thoits. She finds that in the face of distressing emotional experience, women and men choose different emotion management strategies. Women tend to seek catharsis experiences and social support, try to see the situation differently, and gain perspective by writing about the distress in diaries, letters, poetry, and the like. Men tend to try thinking through the troubling situation, engaging in hard exercise, or simply accepting their distress.
Kemper notes that, to a lesser extent, Denzin, Clark, and Gordon also examine emotions through a gender lens.
Managing vs. Accounting for Emotions: Indisputably the most popular perspective in the sociology of emotions deals with how emotions are managed. This contrasts sharply with the approach that attempts to account for emotions in the first place (e.g. Collins and Kemper). "One reason for the popularity of the management approach is its compatibility with a generally antipositivist stance, derived in particular, here, from symbolic interactionism."
The management approach to emotions is based on the foundations developed and elaborated by Hochschild, Thoits, and Clark. First, this asserts that emotions are socially constructed. This means that emotions are not irrevocable, biologically guided, natural phenomena that simply happen to people. Rather, they are amenable to social direction, enhancement, and suppression. Second, social construction is mainly accomplished via norms or feeling rules that inform individuals about which emotion is suitable in which situation. Third, the social constructionist position asserts that emotions can be managed. This means that when a deviant emotion is experienced, the individual who is cognizant of the norms can take measures to reintegrate his or her emotional experience with the normative requirement.
We have already discussed Hochschild's position that in various occupations, emotion management is one of the principal requirements, since the job consists mainly of emotional selfpresentation. Similarly, Peggy Thoits explains in some detail how management of emotions can be accomplished. First, like Hochschild, she postulates that emotion is understandable as a complex consisting of situational cues, physiological reactions, expressive gestures, and an emotion label. These are interlinked in such a manner that changing one of these elements has a potential domino effect on the others. Thoits further postulates two modes of emotion management: behavioral and cognitive. That is, those who find that they are emotionally deviant and want to accommodate to the relevant emotion norms may work on the four emotion elements either through behavior change or cognitive change.
Behaviorally, one may change situational cues by leaving the situation, or by rearranging it. One may change one's physiology by breathing more slowly, or by charging up one's system through heavy exercise or by ingesting drugs. One may change expressive gestures directly by self presentation. Cognitively, on the other hand, one can change the situation by reinterpreting it. One can apply cognitive leverage to physiology by monitoring and concentrating on one's physiological signs. Finally, one can relabel the emotion in the light of other considerations.
Emotion management (of both own and others' emotions) is also the focus of Candace Clark's exploration into the micropolitics of emotion. Clark wishes to explain how people come to know, defend, or extend their "place" in social relationships (i.e. their composite rank vis-a-vis another on the dimensions of power, status, and intimacy). Those with higher place operate more freely; those with lower place are more constricted. According to Clark, although social interaction directly establishes one's place, one's place claims are limited by the self-concept, which may dictate more or less place as proper or deserving. But emotions also have a hand, for they operate in a number of ways to determine the place one claims, and the methods by which the claims are made. First emotion serves a signal function with respect to place, for example, acting in an embarrassing manner directly affects one's place in the interaction. Secondly, emotion is a place marker, both intrapersonally as well as interpersonally. In the former, emotion demarcates one's place, or makes one pliant to the sometimes egregious place claims of others. In the interpersonal mode, emotions are often the message of a place claim addressed to others, altercasting them for certain responses.
Clark proposes five ways in which emotions can be used as tactical weapons in the micropolitics of claiming or maintaining place. First, expressing negative emotions to another or withholding positive emotions is intended to induce the other's fear, or shame, hence "putting the other one down." Second, expressing positive emotions to another, or withholding negative emotions, is intended to induce the other's liking or solidarity sentiments. Third, controlling another's level of emotional arousal, as, for example, making the other lose his "cool" while remaining calm and collected oneself is intended to "displace" the other. Fourth, eliciting the other's feeling of loyalty and obligation assures that the other will not reduce one's own place. And last, one can patronize the other by expressing positive emotions that mark one's own superiority and the other's inferiority, as for example, expressing sympathy for a superior, or pointing out the other's problems. According to Clark, these tactics of emotional micropolitics serve to create and maintain hierarchy in social relations.
Kemper notes, finally, that the affect control theory of Smith-Lovin and Heise can generate predictions about when emotion management is likely to occur, namely, when behavior has led to a transient identity which is at odds with the fundamental identity of the actor. Gordon's work also considers the management of emotion as significant.
Prediction vs. Description: Kemper notes that the beginning of all theory is description. The debate here is whether description can be set aside at some point and more formal operations can be undertaken, such as creating theoretical propositions and testing predictive hypotheses. This issue partakes somewhat of the positivist-anti-positivist polarity, save that most positivist- oriented investigators recognize the importance of description as much as do the anti-positivists. Hence, Kemper, Smith-Lovin and Heise, Thoits, Hammond, and Collins, while not rejecting description, either explicitly or implicitly favor a predictive approach where description can eventually lead to predictive hypotheses. Gordon, Hochschild, and Clark, in contrast, shun the predictive approach in favour of description. To some extent this is due to their sense that insufficient data exist for the formulation of testable theoretical propositions. But their approaches, which owe something to symbolic interactionism, also derive from it the premise that we cannot know in advance how individuals will construct their lines of action. Therefore, to seek predictive theory is inherently problematic.
Biosocial vs. Social Construction: The final issue separating different approaches in the sociology of emotions entails the role of biological and physiological influences. In the sociology of emotions, the confrontation of the biological and the social is both more focussed and more heated than in most other sociological subfields. Virtually every sociologist of emotion acknowledges a physiological substrate to emotions. The debate turns on how important it is.
Gordon denies its significance for the sociology of emotions, affirming rather that the sociologically interesting emotions - he calls these sentiments - are socially constructed derivatives of raw emotional arousal (i.e. if anger is raw emotion, social mechanisms create sentiments of annoyance, rage, bitterness, and jealousy. According to Gordon, it takes social construction and emotional culture to make these variants possible. Hochschild, in contrast, prefers an "interactive" approach, in which physiological reactions are part of the emotion complex, but subject to a significant degree of social management. Finally, there are theorists such as Kemper who emphasize the links between sociological and physiological processes in the formation of emotion, arguing that emotion is biologically rooted regardless of the degree of social conversion, construction, or management.
In closing his survey of current theoretical positions in the sociology of emotions, Kemper concludes that the field is extremely broad and accessible from virtually any sociological persuasion, and argues that these differences provide opportunity for useful research, debate, and, eventually, perhaps theoretical syntheses.