Goffman has carried the conceptual heritage of Dewey and of Gerth and Mills as far as he can without leaving his behavior­ism and his “moments and their men” perspective. But now we need a theory that allows us to see how institutions—such as corporations—control us not simply through their surveil­lance of our behavior but through surveillance of our feelings. Such a social theory of emotion must have both a social and a psychological side. It can start by extending the ques­tion Gerth and Mills ask: How do institutions influence per­sonality? But we may specify that question: How do institu­tions control how we “personally” control feeling? In pursuit of an answer to this question, I draw, as Gerth and Mills did, on Weber’s appreciation of the power of bureau­cracy and on Marx’s sense of the interests that a bureaucracy actually serves. I also draw frequently on C. Wright Mills’s focus, in White Collar, on the “sale of personality.” But I add to Mills the notion that a personality is not simply “sold”; people actively manage feelings in order to make their per­sonalities fit for public-contact work. I also add three ele­ments found in Goffman: the focus on rules, the perspective of the affective deviant (the worker who is not obeying the feeling rules of the workplace), and an awareness of the ef­fort it takes to pay our “emotional dues” to an occasion.

On the psychological side, a social theory of emotion must take into account that these emotional dues can be costly to the self. Institutional rules run deep but so does the self that struggles with and against them. To manage feeling is to ac­tively try to change a preexisting emotional state.

But then we must ask: What is emotion? Emotion, I sug­gest, is a biologically given sense, and our most important one. Like other senses—hearing, touch, and smell—it is a means by which we know about our relation to the world, and it is therefore crucial for the survival of human beings in group life. Emotion is unique among the senses, however, because it is related not only to an orientation toward action but also to an orientation toward cognition.

The connection of emotion to an orientation toward action was key for Darwin. Indeed, he defined emotion as something quite close to this: as a protoaction, as what oc­curs instead of or before an action, as an action manque. An­ger, Darwin suggests, is the preact or prelude to killing, and love is the prelude to copulation; and we may add that envy is the prelude to stealing, gratitude the prelude to giving back, and jealousy the prelude to excluding. Emotion, there­fore, is our experience of the body ready for an imaginary action. Since the body readies itself for action in physiologi­cal ways, emotion involves biological processes. Thus when we manage an emotion, we are partly managing a bodily preparation for a consciously or unconsciously anticipated deed. This is why emotion work is work, and why estrange­ment from emotion is estrangement from something of im­portance and weight.

From the interactional theorists, then, we learn what gets done to emotion and feeling and also how feelings are a pre­amble to what gets done to them. From Darwin, as from other organismic theorists, we gain a sense of what, beneath the acts of emotion management, is there to be managed, with institutional guidance or in spite of it. Yet this is not the whole story. It is not simply true that the malleable aspect of emotion is “social” (the focus of the interactional theorists) and that the unmalleable aspect of emotion is its biological link to action (the focus of the organismic theorists). Rather, the unmalleable aspect of emotion (which is what we try to manage) is also social. This point could be analytically sepa­rated from the rest of the thesis with no harm done, but I add it because I think it introduces still another avenue through which to develop a social theory of emotion. And for this account of the social influence on the unmalleable aspect of emotion I move to Freud’s notion of the signal function of emotion, and from there to the influence of our prior expectations about how signals “signal.”

I have said that one reason emotion is unique among the senses is because it is related to cognition. Broadly inter­preted, cognition is involved in the process by which emo­tions “signal” messages to the individual. Freud wrote about the “signal function” of anxiety; anxiety, according to Freud, signaled the presence of a danger from within or outside the individual. It was a means by which the individual told of an apprehended danger. Similarly, other emotional states— such as joy, sadness, and jealousy—can be seen as the send­ers of signals about our way of apprehending the inner and outer environment. Thus to Darwin’s idea of emotion as an action manque, we may add Freud’s idea of the signal func­tion; they are two elaborations on how emotion, as a sense, differs from our other senses.

But signaling is complex —it is not the simple conveying of information about the outside world. It is not a telling. It is a comparing. When an emotion signals a message of danger or safety to us, it involves a reality newly grasped on the tem­plate of prior expectations. A signal involves a juxtaposition of what we see with what we expect to see—the two sides of surprise. The message “danger” takes on its meaning of “danger” only in relation to what we expect. (Sartre develops this point further, 1948.)

In this regard, expectation enters into the signal function of feeling even as it enters into the signaling of other senses—sight, for example. What we see is known to be me­diated through our notions of what we expect to see. As the classic experiments of Solomon Asch have shown, a person who expects to see a long rod on a screen because others around him say they see a long rod reports that he “sees” a long rod even when the rod is short and what the person “sees” is short (1952).

Prior expectations are part and parcel of what we see, and in the same way they are part of what we feel. The idea of prior expectation implies the existence of a prior self that does the expecting. For example, when we feel afraid, the fear signals danger. The realization of danger impinges on our sense of a self that is there to be endangered, a self we expect to persist in a relatively continuous way. Without this prior expectation of a continuous self, information about danger would be signaled in fundamentally different ways. Most of us maintain a prior expectation of a continuous self, but the character of the self we expect to maintain is subject to profoundly social influence. Insofar as our self and all we expect is social—as by the time of adulthood it inevitably is—the way emotion signals messages to us is also influenced by social factors.

Mechanisms of defense are ways of altering the relation of expectation to grasped fact as well as ways of altering each in itself in order to avoid pain. For example, if a woman sud­denly learns that her life partner has been killed, she may alter the character of her understanding of this event so as to keep it in line with what she expects—that he will still be living. She may defend against the self-relevance of the event: “This isn’t happening to me’.’ Or she may defend against the event itself: “He’s still alive. I know he is. I don’t believe he’s dead.” In these ways she holds prior expectation and current perception in a relation to one another that avoids pain.

When we finally go on to make inferences from our feel­ings to “how I must be interpreting this event” or “what must be happening,” we seem to presume that our emotion signals not simply our apprehension of the world but our prior ex­pectations about it. It signals the relation between the two. As practical actors in the world, if not as theorists, we seem to read feeling as a tell-tale sign of “what we must have ex­pected or wanted” as well as a sign of “what was going on.”

To sum up, I am joining three theoretical currents. Draw­ing from Dewey, Gerth and Mills, and Goffman within the interactional tradition, I explore what gets “done to” emo­tion and how feelings are permeable to what gets done to them. From Darwin, in the organismic tradition, I posit a sense of what is there, impermeable, to be “done to,” namely, a biologically given sense related to an orientation to action. Finally, through Freud, I circle back from the organismic to the interactional tradition, tracing through an analysis of the signal function of feeling how social factors influence what we expect and thus what feelings “signal.”