Abstract
This article analyses the orientation of gaze as a significant communicative resource in televised political interviews. The study explores how interviewees use their gaze, in coordination with talk, in receiving and answering adversarial questions. It is guided by conversation analysis (CA), Goffman’s work on gaze in interaction, and the approach on embodied actions developed primarily by Goodwin. Gaze is described as a flexible recipient and speaker resource available for stance-taking, the downgrading and upgrading of actions, and the claiming of the floor. The study is based on taped and transcribed data from two formats of election campaign interviews on Swedish television, including 350 question and response sequences.
Keywords
Introduction
This article analyses the orientation of gaze as a significant communicative resource in televised political interviews. More specifically, the study focuses on actions by interviewees in interviews where adversarial questions play a key role (Clayman and Heritage, 2002; Montgomery, 2008). Two questions are examined: 1) How do interviewees use their gaze as a resource when receiving adversarial questions?; 2) How do interviewees use shifts in the direction of gaze, in coordination with talk, to perform various actions within the overall project of answering questions?
Political media interviews have been explored in detail, primarily in the context of discourse and conversation analysis (Clayman and Heritage, 2002; Ekström and Patrona, 2011; Montgomery, 2008). Today we have detailed knowledge about the specialized organization of turn-taking in this context of communication, as well as the methods used in various practices (Heritage and Clayman, 2010). The research has focused on how institutional identities are negotiated and displayed in political interviews (Heritage and Clayman, 2010; Thornborrow, 2002); how critical journalism and political accountability are accomplished in the organization of talk and the design of questions (Clayman, 2010b; Heritage and Roth, 1995); and also how interviewees employ resources in the service of various actions and deal with challenging situations created in adversarial interrogations (Bull, 2003; Clayman, 2001, 2010a; Ekström, 2009). The detailed analyses of interaction have provided a deeper understanding of the dynamics and activities involved in mediated political communication.
Gaze has proven to be a basic resource in the coordination of talk and the management of engagement in interaction (Goffman, 1963; Goodwin, 1981; Kendon, 1990; Robinson, 1998). However, studies on gaze work in political media interviews seem to be almost non-existent. To some extent, the rhetorically oriented research on body language in political performances is an exception (e.g. Cienki, 2004). The concentration on the verbal design of turns in talk has proved an exceptionally productive line of research. But this focus has its shortcomings, as there is reason to believe that other modes of embodied communication, such as the gaze, are significant for both the overall organization of media talk and the specific actions performed.
It is my aim to contribute to the ongoing research on media talk (Heritage and Clayman, 2010; Hutchby, 2006; Tolson, 2006) and the research on political communication in journalistic contexts (Bull, 2003; Corner, 2003; Ekström and Johansson, 2008; Ekström and Patrona, 2011), by examining gaze work in televised political interviews. The article will explore the organization of gaze as an aspect of what makes an adversarial interview into a specific form of interaction (Ekström and Kroon Lundell, 2011; Montgomery, 2008). The more detailed analyses focus on how interviewees use gaze shifts in dealing with adversarial questions and related communicative challenges. Following Goodwin (2000), gaze is understood as one of several flexible resources available for building actions in interaction.
Questions asked in political interviews often have face-threatening implications (Bull, 2003; Clayman and Heritage, 2002). They suggest interpretations of activities and identities that are not in line with how the interviewees want to be perceived. In the answers, challenges are dealt with by means of what Goffman (1967) describes as face work and face-saving practices. Concrete forms of gaze work may be significant in performing such face work.
Data and method
The study is guided by conversation analysis (CA), and at the same time inspired by Goffman’s work on gaze in interaction, although we should bear in mind that these approaches are not consistent with each other in all aspects. The study is carried out through detailed analyses of transcribed actions in interaction. However, I do not claim that all analyses of what might be achieved in the gaze shifts can be proven just from observations of the exact sequential positioning of gaze shifts, or the understandings displayed by co-participants. This will be further discussed in the Conclusion.
This study focuses on live studio interviews broadcast on Swedish public service television during the 2006 election campaign. In Sweden, the tradition of broadcasting interviews with the party leaders during the election campaigns goes back to the 1960s. These interviews are recognized as a very important form of political communication during election campaigns, and they receive much attention in the media. To some extent, this makes the interviews extraordinary. However, the various forms of adversarial questions, and the situations created for the interviewees to deal with, are very much the same as in other live broadcast political interviews.
In order to have a reasonable corpus to work with, I have chosen to focus on interviews with the two main challengers for the post of prime minister. To avoid tying the analyses to a specific interviewer and interview, I decided to include interviews in two different programme formats. The data thus consist of transcribed interviews with two party leaders in two different programme formats; a total of about two hours of transcribed interaction which included 350 question–response sequences. The gazes examined are the same as those visible to the ordinary television audience. This means that not all gaze shifts are included in the analyses. However, in this study the main focus is on the interviewee, and in all the programmes a close-up of the interviewee is a dominant framing. This is not only the case while the interviewees deliver their answers, but also while they are receiving the questions. Commonly, the producer switches views from the interviewer (IR) to the interviewee (IE) in the middle or near the end of a question turn. In this study I focus on two dominant orientations: looking at the interviewer, and looking down. It should be emphasized that this study does not claim to determine whether participants have eye contact. The data do not permit such analyses and it is not important for the issues to be examined. The concept of mutual gaze is used in situations where the participants look at each other’s faces (see Exline and Fehr, 1982).
This study is explorative. Based on detailed analyses of a number of cases, the article discovers and describes how gaze shifts are used as resources in the service of different interviewee actions. The examples are drawn from a larger corpus containing similar examples, but it remains to be investigated how general the patterns are that these examples represent.
Analytical approach
Gaze has been researched from a variety of perspectives (Exline and Fehr, 1982; Harrigan et al., 2008; Kendon, 1990). This study is anchored in the approach to embodied interaction originally developed primarily by Erving Goffman (Drew and Wootton, 1988; Goffman, 1963, 1967, 1983). The focus of analysis is on the organization and management of situated face-to-face interaction, and the methods used by participants in such situations (Goffman, 1967). Goffman suggests several conceptual distinctions that can help us to specify the analytical focus of such a research endeavour.
A general observation in Goffman’s Behaviour in Public Places is that the orientation of gaze plays a crucial role in the management of involvement and the orientation towards social orders of interaction. Goffman (1963) distinguishes between unfocused and focused interaction. The typical gaze in the management of unfocused interaction is the glance. Focused interaction is defined by Goffman (1963: 24) as ‘the kind of interaction that occurs when persons gather close together and openly cooperate to sustain a single focus of attention, typically by taking turns’. Mutual engagement in a shared focus of attention is a key aspect of focused interaction. This is partly what makes the broadcast political interview into a specific form of interaction.
Goffman (1963) shows that eye contact as well as its avoidance are significant regulating mechanisms in focused interaction. The basic social interactional aspects of gaze observed by Goffman have subsequently been examined in more detail in different research traditions (Goodwin, 1981, 2000; Heath, 1984; Kendon, 1990; MacMartin and Le Baron, 2006; Robinson, 1998). A great number of observations have been presented, but I will argue that most of them can be understood as detailed elaborations of two basic aspects of gaze discussed by Goffman: 1) gaze in reciprocity, expressiveness, and stance in social interaction; and 2) gaze in the regulation and coordination of encounters (see Kendon, 1990).
When it comes to the first category, Goffman refers in particular to Simmel (1969 (1924)) who argues that ‘the eye has a uniquely sociological function’ (in Goffman, 1963: 93). When people look each other in the eye, nearly pure forms of reciprocity and unity can be created. It is about much more than just observing each other. Eye contact can signify mutual relevance, affiliation, recognition and solidarity. These very basic aspects of the gaze have been explored in research that shows that changing gaze direction is a way for listeners to express attitudes toward what others are saying and doing in focused interaction (Goodwin, 1981; Kendon, 1990). In this article I will analyse the coordination of talk and gaze in order to explore how shifts between a mutual gaze and looking down are used by interviewees to take a particular stance towards both the questions asked and the answers delivered.
The significance of the gaze in regulating and coordinating encounters is a central topic of Goffman’s (1963, 1967) work. It is common knowledge that encounters are initiated, participants move from one activity to another, frameworks change and so on, in part through eye-to-eye activities. This general perspective has subsequently been adopted in detailed empirical studies of the turn-by-turn organization of talk and interaction. Gaze direction has been observed to be a principal way of displaying an orientation toward other speakers and negotiating what have been described as engagement frameworks in conversation (Goodwin, 1981, 1984; Heath, 1984; Kendon, 1990; Mondada, 2009; Robinson, 1998; Stivers and Rossano, 2010). For example, Heath (1984) shows that it is partly through gaze work that participants indicate that they are not only present and available, but also prepared to listen and start a sequence of interaction.
Another important source of inspiration for this study is the action and embodiment approach developed primarily by Goodwin (2000). Actions are built, Goodwin argues, through the flexible and dynamic use of various embodied semiotic resources such as talk, body posture, gestures and gaze. These have different properties and encompass different sign systems. In several studies, Goodwin demonstrates how actions might be accomplished through what he describes as the ‘temporally unfolding juxtaposition’ of a number of different resources (2000: 1492).
This study is more limited in that I focus on two resources – talk and gaze – and how they are used by interviewees in the context of adversarial political interrogations. This is an analytical delimitation justified by the fact that I have found that the coordination of talk and gaze is of great significance when it comes to how the interviewees display a reflexive stance in relation to both the questions asked and their own actions in progress. This does not mean that other resources are not at all relevant. Goodwin also shows how the building of actions is related to shifts in participant frameworks and quite complex orientations to relevant phenomena and material structures in the surroundings. This study is also analytically delimited in this aspect in that it focuses on interviewee actions within a rather fixed participant framework, and gaze shifts within a limited frame space. This is explained in more detail below.
The setting and dominant orientation of gaze
In live media talk, gaze work is partly an adjustment to a media format, and what Ytreberg (2004: 682) describes as ‘a process of formatting’. The participants are instructed how to do what the format requires – for example, to look at the interviewer, not into the camera. Their mutual involvement is pre-arranged in relation to the format and the discursive roles of interviewer and interviewee (Ekström and Kroon Lundell, 2011; Montgomery, 2008). The live interviews are also scripted in detail (Kroon Lundell, 2009). However, as is true for the orientation in space, the organization of turn-taking has to be achieved in interaction. In previous studies of the election campaign interviews, Kroon Lundell (2009) has analysed in detail how the interview script is talked into being, and Ekström (2009) has studied the reflexive orientation towards institutional norms and identities, and how rightness and wrongness of conduct is negotiated in the interviews.
The aim of this section is not to analyse in any detail the complex relations between embodied actions, spatial orientation and camera work in live studio TV productions (Broth, 2008), nor to analyse the participants’ reflexive orientation to the context; but rather it is to describe the settings and dominant orientation of gaze as important contexts for the analyses to come. The first programme, Inquiry, is about 58 minutes long. The main part of the programme, about 44 minutes, is organized as an interview with the politician and the two interviewers sitting opposite each other at a fairly high semicircular desk in the studio (Figure 1). This is the interaction analysed in this study. The second format analysed is a 14-minute interview on the Morning Show. This is a distinct section of the show organized exclusively during the election campaign. The interviewer and the interviewee are seated at a small table in a specially designed studio setting (Figure 2). In both cases the interaction takes place in a constrained physical setting and the participants are seated in fixed positions quite close to each other. Distance is an important factor that circumscribes interaction and one’s orientation to others, not least when it comes to the management of the mutual gaze and looking away (Goffman, 1963).
G. Persson interviewed in The Inquiry (the setting).
F. Reinfeldt interviewed in The Morning Show (the setting).
There are several cameras in the studio, and the participants are expected to avoid making eye contact with any of these. At the same time, the mediation and camera work make the interaction into a staged performance in which the face engagement is more in focus than in many other situations of interaction. Close-ups are regularly used (see Figures 3–6). Details of facial expression, gaze and posture are communicated with heightened public visibility.
G. Persson interviewed in The Inquiry (looking at the interviewer).
G. Persson interviewed in The Inquiry (looking down).
F. Reinfeldt interviewed in The Morning Show (looking at the interviewer).
F. Reinfeldt interviewed in The Morning Show (looking down).
The fact that a media interview is an institutionalized activity taking place within a scripted format and a pre-arranged physical setting does not mean that the dominant orientation of body and gaze is completely fixed or pre-determined. It is in principle possible for the participants to direct their gazes at various objects and people in the studio, through either brief glances or prolonged focused looks. However, a systematic study of all observable gazes in the two hours of interviews shows that the interviewees almost exclusively confine their gaze to two areas: to the face of the interviewer, and to what I prefer to call a semi-private space, namely a downward look close to their own body (see Figures 4 and 6). The only minor exception is when the interviewee on two occasions briefly looks up in the air with a more unfocused gaze. The same applies to the interviewers except that the two interviewers on Inquiry also occasionally look at each other. There are actually no examples of the politician looking into the camera in the two hours of interaction. With this system of gaze orientation, the participants construct a rather restricted ‘common interactional space’ (see Mondada, 2009).
The look at the recipient is the dominant orientation of gaze in the interviews. This is independent of programme format or the politician interviewed. The interviewer and interviewee look into each other’s face for approximately 90% of the time, both as speaker and listener. The figures are quite consistent, but with one exception. In the Inquiry, Prime Minister Persson looks at the interviewer less frequently as listener/recipient of questions (71% of the total time as a recipient).
A general conclusion from previous research on interaction in both everyday settings and, for example, doctor–patient consultations, is that instances of mutual gaze within a turn of talk are normally brief (Goodwin, 1981; Kendon, 1967, 1990). In these political media interviews, mutual gaze is not only dominant in terms of the total time of interaction, but situations of extended unbroken mutual gaze between the interviewer and interviewee are also common. Goffman (1963) explored the different patterns of gaze in focused and unfocused interaction. I will suggest a further distinction between focused interactions with a strong or weak focus of attention. In the interviews studied, the participants’ orientation to a restricted frame of space, as well as the extended situations of mutual gaze, contribute to a strong focus of attention. The interviews go on for a rather long time without the interviewee moving from focused to unfocused interaction (see Mondada, 2009). Of course, the question and answer format of the turn-taking is also part of what makes this a specific form of focused interaction. In the following I will analyse how gaze shifts, within the restricted frame of space, are used in dealing with adversarial questions.
The look away as a resource in receiving adversarial questions
In this section I will focus on the look away as a recipient phenomenon. The dominant orientation toward a strong focus of attention makes it relevant to ask what the interviewees can achieve by looking away, and how the look away might be coordinated with talk in the political media interview. More specifically, I will focus on three examples of what I have found to be a common positioning of gaze shifts in the data. There are three characteristics which make these cases similar. First, the look down is performed when the interviewer is in the process of delivering an adversarial question turn. Second, the look down precedes the interviewee’s spoken answer. Third, the interviewee keeps his gaze away from the interviewer until the turn shift is resolved. While the look down is performed in relation to the question turn, the look back is coordinated with the design of the spoken answer.
The analysis below will focus on how the gaze shift is performed and coordinated with talk, and what is accomplished by using the gaze shift. As in other situations of face-to-face interaction (Goodwin, 1981; Kendon, 1990; Robinson, 1998), the recipient’s look away is used as a resource for displaying a stance to what a speaker is saying and doing. Two interrelated aspects of the non-verbal stance-taking in political interviews will be explored in relation to these examples: first, how the stance-taking is related to face-saving (Goffman, 1967), and the politician’s efforts to maintain his position and enforce his view of the topic (which has been described in the literature in terms of enforcement strategies; Gotsbachner, 2009); and second, how the politician deals with the appropriateness of conduct in the stance-taking activity. In media interviews, the accountability and credibility of the politicians are at stake. The interviews are not just about the effective enforcement of political views, but also about avoiding the risk of being understood, for example, as too hostile and disrespectful.
In Example 1 the first question concerns what the politician is ready to do in response to high electricity prices and criticism of the electricity market. Instead of answering the question, the politician treats the question as if it is to some extent inappropriate and indicates that the journalist is asking about something that she should already know about as in line 3, ‘ja har ju . . .’ (‘I have actually . . .’), where the Swedish ju appeals to shared knowledge. The journalist regains the critical initiative in asking a rather hostile and face-threatening follow-up question (line 5). Not only is the politician interrupted, but the journalist also uses what the politician said in his previous answer against him, and asks a rather insinuating question (line 5). The question relates to the popular criticism of politicians that they talk a lot without doing enough to solve the problems. What seems to be questioned is whether the politician has actually fulfilled his responsibility. I will here focus especially on the gaze shifts in lines 5–8.
(The arrows mark the exact position in talk when the gaze in the related picture is performed)
The politician shifts his direction of gaze and looks down (line 5, picture 4) when it becomes possible to anticipate the criticism in the follow-up question. The gaze shift seems to display a stance toward the specific questioning in process. By this I am not arguing that the interviewees typically look down at the exact moment when the criticism first becomes foreseeable. The gaze shift is, however, an available resource for different forms of non-verbal stance-taking, with the positioning of the shift in relation to the question being a component of the action performed. In this example, the look down is not performed at (or just before) a transition relevance point (see Example 2), and not in relation to the turn shift and the journalist’s interruption, but within the first turn constructional unit (TCU) of the question; in other words in relation to a question in progress.
By departing from the mutual gaze (line 5), the interviewee positions himself as less reachable and less vulnerable to criticism. He is still a listener, but a less engaged listener. At the same time, he might also display an opinion of what the interviewer is doing as less important or relevant. The interviewee keeps his gaze away from the interviewer until he starts to clearly express his own opinion (line 8). In this example the shift back is simultaneous with the Swedish word inte (English ‘not’ in the ‘I don’t think’). However, it is not the timing with a specific wording that is important here, but the fact that the shift is performed inside his answering turn. While the look away in relation to the question can be understood as an activity which downgrades the significance of the question, the shift back to mutual gaze instead upgrades the enforcement of his own opinion. This is consistent with previous research arguing that mutual gaze in speaking indicates self-assertion and an ‘interest in seeing the effect of the remark’ (Nielsen, 1964, in Kendon, 1990: 86). The gaze shift becomes significant in the interviewee’s attempt to influence the interpretation of the situation and his political identity.
At the same time the gaze shift is also used in claiming the floor. The look down starts as a recipient action (line 5, picture 4), but at once (picture 5) the interviewee also starts to prepare his answer (before a possible turn completion). He opens his mouth (picture 6) and breathes, and then starts talking in overlap (line 7), when a turn completion is foreseeable. The look away not only displays an opinion about the content of the question, but also a preparedness or even an eagerness to answer, both aspects being part of an effort to maintain a position.
Responses to questions do display different degrees of alignment and distancing (Ekström, 2009; Schegloff, 2007). In designing such distancing as more or less hostile, the interviewee orients himself towards an understanding of the social relationship and the appropriateness of conduct. In Example 1, the interviewee, on the one hand, looks away while the journalist delivers the main part of the question; on the other hand, the look away makes it possible to display a preparedness to answer without causing an interruption. In terms of posture, the look away is also minimal and closely integrated with the answer. The mitigated form is the typical way of performing distancing in these interviews. There are, for example, no cases of politicians openly and obviously turning away from the interaction with the interviewer. However, the look away can be differently positioned in relation to the interviewer turn, in a way that is relevant to what is achieved. This is illustrated in Example 2.
In Example 2 the interviewer delivers a follow-up in which he questions what the politician has argued in his previous answer. What the journalist proposes is that the political orientation described by the politician (not in the extract) does not apply to the joint manifesto agreed with three other parties in the opposition. Questions suggesting contradiction are rather common in these adversarial interviews.
In this example the politician does not look down in the first TCU of the question, when the questioning is possible to anticipate, but after two TCUs and in a possible turn completion (line 2, picture 2). Compared to Example 1, the gaze shift is not positioned as a clear stance towards the specific questioning in process, but is related to a claiming of the floor. However, also in this example the look down makes it possible for the interviewee to accentuate his own answer in a shift back to mutual gaze (picture 4). The look down is indeed a recipient phenomenon, but at the same time it is part of the overall activity of answering the question.
In Example 3 (as in Example 1), the look down is performed in close connection with criticism delivered in the first TCU of the question turn (line 4, picture 2). The gaze shift is done during the preface to the question, at the exact moment when it becomes clear that the poles show negative results for the politician.
In adversarial interviews, question prefaces are frequently used to deliver criticism and to make the question more problematic for the interviewee (Clayman, 2001; Clayman and Heritage, 2002). In this case, part of the problem is that a preferred response, namely an answer to the question, will mean accepting the criticism at the same time. The pause and the ja:: at the beginning of the answering turn (line 5) indicate some problems with answering (typical of a dispreferred response). The politician then copes with the situation by making an announced refusal to answer (Ekström, 2009). But the politician deals with another dilemma as well. An immediate verbal response to the criticism in the preface to the question might, on the one hand, be the most effective way of preventing the criticism from being established as a precondition for the question (Bull and Mayer, 1988). On the other hand, delivering such a verbal response might require an interruption, an action that can be understood as hostile, in other words something that politicians on TV may have good reason to avoid (Ekström and Bérczes, 2008). As in Example 1, the politician uses the look down as a resource to give an immediate response to the preface criticism without interrupting the interviewer (line 4).
Example 3 is also another example of how the look down is dynamically related to a claiming of the floor. While the question is being delivered (line 4), the interviewee keeps his gaze down and leans slowly forward. At the precise moment he starts answering (line 5, picture 5), he turns back and straightens up his body. It is reasonable to understand the shift in posture as part of the non-verbal questioning of the criticism, and at the same time a claiming of the floor (or at least a display of preparedness to answer). The shift in gaze and posture are coordinated in a direct response to the preface criticism, without interrupting the interviewer. At the same time, the non-verbal response gives the interviewee time to prepare his verbal answer.
These examples illustrate how gaze shifts within a restricted frame of space are dynamically coordinated with talk in a response to an adversarial question. The gaze operates in close relation to talk, but as a self-governing resource not restricted to the organization of turn-taking. In Example 3, the interviewee uses the look down to downgrade a criticism directed at him personally (line 4). By keeping his gaze away during two turn shifts (lines 4–6), and throughout his answer, he is able to prolong the downgrading activity. The extended look away is combined with a verbal answer, designed as a downgrading of the opinion polls (line 5). Apart from a quick look back at the interviewer in the follow-up question (line 6, picture 8), the politician keeps his gaze away until he starts to talk about his own policy. The gaze work is used by the politician not only to deal with a face-threatening criticism, but also to enforce his view on the question at the same time.
Again it is important to note that the gaze shifts are significant in how the politician questions the questioning and enforces his own position, while at the same time managing the appropriateness of conduct in a public setting. He downgrades the question with an extended look away, but avoids being too hostile or dismissive. With the very short look back during the follow-up question (line 4), the politician displays a minimal recognition of the critical question, and an orientation towards the question that is neither very enthusiastic nor overly ignorant.
Looking away when answering adversarial questions
In the interviews analysed in this article, it is rather common for the interviewees to look down as a part of the answering turn. As in the previous section, I will focus on examples of a similar positioning of gaze shifts in relation to the organization of turn-taking. What these examples have in common is that a look down is performed in close association with the opening of an answering turn, and that the interviewee shifts back to mutual gaze while he is still in the process of designing the answering turn.
The question then is how such gaze shifts can be coordinated with talk in the turn construction, and what they might accomplish. The analyses suggest that the gaze shifts are used as flexible resources in stance-taking and in dealing with the relative significance of what is being said (whether accentuating it or treating it as peripheral), and that these are important activities which form part of the management of adversarial interviews.
In Example 4, the interviewer brings up an old conflict between the politician and one of his predecessors as party leader (Carl Bildt). This conflict was well known to the media. What makes this conflict particularly interesting from the journalist’s point of view, and conversely less attractive for the politician to talk about, is that in the event that the interviewed politician’s coalition were to win the election, Bildt would be the main candidate for a ministerial post in the government led by the interviewee.
In his first answer (lines 3–5) the politician evades the question concerning his own opinion and feelings. He keeps his gaze on the interviewer. The interviewer treats the answer as unsatisfactory (line 6). He interrupts with a follow-up which narrows the question and emphasizes that it is about his feelings. The follow-up is reasonably understood as face-threatening as it, in a rather blunt way, urges the politician to comment on something personal that he obviously does not want to talk about. I will here focus on how the interviewee deals with this situation and what the gaze shift might accomplish (line 8).
The interviewee receives the entire follow-up looking at the interviewer (line 6, picture 3). The look down (line 8) is then exactly coordinated with the opening of the answering turn. The politician produces a rather short confirmation of his own feelings (‘yes I right’) before going back to his colleague’s feelings. The interviewee keeps the gaze down during an answer that he obviously has some problems producing (‘so so u:h’). He orients himself back to a mutual gaze at the very moment that he clearly states his opinion (‘I don’t know’) in the third TCU of the answering turn.
Compared to Example 3, in which the interviewee uses the gaze shift as a recipient resource for downgrading an adversarial question, the look down in this case is part of the politician’s way of dealing with a face-threatening follow up that he has already received with a mutual gaze. The look down and the verbal design of the answer are coordinated when doing what can be described as a minimizing avoidance in response to a face threat. While keeping his gaze away from the interviewer, the interviewee produces a rather short confirmation, and then cautiously goes back to talking about the other person (lines 8–9); in other words, it is an evasive answer that was previously interpreted by the interviewer as unsatisfactory. The look away downgrades the significance of his answer rather than the question asked (Examples 2 and 3). At the same time he manages to keep himself less engaged with the interviewer when producing the response. While in Example 3 the interviewee challenges the question, in this case he tries to avoid it instead. These are two different ways of dealing with a situation in which the politician’s identity and positive face are challenged (see Bull, 2003; Ekström, 2009; Goffman, 1967).
In Example 5 the question asked to the Prime Minster of the Social Democrat Government relates to a proposal to lower wages and limit employment security for young people. This was suggested by right-wing politicians but also by some influential Social Democrats, and it was highly controversial within the Social Democrat Government.
The politician is asked a yes/no question, in other words whether or not he agrees (line 3), but he is obviously not willing to give a direct answer (line 4). He looks down immediately after the question is asked (picture 2). The gaze shift precedes his verbal turn construction. There is a pause before he starts talking. The look down might be understood as a signal that there is an answer to come, and a way to keep the floor while searching for an answer. However, this example also shows how the look down helps the interviewee to downgrade the seriousness of the question, and the conflict suggested by the journalist. Instead of taking a clear position in a concrete controversial question, the interviewee says that it ‘can be reason to discuss’. The ‘can be’ is done with a slight head movement (picture 3). In keeping his gaze down he constructs a situation of less engagement. He then shifts back to the mutual gaze before he verbally stresses the significance of the discussion ‘that’s certainly true’. What I suggest is not that the verbal design and gaze shifts are intimately linked in these examples. In Example 5 (picture 6) the look back is performed within a TCU, while in Example 4 (picture 9) it is coordinated with the beginning of a new TCU. The shift back to mutual gaze can been undertaken in relation to different parts of the answering turn, but still be a resource in the service of accentuating the politician’s message.
The look away is obviously significant in stance-taking in political interviews. The orientation of gaze expresses differences in level of engagement. A look away is not, however, related to generic meanings. It is rather a complex and context-dependent semiotic resource which it is possible to employ when performing different actions. While in Example 4 the activity in the answering turn is downgraded, in Example 6 the gaze shift helps to upgrade a disagreement. This example also seems to suggest that it is not only the orientation of the gaze within a specific frame of space that is significant, but also the shift in itself.
In Example 6 the interviewer relates the politician’s answer to a previous question and criticizes him for not acknowledging his responsibility for the consequences of his government’s own policy. It is an example of how accountability is negotiated in political interviews. The interviewee coordinates a look down with the opening of his answer (picture 2), and he keeps his gaze down while sharply disagreeing and questioning the relevance of the criticism delivered by the interviewer (lines 6–7). The gaze shift expresses a distancing in embodied interaction, which is combined with a verbal rejection of the criticism. He then uses the opportunity to shift back to mutual gaze while further emphasizing his opinion (picture 5). While the look down is precisely coordinated with the beginning of his answer, the second gaze shift is coordinated with a change in speech delivery, emphasizing the word handel (English ‘trade’) (lines 6–7). The orientation toward mutual gaze is significant in this context partly because it is related to an increase in the focus of attention. Part of the dynamic in the orientation of gaze is the simple fact that, at the same time as a shift is used to construct an action (e.g. a disagreement), a new position of the gaze is established from which another shift is possible, in this case a look back.
Conclusion
Gaze work is part of what makes a political media interview into a specific form of interaction and political communication. Through their orientation towards a restricted frame of space in extended sessions of adversarial interviews, the participants together create a strong focus of attention. Within this common frame of talk there are different projects. The journalist’s extended gaze at the interviewee, while asking face-threatening questions, helps make the interview an occasion of serious and challenging interrogation. By looking back at the interviewer when delivering his answers, the interviewee might be able to enforce his opinions and demonstrate answerability. In this article, the more detailed analyses have focused on how interviewees use gaze work in coordination with talk when dealing with adversarial questions.
Before I summarize and explore some implications of the results, it seems appropriate to comment on the methodological approach. The CA approach has helped this study to analyse in detail how gaze shifts are used as resources in the design of actions in interaction. In order to understand, for example, how gaze shifts display a reflexive stance toward the questions asked and the answers in progress, detailed observations of the position of such shifts in relation to talk are necessary. However, the interpretations of the gaze shifts and what they might accomplish are not based on the principles of CA in all cases. There are two reasons for this. First, I have found that it is fruitful to take account of theoretical ideas of gaze work in interaction, developed outside CA. This relates to Goffman’s analyses of the role of gaze in the management of engagement and, for example, Kendon’s (1990) argument that gaze is a way of expressing attitudes and making self-assertions. Second, and a more methodological argument, what gaze shifts signify and express, and in that sense also might accomplish in the interviews studied, cannot be fully observed in either the positioning of the shift in relation to the talk, nor in an understanding of the shift displayed by the co-participants. This is partly because the data do not include all responses from the interviewer, as I have chosen to analyse the interview as it appears on the TV screen. More importantly, the actions performed by the politicians, including the gaze shifts, are also designed for an audience whose understanding is not part of the interaction. This is why I have found a combination of CA-based empirical observations, and interpretations based on the researcher’s understanding of what is expressed by a certain gaze shift, as the most fruitful approach.
This study has focused on two common ways of positioning gaze shifts in relation to the organization of turn-taking: 1) a recipient look away performed in relation to the interviewer turn and maintained until the turn shift is resolved; 2) a look away in close connection to the opening of an answering turn. I have explored the dynamic use of gaze shifts within a restricted frame of space. The politician’s shift from mutual gaze to looking down is not just a way of resting from the concentration required to maintain a focus on the interviewer during an extended interview. It is a flexible recipient and speaker resource available for stance-taking, the upgrading and downgrading of actions, the regulation of engagement, and claiming the floor – actions that really are significant from the interviewee’s point of view. In the analyses, I have for example shown how shifts in gaze are used as a dynamic resource to downgrade face-threatening aspects of the questions and upgrade the interviewee’s own opinion. As a non-verbal form of stance-taking, the look down makes it possible for the interviewee to immediately respond to criticism without interrupting the interviewer. The gaze shifts are used in undertaking mitigated forms of distancing and questioning of the question.
As Goffman (1963) notes, individuals may have good reasons to present themselves as available or unavailable for face engagement. Remaining outside can be a way to avoid being exposed to criticism or asked to do what one does not want to do. In a media interview it is most often not a reasonable alternative for the interviewee to even temporarily leave the interaction. This analysis shows, however, how gaze shifts are used to regulate the level of engagement even within an interaction characterized by a strong focus of attention. This is consistent with Goffman’s (1981) general approach to the multi-dimensional role of the listener. Different kinds of listeners (more or less engaged, available or affiliated) are partly established through gaze work.
Performing as a competent interviewee in live political interviews is a communicative challenge. The mediation puts the embodied activities in sharp focus. Questions are rigorously planned in advance, and in many cases the overall rationale is that they should be as tricky as possible to answer. Previous research has described in detail how communicative conflicts are managed by interviewees in the verbal design of various forms of answers and non-answers (Bull, 2003; Clayman, 2001; Clayman and Heritage, 2002; Ekström, 2009). In this article I have tried to demonstrate the importance of expanding the research on media interviews to include relevant aspects of body language. As Goodwin (2000) argues, actions are built through the simultaneous, flexible and dynamic use of different semiotic resources. Part of the flexibility of gaze shifts is that they are self-governing, but it is possible to combine and coordinate them with talk in many different ways. This study shows how gaze is used in building interviewee responses, which start much earlier than the beginning of the verbal answer.
While this study has been focused on the coordination of talk and gaze, in future research we will also need to explore other resources (gestures, posture, etc.) utilized in the various activities during political media interviews. There is also much more to explore when it comes to the significance of the gaze in political media interviews. This study has focused on a limited number of situations and activities in a specific form of adversarial interview. Comparison between different forms of interview formats would also increase our knowledge.