Keywords

Introduction

Emotion plays an essential role in many social contexts where group decision taking and negotiation are important, such as jurisdiction, science, politics, business, medical care, etc. Despite that, there are few discourse analysis studies dedicated to the understanding of the discursive function and linguistic realization of emotion in group decision and negotiation (GDN). We often experience number of emotions at the same time or in short sequences, some we are conscious of others—not. This is one of the reasons why emotions are seldom explicitly referred to in interaction. Instead, they are mostly indicated by intonation, tone of voice, choice of words, and nonverbal behaviors. Therefore, face-to-face interaction is the most suitable modality for the analysis of emotions in group decision and negotiation contexts. Since discourse entities are multi-functional, qualitative ethno-methodological analysis of recorded interactions is suitable and necessary. GDN involve argumentation. However, following an ancient dichotomy between the emotional and the rational, argumentation models seldom address and involve emotion.

The present paper explores the relation between emotion and discursive concepts and theories such as Reciprocal Adaptation, Interactive Alignment, and Theory of Theory of Mind (ToM), the definitions of which are discussed in the next section. It studies if and how activity and interaction affect the role of emotion in group problem reframing and group problem solution. For that purpose, it seeks evidence for and against Interactive Alignment theory and theory of ToM by observing Reciprocal Adaptation in authentic data. The importance of ToM-based reasoning affects both understanding of Human–Human interaction and modeling of Virtual Human cognition.

The study explores the linguistic realization and discursive function of emotions as they occur in authentic GDN. The observed here discursive features and functions aim to contribute to the understanding of face-to-face negotiation as a process and to research on the influence of institutions and activity types on the participants’ ability to express, perceive, and elicit emotions. The presented model of emotion in GDN reflects the active role emotions play in decision taking as modifiers of theory-of-mind models, goals and strategies. The model is based on empirical studies of emotion in human interaction in different activities such as plea bargains, simulated negotiations, and doctor–patient consultations. Empathy is used as an example of a complex emotion, which has a natural and powerful function in the shaping and re-contextualization of decision processes.

Theoretical Foundations

Research on emotion relates it more intimately to cognitive processes such as memory and planning. In this section, we explore the current views on emotion, argumentation and ToM.

Theories of Emotion

Three hypothetical descriptions of the relation between emotion and cognition have been discussed through the centuries, which, as Scherer (1993) suggests, could be summarized in the following way:

  • Emotion is a separate system related to two other systems in an organism, namely cognition and will (Plato, Kant, Mendelsohn, Leibniz etc.)

  • Emotion is a grand system, a coordinator of all developing subsystems in an organism (Freud, Descartes)

  • Emotion is one of many components in a complex organism, which are in constant dynamic interaction with each other (Aristotle, Spinoza)

The dichotomy between emotion and cognition as well as this between irrational and rational stems back from Plato’s political doctrine in “The Republic” where he claims that human and political well-being depend on the harmony between three separate units of society and soul: cognition (ruling class/thought, reason, rational judgment), “thumos” (warrior class/higher ideal emotions), and motivation (lower class/impulses, instincts, low desires). The Aristotelian tradition questioned this dogma by saying that desire can be found even in motivation and in cognition and that there could be many other components in the soul. In the context of Darwinism, emotion has a roll in adaptation in the course of evolution, emotion is universal, and expression of emotion is found in other species (Cornelius 2000). In Descartes’ era, emotions intertwined with cognition of stimuli. Freudians called for exploration of emotion as a basic condition influencing the conscious and unconscious. William James (Myers 2001) introduced the role of the body in the cause and effect chain: The mind perceives the reaction of the body to stimuli, e.g., increased heartbeat; the sensation of the physiological response is a feeling which mental representation is an emotion, e.g., fear. In appraisal theory, which is a form of cognitivism, emotion is seen as something automatic, non-reflective, and immediate and at the same time, cognition leads emotion, i.e., the way we cognize events influences our emotions related to them, not the opposite. In this sense, emotions become and involve coping strategies (Lazarus 1999). According to the social and anthropological constructivist theory it is the sociocultural interpretation, which determines emotions and body reactions, e.g., attitudes to language variations such a dialect (Cornelius 2000).

Contemporary neuroscientists report evidence for the involvement of emotion in so-called rational cognitive processing. Neuroscientists such as Von Uexkull and Kriszat (1934), Fuster (2003), and Arnold Scheibel (personal communication) observe that evolution gave privilege to the limbic system: Emotional feedback is present in lower species, but other cortical cognitive feedback is present only in higher species. In that sense, emotion functions in evolution as a coordinator of other cognitive and non-cognitive functions.

Damasio (1994) suggests that the state of the mind is identical to the state of feeling, which is a reflection of the state of the body. He explores the unusual case of Phineas Gage, a man whose ability to feel emotion was impaired after an accident in which part of his brain was damaged. Damasio finds that, while Gage’s intelligence remained intact after the accident, his ability to take rational decisions became severely handicapped because his emotions could no longer be engaged in the process. Based on this case, the neurologist comes to the conclusions that rationality stems from our emotions and that our emotions stem from our bodily senses. Certain body states and postures, e.g., locking of the jaw, would bring about certain feelings, e.g., anger, which in turn will trigger certain thought and interpretations of reality, a thought, traced back to William James.

All approaches to emotion underline the major role emotions play in cognitive processing, yet cognitive models of argumentation and negotiation processes exclude the involvement and the effect of emotions.

Argumentation Theories and Emotion

Historically, one of the most influential accounts of argumentation is called Toulmin’s model (1958), which analyzes six features of an argument: data, claim, warrant, backing, qualifier, and rebuttal.

van Eemeren and Grootendorst’s argumentation theory (2004), the pragma-dialected approach, is currently most popular. They define argumentation as a verbal, social, and rational activity aimed at convincing a reasonable critic of the acceptability of a standpoint by putting forward a constellation of propositions justifying or refuting the proposition expressed in the standpoint.

Walton (1989, 1996) studies negotiation and argumentation by means of informal logic and critical thinking. He offered an account of argument schemes for presumptive reasoning which constitutes the majority of reasoning and argumentation people engage in. Argument schemes are structures or forms of argument, which are normatively binding kinds of reasoning and are best seen as moves, or speech acts in dialogs (Walton 1996).

In legal argument and legal reasoning, case-based and logic-based approaches (e.g., non-monotonic logic) have been applied to study legal argumentation, supplemented with an argument-scheme approach (McCarty 1997; Prakken and Sartor 2002; Prakken 2005). Meanwhile, in artificial intelligence and multi-agent research community, researchers have built computational models for multi-agent negotiation and argumentation-based systems (Parsons et al. 1998; Kraus 2001; Traum et al. 2003).

With the exception of Walton (1992), these theories did not address the role emotions play in argumentation and negotiation. Gilbert (1995) pointed out that emotional, intuitive (kisceral), and physical (visceral) arguments ought be considered legitimate and studied just as much as logical arguments. However, neither Walton nor Gilbert has come up with a model of how emotions alter negotiation and decision-making process.

Theory of Mind and Emotion

Cognitive-emotive processes that support group decision making and negotiation require a capacity for understanding and empathizing with others. This capacity involves the understanding of differing beliefs, intentions, emotional and visceral states, ability to react and to draw necessary inferences, to predict and plan given these concerns. The term Theory of Theory of Mind (ToM) was coined by Theory of Theory (Morton 1980) and refers to the abilities humans and other higher species have to perceive and reason on the mental/emotional states of others but also of their own. According to some studies, ToM processes provide a special kind of context: the minds and emotions of others (Martinovski and Marsella 2003; Givón 2005). In interaction, people learn to act within these contexts. Beliefs about age, gender, language, environment, and so on contribute to the models that individuals form and keep of each other’s intentions. ToM explanations have importance for the interactive realization of emotion, i.e., the way we understand our own and others’ emotions in negotiation.

Three mutually exclusive theories have been suggested to explain how we relate to others: by imitation (e.g., Iacoboni 2005), by simulation (e.g., Gordon 1986; Stich and Nichols 1992), or by representation (e.g., Hobbs and Gordon 2005).

Originally, the main process for establishing and communication of ToM models was and still is thought to be imitation. There is increasing evidence from neurosciences “that the neural mechanisms implementing imitation are also used for other forms of human communication, such as language … Functional similarities between the structure of actions and the structure of language as it unfolds during conversation reinforce this notion … Additional data suggest also that empathy occurs via the minimal neural architecture for imitation interacting with regions of the brain relevant to emotion. All in all, we come to understand others via imitation, and imitation shares functional mechanisms with language and empathy” (Iacoboni 2005; see also Chap. 1 in this Volume).

According to “simulation theory,” we think of the other’s experiences by use of mental and even somatic simulation of, e.g., our own experience of the same kind (Gordon 1986). Thus, if someone has a stomach ache, instead of imitating his/her experience of a stomach ache one can simulate the psycho-somatic processes related to one’s own previous experiences of a stomach ache and that way form an understanding and a reaction to his/her state.

Yet a third idea is that ToM is the application of commonsense inferences about the way people think (Hobbs and Gordon 2005). Here, if someone has a stomach ache one can understand her/his state based on ready-made mental representations, which describe what it is to have a stomach ache, without going through somatic imitation or mental simulation.

The last two explanations seem mutually dependent. In order to simulate a stomach ache, one must have some representation of what that is. In order to make inferences about mental representations, one may have to play “as if” games. Martinovski (2007a, b) has suggested that imitation, simulation, and representation are cognitive-emotive processes developed in evolution, all equally available for homo sapiens sapiens.

Researchers have suggested different mechanisms for dealing with ToM’s complex processing. Baron-Cohen talks about “mindreading” or the ability to monitor others’ intentions. He goes to an extreme, claiming that successful communication entails a constant feedback-check between communicators to verify whether the listener’s interpretation corresponds to the intended interpretation. In discourse analysis, this feedback-checking is reflected in the concepts of grounding and feedback (Allwood 1976). In computer science, the concept of grounding has been used for the design of computational models of dialog (Traum 1994).

In the model developed here, ToM plays an important role, parallel to emotion, as it enables reasoning about own and others’ goals and strategies and changes thereof.

Interactive Alignment and Reciprocal Adaptation

Negotiation and group decision making are defined as processes of problem reframing and problem solution. Since negotiation and group decision taking are most often conducted interactively, one needs to redefine the concept of framing by adding the socio-interactive aspect of its meaning. Thus by problem reframing, we mean not only cognitive-affective reframing but also socio-interactive restructuring of a problem through discursive grounding. Prospect theory (Kahneman and Tversky 1986) did not offer understanding on decision taking as socio-interactive reframing, but only as a cognitive-affective process. In order to understand negotiation and group decision taking we need to understand how the socio-interactive framing affects the decision process.

The concept of framing has different definition and application in the fields of decision studies and in social studies. Related to decision taking, framing is defined as a cognitive process, which limits and directs the interpretation of and the emotional relation to the perceived and imagined loss and gain. In other words, here the process is cognitive and emotive. One of the ways in which people frame, for instance, losses and gains, is that losses regularly hurt more than gains feel good (ibid.). The framework or the structure that maintains this state of preferences affects decision taking.

In interaction studies and sociology, the concept of frame is different. Goffman’s (1967) original description and understanding of framing is described in short as “organization of experience” in interactive and social order. Frames are not only cognitive but also social interactive constructions, i.e., the frame is set also by the activity and the situation as such, including roles, settings, goals. For instance, in a courtroom, the judge sits in the middle of both parties, independently of how the defendant understands the situation. It follows that in Goffman, a situation is not created by the interactants but a frame is.

Gumpertz (1982: 13) definition of Reciprocal Adaptation is also related to framing: “the procedure … where each participant gradually learns to adapt and to enter into the other’s frame of reference.” In his view, Reciprocal Adaptation is involved in interactive reframing of situations, knowledge, and arguments and it does not presuppose conscious or less conscious processing of information. This communicative, cultural and learning mechanism is not only cognitive but also linguistic, emotive, and behavioral, i.e., speakers adapt to each other on different levels: lexical and semantic choice, syntax, posture, gaze, proximity, orientation, tone of voice, etc. It is a mechanism behind linguistic phenomena such as creole-like varieties of languages and interactive emotions such as empathy and rapport (Martinovski 2010). Similarly, human users adapt even to the speech and behavior of the Virtual Agent (Bell 2003; Martinovski and Traum 2003). Thus, there are basic interactive mechanisms, which characterize human interaction and affect participants’ frames of reference, such as Reciprocal Adaptation. Reciprocal Adaptation can be seen as the discursive correspondent to Moore and Iacoboni’s neurological concept of neural resonance (see Chap. 1 in this Volume).

According to Pickering and Garrod (2006), communication in discourse is accomplished through an interactive process they call alignment and successful communication through good alignment: “the development of similar representations in the interlocutors … interlocutors align situation models during dialogue.” (ibid., p. 1). The main claim of their theory is that “automatic processes play a central role and explicit modeling of one’s interlocutor is secondary” in communication (ibid.). The alignment involves situation models and non-situational knowledge such as language knowledge. Interlocutors align their situational knowledge, but they also align knowledge of situation and language (for instance, what they think “right” means with the word “right”). The situation models include notions such as “space, time, causality, intentionality, and reference to main individuals under discussion” (ibid., p. 2). Alignment is based on willingness for cooperation and on mechanistic automatic imitation (of lexical choices, syntax, tone of voice, etc.): “Our underlying conceptualization of conversation is collaborative, in that we treat it as a ‘game of pure cooperation’ … in which it is in both interlocutors’ interest for it to succeed for both interlocutors” (ibid., p. 22) and “the interactive-alignment account proposes that alignment is primitive. It is a form of imitation and drops out of the functional architecture of the system … In these accounts, imitation is an automatic, non-inferential process and is in some sense the default response. Generally, imitation does not appear to require any decision to act” (ibid., p. 18). Thus, alignment does not involve building of entire theory of the other but a primitive turn-to-turn alignment on different linguistic levels of the message: phonetic, syntactic, semantic, etc. Each level is processed and aligned for itself and misalignment on one level enhances alignment on another level. Garrod and Pickering point out also that children cannot inhibit alignment, which speaks for the forcefulness of this interactive mechanism. They base their view on situated interaction where participants have to find interactively each other’s position in a maze without being able to see it and assume that the same mechanism works on everyday conversation. “Such models are assumed to capture what people are ‘thinking about’ while they understand a text, and therefore are in some sense within working memory (they can be contrasted with linguistic representations on the one hand and general knowledge on the other). Successful dialog occurs when interlocutors construct similar situation models to each other” (ibid., pp. 1–2). They point out “that this account differs from Clark (1996), who assumes that speakers carefully track their addressees’ mental states throughout conversations” (ibid., p. 10) and that “the important point is that effects of partner specificity do not imply that interlocutors need employ complex reasoning whenever they produce an expression. Instead, they have a strong tendency to employ the form that they have just encountered” (ibid., p. 20).

Interactive Alignment is similar to the concept of Reciprocal Adaptation in that they refer to framing in terms of similar discourse processes and do not demand conscious processing during interaction, although the connection has not been made nor explored yet in the literature.

Reciprocal Adaptation, developed around 1982:

“the procedure … where each participant gradually learns to adapt and to enter into the other’s frame of reference.”

Interactive Alignment, developed around 2004:

“the development of similar representations in the interlocutors … interlocutors align situation models during dialogue.”

In that sense, Garrod and Pickerings response to Goffman’s definition, is: Interactants use Interactive Alignment to frame their activities in order to make sense of a situation, where alignment is primitive, automatic, based on imitation and realizes on both production and comprehension level. However, it is hardly the case that alignment is the process through which interactants make sense of a situation since by definition this process is primitive, i.e., does not involve complex reasoning or active long-term memory-based monitoring. Rather it is more likely to see it as a way of preliminary or first layer framing. Gumpez’ response to Goffman, is then: Interactants gradually learn to frame their activities through Reciprocal Adaptation in order to make sense of a situation because Reciprocal Adaptation is defined as a more general phenomenon than Interactive Alignment which does not pose a condition of automaticity and short-term memory basis.

The mechanistic aspect of communication, of actual meeting between people, does bring changes into negotiation, which is otherwise characterized by pre-set values, preferences and strategies, therefore emotive adaptation in negotiation is of great interest and it has not been studied sufficiently. However, even complex cognitive-emotive processes such as some forms of ToM building can bring changes to negotiation and decision taking. For example, Martinovski (2006) found that empathy and rejection of empathy involve ToM modeling. Martinovski (2011) found that emotions function as engines in conflict management and involve opposite Reciprocal Adaptation.

Empathy

Empathy is a complex emotion defined by Mead (1993) as the “capacity to take the role of the other and to adopt alternative perspectives vis-a-vis oneself”; and by Hogan (1969) as the ability to take “the intellectual or imaginative apprehension of another’s condition or state of mind.” Reik (1949) describes four stages of the empathy process: (i) identification—projecting self into the other, (ii) incorporation—introjecting the other into self, (iii) reverberation—interplaying of own and other experience, and (iv) detachment—withdrawal from subjective involvement and recourse to use of methods of reason. In that sense, empathy points to an interesting phenomenon of communication, namely uncertainty or rather openness to unknown outcomes, to the possibility of change of goals, needs, and behavior in the course of communication. In the case of negotiation, empathy may contribute to the unpredictability of the planned or desired outcome.

Davis (1994) distinguishes between cognitive and emotional (or emotive) empathy, which refers to empathy as attitude or taking the perspective of the other and empathy as emotional response to the emotions of the other, respectively. He also suggests a distinction between two types of emotional empathy: parallel empathy (PE) or empathy related to the other’s feeling directed to a third person and reactive empathy (RE) or empathy to the other’s feelings oriented toward it/her/himself.

Under stress people seek what Lazarus (1999) calls problem-focused and emotion-focused social support. Empathy is one of the resources available in the process of seeking and giving social support, whether it is predominantly emotional or predominantly problem-or2005iented. In this sense, empathy is a form of a coping strategy.

Empathy has been studied experimentally and theoretically, with short- and long-term perspectives (Bussman and Muller 1992). Heritage () uses a conversation analysis method in his study of social empathy interplay. Goodwin and Goodwin (1987) use ethno-methodology to study the realization of assessments in talk. Jefferson et al. (1987) examine the pursuit of intimacy. These studies, however, are not oriented toward situations of negotiation per se and are concentrated on certain specific feature rather than looking for descriptive features.

Martinovski et al. (2007) indicate that empathy (elicitation, acceptance, rejection, refusal of empathy) may be seen as a general cognitive-emotive capacity necessary for successful human interaction. This view relates to the Theory of ToM, which claims that interactants consciously or less consciously build theories of each others’ and own goals, knowledge, characteristics, social and emotional status, facilitated by specific neural resources and processing. Three competitive theories explain how ToM building occurs: by simulation, by imitation, or by representation. Martinovski (2007a, b) finds manifestation of all three cognitive-emotive processes in discourse.

Definitions of Basic Concepts

Section “Theoretical Foundations” above discussed analytical concepts such as Reciprocal Adaptation, Interactive Alignment, and ToM-based reasoning, which are used in the empirical analysis in the next sections of this chapter. The present section states the choice of working definitions of main concept used in the study such as negotiation, group decision, and emotion. Negotiation is commonly defined as the communication process through which a group of people or agents try to reach a mutually acceptable agreement on some matter (Bussman and Muller 1992). Typical examples are business negotiations, labor negotiations, salary negotiations, negotiation in courtrooms, diplomatic negotiations. Negotiations can be seen as bargaining or as problem solving or as dispute resolution. One may also use negotiation in a broader social sense as the communicative process, through which social values are discussed before and as they are shared in the community. Negotiations thus involve not only problem solving skills but also general communication skills, including management of emotions. The involvement of empathy in the discussion of the process of negotiation may encourage a more collaborative rather than conflict-oriented conceptualization of negotiation. The group decision is defined by its main goal, which is taking a decision, whereas negotiations may not aim nor end up with a decision (see also Meerts and Vukovic in this volume).

I define emotion in a broad sense as a physiological, cognitive, and discursive experience, which is affected by and affects processing and outcome of cognitive and discursive activity, i.e., emotion is defined as recognition of sensory feedback and as coping strategies, which coordinate decision making not only on personal level but also in interaction. Thus I include as emotions not only experiences of joy, sadness, anger, and fear, but also complex emotive-cognitive experiences such as trust and empathy. Emotions are not static. They are processes on a neurological, biological, and expressive level. One and same stimuli can cause a chain of different physiological reactions, emotional sensations, and cognitive appraisals, each of which can influence the other in time.

Data and Method

The present study uses discourse and conversation analysis methodology in order to observe the relation between linguistic and discourse features and functions of emotion in GDN. Conversation analysis uses detailed analysis of specific and representative dialog to isolate generalizable or not generalizable features and processes. Before conducting any meaningful statistical analysis or modeling we need to understand the phenomena we are dealing with and one way to do that is through a sufficiently rich analysis, which will help us to extract the focal features, their function and organization (Martinovski 2000). The aim of such analysis is to observe the small-scale and local interactive events and processes, which cause social change rather than to infer how institutions determine the interaction. Certainly, many factors shape the emotion episodes and their linguistic realizations, e.g., institutions, biological states, temperaments, cultures, etc. Institutional settings structure the roles of the speakers with respect to their rights and obligation to elicit, give, and respond to emotion. For instance, in courts, emotion elicitation is less successful and welcome; in the doctor’s office and between friends all functions of empathy are quite expected; on a war field the wounded may have the advantage of empathy, etc.

The present study uses data from several different genres of negotiation dialog, including Talkbank, Role Play, an informal conversation, and a plea bargain. The Talkbank Clinical data involve interactions between a healthcare provider and a patient: http://xml.talkbank.org.:8888/talkbank/file/talkbank/Clinical/Holland/. The nurse tried to negotiate with the patient to participate in the scheduled treatments without creating conflict.

The other set of data consists of transcriptions of role-playing dialogs between a US captain and a lay-person playing the role of a Spanish doctor from a non-governmental medical organization in Iraq. The captain has been instructed to negotiate with the doctor to move his clinic without exposing secret information and the doctor has been instructed that in his negotiation he must care mainly about his patients and that he is representing an independent non-militant organization.

The audio-recorded and transcribed informal conversation comes from John Heritage’s corpus. The audio-recorded and transcribed plea bargain is a part of Douglas Maynard’s corpus.

Transcription Standards

The utilized transcription conventions are: “[]” stands for overlapped speech; “:” stands for prolonged vowel; “=” stands for latching speech; “/” indicates pause; capital letters indicate emphatic speech; “+” indicates cutoff; “()” stands for inaudible speech; “?” stands for rising intonation; “.” stands for falling intonation; “,” stands for continuing intonation; “=” stands for latching; “_” stands for emphasis; “{0.9}” stands for seconds of pause. Each line in the transcription indicates an intonation unit.

Structure of the Plea Bargain

The plea bargain has a particular sequential structure. Sitting in a room with a judge, we have a defense attorney and a district attorney. The discussion is whether the accused should get jail and for how long or a fine and in that case of what amount. In general, the parties have to agree first that they are willing to settle the case, then to establish the Penal Code provision that applies of the crime and at last, agree on the settlement value. This particular instance of a negotiation involves sequences and phases of main activities and different kinds of subactivities and topics:

Main activities and subactivities/topics and their initiators (subactivities in italics; major negotiation accomplishments in bold):

  1. 1.

    Brings up Frank Bryan’s case—Judge (Jge)

  2. 2.

    Inserted talk about a different case procedure referring back to a topic discussed before line 1 where the judge brings up Frank Bryan’s case—Prosecutor (Prs)

  3. 3.

    Return to the case topic—Jge

    • Parties present their interpretation of events

    • Defense offers settlement and reference to Penal code, insists that this is a case of disorderly conduct (CPC: 647f) rather than Arrest Resistance case (CPC: 148).

  4. 4.

    A meta-comment on the origin of his settlement strategy—Defense (Def) to Jge

  5. 5.

    Agrees to settle, suggest a type of crime, 148 rather than 647f—Prs

  6. 6.

    Discussion on events, type of crime, and arrest period—Def and Prs

  7. 7.

    Didactic instruction—Jge to Prs

  8. 8.

    Aggressive refusal to involve defendant’s prior criminal history—Def

  9. 9.

    Side talk about rain—Jge

  10. 10.

    Plea Bargain Agreement—Prs, Def

Each one of the phases in the negotiation has particular initiation signals and initiators. The order of the phases provides context and grounding for the rest of the phases, i.e., this sequential order provides the organic structure of the interaction. Phases are defined as larger units of talk distinguished by topic, activity, and location in the conversation. Sequences are units of talk, which involve at least an adjacency pair and which build up phases in conversation. They are often used to jointly accomplish a communicative act/project.

There are number of concrete facts, which are considered by the parties in order to apply relevant provisions, establish settlement value, provide substantial justice, and eventually reach a plea bargain agreement:

  1. 1.

    Did the defendant resist arrest?—yes/no

  2. 2.

    Did the defendant strike an officer?—yes/no

  3. 3.

    Did the defendant cause disorder?—yes/no

  4. 4.

    Did the defendant spent time in jail already?—yes/no/how long

  5. 5.

    Does the defendant have prior convictions?—yes/no/what kind

The defense counsel’s arguments mitigate each stance based on the above questions:

  • Defendant did not resist arrest other than verbally and if he did it just looked like resistance but it was not because he was drunk;

  • Defendant’s character when not drunk is a very peaceful and sweet;

  • There is no evidence that he stroke an officer;

  • Defendant caused disorder but it is a minor family thing thus trivial, in fact he was probably even justifiably angry since “what kind of mom calls the police on her son”;

  • Defendant was drunk and if he was not he would not do what he did;

  • Defendant is black and if he was not it is less likely someone would call the police, even his mother.

Prosecutor’s arguments refer to police report and legal provisions texts:

  • Defendant resisted arrest but not only verbally: he tried to escape;

  • There is not evidence he stroke an offices but the report is not full;

  • Defendant caused serious disorder to this extent that his own mother called the police, which points to 647f provision related to disorder conduct;

  • Defendant has spent time in jail justly since he did resist arrest although not clear for how long;

  • Defendant has prior convictions related to disorder conduct and violent resistance to arrest, including striking an officer, thus the most relevant and urgent provision is CPC: 148, which provides jail in order to reach substantial justice.

Linguistics Manifestation and Functions of Emotions in GDN

Emotion occurs spontaneously, but in negotiation it can be used strategically as argumentation and it can be manifested in various ways. This section studies linguistic manifestations and functions of emotion and their relation to cognitive appraisal and coping processes, which go parallel with interaction.

The first 7 subsections use examples from a plea-bargain conversation in chronological order, followed by a subsection which concentrates on linguistic realization of empathy using friendly conversation between two women and role-play negotiations between a Doctor and a Captain (see also Section “Data and Method” above).

Entertainment

The prosecutor expresses intention to settle rather than continue to trial, what needs to be decided is the settlement price. Instead of acknowledging that the defense attorney (Def) initiates one side talk as a form of entertainment and a show (italized) of his experience and friendship with famous and successful lawyers.

In this embedded sequence, the Def manages to entertain both the Judge, who often laughs, and his opponent with a subnarrative. He also points out that he behaves within a context and with a strategy, that he is playing a role as prescribed by the best in his business. The linguistic means he uses to accomplish emotional experience such as entertainment are:

  • Narrative

  • Slang imitation

  • Feedback elicitors (“right?”)

  • Lexical choices (“new kid,” “old pro,” “that trade”)

  • Tone of voice

  • Intonation

There are many and different kinds of communicative acts involved in this narrative, all carrying a joyful emotion but the context changes illocutionary force of these acts into mitigated forms of threat, i.e., expressing and evoking emotions of fear and humiliation.

Empathy Evocation

Number of the defense’ arguments build on and aim to evoke empathy: Being black is a disadvantage therefore an excuse; being drunk provides an excuse too, as well as having one’s “mom” call the “cops.”

Empathy evocation is signaled by number of linguistic devices, which remind of adolescent speech style thus pointing to the person’s immaturity, reaches also to personal association with own family history such as:

  • Tone of voice

  • Intonation

  • Lexical choices (“mom,” “cops”)

  • Gesture

Anger

The entertaining and ridiculing style is interchanged with demonstrations of helpfulness and claimed agreeability, which later develop into aggression and disgust (italized). The reaction of the prosecutor (Prs) is again self-explanation presented in an even weaker manner as he stutters and has a problem formulating a sentence. The Def continues his ridicule by use of mocking back channels, initial interruptions, latching, ridiculing mocking repetitions, etc. In this manner, the Def gains a dominant emotional role in the conversation, wins the floor, and presents his personal hypothetical interpretations as arguments.

Emotionally loaded imperative expressions such as “knock it off” and “come on” and throat clearing act as more powerful persuasion devices than the arguments, which are inferential and unmotivated:

  • Tone of voice

  • Intonation

  • Sentence modality

  • Turn taking—interruption, latching

  • Lexical expressions

  • Sequential timing of aggression

It is the interplay between context and emotion that changes the illocutionary force of Def’s communicative acts, i.e., the question “You wanna make a federal case of this” in combination with the underlying emotion of anger and aggression carried out by tone of voice, intonation and context turn the interrogative illocutionary force into a threat, evoking fear.

Ridicule, Sarcasm

As the prosecutor has agreed to settle he proposes a settlement value. He is joining the Def in his playful colloquial style of speech, which is evident in lexical choices such as “dandy,” “wanna,” “probably.” Def objects to the suggestion starting with an interruption and an initial “wull” discourse particle. Def has no good argument other than reasoning based on his personal hypothetical interpretation of events. Def interrupts again this time the very beginning of the Prs attempt to take the floor and present his own objection. Def objection is again underlying his personal view in a categorical manner, which involves even sounds such as garbling, signaling ridicule or his personal view of discontent. In response to the Prs is defensive and presents an explanation of his initial settlement suggestion which more or less cancels it and expresses his own uncertainty. When he tries to present his view of the situation, starting with a Theory-of-Mind expression such as “I think” the Prs is again interrupted by the Def. This time the Def continues the ridiculing strategy vocalizing a mocking reaction (italized) of surprise with a single discourse particle or exclamation “oh.”

Having put the Prs in a discursive corner, made him abandon his own judgments and after vocalizing ridicule of the Prs the Def suggests his own version of a settlement value which is of completely different kind, not jail but a very low fine. He does that by following the entertainment line of argument, where he invents a new word and then playfully offers a mocking apology. In that sense, he combines entertainment and ridicule of the Prs by playfully and subtly suggesting that he is too narrow-minded and works only with aid of books, laws, and dictionaries.

  • Repetitions

  • Turn-taking—interruption, latching

  • Rhetorical questions

  • Throat clearing

  • Tone of voice

  • Intonation

The feedback utterance “oh” serves often as expression of surprise, thus in combination with its context and the emotions expressed by its intonation the illocutionary force changes into expression of ridicule and evocation of shame.

Apology

The apology is a communicative act with emotional content and function. It aims to express emotions such as sadness, disappointment, and shame directed toward oneself not others. It intends to evoke satisfaction and empathy. In the example below, we have an explicit linguistic reference to this communicative act, “I’m sorry,” but its intent and effect is changed by the lack of the expected emotions, which is indicated by tone of voice and intonation. Playfully sweet and charmingly apologetic (italized), Def is playing with words (“justicy,” “justiciabe”) used earlier by the Prs thus diminishing his importance and in effect mitigating the effect of his claims for justice.

Thus instead of expressing sadness the apology is used to express joy, fun and this is how we get ridicule of the opponent. Locally, it follows the argumentation line directed toward the judge, namely the line of entertainment.

Flattery

By emphasizing his own professionalism (see Section “Empathy Evocation” above) Def is also using professionalism as a compliment to his opponents.

Flattery is a well-known communicative emotion elicitor, presented here in a more serious and structured language, in difference from other moments of entertainment, sprinkled with casual colloquial mannerisms.

Agreeable and Helpfulness

The Def presents himself as helpful when the Prs lacks information on important issue such as how long the defendant spent in jail already. In parallel with the entertainment and ridicule, Def presents himself also as an agreeable negotiation who does what the Prs wants. This agreeable persona is expressed with a reference to the personal name of the Prs who was just made fun of and put in a corner.

This helpfulness is again dominated by the playful entertaining tone, which mitigates the seriousness of the offense and thus works toward minimal judgment. The contrast between this emotion and the aggression and ridicule expressed earlier illuminates the manipulative character of the expressed emotion.

Empathy

Empathy does not always smooth over negotiations. It may also be rejected and thus complicate negotiations. Most of the research on empathy is focused on the ability or skill of giving empathy, but reception of empathy can also be described as a skill and ability since both acceptance and rejection of empathy function as coping strategies.

In this section I explore the linguistic and discursive realizations of empathy with a special emphasis on rejection of given empathy in order to understand its mechanisms and functions in negotiations and thus contribute to the planning and design of empathy training.

Before we examine the function of rejection of empathy giving in negotiations, we will observe shortly the general realization of empathy in a casual conversation between friends who do not have conflicting interests and use that in comparison with negotiation situations where the parties try to overcome differences in interest.

Like any other communicative act, the act of empathy can be elicited (E), given (G), and received (R). The reception may be either acceptance or rejection. One may reject an act of giving of empathy or reject an act of elicitation of empathy. We will study the first case. These functions of the empathy signs may be realized in phases and different degrees. For instance, one may expect the default formulation of a “fulfilled” empathy episode to start with elicitation of empathy, continue with empathy giving followed by empathy receiving (see example 1). One and the same utterance can have all three functions at the same time: It could be an elicitation, an expression, and a response (e.g., line 35 in example 1).

Ritualistic Exchange of Empathy

The following example illustrates the above mentioned distinctions as well as a situation of successful “empathic moments” (Heritage 2005).

  1. (1)

    [Holt Xmas 85:1:4]

The empathy episode starts with an announcement of trouble on line 2. It is welcomed and elicited on line 3. This is followed by a narrative background on lines 4–13. Turn 14 gives the punch line, which elicits empathy, both parallel and reactive, cognitive and emotional. Joy gives her rather emotional empathy on line 15, and Les implicitly accepts it on line 16. Then starts the separation of parallel and reactive empathy. On line 17, 19, 21, 23, 27, and 29 Joy gives a clear example of what is meant by parallel empathy, i.e., she expresses a disapproval of the person by whom Les feels hurt in that way mirroring Les’ dislike of this person’s actions. These expressions of parallel empathy have also degrees; first it starts with a rhetorical question on lines 17 and 19, then the degree rises to clear assessments such as on line 21, and at last we have a assertive (e.g., “honestly,” “I do think”) and explicit formulations of subjective opinion, e.g., lines 23 and 27. Joy’s parallel empathy is predictable and predicted by Les, in fact she motivates (line 24) her expression of a need of emotional support by pointing to Joy’s disposition to the negative feelings they both share against mister R. At that point it is not even clear who gives the empathy, Joy or Les. On line 30 Les expresses her internal distress, which changes the character of the elicited empathy: On the next line 31 Joy performs a good example of the so called reactive empathy. This empathy type is realized here by the use of the generalizing pronoun “you” and by a tag question followed by a confirmative assessment. The tag question is an elicitor of consent, which again turns the roles around: Joy is supposed to be the empathy giver but she often becomes the empathy elicitor as a form of empathy giving. Thus, being both the “empathizer” and the “empathee” is an important capacity in the process of informal discussion of social values and attitudes, all intertwined with associated and even negotiated emotions. On line 32 Joy exchanges the general “you” with a reference to herself, which in a sense functions as voicing Les’ internal discomfort and embarrassment for which she seeks empathy. This voicing is expressed as a quotation of internal dialog. Thus Joy internalizes Les’ inner state, i.e., she displays reactive emotive empathy. On line 37 Joy has completely taken Les’ internal position and talks about her own experiences of the same state of mind Les complains from. Les now functions both as a receiver and a giver of empathy, the process has reached its climax and suddenly on line 39 Joy announces a completely new topic.

The empathy process in example 1 is fulfilled: There was elicitation, giving, and acceptance of empathy, and there was also identification (e.g., line 31), incorporation (e.g., line 35), reverberation (e.g., line 37), and finally detachment (line 39). The verbalizations are at first more emotional and then become more cognitive as they turn to comparisons of experiences. In this empathy process both speakers verify, confirm, and reconfirm for each other the legitimacy of their experiences, values, and attitudes. The sudden change of topic at the end of example (1) and the repetitive turning of the roles in the process of empathizing suggest that the empathy process is rather ritualistic.

Rejecting Empathy

One may expect empathy to always be as successful as in example 1, but empathy is not always accepted which may be as much a source of trouble as lack of empathy. The next three examples illustrate different ways of rejecting empathy in negotiation. In the first example, we have a role-play, in which an US captain is in negotiation with a Spanish-speaking doctor representing a non-governmental medical organization. The captain has to convince the doctor to move his clinic. The captain (C) has introduced the request and now he has to deal with the reaction of the doctor (D).

  1. (2)

    RPSASO’04.1b

In utterances 15, 17, and 19 the doctor repeatedly takes the turn without waiting for the captain to finish his turn; he verbalizes a list of issues and questions which need to be addressed and/or which make a decision difficult for him. At first, on line 14 and 16 the captain tries to address the questions, but in utterance 18 he signals understanding of the function of the questions without awaited answer as a call for display of empathy which he verbalizes in utterance 18. However, even this display of problem-focused (cognitive) empathy is ignored. In 19 the doctor starts right after the captain’s continuous intonation and overlaps with the captain’s continuation. Does the doctor react to the attempt to add a qualification (“but”) or does he react to the expression of empathy? He might anticipate an argument and try to cancel it before it even starts. His utterance on line 19 expresses reasons to mistrust the captain’s empathy giving expressions by pointing to inconsistency of behavior. Also, in the same utterance the doctor himself elicits empathy by reformulating the captain at the end of his utterance “you understand the position I’m in.” This elicitation is more of a response to or a reception of the empathy given on 18 because it is formulated as a declarative sentence with falling intonation. It functions as an argument in the negotiation, as a motivation of reluctance to accept suggestion. In that sense it is a way of facilitating negotiation because it displays desire to be understood. Thus we may tend to believe that the overlap in utterance 19 is a reaction to the display of empathy rather than to the anticipation of an argument. On line 20 the captain responds to the elicited empathy by repeating the elicitation expression of the doctor and reformulating his own formulation in utterance 18. In this way, he attempts to create greater similarity of positions on the negotiation floor. However, he is again overlapped and in utterances 22, 23, 24, and 25 we have simultaneous speech: The captain most probably continues his argument (this part not audible) whereas the doctor signals desire to walk out from the negotiation in utterance 25 by repeating the same utterance twice, once as simultaneous speech and once after winning the turn. This rapid removal from the negotiation is met by the captain with continued display of empathy, which is again completely overlapped by the doctor’s expression of lack of trust and direct criticism in utterance 26. This last utterance is complex because it contains change of strategy and change of phase in the negotiation. The doctor rejected empathy (utterances 15, 17, 19, 21, 23, 25, 27), motivated why (19, 27), displayed desire to walk out (25, 27), threatened with intentions to refuse cooperation and damage planned operation (27) and at last, staring with a topic initiating “now” he stated conditions for further negotiation (27). In this sequence the rejection of empathy functions as a display of lack of trust, as a display of lack of desire to be locked in a disadvantageous negotiation and as a bargaining method.

Eliciting, giving, accepting, and rejecting empathy are thus strategic resources in a negotiation. The actual realization or style of empathy exchange can be part of the strategy as such. In the above example (2) we had a rather aggressive doctor who used time pressure, listing of issues, interruptions etc. to realize his strategy of rejection of empathy which would give him stronger positions in negotiating benefits for his party, because he has no interest in moving unless he manages to gain something substantial for his patients.

Polite Rejection of Empathy

In the next example (3) we have another pair of role-playing doctor and captain, where the doctor is realizing the same strategy (i.e., rejection of given empathy) for the same reasons (namely, increase own benefits and avoid being locked in a disadvantageous agreement). However, here the doctor realizes the strategy in a more polite and evasive manner with even high degree of success because the captain is now truly anxious to satisfy the doctor’s needs.

  1. (3)

    RPSASO’04.1a

The captain is not in a position to get empathy whereas the doctor is. These positions are reflected in the dynamic of the negotiation: The captain gives repetitive displays of empathy and the doctor politely acknowledges them and rejects them at the same time. The captain repeats with increasing intensity the claim that he has “concerns” (line 66, 82, 131). His first concern about accessibility to supplies is performed first by a form of reactive empathy starting with a self-report “I imagine” and guessing the other’s situation utterance and ending with a wh-question which shows care. This first concern and proposal is accepted by the doctor. The next expression of concern occurs after a weak promise to do a personal favor to the doctor: The captain refers to “i i think I have some avenues”; also on line 38 the captain rephrases his utterance from “we can certainly” to “I can certainly” which shows that he is in fact not so certain how he can get supplies and that this proposal is his personal strategy for establishing trust before presenting the real cause of his visit, the mission to remove the clinic. Line 52 presents his purpose and request staring with an expression which mirrors the doctor’s formulation of his needs “well to be honest captain …” namely “and quite frankly.” Thus the captain formulates a bargain in which even the linguistic expressions are even: I can find you supplies if you move close to us. However, if the promise of supplies is formulated as a personal favor the demand to move and “keep save” is formulated as a collective gain: “much easier for us to KEEP you supplied if you were in a safer area among our troops”: The personal pronoun “I” is changed to plural “us” and “our.” At that point, line 54 above the doctor performs the first interruption by voicing with emphatic idiomatic expression “I SEE” his realization of the captain’s negotiation strategy, i.e., realizing that the supplies he agreed to accept at the beginning are conditional and the expressed concern and empathy were strategic, not authentic. At that same point the doctor starts to walk out of the negotiation. His initial trust signaled by acceptance of empathy and offer for help has been broken; from now on every following utterance by the doctor expressed his need to remove himself from the conversation. The captain realizes the lost trust with respect to his real purpose, and he tries to fight for it by explicitly stating his concern and emphasizing the authenticity of his intentions and empathic feelings on line 66 “but REALLY my major concern is is the safety of YOU and your STAFF.” This effort is acknowledged by the doctor in the following way: he does not interrupt; he also explicitly states his acknowledgement and also emphasizes his feelings “APPRECIATE”; he gives authentic reasons why the captain’s demands are problematic by demanding reactive cognitive empathy using strong deontic modal verb expressions such as “you must understand,” after which he implicitly points out what was missing in the captain’s concern, namely the condition of the patients (on line 66 the captain emphasized concern for the staff and the doctor but not the patients, who are the doctor’s main concern and purpose) and in that way again appealing for empathy toward the patients; at last he opens the possibility of future negotiation by promising to discuss the issue with others.

As a response to that on line 82–3 and 115–18 the captain starts to display reactive and parallel empathy (which reminds of Joy’s empathy giver on line 31, example 1) after a sequence of unsuccessful problem-solution-oriented empathic displays (e.g., line 38, 50, 66, etc.). The emphatic “i CERTAINLY understand” on line 82 is a response in strength to the doctor’s deontic appeal “you must understand.” The captain again states cognitive empathy for the doctor’s situation, but he exchanges the term “patients” for the more militant and mitigated term “casualties.” This time he is met by a repetition of the deontic appeal for cognitive empathy presented in an interruption (line 87) followed by a reformulation of the previously stated reasons (line 88). When the captain does not show more empathy but gives promises the doctor evades the negotiation by an implicit rebuttal through the initial “well” followed by a polite dismissal: An expression of gratitude (line 108) and a polite need to leave (line 109). The captain again perceives a need to express empathy, and his following lines are tribute to that: he starts with impersonal formulations of understanding which are lexically emphasized with qualifiers such as “very,” then uses self-report formulations to voice the other’s mental states thus expressing reactive empathy: “I think …  that you,” “I certainly know that you probably ….” The modally strong adverb “certainly” and verb “know” here seem not to be meant to state certainty in the other’s state because they are followed by a modally weak adverb “probably,” but they seem to be meant to amplify the expression of empathy as such. These are followed by declarative guessings of the other’s private desires (line 118) which function as invitations for confirmation of the guessing which is given space in a separate intonation unit on line 119 and filled with a hesitation sound. The doctor does not join this expression of empathy as Les did in line 32, example 1 thus the captain continued by stating new conditions of negotiation. This is met by a next repetition of polite acknowledgment of empathy (“I i again appreciate your CONCERN”), polite dismissal (“so I will look forward to meeting you again”), and polite expression of desire to leave the negotiation (“now uh if you excuse me”) with a new implicit lexical emphasis elicitation of empathy (“very sick little girl”) toward the patients and the time pressure.

Thus in this dialog, the doctor practiced the strategy of politeness, elicitation of empathy and walking out whereas the captain was using the strategy of giving empathy, presenting demands, deadlines, extending his authority, and at the end urging for caucusing.

Antagonistic Style of Giving and Rejecting Empathy

Empathy can be rejected in a more explicit way. In the following example (4) we have an excerpt from a conversation between a patient (P) who suffered a stroke and a nurse (N). The patient has demonstrated anger especially before doing therapy, which he refuses to do. The patient suffers loss of memory, general discomfort, worry for his life, and quality of life. The nurse deals with the patient’s uncooperative behavior. She intends to ensure the patient’s cooperation with the medical personal in the future which she explicitly states in a few occasions during the long conversation. She has introduced the issue after an initial polite empathic chat and on line 65 below we see part of the patient’s explanatory response.

  1. (4)

    Whocares.TALKBANK’04

The nurse is faced with an angry avoidance and rejection of empathy. The rejection here is not realized with interruptions and cutoffs, but it is verbalized as cutoff and explicit rejection (utterance 69 above), confirmations of rejections (71), imperative orders and swear words (73), and rhetorical questions (75) and ironic signals of elicitation of empathy (e.g., initial reference “honey” preceding rhetorical question). The nurse is not offering emotional empathy, and she is not giving cognitive empathy as the captain in example (2). She does not use any of the parallel or reactive empathy expressions we observed in example (1) above. Instead, she uses devices such as ritualistic questions (utterance 66 is an question which promises introduction of news or surprise, prepares the mind of the listeners to something unexpected or undesired but still true), guessing of mental state (“I have a feeling,” “you don’t care,” “I think probably all they want …”), acceptance (utterance 72), personal formats and modal expressions (“I think,” “I know”), mitigators or “softeners” (such as “probably,” initial “well,” final feedback requests such as “huh”) and even rebuts (76). The initial “well” in utterance 76 is typically used preceding partial disagreement and qualification of statement, which has been provoked by other’s utterance and/or understanding of an attitude. Thus the nurse’s display of empathy is antagonistic which reflects her position as a caregiver: she needs to display empathy with the patient’s state but also needs to display commitment to the patient’s medical treatment. The patient’s rejections of empathy are also antagonistic and at first seem to have no bargaining purpose. The patient displays lack of desire to negotiate but also lack of belief in sincerity and true care or at least lack of desire to display trust. In contrast to the previous negotiation where the doctor takes over control of the negotiation, here the patient rejects empathy as a rebuttal but does continue to engage in the conversation (the continuation is not displayed above) and does not interrupt the nurse, which contradicts his linguistic display of no desire to talk. In fact, this conversation continues for quite a while despite the explicit refusals, which suggests that the rejections of empathy do have some strategic value for the patient (which might be the reason why the nurse is reluctant to engage in a more emotive empathy episode).

Linguistic Manifestations of Empathy

Giving empathy is not sufficient to realize empathy. There must be also willingness, ability, and even skill in receiving empathy. In the analyzed examples, rejection of empathy is associated with lack of trust, lack of desire to engage in negotiation and/or with desire to gain control over the negotiation conditions, i.e., as a bargaining strategy. The less trust there is between the negotiating parties, the more unreceptive they are to expressions of empathy and the more strategic for the negotiation the functions of empathy become, as observed in examples (2), (3) and (4). And the opposite, the more receptive the speaker is to empathy the more trustful and smooth the negotiation is, as we could see in example (1).

As a discourse phenomena empathy is complex: It is hard to pick one linguistic feature and tie it uniquely to one function only, but one can observe co-occurrences and patterns, in which multiple linguistic features realize multiple functions in particular sequences.

Rejection of given empathy is realized linguistically by discursive features such as refusal to release the turn, overlaps, interruptions, cutoffs, and simultaneous speech as well as by communicative acts such as explicit rejections, confirmations of rejections, rhetorical questions, imperative orders, irony, swearing, “walking out” moves but also display of reception of given empathy followed by rejection.

Giving empathy, on the other hand, is realized by communicative acts such as answering questions, display of non-elicited empathy, repetitions of elicited empathy, ritualistic rhetorical questions, guessing of mental state, acceptance, rebuts. All these are realized with the help of discourse devices such as personal formulations of modal expressions, quoting, and mitigators or “softeners.” In our data exclamations, extra-linguistic emotional expressions, rhetorical question, assertions, and assessments realize the displays of parallel empathy. Reactive empathy is verbalized as voicing of other’s mental states, comparing inner experiences, and exchanges of generic and personal pronouns.

Elicitations of empathy are realized by narratives, “walking out” moves, repetitive deontic declaratives, quoting, exclamations, laughter, rhetorical questions with prolonged such as “what do you say.”

We may observe sequences of features such as:

  • Rejection of empathy = final – initial overlaps + enumeration of questions → contrastive narrative of other behavior → topicalized declarative descriptions of other’s actions (see example 2)

There are also degrees for realization of empathy in, e.g., giving of parallel empathy:

    • 1st degree: rhetorical question (line 17 and 19, example (1))

    • 2nd degree: assessment (line 21, example (1))

    • 3rd degree: assertive with self-report (line 23 and 27, example (1))

References to authenticity of feeling or intention such as “well to be honest uh” (line 41, ex. 3) and “and quite frankly” (line 5, ex. 3) produced one after the other by both negotiators contribute greatly to the ritualistic proximity searched in negotiation which takes even linguistic expression, i.e., the negotiators tend to repeat each others expressions and even communicative acts, which is one way of signaling closeness or similarity. This proximity would then become a basis for more trust between the parties and thus facilitate acceptance of both empathy and propositions.

Reciprocal Adaptation, ToM Vs AI, and Cooperation Versus Collaboration

This section explores the way in which activity and interaction affect the decision process. In particular, I seek evidence for and against Interactive Alignment Theory and Theory of ToM by observing authentic data. I use again two types of activity, one is an everyday interactive sharing between two friends and the other is an example of strategic interaction, namely a plea bargain negotiation, both from Heritage’s and Maynard’s corpora.

Reciprocal Adaptation in Empathy Exchanges

Emotive Reciprocal Adaptation is manifested linguistically most expressively when Joy offers elicited empathy on line 15 and Les implicitly accepts it on line 16. The example is quoted again for the sake of clarity.

Example (1) [Holt Xmas 85:1:4]

The exclamations on lines 15 and 16 are similar; Les starts her exclamation with a similar sound to this of Joy “oo” and continues with a slight differentiation. In that sense, the speakers align with each other in tone and sound, starting with an imitation, although the functions of the utterances are different. On lines 17, 19, 21, 23, 27, and 29 Joy expresses a mirrored feeling of disapproval of the person by whose actions Les feels hurt, i.e., this is a parallel form of entering into each other’s frame of reference, i.e., of Reciprocal Adaptation through emotional alignment. Line 31 illustrates role-play simulation, which is another linguistic formulation of the Reciprocal Adaptation mechanism. Exchange of roles and positions is part of the reciprocity ritual (see Section “Linguistic Manifestations of Empathy” above).

Line 35, however, is an example of verbalized simulation process, which is in the lines of theory of ToM rather than Interactive Alignment theory. On line 37 Joy has taken Les’ internal position and talks about her own experiences, which is another example of cognitive Reciprocal Adaptation.

The adaptation is at first more somatic, uncontrollable and then becomes more cognitive as they turn to comparisons of experiences and mental representations of experiences. In this empathy process both speakers verify, confirm, and reconfirm for each other the legitimacy of their experiences, values, and attitudes and in the processes they often mirror each other’s verbal actions. We observe Interactive Alignment but also ToM building processes during empathy exchange. Interactive Alignment has diverse linguistic manifestation, e.g., exclamations, tone of voice, tag-questions, parallel and reactive empathy forms. Interactive ToM building manifests at the incorporation and reverberation empathy stages and take the form of explicit reasoning from the other’s point of view through generalized pronouns. Thus, Reciprocal Adaptation in ritualistic informal empathy exchange is realized by frequent Interactive Alignment and a few final more complex interactive ToM processes. The next section studies if the same format of Reciprocal Adaptation takes place in a more strategic and formal activity such as a plea bargain.

Interactive Alignment Versus Complex ToM Reasoning

In comparison with friendly sharing and empathy exchange, a plea bargain is a more formal activity, which involves strategic and tactical interaction. It is a negotiation where participants have conflictual roles and goals. Garrod and Pickering predict manifestations of misalignment in such activities. Would there also be ToM building processes? To answer this question I first outline the structure of the plea bargain and then analyze manifestations of both alignment and ToM processes.

I have studied this plea bargain earlier. Here I will concentrate on the occurrences of Interactive Alignment and interactive ToM reasoning in relation to Reciprocal Adaptation and emotion in negotiation.

Introduction Phase

Already at the very start of the plea bargain the incompatibility of Interactive Alignment theory or theory of ToM is questioned. The judge introduces the topic, asks about the identity of the defendant (line 3) and the defense attorney responds (line 4–6). The prosecutor is not joining the topic yet and throws in another topic, as a subtopic (lines 7, 8).

Example (2) [Frank Bryan]

The question of the judge expresses abundance of empathy: “poor chap,” “all by himself,” light tone of voice, positive rhetorical intonation. The response of the defense counsel aligns with this choice of description by adding sympathy and kindness to the character of his client: “sweet man,” “nice smile,” mild tone of voice. (It is possible and common that perpetrators of crime do not look aggressive, thus how the defendant looks is the least important evidence related to the case.) This is a clear Interactive Alignment and interactive framing of the case, but it is hardly likely that this alignment is primitive, unconscious, mechanistic, and imitative. Just the opposite, if the defense attorney aligns so obviously to the judge’s formulation, it is a clear indication of their ToM of the case and its outcome. Utterance on line 6 topicalizes the preferable “disorderly conduct” (“six forty-seven ef”) interpretation instead of the “arrest resistance” (“one forty-eight”) interpretation and adds a throat clear, which function seems to be to nonverbally mark or problematize the “arrest resistance” story. On the surface, this is grounding though Interactive Alignment using a question–answer pair, semantically similar lexical choices producing cooperatively achieved orientation in the case. However, the emotionally charged formulations by the judge and the defense counsel and the similarity of their descriptions manifest a case-theory-building intentionality or intentionally synchronized framing of the case. Each formulation directs the outcome of the plea bargain, together they are even stronger. The prosecutor is met from the start with two opposing parties, and he is 1:2. This explains the fact that the prosecutor does not align at first, but he does not misalign either; later (see Example (6) below) he monitors the story told by the defense and only at the end of the plea bargain clearly aligns but negatively (on line 211 below) and addresses issues of importance to law, namely that the defendant has a prior related to both resistance to arrest and disorderly conduct. In that sense, already at the start of the plea bargain, we see manifestations of both Interactive Alignment and complex ToM reasoning at the same time, which in their combination express the strategic character of the activity.

Interactive Story Telling and Leading Questions

Story telling is an interactive way of framing, of making sense of a situation or a case. Interrogation is another. In a plea bargain, there is no interrogation of witnesses, but the defense takes the opportunity to frame the problem at hand by telling a story. Below I give the entire phase of the story telling which ends with the decision to settle the case rather than go to a trial. On line 31 below the defense counsel opens the plea bargain with a yes–no question, initiated by a lip parting gesture, emphasizing “s” in “strike”, pronounces “actually” rather informally, and ends with a rising intonation. Then he waits for an answer for a long time, even after the prosecutor’s “hmm:” on line 33, which is not a sign of misalignment; of its tone we understand it expresses lack of knowledge. Since one is asking and the other has no answer, both align in lack of knowledge, interactively. However, when the defense attorney establishes that the prosecutor is not prepared to answer this crucial for the final verdict question the defense attorney takes the privileged opportunity to frame the case through a story, the way he wants to see it, i.e., suitable for his client. Now, did the defense counsel ask an honest question or did he trigger on purpose the fact that there is no violence? The fact is that he repeats the same question one more time on lines 58–59. There the question is formulated as a leading tag-question, it starts with a volume escalation, rather colloquial humoristic lexical choice, it is uttered in overlap, and this time it does not allow or await an answer. This indicates that the defense counsel did not ask a question he did not already have an answer to. On the surface we have alignment of “no knowledge,” but actually we have a planned and explicit modeling of the interlocutor and case framing. The defense uses such a leading question again on line 71, with a tag question and ample time for a response, thus emphasizing it. The fact that the defendant has been in jail already is an important fact for the defense, because it claims that time has been served already although crime is not clearly established. It is important for the judge to memorize that fact as well, i.e., it informs the judge that the man he found to be a “poor chap” has already sat in jail for unknown amount of time, increasing the reasons for empathy. That is, again we see that Interactive Alignment realizes complex reasoning rather than automatic processing, i.e., this question does not seek information, it seeks to strategically show and influence a ToM process, not just Interactive Alignment of information.

Example (3) [Frank Bryan]

The defense attorney’s story telling is not a monolog. It involves: intonation shifts, style shifts, volume escalations, laughter, lexical jokes, rich adjectives, pointed reformulations, other repetitions, emphasis and repetition of crucial for the story aspects (“own house,” “own front yard,” own home, “his castle”), interruptions, latching, initial feedback givers and elicitors, etc. All these are linguistic-discursive means for framing or reframing a problem. The prosecutor adds and corrects, contributing to the interactive framing of the case, which ends up with agreement to settle. However, the moments of clear alignment, such as this on lines 40 and 41 do not support the primary role of mechanic alignment. On these lines, the defense repeats literally the prosecutor’s addition that it is the mother who called the police. He does not do such an other repetition before or after during the framing of the case. He chooses to align there and uses that as a support of his framing, but this is not simply because it is easier to imitate but because it fits his framing of the story, his goal, after explicit modeling of the interlocutors and the effect of the fact, not only the prosecutor but also the judge. He makes a joke out of this fact, which entertains the judge, makes him laugh, thus mitigates the severity of his client’s actions. Thus alignment here realizes a complex strategy using ToM modeling, not automatic processing. The final lines above 94 and 95 exhibit alignment and also realize agreement through other repetition, but again the decision of agreement has come first and then the decision of alignment through repetition, not the other way around.

Meta-Comments as Interactive Manifestations of ToM Processes

Evidence for the interactive ToM building is the “intermission,” the subjectivity directly after the agreement to settle:

Example (4) [Frank Bryan]

The subjectivity is initiated by the judge’s laughing comment on the settlement, line 96. As a response, the defense lawyer starts a story, which explains and describes what happened in lines 1–95 and what will continue for the rest of the plea bargain. He tells a story of a colleague, a famous lawyer, who strategizes interactively in such a way that he always makes the opponent agree to settle after some time of mutual challenging. This meta-comment reveals that the defense counsel’s contributions build on chess game-like premeditated processes, which are often repeated and passed on:

Example (5) [Frank Bryan]

In that sense, this meta-explanation grounds the judge’s laughing meta-comment, but it also frames the plea bargain as a routine, however, not a routine of automatic alignment but a routine of strategic planning based on complex interactive ToM modeling.

Emotion as Manifestation of ToM-Reasoning-Based Strategies; Positive and Negative Reciprocal Adaptation

Reciprocal Adaptation or the procedure by which participants gradually enter each other’s frame of reference may realize in different ways. Rather than claiming that there is resistance to adaptation, the data indicate that there is adaptation even in negative contexts. For instance, in Example 6 below the prosecutor offers facts, which aggravate the guilt of the defendant and indirectly suggest a harsher verdict. He does that after refusing to respond to empathy elicitation by the defense counsel. In effect, the defense counsel interprets the prosecutor’s stance taking as a challenge (204–6) and responds with a sudden explosive expression of anger, contempt and a threat (line 207). This emotional Reciprocal Adaptation takes a form of mirroring: A calm and sober threat to his client’s interests is met with an emotionally loaded counter threat. On top of that, the defense repeats lexically the prosecutor on lines 216 and 220 with mocking intonation. This is good lexical alignment but no cooperation and no success in communication, in the sense of communication as alignment. In the sense of communication as a meeting with otherness this is good communication, because it expresses otherness. Rather than getting into the other’s frame of reference and accepting it the defense counsel gets into the other’s frame of reference, rejects it and, with the emotional display, blames the other party. Thus there is adaptation in negative terms, but it is not in the mechanistic alignment format but rather as a well-planned ToM process, predicting and preventing other’s interactive moves and reactions (as described by the defense counsel himself in the “intermission” phase).

Example (6) [Frank Bryan]

To this explosion of anger, the prosecutor reacts with self-explanation and he stutters and has difficulties formulating a sentence (line 209). The self-explanation on line 209–211 consists of a rejection of the defense counsel’s story or ToM of the defendant using a quotation (“happy go lucky chap”) of the defense’s own formulation on line 51–52, in the beginning of the plea bargain. At the time, the prosecutor did not say much as the defense counsel was telling his view of the story, just answering leading questions and adding details that could change the story but nothing explicit. All of a sudden now, the prosecutor directly rejects the story and the picture the defense counsel drew of the case based on the available evidence. According to Interactive Alignment theory the alignment happens from turn to turn but here we have an Interactive Alignment using negative lexical quotation, which jumps over 100 turns in the conversation. Such instances are rather manifestations of the strategic character of the interaction in this activity, which presupposes conscious ToM building and monitoring. The ToM processing is calculative even on the level of timing: When the relevant evidence if to be mentioned in the conversation appears crucial as to its effect to the goal of the opposing parties.

Locally, on lines 211–213, we do not have adaptation in the form of mirroring but in the form of opposite reaction: The prosecutor adapts to threat and blame by rejection, confusion, defense, reference to contradicting evidence, increase of self-repetitions, pauses, hesitation sounds, and final silencing, i.e., by reframing the problem through an opposite story or ToM frame plus reactive emotional alignment.

The defense counsel continues his ridicule by use of mocking back channels, initial and final interruptions, latching, ridiculing, mocking repetitions, etc. (lines, 207, 208, 216, 220). In this manner, he gains once again a dominant role in the conversation. Emotionally loaded imperative expressions such as “knock it off,” “come on” and vocal gestures such as throat clearing act as more powerful persuasion devices than the arguments. Defense’s sudden and timely anger display on line 207–219 has a successful strategic effect because 10 turns later on line 239 the opponent himself suggests to settle on terms preferred and suggested by the defense.

The linguistic manifestations of the ToM models as minds-of-others context (Martinovski and Marsella 2003; Givón 2005) in this plea bargain are summarized as follows (Table 6.1):

Table 6.1 Linguistic manifestations of interactive contexts as minds-of-others in a plea bargain
Full size table

Similarly to the finding in Chap. 3 in this Volume the juxtaposition of the discursive manifestations of the ToM contexts of the two main participants shows that there is emotional coherence in the performance and experience of each party and that this need of emotional coherence leads argumentation and negotiation, which may go through distortions. While the defense counsel is staging an entertainment performance relying on prior contact with the judge, the prosecutor reacts to this emotional domineering by bringing in the external context of the law. The linguistic features clearly exhibit this ToM contextual contrast. The negotiation goes through number of stages, which are driven by dynamic re-contexualizing of the other’s mind: As the defense attorney presents his client as “a good guy in trouble,” the prosecutor refers to previous record; as the previous record is mentioned, the defense counsel ridicules the idea of a jury trial for “such as small thing,” etc. After a few cycles of strategically emotionally loaded interactive duel the parties end in silence with no resolution:

This sudden interruption brings an unexpected reframing of the situation outside of the judicial and personal/emotive context. Instead of directing attention to the other’s mind as a context, the participants are asked to shift mental attention to a larger context, in which they are all embedded. The explicit shifting of attention from contexts of other’s minds and emotions to larger inclusive contexts offers relief in face-to-face negotiation, which brings out collaborative re-interpretations of the situation at hand.

Based on the empirical observations presented in this section, this chapter continues with a formulation of a model for dynamic re-interpretation and re-contextualization of emotion in GDN.

Model of Emotion in Negotiation and Decision Taking

The Model of Emotion in Negotiation and Decision Taking (MEND) developed here suggests a process-based representation of emotion in decision taking and negotiation and offers operationalization of the process components. Following the definition of emotion presented in Section “Definitions of Basic Concepts” above, physiological reactions may bring about an emotion, which can influence cognitive appraisal, but this appraisal can in turn bring about coping strategies, which generate other emotions.

In this model (see Fig. 6.1) emotions and ToM beliefs (the later summarized as “self” and “other”) can be instrumental in alteration of goals and strategies in the decision-taking process. Their importance is a derivate not only of visceral reactions but also of planning and ToM processing. Each negotiation situation starts with some set of ToM beliefs and goals associated with Self and Other, which relate to a choice of negotiation strategy and tactics realized in the conversation. The decision making is analyzed into negotiation strategies and transaction and interaction goals. These influence the communication process, which is lead by Reciprocal Adaptation, which, as suggested by the theoretical analysis in Section “Theoretical Foundations” and the empirical analysis in Sections “Linguistics Manifestation and Functions of Emotions in GDN” and “Reciprocal Adaptation, ToM Versus AI, and Cooperation Versus Collaboration” above, comprise both Interactive Alignment and complex ToM-based reasoning.

Fig. 6.1
figure 1

A model of emotion in negotiation and decision making

Full size image

During communication feelings and appraisal of gains and emotion bring about coping strategies. These trigger re-evaluation of ToM models (ToMM), goals, beliefs, and strategies, which might be changed. Besides the particular goals, ToMM and beliefs, each negotiation is embedded in a larger existential context, which wraps in all human and other activity. Studies suggest (e.g., Martinovski 2007a, b) that awareness of and reference to co-existence in a larger context facilitate group decision making and negotiation.

Goal Structure

The goals can be interactional, i.e., related to ethics, face saving, ego (see also Chap. 7 in this Volume), and transactional, i.e., related to issues at stake. The interactional and transactional goals can be subdivided into cooperative (win–win), combative (win–loose), and non-cooperative (loose–loose). The negotiation strategies are designed for accomplishing goals and could be avoidance, demand and consent (the Dual Model of Negotiation adds even other goals such as conceding and compromising). The goals are communicated and in the process gains and emotions arise and are appraised, consciously or subconsciously, followed by coping with gains in status and emotions. Coping may result in evaluation of need to change goals and/or negotiation strategies. In the following turn we show how one can distribute goals, tactics, and strategies within MEND (I use as an example of Doctor/Captain negotiation, see Section “Data and Method” and Example (2) RPSASO’04.1b above)(Table 6.2).

Table 6.2 Types of goals and subgoals: functional, strategic and tactical
Full size table

  • Example 0:

  • A: you have to move the clinic, it is too dangerous here.

  • The MEND analysis is as follows:

  • interactive goal: show desire to help someone or/and help someone

  • transactive goal: move the clinic

  • tactics: bonding

  • negotiation strategy: direct demand

  • emotion: empathy giving

Each speaker has a particular set of interactive and transactive goals, which might change during communication. Since the interactive goals often determine the choice of tactics, they are not always stated. In addition, in certain utterances, the transactive goals are more salient (ex. move the clinic), in other’s it is the interactive goals, which are more prominent (ex. it is too dangerous here). There are different types of goals. We choose to model functional, tactical, and strategic goals each of which can acquire different values, i.e., different types of subgoals.

In each interaction, one is dealing with the model of the other’s goals one has rather than with the actual goals of the other. This is true not only for the cognitive organization of a virtual human but also for interaction between humans. The communicative exchange and feedback system involved in it serves to resolve mismatches due to this ToM character of communication. According to MEND emotions can be instrumental in alteration of goals and strategies in the decision-taking process. Here are a few examples of how the model can play out in relation to empathy in an artificial intelligence simulation of a negotiation between a virtual human doctor and a human captain. In that sense, what is available during decision making is only the internal “mental” and “emotional” process of the modeled agent, including its represented interpretations of the other. The ToM goal model is dynamic, i.e., it is generated and created via and during interaction.

Operationalization of the Model of Emotion in Negotiation: Empathy

This section goes through an operationalization of the model of emotion in negotiation in terms of preparation of a rule-based system. Different emotions have different effect on the decision-making process. This operationalization takes as its object empathy because it stimulates negotiation and has great importance for social harmony (Stephan and Finlay 1999; Davis 1994; see also Chaps. 1 and 3 in this Volume), and as it is one of the complex cognitive processes that involve emotion, rational reasoning, understanding, and feeling of the other also on a visceral and somatic level. Similar to other discursive phenomena empathy realizes under certain conditions and has three main functions in discourse: giving, eliciting, and reception, as well as their negative counterparts, namely, refusal to give and rejection to accept empathy.

Conditions for Realization of Empathy

Empathy is realized under certain conditions, which involve factors such as power, credibility, and bonding. Even in the weakest form of credibility, trust is a necessary condition for empathy. Others’ empathy is empowering; this is expressed by the power condition—an aspect of communication, which is not necessarily determined by social status but by general human co-being.

  • POWER:

  • If A feels empathy for B then B has power over A

  • If A feels empathy for B then A has power over B

  • CREDIBILITY:

  • A finds B credible iff A beliefs that B says the truth and A beliefs B is competent.

  • If A feels empathy for B then A finds B credible.

  • CLOSENESS, BONDING:

  • A is close to B iff A shares N values, goals, interest, and/or experiences with B.

  • If A feels empathy for B then A feels close to B.

  • In the following example, A exercises power over B, displays closeness and credibility to B by expressing empathic feelings on personal matters. By accepting them B allows the exercise of power and closeness and implicitly confirms the credibility of the communicated state of affairs:

    • A: I am sorry you could not meet your mother before she died.

    • B: Thank you, dear.

  • Power, credibility, and bonding are disjunctive conditions for empathy, but they are also results of realized empathy. By giving and accepting empathy one increases power, credibility, and closeness.

Functions of Empathy in Interaction

Empathy has different functions in discourse. It can be experienced, given, elicited, accepted, rejected, and refused. Having in mind Allwood’s (1997) conditions for ideal cooperation, namely cognitive and ethical consideration and trust, and the conditions for realization of empathy described below, one may conclude that according to a wider definition, empathy or lack of it may be characteristic of any interactive situation. In the present analysis, one may feel empathy without communicating it intentionally. Lack of empathy does not imply lack of cognitive consideration, but it does signal a lower level of cognitive consideration.

  • EMPATHY:

  • A feels empathy for B iff A adopts B’s ToMM (emotions, goals, feeling, mental states, beliefs) and A feels close to B and/or A finds B credible and/or B allows A to exercise power over B.

  • Giving (EG): When one feels and appraises empathy for the other that may mean that if the goals were combative, one may re-evaluate to change, e.g., permanently or temporarily suspend some or all transactional goals, which are not aligned with the good of the other and change or not some or all interactional goals in such a way that they benefit the other. Empathy giving could be displayed insincerely in case it has a strategic value.

  • GIVE EMPATHY:

  • Acceptance (EA): When during communication phase one feels and appraises that one is accepting empathy, one may also go through goal and strategy revision in such a way that they benefit the other.

  • ACCEPT EMPATHY:

  • Eliciting (EE): When in communication one appraises and feels the need to elicit empathy by the other, if this was not a goal from the start, that means that one changes interactional goals, negotiation strategies and may be even transaction goals in order to receive empathy.

  • ELICIT EMPATHY:

  • Rejection when given (ER): When during communication one appraises and feels empathy one may choose to cope with it by rejecting it which ultimately means that one has added a new interactional and may be even new transactional goals and negotiation strategies and that one is going to keep some or all combative goals. Empathy can be rejected for local reasons but the main causes for rejection of empathy in this model are lack of desire to give power to others over ones self, lack of credibility and/or lack of closeness.

  • REJECT EMPATHY:

  • Refuse to give (RefE): When during coping one rejects to give empathy, one may go through goal and strategy revision in such a way that they combat the other or may keep goals combative.

  • REFUSE EMPATHY:

Levels of Consciousness for Realization of Empathy

Any of emotional process may happen on different levels of consciousness and be realized in interaction on different levels of consciousness: display, signal, and indication (Allwood 1996).

  • Sender:

  • DISPLAY EMPATHY:

A intends to communicate and communicates empathy to B.

  • SIGNAL EMPATHY:

A intends to appear as if A intends to and communicates empathy to B.

  • INDICATE EMPATHY:

  • A does not intend to communicate but communicates empathy to B.

  • Signal is a premeditated level of consciousness in interaction, display is involved in interaction on most common daily level of consciousness, and indication is communication on the lowest level of consciousness.

Summary and Conclusions

Emotion is a complex phenomenon with cognitive and behavioral aspects and manifestations. This chapter studied the role of emotion in negotiation and demonstrated in different ways how discourse analysis can contribute to the understanding of emotion in GDN and how emotion can be an engine of argumentation, driving the wagons of rational thought (Martinovski and Mao 2009). Specific linguistic manifestations of emotional dominance (flattery, sarcasm, ridicule, aggression etc.) function as strategic means for negotiation with different levels of awareness—from lexical choices to tones of voice and paralinguistic expressions. The study suggests that communicative acts can be defined not only in terms of pragmatic meaning but also in terms of emotion, because changes in emotional state bring changes of illocutionary force and the opposite. Analysis of grammatical-pragmatic concepts in term of emotion will facilitate the study of emotion in Negotiation Support Systems, which for the moment do not involve systematic relation between emotion and communicative acts (see Chap. 5 in this Volume).

Problem reframing or negotiation is described as a process affected by interactivity and led by discursive mechanisms such as Reciprocal Adaptation, which can realize as Interactive Alignment or/and complex processing. The type of activity interlocutors involved in predicts the functionality of Reciprocal Adaptation. We noticed that in plea bargains Interactive Alignment realizes complex ToM reasoning rather than automatic processing. We also saw that in the case of empathy we have automatic alignment in the early phase of empathy process and ToM reasoning in the later phase. Discourse features such as other repetition cannot be classified simply as alignment because they can realize complex ToM processes, i.e., understanding of alignment cannot be based on statistical occurrences.

Based on such observations, the paper presents a model of dynamic re-interpretation and re-contextualization of negotiation, MEND (Modeling Emotion in Negotiation and Decision making), according to which emotions contribute to the changes of goals and strategies during negotiation. The purpose of the model is to aid understanding of the role of emotion in negotiation but also to assist artificial intelligence models of negotiation. The operationalization of the model relates adjacent turns and utterances to updates of ToM strategies, transactive and interactive goals, tactics, and interpretations of emotion, either on primary or on appraisal and coping level. A typical example of an emotion, which requires adoption of other’s goals is empathy. Following the empiria, the model suggests that explicit shifting of attention from contexts of other’s minds to larger inclusive contexts offers relief in face-to-face negotiation, which brings out collaborative re-interpretations of the situation at hand. In the MEND model, emotion is a process that regulates Interactive Alignment and the ToM models, which interactants build of each other’s goals, states, tactics, and strategies.

If offered and accepted, empathy may cause adoption of others’ assumed goals or change of own goals and thus enhance decision making (Thagard 2000, see also Chap. 3 in this Volume). The traditional idea of win–win, win–lose, and lose–lose negotiation types is thus put into perspective where these processes are seen as dynamic re-conditioning of negotiation by changes of ToM models driven by emotions. The realization that each negotiation is embedded in a larger context of co-being invites empathy and awareness of common goal/condition. Besides being a cognitive and neural process, empathy is a joint interactive effort in which speakers verify, confirm, and reconfirm for each other the legitimacy of their experiences, values, and attitudes. This verification is of great importance for the development and the function of the individual in the social and discursive world. Being able to take the role of the “empathizer” and the “empathee” is an essential characteristic of the empathic communication. Rejection of empathy may be due to failed recognition of the rejector’s needs and desires, it may have strategic functions gaining momentum in the negotiation or it may be a combination of both. In any case, it is a phase in the negotiation not a breakdown. Elicitation, giving, acceptance, and rejection are functions of empathy any of which could be eliminated of the dialog with consequences. The style of empathy exchange, e.g., antagonistic or polite can also be a strategy in negotiation.

Future applications of discourse analysis of GDN would benefit the understanding of emotion in GDN if they include rich information about nonverbal aspects of face-to-face GDN.