Emotion-identity talk in aggressive interactions and in reflexive accounts

Introduction

The culture-specific nature of emotions has long been accepted, leading to viewing emotions as culture-determined, multi-systemic-shared dispositions towards action (Gross & Barett, 2011). Accepting the preeminence of culture in shaping emotions, cultural psychologists and anthropologists seek to establish a nexus between them and group identities, conceptualizing emotions as interpretation schemata or scripts (Frijda & Mesquita, 1994; Menon, 2000; Ratner, 2000; Shweder, Haidt, Horton, & Joseph, 2008), or cultural models that correspond to collective self-construals (Kitayama, Mesquita, & Karasawa, 2006; Mesquita, 2001). In this context, emotions are discussed with reference to grand-scale identities, such as nations or ethnic groups. However, evidence show that diverse and dynamic emotional repertoires are at play even within what is believed to be one and the same socio-cultural setting. Moreover, such diversity is often also at play within the emotional repertoire of one and the same person (Lahire, 2003). This inevitably calls attention to emotions as dynamic and flexible competencies, learned and transformed through diverse socialization processes (Cahill, 1999; Pollak & Thoits, 1989), and employed by individuals as contingent on their social relations in their immediate surroundings (Scheff, 1990).

From this perspective, the focus of attention shifts to the role of emotions in micro-scale identity work (Snow & Anderson, 1987) of individuals in everyday encounters, where questions of personal interrelatedness and status are central. Following Goffman and the interactionist approach, the notion of identity work was proposed by micro-sociologists to account for the complex strategies people use, if unintentionally, to demonstrate their sense of self and self-worth (Goffman, 1956). In this framework, emotions are understood in the broadest sense as practices that entwine embodied feelings for others and things and the socially recognized conceptual categories of such feelings, which together constitute experiential patterns (Wetherell, 2012). As such, they are taken to signalize, minute by minute, the way a person's own sense of self corresponds to (or deviates from) accepted social identities (Burke & Stets, 2009). By and large, it is agreed that negative or positive emotions direct the individual's demeanor in given interactions (Turner & Stets, 2006) so as to maintain their desired self-image and position vis-à-vis their counterparts. Accordingly, emotion management is a key psycho-social notion, referring to the individual's modulation of response in accordance with socially accepted feeling rules in various environments (Hochschild, 1983). The assumption here is that emotion management is embedded in the diverse, fuzzy self-images intuitively mobilized by individuals in their different cultural arenas, be they workplaces and professions (Frewin, Stephens, & Tuffin, 2006; Lively, 2000; Martin, 1999), therapy (Francis, 1997), sports (Vaccaro, Schrock, & McCabe, 2011), or gendered encounters (Schrock, Boyd, & Leaf, 2009). It works to indicate their social worth and status, moving between solidarity and conflict, intimacy and distance vis-à-vis their counterparts (Gee, 1996; Scheff, 1990).

Understood in this way, emotion identity management (Vaccaro et al., 2011) occurs incessantly. However, it is especially tangible in situations of conflict that involve threats to one's self-image and dignity. In such cases, individuals invest maximal efforts in verifying their desired identity (Goodwin & Alim, 2010; Hunt & Miller, 1997; Van Dijk, 1992). Emotional work is then expected to be all the more intense. Moreover, in contrast to studies of aggressive encounters that focus on negative feelings, notably anger (Allred, 1999; Sinaceur & Tiedens, 2006; Van Kleef, Van Dijk, Steinel, Harinck, & Van Beest, 2008), many studies on groups facing status threats point at a greater complexity and dynamism of both negative and positive identity work (Bucholtz, 1999; Sela-Sheffy, 2006; Snow & Anderson, 1987).

Emotion talk: Situated performances and reflexive accounts

Among the manifold behavioral forms of emotion management, talk provides an ultimate site for analyzing this dynamics, on twofold levels. Being a most central interactional channel, verbal behavior is imbued with expressive cues (e.g., prosody, connectivity or contextualization; Gee, 1996) that imply the speaker's emotional response even unintentionally. At the same time, it also often serves to articulate emotions as a conversational theme through explicitly speaking about them (Edwards, 1997, 1999). Studies in Conversation Analysis and Discursive Psychology offer analyses of how individuals account for themselves and others in a desirable way by using recognized emotional terminology—for instance, by portraying a person as insensitive or offensive (Baruch, 1981; Edwards, 1999); irrational or mentally awkward (Smith, 1978); or caring and nice (Li & Arber, 2006; Voutilainen, Peräkylä, & Ruusuvuori, 2010). Other studies discuss related techniques, such as avoiding emotional expressions while accounting for failures (Locke, 2003); using role-related emotional discourses (Coupland, Brown, Daniels, & Humphreys, 2008; Howard, Tuffin, & Stephens, 2000); expressing indignation (Selting, 2010); or interjecting typically gendered emotional assessments (Seymour-Smith & Wetherell, 2006).

Emotion identity talk is thus a pivotal interactional tool that regulates intimacy or distance between individuals, whether implicitly or explicitly, through different types of talk events, including, notably, reflexive accounts. Obviously, while the production of affinity or alienation occurs in all verbal encounters, emotional reflexivity is confined to moments where individuals explicitly ascribe motives (Mills, 1940) to their actions by using emotional discourse (see also Seymour-Smith & Wetherell, 2006), thereby establishing categories of relatedness between feelings and personality models. It follows that while from a Goffmanian perspective, everyday emotion expressions occupy center stage as dramatic performative tools of ideal-selves (verbal performances included; Goffman, 1974), the concern of discursive psychologists lies in how such ideal-selves are constructed through mobilizing feelings as topics of conversations. That is, the focus here is on how a person tells themselves and others who they are through talking about feelings, thereby attributing recognized emotional concepts to their (or their counterparts') personality.

It is the relatedness between interactive affective practice and reflexive emotional discourse (Wetherell, 2012), as two levels of identity construction, which we aim to explore in this study. We examine two types of identity talk performed in two different situations, namely, when the speakers react to direct offense, and in a genial situation, where they are solicited to report this experience. This, with the aim to trace in detail (1) how emotion performances unfold throughout an interaction to credibly maintain the speaker's self under conditions of offense, and (2) how the same speaker retrospectively accounts for their performance in this offensive event, thereby maintaining a self appropriate to the reporting situation.

Evidently, rather than understanding reflexive emotional talk in psychoanalytical terms, we take it to be a creative employment of legitimate discursive practices of self (Billig, 1999), which are cultural resources available to the speakers in their specific social settings (Lupton, 1998; Swidler, 2001). Often, emotional reflexivity is expected to be performed in specific types of interactions, such as psychological therapy, witness's testimony or qualitative interviews. In such cases, the participants' accounts imply their cultural disposition to use emotional reflexivity and the values they attribute to it. From this perspective, a contrastive examination is intended to reveal two interrelated discursive emotional repertoires by which members of a certain group play, namely, their pool of situated-performance strategies and that of account strategies. The inconsistencies that may emerge between a persons' situated-performance and retrospective account are precisely what reveals, in this view, the culturally determined complexity of patterns of action (Vaisey, 2009), including emotional praxis, which allows the flexibility of identity options, while bridging between them in a morally meaningful way.

Method and materials

Our analysis is based on materials from the verbal behavior of young Israeli males. The subjects are men between 18 and 20 years old, with complete high-school education, all before military service or recently drafted combat soldiers. While studies of identity often focus on marginalized or disadvantaged (e.g., gendered or ethnic) groups, mainstream men identities have until recently received lesser attention. Among gender groups, men in general are usually stereotyped as less ‘emotionally skillful’ (Hochschild, 1983). However, recent studies suggest that men's emotional repertoire is diverse and dynamic to no lesser extent than that of women (Connell, 1995; Erickson, 2005; Kiesling, 2005; Lively, 2008). In the Israeli context, male combat soldiers in particular are seen as the icons of forthright, self-reliant and aggressive Israeli manhood (Lomsky-Feder & Ben-Ari, 2012; Sasson-Levy, 2003). As such, a growing scholarly attention is attracted to the complexity of their emotional repertoire (e.g., Kaplan, 2007; Lomsky-Feder & Rapoport, 2003)—in line with current research on men's emotion-identity work (Sasson-Levy, 2008; Vaccaro et al., 2011) and men talk (Coates, 2003; Kiesling, 2001; Wetherell & Edley, 1999). However, unlike discussions that focus on collective gender identities or inter-gender tensions (e.g., Madureira, 2012), our concern here lies in micro patterns of identity work in same-gender encounters. Therefore, without venturing to tackle the intricacies of Israeli male identities in their broader social context, the present study takes the performance of Israeli soldiers and would-be soldiers as a symptomatic case for a micro-analysis of emotion-identity work in interaction. Precisely in view of their highly stereotyped image as emotionally inflexible, we aim at revealing the variable emotional efforts they invest in performing ‘this kind of man’ in specific encounters. We seek to trace the dynamics of these men's use or rejection of emotional talk when facing a massive threat to their sense of self, and specifically their use of aggressiveness when moving between the different encounter types.

We analyzed the talk of 30 subjects during a humiliating interaction and in an interview that followed immediately thereafter, where they recount this event. ¹ The interaction event comprised an ultimatum-game, in which the subject has to accept or reject offers, made by his counterpart, of splitting between them a virtual amount of money. If the subject accepts the offer, the money splits accordingly; in case he rejects, both sides gain zero. Usually, this experiment, comprising one single round, is intended to measure the participants' notion of fairness. In the present study, we designed an elaborated version of the game, which allows an extended negotiation interaction between the players, including verbal exchanges. The game comprised 10 rounds of offers-and-decisions, each ensued by a 30 seconds interval in which the participants interacted, assessing recent moves and negotiating the next steps. These exchanges were purely verbal, since the participants did not see each other during the game. The rival (an experimenter, introduced to the subjects as a student) was instructed to play an unfair, aggressive role and humiliate the subjects. The subsequent individual interview with each subject lasted ca. 30 min. These were semi-structured, conversation-like interviews, in which the subject was encouraged to recount, in their own words, their experience during the game.

The game interactions and the interviews were recorded and carefully transcribed. All the verbal output at our disposal (the ultimatum game exchanges included) was produced as natural spoken discourse. This fact is crucial for reconstructing from the speaker's own speech their nuanced complex emotional work. We proceed from the assumption that how people talk (and not just what they say) is what governs meaning production and impression-management (Boxer, 2002; Cameron, 2001; Gee, 1996; Tannen, 1991). Hence, we use close discourse analysis beyond content analysis, based on maximal cues to the speaker's interactional inferences (Gumperz, 1979) and evaluation signals (Labov, 1972). These include, beyond lexical selection, prosodic and organizational patterns (e.g., repetitions, pauses, corrections, hesitations, disclaimers or simultaneous talk), as well as paralinguistic signals, such as laughs and sniffs.

Analyses of the game tapes and the interview tapes were conducted separately. Each tape was meticulously transcribed by one member of the research team (composed of three researches), then read and coded by the two others separately, using a table to summarize the results. Results were subsequently compared and discussed by the three team members, to sum up the dominant strategies employed by each subject. Analysis was led by the following research questions: (1) Which moral percepts motivate the subject's talk? (2) What emotional discourse does he employ, if at all? (3) Which discursive techniques does he use? Finally, a comparative analysis was conducted for all the transcripts in the samples, to identify prototypical ‘emotional selves’ performed by all the subjects during the game and during the interviews. For length constraints, in this article, we present an in-depth analysis of the verbal performances of one single subject (A), which serves here to illustrate the prevailing discursive patterns performed, albeit with variations, by most of the subjects in our study. We focus our analysis on this specific subject because his performance during the game was overall very confrontational (out of 10 rounds of offers, he rejected 9; while the average was 7 rejections per subject). Consequently, his total scores were very low (while the highest possible total score in each game was 48 points, his was 10).

Analysis and discussion

Our analysis aimed at an anatomy of emotion-identity talk in concrete events. We sought to identify from the transcripts at hand the diverse conversational strategies employable and modifiable by the given group of speakers to co-establish their relationships with their counterpart and to position themselves with respect to this event in retrospect. This, with the view to tracing how these speakers produce and adjust their self along a time continuum determined by the given interaction events.

Emotional situated performances

First, we sought to identify how the speakers created involvement or detachment vis-à-vis their counterpart in the game and the morally-informed type of self they assumed thereby (Goffman, 1957). Findings revealed that most of the speakers perform a detached, tough negotiator, who neither gives-in nor seeks a compromise in a conflict, which, in his view, violates norms of rational reasoning and fair play. In conversational terms, this type of self is constructed through focusing attention on the game rules and overlooking the counterpart's offense, while oscillating between business-like confrontations and avoiding response. Let us examine how all this works, using the example from A's performance. The following exchanges occurred after the fourth round of the game: ²

(4th round; the rival's offer of 16-4 was rejected)

1 R: tell m:e, you want to continue being pigheaded?

2 or you want to go on using ye brains? <louder>

3 A: there's one ((person)) here who's pigheaded

4 (-) eh… it's you!

5 R: no, ma::n, I'm not pigheaded, you are pigheaded <louder>

6 we're not in ((a name of a combat unit)) he::re <scornfully>

7 A: <chuckles>

8 R: give me some credit, start thinking! OK? <louder>

9 A: as long as I don't get more, I reject,

10 I don't care getting zero

11 R: tell me, can you say something else apart from that?

12 A: (-) d.do you know ((how)) to give more?

13 show me! <louder>

14 R: I know lots of things. Give me some credit here

15 and I'll go along with you!

16 A: (-)

17 R: don't be a jerk! <louder>

18 A: (-)

19 R: I give you one more cha:nce, come on?!<louder>

All these conversation techniques produce the speaker's sense of self-distancing without resorting to explicit emotional discourse (Goodwin & Goodwin, 2000). Three main strategies emerge as central in this speaker's impression management: (1) Back attacks and Accusations work to concentrate attention on the improper behavior of the rival and his unacceptable negotiation style. (2) Whereas threats shift the focus of conversation to the rules of the game, attributing responsibility for potential negative implications to the rival's breach of these rules. (3) Finally, exchange-interruptions are attained through chuckles, silence and lesser ‘tying’ (Sacks, 1992, pp. 150–156) between the speaker's utterances and these of the rival. These recurrent conversational routines constitute a performative model of a distanced interaction-partner in a conflict. Using these routines, the speaker persistently calls attention to violations of the game rules and at the same time remains unresponsive to personal assaults. Thereby he actually renders the conflict unresolvable.

Retrospective emotional discourse

Let us now briefly examine the emotional account (Edwards, 1999) of this game experience produced by the same individual during the interview that followed. All the interviews in our sample started with a most general question, “how was the game?” which, save for rare cases, triggered the interviewee's story with but minimal follow-up interventions by the interviewer (Rapley, 2001). We examined the strategies the speaker employs in narrating their relations with the rival and their involvement in, or in self-distancing from this unpleasant past event.

In recounting the game event, the rival's misbehavior constitutes a key element in scripting the violation of the game rules, against the background of which the subject's own moral advantage transpires. Unlike his talk during the game, in his retrospective account the speaker often resorts to an explicit emotional discourse, usually as conducive to highlighting the rival's flaws. However, the overall emotional effect of his accounts is generated mainly by the speakers' complex positioning (Davies & Harré, 1990) of his present (reporting) self vis-à-vis his past (reported) self-in-interaction. Overall, he wavers, if unintentionally, between a ‘rationally objective’ and ‘emotionally subjective’ modes of reporting, thereby projecting alternatively ‘external’ and ‘internal’ perspectives on his past experience as an actor in the game. It is mainly in this nuanced interplay between reporting modes where his emotional identity work emerges and diversifies. In what follows we present two most recurrent patterns of such accounts, as performed by A.

Conclusion

Unlike studies that focus on collective (gender or other) group identities, our analysis addresses the micro-level identity work in interaction. We proceed from the assumption that every interaction type—be it a game negotiation or an interview—imposes certain rules that constraint individuals’ performance of self. Identity performance is, therefore, dynamic and modifiable, depending on one's culturally-available options to act according to their understanding of the situation. From this perspective, we take emotions to be cultural competencies that work to fulfill self-in-interaction tasks, such as managing relations with the counterpart and maintaining dignity. As such, the various forms of emotion talk constitute vital tools of self-management, which allow shifts and ‘corrections’ of performance routines, especially in situations where one's dignity is at stake (Goffman, 1956).

Our analysis reveals that even members of a group extremely stereotyped as ‘emotionally rigid’ and aggressive, such as young Israeli male soldiers or would-be soldiers, creatively employ emotion-identity talk while moving between the different encounter types (the game and the interview). Findings, consistent with the prevalent assumption that men tend to hide emotional vulnerability (Lois, 2003), demonstrate the speaker's skillful ways of using two emotional self-in-interaction types—confrontation or detachment—as two different discursive models, with their own rules and techniques. However, his flexible emotional performance is not arbitrary, but rather creates a coherent self (in this case—a rational, invulnerable and justifiably aggressive man) throughout a given sequence of changing encounters. Specifically, our findings reveal the following points:

The dominant strategy employed by the subject to maintain a dignified self is that of engagement in the interaction through detaching oneself from it (cf., with reference to macro-level identity processes, Sela-Sheffy, 2006). Such a micro-scale emotional identity management of an individual speaker obviously also contributes to understanding large-scale identity processes and the production of group emotional stereotypes. However, whether this strategy, and the discursive techniques attached, are distinguishing of a specific Israeli masculine culture or permeate wider socio-cultural contexts is still to be examined.