THE INTERPERSONAL MANAGEMENT OF CRYING AMONG SURVIVORS OF STROKE
John F. Manzo, University of Toronto
Robin L Heath, University of Kentucky
Lee Xenakis Blonder, University of Kentucky

INTRODUCTION
 This study concerns the social-interactional consequences of crying among survivors of stroke. The episodes of crying analyzed here take place in interviews comprising the patients, the patients' spouses, and an interviewer. This investigation innovates on past studies within the sociology of emotions by concentrating on the interpersonal dimension of emotional displays of brain-damaged persons. This study also supplements research on stroke patients' "pathological crying" from the field of neuropsychology, because it concentrates on the social, and not only the neurological or otherwise individual-level, nature of such crying. We first present overviews of both the sociology of emotions and the neuropsychology of post-stroke emotionalism, and address how our study contributes to both fields. We then discuss our subjects and method of analysis and finally present our findings with respect to the techniques of the management of crying exhibited by the stroke patients' interlocutors as well as the patients themselves.

SOCIOLOGY AND THE STUDY OF EMOTIONS
 In the last decade, the study of emotions has emerged as a fertile subfield in sociology (for a comprehensive review see Cuthbertson-Johnson, Franks, and Dornan 1994). This development has permitted sociologists to uncover the social character of phenomena that investigators in other fields had previously studied as individual level, but that are now seen as appropriate for sociological analysis. The perspective of sociology has similarly been brought to topics such as the production of narratives of personal experience (Manzo 1993) and even dreams (Fine and Fischer-Leighton, 1993), which are construable as referencing social reality and as collectively interpretable. Emotions are thus part of the range of "private" experience that can be subject to sociological inquiry, and although we accept the caveat that sociology can never provide the only basis for theoretical and empirical works on emotions (see Craib 1995), sociology does provide an important strain in the study of emotions, one that attends to previously-neglected social aspects.
 Reviews have defined the sociology of emotions as divided into two orientations, the "positivist" and the "constructivist," which see emotions as rooted in biology and in social experience respectively (see Franks, 1987). We reject the labels "positivist" and "constructivist," because they reference stereotypical caricatures that we have not found to exist among real practitioners in the field. Our review of the literature suggests that no sociologists have ever asserted the purely positivist (biological) view, and no sociologists reject, in toto, the idea that emotions have (on at perhaps a superficially fundamental level) a biological component. Thus, descriptions of the field as divided along "positivist" and "constructivist" lines have failed to capture the state of the sociology of emotions.
 We propose that there are indeed two camps in the sociology of emotions. We see a divergence more accurately, however, as between those who emphasize the place of emotions in larger social-structural context, an approach we label "variable-analytic," and those who examine persons' interpretation of emotions in interpersonal encounters, which we call the "interactionist" perspective. The former approach views emotions as resources in designs of research that attend to "larger" sociological issues, and the latter construes emotions as topics in their own right.

 Variable-analytic Approaches
Central to a "sociology" of emotions is the idea that emotions can in fact be treated as social phenomena. One obviously "social" aspect of emotions is their distribution and experience among different groups of persons, and so researchers have examined the relative experience of emotions among persons in various social categories. They have thus delineated emotion(s) as either a dependent or independent variable that is correlated with, or consequential for, individuals' social-structural locations, including (but not limited to) their social class (Barbalet 1992; Correno 1992; Rubin 1976), gender (Hochschild 1983; Lyman 1990; Peplau and Gordon 1985; Sprecher and Sedikides 1993), and age (Lin, Ensel, and Dean 1986). This orientation is paradigmatic of sociological "variable analysis" (Blumer 1956) which examines associations among different variables or classes of variables, and for that reason we see the label "variable-analytic" as most appropriate to define this approach. In these studies, analysts envision social-structural characteristics as factors that influence emotional experience, or as conceptual anchors that delimit the analysis to emotional behavior experienced by a specific social category, such as "men," "women," "working-class families," and so forth. This perspective construes emotions as resources for study: Variable-analytic researchers use emotions to address the larger issue of differences among social groups. The variable-analytic approach does not study emotions as topics in their own right. The predominant alternative perspective, the interactionist approach, does, on the other hand, topicalize emotions as such.

 Interactionist Approaches
The "interactionist" variety of analysis examines the production, interpretation, and discovery of the meaning of emotions by persons in the course of social encounters. It is an approach that "unpackages" the social work that inheres in emotional displays, work that much of variable-analytic study takes for granted. For example, Johnson (1992) has proposed a symbolic-interactionist model of the "emergence of the emotional self," using an approach that makes emotion a focus of analysis instead of employing emotion as a resource for larger social study. Consistent with this approach are theoretical works by Collins (1986), Denzin (1984), and Harré (1986).
 One theme joining each of these works is that emotions are, in the first place, indeterminate biological phenomena that are defined through processes of social learning. Humans learn to name their feelings, and their instruction takes place in processes of social interaction. Thus, the interactionist approach holds that there is a biological substrate of emotions, but that emotions are only understandable through an interpretive lattice which socialization provides, a framework that varies with larger historical and cultural change (McCarthy 1989). However, that biological substrate, even where it is emphasized (see Franks 1987), remains undefined.
 This study is more closely allied with the interactionist tradition than with the variable-analytic, and it concerns a topic that has not yet been studied by sociologists. This study examines the social and interactional management of crying among survivors of stroke. Our study acknowledges thoroughly the biological (in this case, the neurological) substrate of emotions, and examines specific techniques through which these biologically-grounded emotional displays come to be understood. Our investigation asks how interactants (in our study, the patient, her or his spouse, and an interviewer) collaboratively interpret episodes of crying and render them socially accountable, that is, how they render crying episodes as explicable and even as normal.
 We begin with a brief overview of past research on what some refer to as "emotionalism" following stroke. We will then discuss how we address this topic in a way that exposes its social aspect, and we will finally present our findings and their implications for the sociology of emotions and for the experience of stroke survivors and their significant others.
 
"EMOTIONALISM" AND STROKE
 Previous Research
 A British study using a probability sample of survivors of stroke (House et al. 1989) found that 20% experienced "emotionalism," or a heightened tendency to laugh or (much more commonly) to cry following their stroke. In their study of 30 stroke patients, Allman, Hope, and Fairburn (1992) found that subjects reported between one and 308 episodes of crying in the month before being interviewed. Subjects also reported a tremendous range of social, affective, and cognitive events associated with the crying. Allman et al. (1992:321) thus suggest that "crying following stroke can be seen to have a number of components each of which varies widely." Their study calls into question the existence of a distinguishable condition, labeled "pathological crying," of stroke survivors as defined by Poeck (1969); Allman et al. did not expose any unambiguous instance of "pathological crying." Their observations suggest instead that there is no unitary clinical phenomenon to warrant the label "pathological crying" that denotes a specific, definable medical condition.
 Although we find Allman at al. (1992) instructive in their discovery of post-stroke crying as a complex topic, we would suggest that it, and the study of post-stroke "emotionalism" as a whole, would benefit from direct observational study of crying as opposed to reliance on interviews that ask patients about their crying (see Allman et al 1992; Robinson et al. 1993). Robinson et al. (1993), in constructing a measurement scale for determining the presence of "pathological" laughing or crying, rely not on the subjective experience of the patients but rather the ratings of interviewers and the patients' caregivers. They do not examine crying directly. Allman et al. do observe nineteen cases of patients crying in their interviews, but describe these cases only in terms of patients' facial expressions and the sounds they make while crying. Studies such as these fail to attend to the social aspects of these emotional displays, especially the observation of post-stroke crying in social contexts. This study entails precisely this sort of examination.

 This Study
 This study innovates on other sociological studies in several ways. First, although it treats emotion as socially organized as do past works, we do not attend to this social organization with deference to "social structure." Instead, we are concerned with the endogenous social organization of emotion. In this report our specific concern is how persons party to episodes of crying by survivors of stroke respond to it in order to render that crying interactionally seamless and "normal."
 This study also innovates on past treatments of post-stroke emotionality. Whereas such studies have examined patients' crying by either asking patients or their caregivers about the crying, or, much more rarely, by observing patients' crying in isolation, we are inspecting instances of crying in social context. The following discussion of our data and method will clarify the novelty of our data and our analytic approach.
 

DATA AND METHOD
 Subjects
 Fourteen stroke survivors were interviewed in this investigation. The subjects included seven right-hemisphere stroke patients and seven left-hemisphere stroke patients; none of the subjects had experienced bilateral damage. Five were female. The average age of the interviewees at the time of  stroke was 57, and their ages ranged from 34 to 77. All interviews were conducted approximately one month following the stroke. All patients were married.

 Interviews
 The interviews were conducted, wherever possible, in patients' homes to render the interview situation as comfortable, realistic, and spontaneous as possible. Spouses were present in all of the interviews. Each couple was interviewed approximately one, six, and 18 months following the stroke. We have limited our analysis to the first set of interviews only, because patients are more likely to experience heightened emotionalism soon after strokes, and because as of this writing not all follow-up interviews have been completed.
 The interview schedule consisted of twenty-two open-ended questions, some directed to the patient, some to the spouse, and some to both. We have provided the complete interview schedule in the appendix. The questions often pursue the topic of emotion (for example, "Tell me about a time when you felt sad"), and some are specifically intended (with varying success) to elicit displays of emotion. Which questions "work" in this respect is highly variable, however, and we are not in this report evaluating the interviewer or the interview schedule on that basis. We are instead motivated to report on what happens when crying does occur, regardless of what elicits it.

 Data and Investigative approach
 Data for this study consist of the videotapes and resulting verbatim transcripts of the interviews. After the interviews were transcribed by support staff, the first author decided to treat crying, a prominent feature in several of the interviews, as this paper's topic. The second author then scrutinized each interview and extracted every instance of crying. The first author examined the "natural history" of each episode to determine how the patient and their interactants dealt with crying episodes. Our findings comprise different varieties of management of the crying.
 The investigative approach used in analyzing the data did not entail the development of a priori analytic frameworks. The approach is instead associated with that of Conversation Analysis, or CA (Heritage, 1984:232ff). Analysis in CA is an inductive enterprise in which the investigator uncovers patterns of talk that are sequential in nature. Doing CA entails considering speakers' turns of talk in light of the accompanying talk and related activities of other speakers, and it attempts to specify phenomena that are relevant for and created by the speakers themselves. The topic of this paper was discovered following scrutiny of all the data; it was not specified beforehand in hypothetical terms. Instead, the data were allowed to "present themselves" in multiple viewings and hearings, and what follows is one result of that examination.

FINDINGS
 General Observations
 Seven of the fourteen stroke patients cried at least once during the interviews. The number of crying episodes ranged from one to fifteen. Table 1 summarizes the number, average duration, and range of duration of crying episodes for each of these seven stroke survivors.

TABLE 1 ABOUT HERE

These data suggest, first, that crying among persons who recently suffered a stroke is relatively common, and, in the case of patient 18, recurrent. Second, the number of crying episodes and the duration of these episodes are widely varied, and this observation suggests that post-stroke crying resists categorization as a unitary phenomenon such as the term "pathological crying" suggests.

 Analysis
 The focus of this analysis is interactants' responses to stroke patients' crying. However, it is fitting to include a few words here about what precedes and elicits episodes of crying. Most generally, crying takes place in the context of communication even when that context seems inappropriate for crying. This observation suggests that crying should, or could, be construed as an act of communication and not only as random act of neurological "misfiring" or something similar. Often crying emerges during reminiscing and is perfectly fitting in context of reports that are themselves emotionally infused. This topic will be explored fully in another piece employing as data these interviews (Blonder, Heath, Manzo, n.d).
 Again, we are in this report concerned with the sequelae of stroke patients' crying. We have uncovered a number of methods that interactants deploy after crying takes place, and each of these responses serves to "normalize" the episode. They are as follows:

  1. Empathy.
  2. Derision of the crying.
  3. Explanations and excuses.
  4. Biographical accounts.

The first two might be termed "emotional" responses, because they are themselves infused with emotion. In addition are the next two "rational" (nonemotional) responses, which provide reasons, accounts, theories, and so forth that explain the immediately prior crying episode. In the following section we provide examples from our data of each of these resources and discuss their unfolding.

 Empathy
 "Empathy" refers to a hearer's displayed capacity to feel, and to express, the same unhappy emotions produced by the speaker. Displaying empathy in stroke patients' crying helps to explicate the crying as warranted by the patient's innate sadness and thus to normalize the crying as justifiable: To cry along with the patient (or to express why the patient would be crying on the basis of empathetic "understanding," as in Excerpt 1 below) establishes the crying as normal and explicable, and as nonpathological. Excerpt 1 contains an example of a spouse's deployment of empathy. In this and all excerpts, "I" is the interviewer, "P" the stroke patient, and "S" the spouse.

EXCERPT 1
Patient 13
1 I: Tell me a little bit, if you can, about your marital
2    relationship. How has the stroke affected it?
3 P: Different.
4 I: Different for you?
5 P: Uh huh. ((starts to cry))
6 I: It's different. Okay.
7 S: And that's probably why he don't want to talk about it.

 The patient begins to cry after being asked about "the marital relationship" (lines 1-2), which addresses the issue of the couple's sexual relationship presumably; upon his crying after assessing the relationship as "different," the spouse's response (line 7) to the interviewer's restatement is "that's probably why he don't want to talk about it." This statement signals a specific interpretation on the part of the spouse in managing the patient's crying. She does not address the crying as problematic itself. Instead, she expresses empathy and understanding concerning "why" he is crying: The "marital relationship" has changed, for the worse we must assume, and "he don't want to talk about it." Even though there is without question a neurological cause for his crying (e.g., damage to his brain has caused him to be disinhibited enough to permit him to cry), the account provided in this context has nothing to do with neurological damage; it has to do with the (justifiable) sorrow he feels about his marriage.
 A clearer example of empathy produced by a spouse would be in instances in which the spouse cried as well. Excerpt 2 contains an example of a spouse (Patient 18's husband) crying in response to the question concerning a time he "felt sad." His response references the stroke as something about which one cries, and in so doing he helps establish his wife's crying as justifiable.

EXCERPT 2
Patient 18
1 I: How about a situation in which you felt sad.
2 S: I felt very sad when she was in the hospital because I
3    felt like I was going to lose her and the first three
4    days she was very bad and I couldn't imagine living alone
5    without Tonya, because we're so close and that made me
6    very sad. But after three days she started to rally and
7    then I gained, I felt a lot more hope that would have
8    her because I couldn't think of going into retirement
9    years without her.  Now we're both going to cry. ((starts
 10    to cry))

 In another example, in excerpt 3, the spouse and the interviewer cry along with the patient, which reinforces the notion that crying is warranted because "normals' engage in it as well:

EXCERPT 3
Patient 14
1 I: ...her mood? How would you describe (your wife's) mood,
2    the last few weeks.
3 P: Assists, assi- her mood has been s::: the accident toward
4    mine, I can't peter the- don't she crying sometimes and
5    that why I think, I think that she crying for me.
6    ((patient starts to cry))
7 I: She goes in her room and you think she's crying for you?
8    maybe she's just got a good soap opera on.
9 P: No.
10 I: Maybe she's just watching Bambi ((laughs; patient
11    stops crying)). Well, has she been supportive
12    has she bitten your head off?
13 P: Oh no no no.
14 S: A few times, a few times=
15 P: =no no no first time. She would be would=
16 I: =so she's a good egg, huh.
17 P: ((starts to cry)) yeah.
18 I: Yeah. How would you describe your mood in the last few
19    weeks? I think we all need a kleenex ((interviewer and
20    spouse laugh while wiping their eyes)).

 One important feature of the strategic use of empathy is its initiation by the spouse or some other hearer; one cannot empathize with one's own emotions. The next technique of management is initiated by the crier herself, and it entails the use of derision, of one's self and of the crying.

 Derision of the Crying
 Patient 18, who cries in Excerpt 3, is a recurrent crier who expresses very moving narrative about how her stroke has affected her well-being. A former musician, she can no longer teach or play, and this loss of agency and meaning in her life emerges frequently in the course of the interview. It is, perhaps, inevitable given her previously independent lifestyle that she employs a great amount of self-derision in her talk, infused with anger and sadness at her new physical limitations and her sudden dependence on her spouse. Excerpt 3 contains a clear example of derision of her crying as a tactic that only the patient herself may deploy:
EXCERPT 4
Patient 18
1 I: Do you notice any changes in the way you feel about
2    yourself?
3 P: Yes, lower self-esteem. ((crying))
4 I: That's okay. Can you try and tell me about it or is it
5    too painful to talk about?
6 P: I hate it that I cry too much. I hate it.
7 I: Did you use not to cry?
8 P: Very little. Oh I thought I was pretty even temp- even
9    tempered and a fairly happy person. So this bothers me.
10 I: So you don't think you're as happy person now as you were
11    before?
12 P: Not when I cry too easily. That upsets me. I would like
13    not to do that.

 The question that the interviewer issues in lines 1-2 is especially productive of emotional responses on the part of the stroke survivor, and she does respond with an emotionally-infused reference to "lower self-esteem" in line 3, crying throughout her delivery of that response. Her reported loss of self-esteem could issue from a number of causes, and we might imagine that the more obvious physical manifestations of the stroke (in this patient's case, partial paralysis in one side of her body, including facial paralysis) might account for such a change. However, what bothers her the most (in this section of the interview) is not any of these more salient physical problems, but the crying itself. This she clarifies in subsequent turns of talk (lines 6, 8-9 and 12-13).
 Excessive post-stroke emotionality is, presumably, disturbing, embarrassing, and disruptive for all survivors of stroke and those with whom they interact. The thesis of this report is that patients and their interlocutors have a variety of ways of managing the crying, to lessen its negative impact on interaction. Patient 18 is one of the few survivors in this sample to topicalize the crying itself, and to explain that she is not crying because she is sad; rather, she is sad because she is crying. In so doing, she helps distance herself from the crying as behavior that is not normal for her and which she acknowledges as abnormal behavior, and paradoxically "normalizes" the crying by accounting explicitly for it as intrusive and an unwanted feature of her post-stroke existence.
 Most of the techniques that we have discovered here do not treat the crying as directly as does patient 18. In the next excerpt we find the initiative for "normalizing" a stroke survivor's crying on the part of the interviewer. In this case, the interviewer implies two possible explanations for the crying, neither of which address the neurological aspect of it.

 Explanation/Excuse
 The previous interview extracts illustrated techniques employed by the spouse and the patient herself in attending to crying episodes. This next category, which we call "explanation or excuse," entails the interviewer offering a candidate explanation, or sometimes a menu of explanations, for the crying episode; we saw one jocular example of this phenomenon in excerpt 3, when the interviewer asked whether the patient's wife was "watching Bambi." In excerpt 5, the interviewer offers two candidate explanations, one "emotional," and one "physical":

EXCERPT 5
Patient 6
1 I: Were you at your mom's house when it happened?
2 S: See, her mother had a stroke.
 3 I: That's right. Okay. Do you remember being at your mom's
4    house now. Yeah? I know and Mom just died a while back. I
5    know it's hard for you.
6 P: ((begins to cry))
7 I: Did you feel- got a kleenex? How's your cold? How are
8    your sinuses? Still bad? Still got the draining?
9 P: Yeah.

 In recounting the history of the patient's stroke, the interviewer asks the patient to confirm that she was at her mother's house when the stroke occurred, and that this patient was, ironically, caring for her stroke-afflicted mother when she had a stroke herself. After the patient's husband notes that his mother-in-law had had a stroke, the patient's expression becomes dour; she is on the verge of crying. In response, the interviewer references the mother's recent demise and acknowledges that it must be "hard for you" (lines 4-5). This utterance proffers an explanation for the patient's forthcoming crying, one that is tied to the local conversational topic (mother's stroke and subsequent death) and thus treats the crying, preemptively, as justifiable.
 After the patient begins to cry, the interviewer asks about problems with her sinuses that seem to propose another explanation for the crying, or at least physical symptoms (such as nasal congestion) associated with crying. This is another approach to "normalizing" the crying, one that attends to the possible association between non-stroke-related physical problems (sinus congestion) and some of the biological components of crying. In her utterance the interviewer manages to render the crying as neither pathological nor as solely relevant to the emotionally-laden topic at hand: People blow their noses when they have excessive drainage from their sinuses. In this excerpt we can see how crying can be "explained," in minute conversational moves, by referencing emotional or (non-stroke-related) medical issues. The stroke itself need not be an issue, as in excerpt 5.

 Biographical Accounts
 Another method available to the crier and his or her interactants to explicate crying episodes is to present them as completely nonpathological by saying that the patient has always been a crier, or to describe the patient more generally as a person who is sensitive, emotional, soft-hearted, and so forth. Excerpt 5 contains an example of a spouse, and not the patient himself, providing biographical information that accounts for the crying:

EXCERPT 6
Patient 11
1 I: When do you become tearful?
2 P: ((begins to cry))
3 S: Tell it like it is. Ah, prayer. He's always been
4    accustomed to the blessing and the prayers and so on. And
5    this is a real part of his life. And uh so he uh and this
6    has been uh I'd say Kevin the only difference in your
7    uh personality and uh he's always been uh I said a very
8    tender-hearted person.

 The question preceding this instance of crying is "when do you become tearful?" We cannot say conclusively that this question causes the outburst in line 2, but it does help set up the spouse's utterance in the balance of the excerpt as responsive to when the patient "becomes tearful." Note that her answer addresses biographical features of her husband that do  not bear in any way on the stroke, but rather on his familiarity with "blessing and prayers and so on" (line 4), and she goes on to describe her husband as "very tender-hearted" (lines 7-8). "Tender-hearted" people cry, especially when thinking about the "blessings" bestowed on them after a stroke or other near-fatal event. The spouse helps to establish her husband's crying as part of his normal behavior by discussing his personal history surrounding emotions.
 
 The Absence of "Clinical" Accounts
 Of course, one "ready-made" account for crying is for patient, spouse, or interviewer to say something like, "The doctor says that crying is normal," but this account emerges nowhere in our interviews. The only instance in which the idea of clinical pathological crying is mentioned is when Patient 18's husband invokes the concept specifically to note that his wife is not suffering from it, but rather exhibits emotions that are "rational":

EXCERPT 7
Patient 18
1 S: ...she cries frequently but it's not uh as a psychiatrist
2    that we have seen says she's not really clinically
3    depressed, it's just that she responds in a normal,
4    depressive reaction to events and occurrences and it's
5    very appropriate I feel in most instances when she cries
6    that it's something either the symphony playing or attend
7    a symphony concert or she sees some of her violin or
8    hears some of her violin students performing and she's
9    aware that she cannot do some of the things she did
10    before. She responds sadly to these things. But it's not
11    as I have heard some stroke patients do where they will
12    cry very inappropriately or laugh inappropriately their
13    emotions are extremely label or labile and I have not
14    found that with her.

 It is a matter of speculation as to why there are no instances of patient, spouse, or interviewer explicitly denoting "pathological crying" as the reason for a patient's emotional displays. Even in the case of patient 18, who, along with her husband, expresses some facility with neurological terms (as with her husband's use of "labile" in line 13 of the previous excerpt, and in another unexcerpted part of the interview where the patient refers to her own "flat affect" in speaking), refuses to account for the crying as a straightforward result of brain damage. It appears likely, based on our sample here, that all stroke patients and those with whom they communicate will steer away from that explanation, and our findings suggest some of the breadth of techniques that are available to obviate references to the stroke as "causing" their crying.

 CONCLUSIONS
 Although the sample of stroke survivors interviewed for this research project is small, we feel our findings are broadly instructive for the study of post-stroke crying as well as the sociology of emotions and also more general sociological issues.
 First, this paper is an illustration of how the "meaning" of emotional displays needs to be defined and disambiguated in social context. Among the episodes of crying analyzed here, the meaning of crying is never obvious but is treated as meaningful through subsequent interactional practices. Furthermore, the episodes of crying we observe are never straightforwardly random or "pathological." They are in every instance contextually tied, but they are made contextual in conversational moves. This study demonstrates how this clarifying, "contextualizing" work is done in specific, studiable, concerted activity on the part of the stroke survivor and the persons with whom he or she interacts. This finding advises that studies of post-stroke emotionality, as well as emotions generally, should be conducted while recognizing that emotions are part of sociality. Even among brain-damaged persons, emotions should be studied as features of social interaction and as subject to social definition.
 This study addresses at least two more general sociological themes. The first of these concerns social responses to deviance. "Pathological crying" is a kind of residual deviance (Scheff 1984), which category refers to non-normative, seemingly random acts that are defined, and interpreted, by onlookers only after the acts have been committed. For example, bizarre behavior, regardless of its clinical pathology or any other "official" explanation, often comes to be labelled as symptomatic of mental illness, but this definition, this sense-making, of the behavior is only decided on after the act has taken place. Residual deviance is not, strictly, "rule breaking" such as deviance is usually defined, because the "rules" that residual deviants "break" are tacit and undefined. Onlookers and interlocutors manage residual deviance by naming it and thus accounting for it.
 Managing stroke patients' crying, crying that often has no clear empirical cause (though this is sometimes far from the case in our data) requires responses on the part of the persons interacting with the stroke patient as well as the patient her/himself. This investigation suggests that "managing" stroke patients' emotions is not a matter of controlling their emotions, and this finding militates against the idea that "emotional management" (cf. Hochschild 1979) is (only) an aspect of coercive social control, whether on the interpersonal or social-structural level. We see "emotional management" more as approaches to the accommodation of behaviors that can disturb the stream of social interaction. This study has uncovered several techniques through which the disruptive, or deviant, or simply atypical, is rendered normal and explicable.
 A final sociological issue addressed here concerns the topic of "meaning" and its discovery and construction among persons in social interaction. We agree with Palmer (1991), who suggests that the process of ascribing meaning often entails ascribing emotions. For example, persons often describe inanimate objects as expressing emotions; they describe the sky as "angry," the sea as "calm," and so forth. Palmer examines anthropomorphism, especially the use of emotional labels on pets, and the process through which nurses ascribe emotions on (human) neonates at an ICU facility. Palmer cites nurses saying that a baby is "angry at me," and that babies "miss me when I am away" (Palmer 1991:223). In both of these examples, the interpretation and naming of emotions is the task of observers, not the entity (human, non-human, or inanimate) to which the emotion is ascribed. In our study, we find a similar process of ascription of emotions, and we also maintain that the interactants are making meaning by asserting the "true" character of behavior that might remain undefined, bizarre, or otherwise "pathological." We believe that we contribute to Palmer (1991) in two ways: first, by discovering that the task of interpreting emotional displays is sometimes assumed by the person producing them, even to the extent that it entails statements that are highly self-effacing; and second, by analyzing the specific interactional moves and strategies that go into the process of constructing meaning. We have, in sum, addressed some examples of the social construction of emotions, and have demonstrated how this process emerges in real time.

Appendix: Interview Schedule

Introduction: I am going to be asking you some questions. Some of these will be directed to the patient [acknowledge with Mr. X or Mrs. X] and some to the spouse [acknowledge with Mr. X or Mrs. X]. Then there will be some questions that I would like both of you to answer. Any questions?

** Q1 in Initial Interview Only **

 1. To Patient: Can you tell me a little bit about the history of your illness? When did it begin? What were the symptoms? What do you think caused your illness? What other things were happening in your life when this problem began?

 To Spouse: Do you agree with that? What do you think caused your spouse's illness?

 2.  To Patient: Please tell us a little bit about how you have been feeling since the stroke/surgery/injury

 [Initial] in the past few weeks.

 [6 months] in the months since you came home from the hospital.

 [18 months] in the last year since our 6-month interview.

  Probes: What kinds of problems or symptoms are you having? Where do you feel ill?

  Further Probes: Let me ask you some specific questions.

  Have you noticed any changes in the way you look?
  Have you noticed any changes in your ability to get around?
  Have you noticed any changes in the way you communicate?
  Have you noticed any changes in the way you feel about yourself?
  Have you noticed any changes in your mood?
  Have you noticed any changes in your memory?
  How have others acted toward you since the stroke/surgery/injury? Are they sympathetic or supportive? Do they ask questions? Do they feel uncomfortable?

 3.  To Spouse: In your opinion, what kinds of problems or symptoms have you noticed

 [Initial] in conjunction with the stroke/surgery/injury?

 [6 months] since s/he came home from the hospital?

 [18 months] since our 6-month interview?

  Probes:

  Have you noticed any changes in the way s/he looks?
  Have you noticed any changes in her/his ability to get around?
  Have you noticed any changes in the way s/he communicates?
  Have you noticed any changes in the way s/he feels about her/himself?
  Have you noticed any changes in her/his mood?
  Have you noticed any changes in her/his memory?
  How have others acted toward her/him since the stroke/injury/ surgery? Are they sympathetic or supportive? Do they ask questions? Do they feel uncomfortable?

 4.  To Patient: What kinds of problems has your illness caused your spouse?

 To Spouse: What kinds of problems has your spouse's illness caused you?

 5.  To Patient and Spouse: What kinds of things do the two of you do in response to this illness? What kinds of adjustments have you made?

 6.  To Patient: How would you describe your mood in the

 [Initial] last few weeks since your stroke/surgery/injury?

 [6 months] months since your stroke/surgery/injury?

 [18 months] year since our 6-month interview?

 7.  To Spouse: How would you describe your spouse's mood in the

 [Initial] last few weeks since her/his stroke/surgery/injury?

 [6 months] months since her/his stroke/surgery/injury?

 [18 months] year since our 6-month interview?

 8.  To Patient and Spouse: Tell me a little about your marital relationship.

  Probes:

  How has the stroke/surgery/injury affected it?
  Do you feel as if your spouse understands you?
  What things do you disagree about?
  What things do you agree about?
  Are there times when you find your spouse's behavior difficult to tolerate?

 9.  To Patient and Spouse: How would you describe your personalities?

  Probe: What type of people are you?

 10.  To Patient and Spouse: What kinds of things do you enjoy doing together?

 11.  To Patient and Spouse: How often do you see friends and relatives? Is that something you enjoy?

 12.  To Patient and Spouse: Can you tell me how you spent your most recent holiday [or birthday, Christmas, Easter, Thanksgiving, 4th of July, Wedding Anniversary, New Year's, etc.]

**Take a short break**

 13.  To Patient: Can you tell us about a situation in which you felt very happy?

  Probes: What happened? Where? Who was involved? What were your specific reactions?

 14.  To Patient: Can you tell us about a situation in which you felt very sad?

  Probes: What happened? Where? Who was involved? What were your specific reactions?

 15.  To Patient: Can you tell us about a situation in which you felt very angry?

  Probes: What happened? Where? Who was involved? What were your specific reactions?

 16.  To Patient: Can you tell us about a situation in which you felt very frightened?

  Probes: What happened? Where? Who was involved? What were your specific reactions?

 17.  To Spouse: Can you tell us about a situation in which you felt very happy?

  Probes: What happened? Where? Who was involved? What were your specific reactions?

 18.  To Spouse: Can you tell us about a situation in which you felt very sad?

  Probes: What happened? Where? Who was involved? What were your specific reactions?

 19.  To Spouse: Can you tell us about a situation in which you felt very angry?

  Probes: What happened? Where? Who was involved? What were your specific reactions?

 20.  To Spouse: Can you tell us about a situation in which you felt very frightened?

  Probes: What happened? Where? Who was involved? What were your specific reactions?

 21.  To Patient and Spouse: Is there anything else you would like to tell us about yourself or your feelings that we haven't discussed?

 22.  To Patient and Spouse: Where do you think you will be six months from now? A year from now?

Table 1: Summary of Crying Episodes

Patient Sex Age at Stroke Site of Stroke*  Number of Crying Episodes Average Duration (m:ss:ss/100)  Range
1 m 74 LH 1 0:27:32 -
6 f 56 LH 6 0:44:01 (sd=0:21:10) 0:09:97-1:07:17
11 m 77 RH 2 0:34:16 (sd=0:02:83) 0:31:33-0:36:98
13 m 52 LH 5 1:05:42 (sd=0:47:90) 0:12:25-2:09:89
14 m 55 LH 6 0:25:64 (sd=0:14:17) 0:13:63-0:50:83
17 f 40 RH 3 0:16.15 (sd=0:12:65) 0:06:89-0:34:37
18 f 59 RH 15 0:30:62 (sd=0:19:33) 0:08:39-1:17:49
means: - 60.88 - 5.43 0:35:50 (sd=0:27:72) 0:13:74-1:06:07

*RH=right hemisphere stroke; LH=left hemisphere stroke.