Ideology and Interpersonal Emotion Management

Social Psychology Quarterly 1997. Vol. 60, No. 2. 153-171

Ideology and Interpersonal Emotion Management: Redefining Identity in Two Support Groups* LINDA E. FRANCIS State University of New York at Stony Brook

This ethnographic study demonstrates the process by which our emotions are constructed not only by ourselves but also by others. Support groups for divorce or bereavement are used to illustrate the process of interpersonal emotion management. The study considers how support groups with differing ideologies produce dissimilar situational definitions for the loss of a spouse. Each of these definitions promotes different cognitive and affective outcomes for the participants. Despite these differences in definitions, the group leaders use a largely identical process of interpersonal emotion management, in which they redefine not only the event of spousal loss but also the sufferer's very identity. These redefinitions encourage understandings and emotions that coincide with the groups' own ideological perspectives. In addition, the results of the study are unexpectedly congruent with affect control theory, thus demonstrating that the propositions of this theory inform qualitative as welt as quantitative research.

Just as others influence what we think, they can also shape what we feel. Drawing on Katz (1980). Hochschild (1983) argues that "what feelings 'signal' to us as sociologists is how culture influences what we feel and how we name it" (p. 224), Though this statement is quoted somewhat out of context, the notion that we look outside ourselves for clues in interpreting our emotional signals is relevant here. Not all signals from others, however, are given off unintentionally (Goffman 1959). As with any hehavioral norm, others may deliberately encourage, discourage, or mold our emotions to fit the situation. This is most apparent in the support group, where emotions and emotional hehavior are the primary focus and where interpersonal emotion management (Francis 1994) is the reason for the group's existence. This fact, however, raises two questions: How do support group leaders choose which emotions are defined as healthy? How do they deliberately generate these emotions (and suppress others) in their group participants? This study is an attempt to further our * I am grateful to David Heise and Richard Adams for iheir extensive comments on earlier drafts of this paper, and to K.T. Amesen for her work on the figures. I would also like to thank the editor and anonymous reviewers of SPQ for their many helpful suggestions. Please e-mail any correspondence to Linda Francis, [email protected].

understanding of the social and interactional nature of emotion. I focus here on the process by which others can shape, work, or manage our emotions. I particularly wish to learn how support groups seek to accomplish emotional and cognitive change in individuals, and how those attempted changes reflect the ideologies in which the groups originate. For this reason I study support groups for bereavement and divorce, events for which the focus is not on problem solving but on managing strong, negative emotions. The resulting project has implications for the study of social support, support groups, and the self-help movement, but focuses primarily on emotion as an interactional phenomenon. The approach is that of grounded theory (Glaser and Strauss 1967; Strauss and Corbin 1990); the study draws theoretically on the work of Hochschild (1978, 1983, 1990), Thoits (1984, 1985, 1986, 1996), Heise (1979, 1987). The patterns that emerged from my analysis, however, are particularly congruent with the relationships described by affect control theory (ACT)(Heise 1979; MacKinnon, 1994; MacKinnon and Heise, 1993; Smith-Lovin, 1990 ), ACT was designed as a quantitative theory; thus the results reponed here demonstrate the productive cross-fertilization that can occur between quantitative and qualitative work.

153

154

SOCIAL PSYCHOLOGY QUARTERLY

Ideology. Emotion Management, and Support Croups Hochschild (1979,1983) was the first to argue that ideology is a crucial influence on the social construction of emotion. How people choose to perceive events or how others choose to have them perceived (Wasielewski 1985), reflects what they believe in general about the way the world works (or should work). Ideology therefore affects the way we define the events of our experience; these definitions outline the rules of behavior, whether that behavior is physical, rational, or emotional (Blumer 1969; Gofftnan 1974; Shibutani 1961). When our emotions do not conform to the feeling norms, we attempt to amend the emotions, at least superficially, in order to avoid sanctions (Hochschild, 1983). This effort is known as "emotion work" (Hochschild 1983) or "emotion management" (Thoits 1985). If the task of reworking our emotions is too large, we may go a step further and attempt to modify or redefine the situation so as to make them appropriate. When the emotion cannot be changed easily, the situation may be an easier target. Thoits (1985, 1986), however, argues that some events are not encompassed by Hochschild's work. A person might be faced with a situation—a death, for instance—that cannot be resolved, and with emotions so strong they cannot be tamed. If those emotions endure too long (or otherwise do not fit the emotion norms of those around the individual), others will begin to question that individual's mental well-being. In such a case, emotion work has failed and the situation that caused the emotion cannot be altered. The individual has reached the end of his or her own resources and is in danger of being labeled as deviant. Thoits (1985) describes this set of circumstances as the "self-labeling process," Individuals in such situations come to believe that there is indeed something wrong with them and they seek professional help. Whether they go to psychiatrists, counselors, support groups, or elsewhere, they are looking for someone to help them "fix,'* cope with, or manage their emo-

tions. Therefore the primary purpose of the therapeutic interaction is quite arguably interpersonal emotion management. Interpersonal Emotion Management According to traditional symbolic inieractionism, the participants must agree on a definition of the situation if interaction is to proceed (Blumer 1969; Stryker 1980). Correspondingly, when people turn to others for support, as in support groups, the supporters then must be able to define the stresscausing situation themselves. To do so, they must have relatively detailed information about the circumstances. This is probably one reason why the best informal support comes from those who have had similar experiences: They are most likely to understand the sufferer's definition (Thoits 1984). In any case, supporters and sufferers must cooperate to build a definition of the situation from which they can work. Unlike members of most informal support networks, professional social supporters (such as counselors, therapists, and support group leaders) are strangers to the person needing help. Therefore I argue that they must negotiate a definition of the situation from scratch, In the process these counselors will introduce the theoretical (and ideological) framework in which they were trained, which will shape the definitions that they negotiate with their clients. Individual psychotherapists may tend to work on definitions by going into great depth with one individual and building a somewhat unique definition (Frank 1974). On the other hand, support groups and counselors who specialize in particular problems have stock sets of "healthy" definitions for certain events (as in twelve-step programs) (Katz 1993). In either case, these definitions will limit the emotion norms that professional supporters suggest to their clients, as well as the coping strategies they recommend. Affect Control Theory and Redefining Emotional Events With the above discussion in mind, we can see that professional supporters, by altering

INTERPERSONAL EMOTION MANAGEMENT the definition of the event, should be able to change both the normative goals for which the sufferers are striving and the emotions they are experiencing, with the intent of bringing the experience and the goal into congruence. This point raises the issue of how one goes about changing these definitions. Indeed, this is the central question of this paper: How is interpersonal emotion management accomplished? Symbolic interactionists believe that a situation can be redefined cognitively by focusing on different aspects of that situation. Affect control theory (Heise, 1979,1987; MacKinnon. 1994; Smith-Lovin, 1990 ) extends Mead's paradigm by demonstrating how the emotional elements of a definition can be altered as well. Heise (1979, 1987) has drawn on Osgood, May, and Miron (1975) to argue that people interpret the world in terms of three "EPA" dimensions: evaluation (good-bad, niceawful), potency (strong-weak, large-small), and activity (lively-calm, fast-slow). People learn to associate certain enduring EPA ratings with certain identities. Thus a mother is rated as quite good, quite powerful, and neither particularly lively nor calm, while a gangster is rated as extremely bad, potent, and lively (Heise 1987). These enduring ratings are called "sentiments." When identities are enacted under various circumstances, events will occur which produce "transient" feelings that either confirm or confiict with the EPA sentiments associated with the identities. For example, a mother saving a baby confirms the sentiments associated with "mother," whereas a gangster saving a baby conflicts with the sentiments surrounding "gangster." The emotion that the agent experiences is due to this confirmation or conflict. According to ACT, when a conflict (or "defiection") occurs the person is motivated to seek explanation and to fmd some means of returning the confiict to a confirmation of sentiments. Thus people seek for ways to qualify the situation and remove the deflection. For instance, Heise (1987) points out that people might look more closely to see whether the gangster actually kidnapped the baby rather than saving it. Alternatively, observers might discover that the gangster

155

had children of his own. so the relevant identity is "father" rather than "gangster." A father saving a baby and a gangster kidnapping one confirm the sentiments respectively associated with "father" and "gangster." By rephrasing the argument of this paper in ACT terms, I suggest how interpersonal emotion management may be achieved. For instance, a sufferer presumably comes to a counseling situation with a ready-made deflection caused by a conllict between a transient feeling and an accustomed EPA sentiment, due to divorce or bereavement. In reference to the idea of emotion management and self-labeling (Thoits 1985), the sufferer has been unable to resolve this deflection, or, alternatively, has resolved the deflection in such a way as to take on a label that harms his or her self-esteem or emotional well-being. The counselor then draws on his or her training (and therapeutic ideology) to qualify the elements of the situation so as to reduce the deflection and transform the experience into a positive confirmation of sentiments. The qualifications may be (in fact probably are) extreme enough to require a new definition of the situation for the event. Thus the situation "husband leaves wife" may produce an unresolvable deflection for the wife; or she may resolve it negatively, as in "husband leaves shrewish wife." The counselor will alter these circumstances into (perhaps) "immature husband leaves wife," thus resolving the defiection and removing fault from the sufferer. By qualifying the elements in this event, the counselor has altered the definition of the situation, has invoked new emotions and new emotion norms, and (one hopes) has made some progress in reducing the conflict. This theoretical rendering of an interpersonal emotion management process creates a mental image of human action. One then can compare the image with the reality to determine whether it "fits" or needs to be altered. As I will show in the following analysis, this model of the process is applicable, but in unexpected ways. I argued above that the support group facilitator changes the frame of the event, which in turn calls up new expectations. In reality, however, the identities of the event participants are redefined; all else fol-

156


lows. Each group facilitator identifies the actors he or she considers relevant to the event and then defines their identities along the EPA dimensions in accordance with the ideology of the group. These identities promote—indeed, create—a particular definition of the event, which in turn encourages in the participants the emotions desired by the facilitator. Such a revision lends itself to an even more extensive application of affect control theory and demonstrates its usefulness for qualitative as well as quantitative analysis. ( I present these additions in greater detail in the discussion.) The redefining process as it emerges from the data is illustrated in Figure 1. DATA AND METHODS The data for this study were collected primarily through participant observation. This material had its source in support groups for the loss of a relationship through separation, divorce, or bereavement. Because of the comparative nature of the study, I include two dif ferent groups, both located in the Los Angeles - Orange County metropolitan area. These two groups varied in format and ideological perspective. The first was a sevenweek divorce recovery workshop, organized through a large evangelistic church and conducted by a marriage and family counselor who described himself as a "Christian thera-

pist." This group began with approximately 70 people but ended with only about 45. The format was half lecture/sermon and half small-group discussion. I refer to this as the "Discovery" group. The second group was an ongoing bereavement group organized and conducted by a widowed elementary schoolteacher with a master's degree in psychology. The group met twice a month in the library room of a church, although the group was actually associated with a local mortuary. Between 10 and 20 persons were present at each meeting during the six months 1 attended. I refer to this as the Rebirth Grief group. The data from these groups include 10 hours of audiotaped sessions from Discovery and 14 hours from Rebirth. For Discovery, these tapes cover the entire seven-week seminar. For Rebirth, they record seven consecutive meetings covering about 3 2 months. In addition, I have fieldnotes as well as written materials that the facilitators handed out to their groups. I transcribed the tapes and fieldnotes and imported them into Ethnograph, a computerized text-analysis program. In the following data excerpts I illustrate the process of interpersonal emotion management used in each group. These examples are not exhaustive because of the quantity of data from which they were drawn. I chose each excerpt for

Seek Profes.si(»nal Help

Professional Training/Ideology

Redefining of Event

Emotions

Feeling Norms

Emotion Management Figure 1 Model of interpersonal emotion management

Actors' Identities

INTERPERSONAL EMOTION MANAGEMENT representativeness of strategies used and for the clarity with which it demonstrates trends in the transcript data, These two groups represent differing perspectives and differing formats. Yet the purpose of examining such different groups is not to test hypotheses about support groups or life events, hut to investigate a process of interpersonal emotion management using two case studies. I make no claims of representativeness of support groups, nor of divorced or bereaved people. Indeed, my purpose is not to compare the experiences of the divorced and the bereaved, or even to compare support groups. Rather, my goal is to compare implementations of a process under conditions in which the ideological starting points are not the same. What is the connection between ideology and emotion? What role does this connection play in managing the emotions of others?

157

nite beginning and end, and the facilitator was able to control the flow and direction of information quite precisely. Thus it was relatively easy to identify the actors who were accorded notice and to discern the changes made in the depictions of these actors. Yet I could hardly determine how the participants were responding to this redefining. Ready access to the participants' response was the key strength of Rebirth grief group, which maintained a traditional ongoing support group format. Thus the data from this group indicate the negotiation of definitions between facilitator and participant. In the Rebirth group, however, the redefining process was much less direct and thus less clear. Indeed, only after analyzing the Discovery group could I detect a similar pattern in Rebirth's ongoing format. The Discovery Divorce Group

RNDINGS: SUPPORT GROUPS FOR LOSS OF A SPOUSE

This group is most notable for the leader's dramatic redefining of divorce: He changes it The Discovery divorce recovery group was from a negative to a positive event. This evaloperated as a short-term workshop; the pri- uative about-face is evident in the topics of mary source of information was a main facili- each week's lectures. Week by week, the tator speaking at a podium. Participants' input group progresses through increasingly upliftwas relatively small. This group was highly ing themes, which are summarized in motivastructured, the facilitator's ideas were well tional handouts distributed at the beginning of organized, and the groups' definitions were each meeting. Most of these handouts also clearly stated. feature short quotes or sayings that serve as My analysis of this group reflects the par- mnemonic devices for the theme of each lecticular clarity of Discovery's defining and ture {see Figure 2). redefming process. The workshop had a defiAs shown in Figure 2, the group begins

Week 2: Sugesof Relttiiiinship Retovcrv

Week 3: Fnreiueiw , iRcdcmpli c Hfrncmhcn 81 rhfiKi.ilnn. •vcti WIIC», fCViK Dl nkine. m l remmlel w

Week 4 LeHmgGo (ind Moving Forwuidl Wrck^

Goal

SellingStrategy [or hicallh

"Ify

SRLF AS CHRISTIAN

J i>i>i>'l t n

Figure 2. Discovery of Group Lecture Topics and Themes, by Week

CclcfuuW New Life ••G I G.O .-n." Giirbun in. Cartnii* Oul, huiGudln.GiHl

158


with an acknowledgment of distress, but with a push toward improvement.' The second lecture refocuses on emotions, with particular emphasis on overcoming negative feelings. In week 3 the group members look backward and rethink {or redefme?) those experiences. At the next meeting they are enjoined to let go of the past and its negative associations, which are holding them back from a positive future. The meeting in Week 5 stresses new directions, and in the final week the group leader specifics the direction he wants the group to take: toward God.' The emphasis changes explicitly from a negative to a positive view of events. This change is particularly evident in Week 1 ("turn your scar into a star"). Week 3 (review, revise, and remodel), and Week 7 (not Garbage In, Garbage Out. but God In, God Out!). With such careful arranging, one can scarcely say that the evaluative redefming of this group is unplanned. This should not be surprising, because it is the explicit goal of all therapy to (at least eventually) make people feel better. The more interesting question, then, is not whether but how the facilitator reaches a new defmition of events. To accomplish this redefining, the facilitator (whom I call Joe) identifies three actors in the divorce: the self, others (mainly the exspouse), and God. Each of these three is described—in fact, defmed—in terms of their identities and the situations in which they enact those identities in the ongoing event of divorce. The divorce, then, is defined as the composite of all these identified actors. The definition of divorce as an experience and a cultural phenomenon emerges from the definition of the actors, not the other way around. The important point here is that the therapeutic ideology of the group determines which actors are accorded notice as part of the divorce. This group's religious ideology states ' The quote for Week 1 is attributed to Dr. Robert Schuller. for Week 2 to Zeke Zigler, and for Week 5 to Dr. Henry Kissinger. ^ Week 6 has no obvious key quote, largely because the motivational handout bore little resemblance to Ihe night's lecture: the topic of the lecture was self-trust. while the handout topic appeared to be on making friends. Simple error was probably the reason for this discrepancy.

that God is an active presence in our lives, so God becomes an actor in the event of divorce. Thus, to redefine the divorce, Joe both defines tbe actors and then gradually changes their identities. Rather than redefining the event first, he changes the actors' identities until they require a new situational defmition in order to make sense. In this gradual redefining process, Joe alters the actors' identities along the three dimensions of ACT: evaluation, potency, and activity. As we shall see, the end result is a transformation from "I'm a failure" to "God is challenging me." Joe's first action in the group is to acknowledge the original definition of self-labeling, which led the participants to seek help (Thoits 1985). Yet he also makes an effort to nonnalize the feelings that people are experiencing, thereby assuring them that there is nothing wrong with those feelings. He takes pains to assure the participants that they are good and worthwhile people, and that although "God hates divorce...1 have never seen anything in scripture that says God hates divorced people." In this first lecture, titled "The Gift of Hope." Joe indeed offers the hope that because they are not bad people, their negative feelings eventually will improve. In ACT terms, Joe is telling the participants that they do not have a bad identity witb which they must continue to associate negative sentiments. Rather, they are good people who have merely had a bad experience, and their negative emotions are temporary (if deep) deflections that can be resolved. By contrast, Joe initially defines the exspouse—and indeed, all otber associated people—as more powerful, more active, and less good tban the self. They control the participant's life and cause much of the participant's misery: "There's always someone willing to stir your mud." "Someone's always around to pull your chain." How many of you have had an experience with what's called "someone else's priorities?" called 'This is imponant for you?" And I love the way they enunciate the you. How many of you have not established your line of priorities? To counter this, Joe begins to position tbe participant as more active. For instance, he

INTERPERSONAL EMOTION MANAGEMENT defines the world in terms of territory, that of the self and that of others, and insists that within the participant's own territory he or she can take control of his or her life. The participants are vulnerable only when they allow thenfiselvcs to be drawn outside their boundaries. Thus it is not healthy to trust others to set one's priorities, because "only you know what is important to you." Joe also is careful to reduce the exspouse's power, stature, and importance: Part of the problem we have has to do with the thing called attitude, because many of us feel that our former spouse is the only one that we could possibly share a life with, or the only person who'd possibly let us share life with them. Well, guess what, folks. There's only sixteen basic personahties. Now there's a lot of subparts within each personality, but I assure you, there's a lot of other people around that we can have marriages with.... In fact, not only are there other fish in the sea, but these particular fish are neither especially strong nor healthy. So if you can identify that situation and see the person for what they really were, and anybody that would hurt you. degrade you. humiliate you, embarrass you, are they secure, are they stable, are they healthy? No, They're just the opposite. So see them for the way they were at the time. Then you no longer have anything to fear, because you become in control yourself. Thus this fish is not even particularly desirable anymore. Forgive them for the damage they've done you in the past and move on to beuer things. This, then, is the close of the first redefining, in which the ex-spouse is relegated to a weak position from which he or she is unable to cause h;'rTn. In contrast to a good (beloved by God), strong (unconfused), and active (taking control) self. Such a definition, however, is not stable in itself. It unlabels the self by turning the negative sentiments of the original definition into a mere negative deflection. Yet the proposed resolution of this situation—forgiving the ex—is not likely to hold unless the participant never is required to interact again with the ex-spouse. For every interaction may arouse the same defiec-

159

tions and eventually may resolve into the same self-label as before. Thus Joe must redirect the interaction into a more stable form. To do this. Joe essentially discards the exspouse as a key actor in the frame and replaces him or her with God. That is, he redefines divorce altogether with new actors: the self and God. This is not to say that the first redefining was not necessary. After all. it established a positive identity for the self—an unlabeling, as it were. TTien, in the new definition, Joe argues that if one cannot trust others (especially the ex-spouse) to set the priorities of one's life, upon whom can one depend? His answer Is twofold: oneself and God, because one's own happiness and health will only be a priority to those two actors. Setting one's own goals and one's own priorities— putting one's own needs ahead of others'—is the "strategy for health" and tbe definition of success. To do so is to "be the person God always intended you to be." According to the Christian tradition, God is omnipotent and nothing is out of his control. Even negative events occur only because God either allowed them or deliberately set them in place, Yet God supposedly has only his children's best interests at heart; he would not hurt them except to achieve some greater long-term benefit. Thus negative events (such as divorce) are trials, tribulations, tests, challenges, or (as Joe describes them) "roadblocks" designed to slow individuals' progress and force them to reexamine their path. So why does God permit roadblocks in our life? Why? Maybe a roadblock is necessary to force us to examine our own perception and perhaps consider [others]. Maybe a roadblock is deliberately placed there, to get our attention. To awaken us. Maybe it's there to stop us to say "'Hey, wait a minute, you are on a wrong course, you need to reexamine the direction you"re travelling in." Maybe that roadblock has a very significant purpose. Roadblocks, then, are God's attempts to make people reexamine their priorities and do what is best for their own well-being. Thus God has been defined not only as very powerful and very good, but also as very

160


active. And the self has been totally redefmed from weak, passive, and bad to powerful, active, and good (see Figure 2). The self, however, is defined as powerful only in comparison with other people; it is weak in comparison with God. At this point we see that the situation of divorce has been redefined entirely. That is, the negative event has been redefined, by virtue of God's omnipotence, as a positive event, at least if one uses it to improve one's relationship with God. Joe ends by reiterating the new definition and offering it to replace the original self-labeling definition of pain and helplessness. I still have bouts of doubts, about listening to ihe old tapes and reactivating the old tiegative thoughts. But,..when the old tapes stan playing, then 1 ask a simple question: What is the possible worst case scenario that could come out of ihis particular situation?...But my God, the God of my heart, is so much bigger than my personal fear of failure or success. And I simply take hold of His hand. His outstretched hand, and allow him to lead me through the valley. I fmaily realized that my way is not always the right way. And he leads me through that challenge. Events now have a purpose and a potential happy ending. Your yesterday is your history. Tomorrow's not here. We only have this one God-given moment. So be willing to make the change, called the celebration. Wherever you are. be there! Rejoice! Say, "Lord, thank You for this opportunity, for getting my attention." Thus the fmal result is a new definition calling up new emotions: challenge, determination. Table I, Discovery Croup, hy Definitions

rejoicing, trust. (Notice the contrast with the ftrst quote from Joe.) This definition has been called up by means of the gradual redefinition of tbe actors in the situation, not merely of the divorce. This point is illustrated in Table 1. Each of the three main actors—God. the self, and the ex-spouse—has been altered along the tbree dimensions of evaluation, potency, and activity.

The Rebirth Grief Group The process in the ongoing support group is somewhat different than in the seven-week workshop. Joe sets out a definition, enumerates the identities, and then alters them until they call up an entirely new definition in order to make sense. In the ongoing group, however, this is not an option because the group has no clear-cut beginning, middle, or end. People enter and leave at different points; thus it is impossible to ensure that everyone begins with the same information. As is evident from the data presented below, however, the redefining of the event is still the goal despite the difference in procedure. Only the point of entry (the point at which the redefining begins) differs according to the format of the group. In the ongoing group the facilitator generally follows one of two procedures: either she makes a statement congruent with her definition of the situation, and then encourages discussion of that statement, or she allows participants to describe their experiences and then provides an interpretation congruent with the definition. In either case the process of redefining the actors in the event along the EPA dimensions

Self-Labeling Definition

First Redefinition

Second Redefinition

:ieit

Bad Weak Inactive

Good Strong Active

Good Strong Active

Ex-Spouse (and Others)

Bad Strong Active

Bad Weak Inactive

N/A

God

N/A

N/A

Very Good Very Strong Very Active

INTERPERSONAL EMOTION MANAGEMENT of ACT holds as well in this ongoing group as in the Discovery workshop. The Rebirth grief group is based primarily on Kubler-Ross's (1969) stages of death and dying; Kubler-Ross in turn is influenced by Ereudian theory. The facilitator of the Rebirth group, whom I call Lydia, thus characterizes her group as largely psychological. She believes strongly that her participants will not recover from their grief until they have worked through all five stages: denial/isolation, anger, bargaining, depression, and acceptance. When Kubler-Ross developed these stages, however, she was concerned with the dying person, not with the survivor. Drawing from her own experience of bereavement, Lydia has recharacterized the final stage, acceptance, as "psychological incorporation of the deceased" into the self Simply using a theory composed of stages implies a transformation, a movement from one emotional point to another, presumably better one. In this sense, such a perspective fits well with the explanation of interpersonal emotion management offered thus far. Yet in itself it does not demonstrate the appropriateness of an ACT explanation. And unfortunately, Lydia, unlike Joe in the Discovery group, does not provide a convenient list of titles to illustrate her purpose and process. In an individual interview, however, Lydia directly supported that view: Lin: We were talking about healthy versus unhealthy emotions, or progress versus lack of progress in grief recovery. What do you think would be the healthy emotions for people to feel, or...maybe a better way to put it would be. What emotions constitute progress? Lyd: Progress is when you can tell your story and cry, but not in a way that it controls you. I think progress in grief work is when you can come to group and say, "I see this person as he really was. I loved him, but I see his shortnesses and his weaknesses." and...then that's okay...I think that's growth. And then...also...for that person to say..."I see the other person in relationship to myself and I see my growth. I see who I am." That's part of the whole business in a relationship over [thel long term is you lose an identity. And so it's like you suddenly say..."I have incorporated some of his identity, and now

161

my identity is different, but I'm feeling pretty good about just that." In the above quote, I still focus on events and emotions rather than on actors or identities. My intent was to find out which emotions Lydia associated with the event of bereavement, and which emotions with the process of recovering from that event. Lydia is intent on talking about people, not events. Even at this late stage {I conducted interviews late in my fieldwork) I had not yet grasped that events do not exist of themselves but only through the people to whom they happen. Also, Lydia defines progress as seeing the deceased as more negative {having weaknesses) and the self as more positive (experiencing growth). In this way this quote foreshadows the EPA redefming of actors used by Lydia in her group. As in the Discovery group, Lydia's definition for the group lists three actors or identities of consequence, in this case the self, the deceased, and others (especially family members). Also like the actors in the Discovery definition, Rebirth actors are redefined twice in order lo reach Lydia's desired goal of acknowledgment and acceptance. Like Joe, Lydia aims in the first redefinition to defeat the self-labeling that the participants bring with them. These participants come to the group with a perception of the deceased as a good person torn involuntarily from his or her family. The bereaved watched the death helplessly, unable to save the deceased or even to ease the pain. Erom her therapeutic perspective, Lydia attempts to move the bereaved out of this perception and into the more empowering (but problematic) stage of anger. In my interview with her, she explains the matter: Lin: [Do] you try and work on different stages in the group? Maybe I misunderstood you when you said [that] you go back to old material when you have new people in the group. Lyd: 1 look at the grief process, and in that process there's a process of denial and there's a process of anger and there's a process of guilt, and there are all of those things, and..when I get a new person into group,...one of the areas that

162


always...comes up is an inability to express any type of anger [hat someone has died. It's toially unacceptable to be angry at someone that has died. But most of all...it is unacceptable in our culture, maybe even beyond our culture, to be angry at God,.,so Ihere's somebody you have lo pass it on to. Most of us lend lo transfer it onto the one that's dead because it's the safest one. But we also tend to deny that we're transferring onto that person because it's also unacceptable. So.,.we're not about to take on God because there seems to be some kind of innate something in us that says there is a higher power, and if there's a higher power then this is an awesome power, and why would I think that I can better an awesome power? 1 can better somebody that's dead because that person isn't here to bat. But then I have to deny that because if I admit it then everyone says, "Oh, my God, what a nasty person you are.,.," So there is a component of anger to having someone die and abandon you, and getting people to accept that, "Maybe I was a little angry." .,.It's very important. But early on there's no way it's going to happen because...you're still wanting to hang onto it. It's the only pan you have left anyway.

Lin: I see that quite a bit in the group, where people are refusing to say [they're mad]: "No, tl'm] not mad at the spouse. [I'm] not really mad at God either" They're not mad,..io quote Vera, "I'm not mad. I'm discouraged," To inspire the anger that she regards as crucial. Lydia seeks to transform the self from a failure to a victim and the deceased from a victim to a perpetrator. To make the bereaved a c k n o w l e d g e anger and conflict with the deceased, Lydia must make that anger appropriate. In an attempt to generate this anger in all members of the group, she tries to alter the perception of the deceased that most members brought to the group, along the lines of ACT. Lydia works to change the participants' view of the deceased from weak, inactive, and good to powerful, active, and somewhat bad. One of the neat things in grief recovery is [to] go to their workbench and wipe that sucker clean! They left you! They abandoned you! What's going on here? Let's be real clear about that. That's what you're feeling right now. That's what you were crying about in the cemetery. That's what hurts. The man abandoned you!

Note the active verbs in this quote. The implication here is that the deceased actively took the easy way out and left the survivor to clean up the mess alone. Thus although the deceased is not evil, he or she is defined as insensitive and selfish, as committing a negative act, and as meriting a certain amount of justified anger. Lydia finds it difficult to gain acceptance for this notion. Most group members think that if the deceased had had any power at all he or she would not have died. Nonetheless, such negation of this element of her defmition does not stop Lydia from attempting to incite anger in her group m e m b e r s . She resorts to other less direct tactics to achieve the same goal of empowering, activating, and desanctifying the deceased. She is more successful, for instance, in combating the participants' guilt in regard to the deceased. Many of the group members feel guilt for not having done something for the deceased while he or she was still alive. One m e m b e r , Gary, is a s h a m e d of not cleaning the garage during the entire 15 years of his marriage, despite urging from his wife. Lydia responds: Lyd: If your garage was a pigsty for fifteen years, then lei it be a pigsty! Who cares? It didn't send away your happiness for fifteen years, did it? h didn't send away your joy? Gar: No, but I know now. looking back, that the journey would have been happier if it had been put in some semblance of order.... And I guess I feel guilty that i didn't do it while she was alive to enjoy the fact that I was doing what I should have done. Lyd: Let's talk about you feeling guilty for a minute. All of the things that I've heard from you about your caring and loving for Jeannie, and the fact that you had found a relationship with a woman that you just adored, and then she died, and now you're going to try and make up with her the things that you didn't do. Gar:

What's wrong with that?

Lyd: You adored the woman, for God's sake! How much more do you owe her? How many

INTERPERSONAL EMOTION MANAGEMENT people in this room were adored by their spouses? Lydia goes to great lengths to reduce the goodness and deservingness of Gary's spouse, even to the point of arguing that she received more love than she deserved. Again, the definition Lydia places on Gary's wife lowers her ranking on the dimension of evaluation. This change is so important to Lydia's definition that she spends an entire day on making the deceased less good. Of course, she does not describe her approach in this way; she refers to this topic as "honest memories": [E]arly on in recovery...we tend to give [the deceased! characteristics that we wanted them to have that they didn't have....[W]hat it Idoesl.. you're going to lose a part of who that person was, and I think that's really a shame. There isn't any need to lose that person, even the things that you didn't really like about them, or the things that were...annoying. Everybody, every mate, there is a mild annoyance about something....that's okay and that's normal. And so after somebody dies you don't erase all of the annoyances and say they didn't occur. What you say is "They were there and I chose to love the person." ...So we're going to work a little bit on honest memories. This discussion also implies greater activity on the part of the deceased. It focuses on things the deceased did when alive, with the implication that these are negative actions for which the deceased requires forgiveness. Thus "honest memories" might he more aptly named "critical memories." Some of the long-term members of the group show signs of accepting this defmition and believing that the deceased was often at fault. For instance, Ron feels bad about not spending money on his wife to huy her things she wanted, but he also sees this as partly her fault; Ron; I never went in for buying a lot of things for her, because....It was sort of her own fault, because she loved for me to get things for her, but on the same token, she was always sort of critical.

Lyd:

163

Not a good receiver.

Ron: Yeah. When Ron then describes an event about which he still feels guilty, he is amenable to redefining his wife, Dorothy, as deserving blame. Ron: And I feel badly, too, about—I hear all our friends taking trips, cruises and so on. She'd dearly have loved that, but we didn't...get to do it. Lyd:

Could you have done it?

Ron: Well, I doubt it...money was the big thing. Lyd: Did you put it off because you couldn't afford to, or [itl just wasn't a plan that you had anyway, even though you both agreed that you'd like to do it? It wasn't really a plan because it wasn'tfinanciallypart of the picture? Ron: You're right there. In other words, if we would have wanted to go bad enough, we would have set aside the money to go, instead of taking in the ballet, or a couple operas, things like that. Lyd: ...[W]e sometimes talk about a dream that we have, but it isn't one that we really ever intend to have happen. And cruises and trips are sometimes like that, because we would rather take the day, which is, an opera's going to be playing... Dorothy was very much into music...So that would have been a good place to spend money, that she wouldn't have been critical of, right? Ron: Right. In this instance, Ron even contributes to redefining the fault as shared. If they had really wanted the trip, both Ron and his wife could have done something about it. But as Lydia points out (and Ron agrees), it was easier to spend money on day-to-day things that Dorothy would not criticize. The cases of the too-iong-postponed cruise and Gary's cluttered garage show how redefming the event has implications for the sufferer's definitions of the self as well as the deceased. By redefining the deceased as less

164


good and as deserving some of the blame, Lydia helps these participants to remove some of the blame from themselves and to make themselves better in their own eyes. Though good, the bereaved members were also weak—especially in their inability to change events—while their loved ones were dying. Even if they know now that a different course of action might have made a difference, their ignorance or inexperience made them weak at the time. One lengthy discussion between Lydia and Lee illustrates Lydia's views on blame. Lee relates how her busband's doctor sent them directly to the hospital, telling them that a specific doctor would be ready and waiting for them. Upon arrival in the emergency room, however. Lee and her husband were shuffled off to the side while the "more urgent" heart cases were seen first. They waited hours while Lee's husband's condition (and pain) grew steadily worse. Lee bitterly relates how she was told, after her husband's death, that if she had taken the ambulance her first doctor had offered rather than driving her husband herself, her husband's arrival would have been treated with more urgency. She blames the first doctor for withholding information that could have made a difference, she blames the hospital for not properly investigating the seriousness of his condition, and she blames herself for not accepting tbe ambulance. Lydia responds by trying to refocus blame on Lee's husband: Lyd: Do you feel some anger at yourself as well in this situation? Lee: I thought about,..people who said, "If it's bad enough to want an ambulance, then it's something you have to do." I have heard about that before, of course. I just didn't think about it with Pete.... Lyd: This was a crisis. And how you respond in a crisis is you try to attend to someone, right? Your intent was to get him help immediately....Who would you say was the leader in your family? You or Pete? Lee:

Well, he was my driving light....

Lyd: Now...you're taking him and he's directing you....But as you're driving there, there wasn't a thought in his mind about why didn't he suggest taking an ambulance? I

Lee: He was so dam independent that I would say due to his independence, he just assumed ihat we'd get there on our own. Lyd: It's really important to remember that. Because I think that for people that live with very strong-willed people.,.our tendency after death is that we look around and we say, "Boy. how did we get into that mess?"

The sufferer's weakness is not badness, however, and the sufferer is not to blame for the death: Lyd: You did the best you could. And it takes a while to learn to just look at it and say. "Pete and I did the best we could and it just didn't work out very well for me. It seemed to work out fine for you. you son of a gun." Sometimes you feel like that. He bailed out and here you are. Lydia slips in a reminder that Lee's husband did not simply die, he deserted her. By Lydia's argument, if anyone could be considered blameworthy for the death in this situation, it was the deceased. Again in this definition, just as the deceased is the only negatively evaluated actor, he or she is also the only powerful one. Bonnie, one of the long-term members of the group, shows how she has internalized this redefinition of her husband, John: 1 wish I had known that my husband wasn't being completely honest with me,,. He had high blood pressure [and was taking] medication for that. But after talking with the doctor, I can understand where John was coming from, and where the doctor was coming from. Because the doctor really couldn't call me and say anything that, if I had known [it], maybe I would have gotten him there for a treadmill test. But he wouldn't go in for any of the stuff, John just denied it. He was a very proud man, and to acknowledge that there was something wrong with him...was very hard for him....[A]fter being in this group for a while I tried to figure it out....John was a grown man, he had his choices, and unfortunately that choice was not one that I liked....

INTERPERSONAL EMOTION MANAGEMENT The deceased was often strong and directive, and made decisions that led to his or her own demise. The appropriate emotion regarding the deceased, then, is anger. Another one of Lydia's main concerns is to be sure the bereaved person's anger remains directed at the deceased and is not deflected onto others. Sometimes this is a difficult task because many of the sufferers see others as more appropriate targets and will admit to anger and frustration only toward them. This is particularly true of the medical profession and the physicians who treated their spouses before death. Lydia. however, is unwilling to accept doctors as targets. Although she is never explicit, her reactions show that she views doctors merely as targets of projected anger that the bereaved cannot admit really pertains to the deceased spouse. If you can find someone to blame and build it up into something where it becomes a crusade, then you don't have to deal with your pain. You avoid it. Well, if I could really go and just say that all the medical profession is just a bunch of bums and they didn't do a good job. and I could do this over and over and over again, then I would just be putting so much energy into that part of my anger that I wouldn't be dealing with the pain of my loss....I've had people hide behind their anger only because they didn't understand who they were angry at. As a result, Lydia continually attempts to defme doctors as not meriting anger. But you know what medical people are. They're the people that we expect to perform miracles for us when the people we love are dying. That's exactly who they're supposed to be. [But] they are not miracle workers. They go at it, and some of them botch it. Some of them do the best they can, and the end result is they've given back a product of some kind, And sometimes the gift they give is life for two months, life for a week, or no life at all. But...sometimes...with someone we really love, we tend to go through our lives a long period of time, and we never had to call on those folks. And then one day. something really critical happens in our lives, and we go to these folks and we honestly...believe now they will do the big miracle. And they're like you and tme]. They're educated to do something, but not to [perform miracles].

165

At worst, physicians are well-intentioned but mistaken, and very definitely powerless against the inevitable. In the story of Lee and her husband waiting in the emergency room. Lee directs much of her anger toward the hospital staff and doctors. Lydia attempts to redirect that anger toward Lee's husband for making the wrong decision. Thus Lydia always seeks to make the deceased the only person truly deserving of anger, even in this rather unlikely circumstance. She relegates others, even medical professionals, to the auxiliary role of trying to do good but being ultimately impotent, no matter how active or inactive they may be. The first redefmition, then, is retrospective: The deceased was a strong, somewhat bad (blameworthy) actor who perpetrated a negative act on the faithful but powerless and stunned bereaved, while others made futile but well-intentioned attempts to help. In other words, the deceased (and no one else) abandoned the bereaved. Now that the death is past, however, Lydia argues that the bereaved must take on a new definition of the situation in order to heal. By acknowledging the anger and conflict with the deceased, the bereaved transforms himself or herself into a stronger and more active person. The emotions are out in the open; now the sufferer can direct those emotions rather than letting unconscious anger (and thus the deceased) take control. A couple of years ago. I had a woman in group and she was absolutely furious, and it took about a year and a half to finally detach what she was furious about. She didn't even know....And finally we discovered that she still had the ashes of her husband on her mantle for three and a half years, and no one had given her permission to do anything with them....And she took them within a week or two and had them interred,,..And then she could discuss her anger about her husband dying just before retirement, when they had all these wonderful plans. But before that, as long as his presence was there, she could never be angry....tW]hen she found out she had permission to get rid of [the ashes]...let him go and put him in the appropriate place, then she could vent her anger, and she was okay.

166


Once they acknowledge their negative emotions, the bereaved can express them, thereby purging the associated emotional energy and releasing themselves from the power of the deceased. Once the anger is expressed, the bereaved can allow the deceased to be dead and can say good-bye. In Lydia's words, the bereaved reaches a point where he or she can say, "I forgive you. I release you. Go in peace." This, Lydia argues, is the first step in acceptance, where one has worked through both the good and the bad and has begun to incorporate the deceased into oneself. I think when you'rereadyto tell someone..."You have permission to leave, I'll be okay," then you can finish your grief with that person, and that person literally becomes a pan of you. And that's really the end of the alienation because our grief work is alienation. Before, we were tight with that person, and then they died, and then we became estranged from them again. But I think saying good-bye brings you back together again, just as I say to you, "You have permission to go take another walk, and go in joy," and I mean that.... And when you can say it to someone who has died, "You have permission to go. I'm going to be okay. You are a part of my life, and...I have to stay here in this unpredictable, unholy place, but I'm okay."...now you bring it back together and you say, "Well, I guess you're going to have to become a part of me so I can move on. So wherever I go, 1 just take that part of you with me and we're together." In recovery, then, the deceased ceases to be an actor and becomes part of the self. Thus, as in the Discovery group, the deceased spouse is reduced in power and then removed altogether from the situation. This process makes room for a current definition of recovery which subsumes the deceased into the self and makes the self strong, good, and active. In addition, as in the Discovery group, the second goal of redefinition is not merely to heal a wound but also to achieve personal growth. One's old world has been torn down by a traumatic event beyond one's control, but as survivor, the self is stronger, wiser, more compassionate. Much like clay fired in a kiln, the survivor is stronger for the pain. The goal is now to rebuild, to start anew, and to create a new and complete self:

...Carl died. He'll never be here again, and now my job is to find out what my mission is. What am I supposed to do with my life from here on as an individual person, just me alone in this world, in a place that's unsafe?...There's no security, it's not fixable...,what a hell of a place to leave somebody. And anyway, now my job is tofindout what my mission is. And so I am now ready to...reorganize my ideas and my attitudes. This reorganized self-identity must be maintained by interactions with other good actors. The actors whom Lydia proposes to fill this void are the children of the bereaved; they become part of the "mission" of the recovery definition. All of the long-term members of the group have grown or nearly grown children. How their spouses' deaths have affected their relationship with their children is a repeated topic in the group. Members often lament that their children do not understand and are no longer able to interact with their remaining parent in the same way. Lydia's response to Arlene, for instance, is that this is part of bereavement. Arl: [crying] I don't think my daughter is the same way as she was before. Lyd;

No, she isn't the same as she was before.

Arl:

And it seems like I lost her too.

Lyd: What does your daughter want from you, mom? She wants you to be happy. Isn't that true? Arl:

I guess so.

Lyd; And so if I am your daughter and I say to you, "Mom, I really have to talk to you about Dad. I just miss him so much. 1 don't feel like 1 can say that to you because I'm afraid you're going to fall apart. I already see that you're unhappy.... And you know what happened here is that you haven't given me information as a daughter to come in and be sad with you. Yes, I'm different because I lost my dad, but I am also very, very afraid that I'm going to lose you. Dad has died, and I'm quite certain that you could die also. And I just don't know how I'm going to stand that." So she isn't the same. The bereaved can no longer have the same

INTERPERSONAL EMOTION MANAGEMENT Emm: Yes.

relationship with their children as before the death because that relationship was based a set of interrelated roles, part of which is now gone. This situation requires a reestablishment of role relationships. The parent must become a role model in grief for children because they don't know how to grieve and recover on their own. Your children...don't understand. And maybe...your job is to teach them how to understand. That you're now the role model in this whole process....My guess is probably what you're supposed to do is teach them the process. I'm almost convinced...that our role now becomes, "Okay, I saw it. I was in an enormous amount of pain. I'm going to try to teach you the process so you won't be as surprised or amazed,",..You came into it totally cold, because [the previous] generation did not teach us anything about emotions. But now you can teach your children.... Thus children, even adult children, are weak and good, but somewhat inactive because they need to be shown or led in grief. Because the survivor's children—the new key actors in the definition—are so weak, maintaining positive interactions entails helping, teaching, and other benevolent acts. These acts are the survivor's new "mission"—to become a role model for others in grief. Lyd: So after disorganization comes this reorganization to redefine yourself Isn't that scary? ...We're not exactly 21 anymore,...so...isn't redefming yourself at this age...kind of an odd thing to be doing? So why are we doing it? Or should we do it?

Lyd;

Why?

Emm: Because it's healthier. Much better than sitting and being depressed. Lyd: Will I be happier if I redefined myself? Ron: You have. Lyd; Yes, I have redefined myself, and, yes, I am happier. But the fact of the matter is in grief recovery that [it] isn't our goal necessarily to be...happier. I think our goal is to be healthier. To be more functional. To be more giving. To be more caring...to help other people go through what we have gone through with less pain than when we went through it. Thus the redefined self becomes someone who is growing, helping others to grow, and seeking a new mission or purpose in life. Lydia calls this final identity a "survivor" but with the stress on teaching others to survive as well. As shown in Table 2, the self in the second redefinition is good, strong, and active in both healing oneself and helping others to heal as well. The process of redefinition in the Rebirth group is remarkably similar to that in the Discovery group. Yet each group produces a definition that is consistent with its stated ideology and differs in fundamental ways from the other group's definition, The EPA ratings of the key actors (especially the self) are quite comparable, but the ideologies create identities with very different expectations for action.

Tabie 2. Rebirth Group, by Definitions Self-Labeling Definition

First Redefinition

Second Redefinition

Self

Bad Weak Inactive

Good Weak Inactive

Good Strong Active

Deceased

Good Weak Inactive

Bad

N/A

N/A

ti/A

Children of the Bereaved

167

Strong Active Good Weak Inactive

168

SOCIAL PSYCHOLOGY QUARTERLY DISCUSSION

Interpersonal Emotion Management and Affect Control This tie between identities and actions is largely the reason why the process of interpersonal emotion management lends itself so well to explanation by affect control theory. No other framework focuses so directly on the identity-action connection (Heise 1979, 1987; MacKinnon. 1994; Smith-Lovin and Heise, 1988). The results of this study agree well with the 24 propositions that form the basis of ACT (MacKinnon 1994:338-40). Propositions 16 through 21 (shown in Table 3) are particularly relevant to my fmdings. In ACT terms, the support group defines the participant's persistent negative sentiments as problematic because they indicate that the person has developed a negative identity. (Thus the attempts in both groups to "unlabel" the self) Specifically, the person's suffering has been caused by a deflection between a transient feeling associated with an event (death or divorce) and the EPA sentiments customarily associated with the identities of the people involved (self and spouse). The sufferer has resolved the deflection so as to give himself or herself a label (a negative identity) that harms self-esteem or emotional well-being; for example, a bad act was done to a had person. This point accords with Proposition 19. Therefore, the task of facilitators is to use their expertise (and ideology) to qualify the elements of the situation. The intent is to reduce the deflection and to transform the

experience into a confirmation of positive sentiments, To create enduring positive affect, the facilitator must treat the sufferer as having a positive identity. Then (in accordance with Proposition 18) the facilitator can argue that the current negative emotion arose from the deflections produced hy the had act (death or divorce). But because the sufferer now has a positive identity, he or she could not be responsible for that bad act (Proposition 16). Therefore the deflection must he due to a had act by the other party (the spouse) in the event (divorce or bereavement). Thus, according to Proposition 21. the spouse must be the one with the bad identity in order to have committed the bad act. In this way the self-labeling and persistent negative emotions are defeated. This redefining, however, does not completely accomplish the goals of the support group, because a bad person committing a bad act upon a good person does not engender positive emotions; nor does it remove all deflections. (Indeed, in my analysis this realization sent me back to the data to discover the second redefinition in each group.) Thus the powerful, good self mnal forgive the weak, bad person and the bad act and must go on to interact positively with good others. As stated in Proposition 17, this will help to establish and maintain enduring positive selfidentities and positive sentiments. By qualifying the elements in this way. the facilitator has altered the defmition of the situation and has called up new identities and actions, new emotion norms, and (one hopes) new emotions as well. This process, according to Proposition 20, generates much of the

Table 3, Six ACT Propositions Illustrated by This Study* 16: The likelihood that a person will engage in one feasible behavior rather than another is inversely related to the affective disturbances which the behaviors produce. 17: In the course of validating social identities, people engage in role-appropriate acts18: An interactant's emotion following an event reflects the outcome of the event and also the identity thai the person is maintaining. Specifically, the emotion is a function of the transient impression of the interactant that was created by the event, and the discrepancy between this transient impression and the fundamental sentiment associated with the interactant's identity. 19: People tend to maintain emotions that are characteristic of their salient identities. 20: Emotion displays facilitate intersubjective sharing of definitions of siluations and of the operative social structures that are implied by definitions of the situation, 21: Social labeiings render past events more credible by assigning interactants new identities that are confirmed by past events. •Propositions 16-21 of the 24 propositions stated by MacKinnon (1994:338-40).

DSfTERPERSONAL EMOTION MANAGEMENT power of the support group in interpersonal emotion management. Redefining Identities The significance of this redefinition of identities becomes clearer when the process is stated in terms of Stryker's (1980) identity theory, the framework for affect control theory. Stryker argues that people have "separate selves" for each position they hold in society—spouse, parent, worker, neighbor, and so on. Thus the variety of selves reflects the multifaceted nature of society (Stryker and Serpe 1982). These selves are expressed by the concept of identity, and are organized in a salience hierarchy with the most salient identities at the top. Stryker defmes salience not as a person's preference for an identity, but as the probability that a particular identity will be invoked in a given situation or in many situations. In addition, Stryker and Serpe capture the multifaceted nature of society through commitment, which refers to the pattern of relations to others. Because identities reflect social relationships, and salience reflects the probability of invoking a given identity, commitment and salience are mutually influential. In a support group, by identifying the other relevant actors in the event, the facilitator in effect is changing the pattern of available relationships. This process in turn increases the salience of the chosen identity. By providing each situation with a "defining actor," the facilitator increases the probability of invoking the sufferer's new situational self-identity. According to Burke and Reitzes (1981), these are "counter-identities," actors in relation to whom the individual defines himself or herself. Thus the identity appropriate to the relationship with that actor becomes the identity appropriate to the event (spousal loss). Invoking this identity then presumably creates in the sufferer the emotions desired by the facilitator. In ACT terminology, enacting this identity in relation to the defining actor confirms the associated sentiments. In this way the facilitator provides a stable resolution of the deflection that the sufferer brings to the group as a result of a loss.

169

The defining actor in each of the groups described here is unmistakable. In the Rebirth group, the emphasis on the relationship with the children of the bereaved determines the salient identity—the identity to be enacted as a role model/teacher. In the Discovery group, the defming actor is God; the resulting salient identity is "Christian" or, even more specifically, "child of God." These identities call for very different behaviors and emotions on the part of the group members. The enactment of these positive identities will confirm the situational definitions (and thus the ideologies) of the respective groups. In fact, an affective resocialization is being accomplished, with new identities (Heiss 1981), new role-taking others (Mead,1934), and new means of acquiring or appropriating emotions to the self (Averill 1986; Gordon 1989.) CONCLUSION To summarize, the key argument of this paper is that facilitators in support groups accomplish interpersonal emotion management less by addressing the emotions themselves than by addressing the identities of those feeling the emotions. The facilitator alters the emotional situation and the emotion norms by altering the identities of the relevant actors. This is not to say that every act of emotion management involves a fundamental change of identity; only the most traumatic emotions are likely to require so drastic an alteration. Nonetheless, interpersonal emotion management perceived as an interactional process based on ideology and culminating in identity, has implications for other research. Hochschild (1990), Stearns and Stearns (1986), and Wasielewski (1985), for instance, might move from asking "How are identities redefined?" to "Whose interests are being served by this redefinition?" Collins (1990) and Scheff (1988, 1990), on the other hand, might see the outcomes in a more positive light. Collins might view the support groups as sources of emotional energy, which the participants then can use in other encounters. Scheff might consider the new positive selfidentities as a means of suffusing the partici-

170

SOCIAL PSYCHOLOGY QUARTERLY pants with pride, on which they can build other interactional bonds. Thus conceiving of interpersonal emotion management as an interactional and cultural process is a useful empirical step in the sociology of emotions. My purpose here has been to investigate how others influence our emotions. I pursued this investigation in two support groups for people who had lost a spouse through divorce or bereavement. Despite differences in ideology, format, and even event, the processes in the two groups proved remarkably similar: The actors in the event were defmed and then redefmed along the EPA dimensions of affect control theory and in accordance with the group's therapeutic ideology. The results of this study, I hope, will provide useful analytical tools for studying emotion as a social process, not only in the context of interaction but also in its structural, cultural, and ideological milieu. REFERENCES

Averill, James R, 1986, "The Acquisition of Emotions during Adulthood." Pp. 98-118 in The Social Construction of Emotions, edited by Rom Harre, New York: Blackwell. Blumer, Herbert, 1969, Symbolic Interactionism: Perspective and Method. Englewood Cliffs, NJ: Prentice-Hall. Burke. Peter J, and Donald C. Reitzes. 1981. "The Link between Identity and Role Performance," Social Psychology Quarterly 44:83-92, Collins, Randall. 1990, "Stratification, Emotional Energy, and the Transient Emotions," Pp. 27-57 in Research Agendas in the Sociology of Emotion, edited by Theodore D. Ketnper, Albany: SUNY Press. Francis, Linda E, 1994. 'Laughter, the Best Mediation: Humor as Emotion Management in Interaction." Symbolic Interaction 17:147-63, Frank, Jerome D, 1974. Persuasion and Healing: A Comparative Study of Psychotherapy. New York: Schocken. Glazer, Barney G, and Anselm L, Strauss, 1969. The Discovery of Grounded Theory: Strategies for Qualitative Research. New York: Aldine, Goffman, Erving, 1959, The Presentation of Self in Everyday Life. New York: Doubleday Anchor, 1974. Frame Analysis: An Essay on the Organization of Experience. Boston: Northeastern University Press. Gordon, Steven, 1989. institutional and Impulsive Orientations in Selectively Appropriating Emotions to Self." Pp, 115-135 in The Sociology of Emotions: Original E.