Threat & Equivocation in Political News Interviews: An Analysis of Interviews by Fox News with Obama & Trump

(1)

Threat & Equivocation in Political News Interviews:

An Analysis of Interviews by Fox News with Obama & Trump

MA Language & Education

30 June, 2017

(2)

Abstract

The present study investigated equivocation in political news interviews by analyzing interviews by Fox News with Obama and Trump. This study has specifically focused on the ways in which Fox News questioned its interviewees and the ways in which the interviewees responded. This will provide more insight in a possible relationship between the way of questioning and the political affiliations of the interviewer and the interviewees. It will also show whether there is a relationship between the levels of threat questions can pose and their following replies. Additionally this study investigated the questions and answers before Obama and Trump were elected and after they were elected to see if there were any differences in the types of questions before and after having been elected. This study found that overall, Obama received more threatening questions than Trump, while Trump equivocated more. This study also found an increase in equivocal replies after the elections for Obama and a slight decrease in equivocal replies of Trump after the elections. This study found evidence that suggests that the political affiliation of news interviewers affects their way of questioning, and that they ask more threatening questions to interviewees with opposing political views. Although some evidence was found for a relationship between the level of threat of a question and the reply that follows, Trump unexpectedly equivocated more than Obama. A possible explanation for this could be that levels of threat are interpreted subjectively, and that Obama felt less threatened than Trump, for instance. No evidence was found that suggests that there are differences between interviews before the elections and the interviews after the elections.

(3)

Acknowledgements

I would first like to express my gratitude to my supervisor Manon van der Laaken. Without her comments and support I would not have managed to complete this study. I would also like to thank Suzanne and Michel for their unwavering support, and Frank for keeping me fed, caffeinated, and sane during the last few weeks.

I hereby declare that I have read all of the UvA guidelines on thesis requirements and plagiarism. This work is all my own and all the works that have been used are accounted for in the reference list.

(4)

1. Introduction

Although much research has been done on equivocation in politics (Bavelas et al. 1990; Bavelas et al. 1988; Bavelas, 1983; Bull et al. 1996; Bull, 2008; Bull & Mayer 1993; Bull & Simon-Vandenbergen, 2014; Clementson, 2016; Clementson & Eveland, 2016; Wodak et al. 1999/2009; Edwards & Bello, 2001; Ekman, 2009), little is known about what types of questions cause politicians to equivocate. Much research exists on the potential threats questions can pose in political news interviews (Clayman, 1988; Gnisci et al., 2013; Strömbäck & Shehata, 2007; Rendle-Short, 2007; Clementson, 2016; Clementson & Eveland 2016; Bull et al. 1996), however, this study found no previous research that combined an investigation of the questions and the replies in political news interviews. The present study therefore aims to fill this gap, and combine an analysis of the questions and the replies in political news interviews to see if a relationship exists between them.

This study investigates threatening questions and evasive replies in political news interviews with Obama and Trump. The main interests of this study are finding out whether Obama and Trump receive different types of questions and whether these questions differ in levels of threat. Additionally this study aims to find out if there are any differences between the types of responses Obama and Trump provide. This study looks specifically at equivocal responses, and aims to find out what types of equivocal replies Obama and Trump provide, and what types of questions cause them to equivocate. Additionally this study looks at differences in equivocal replies in terms of frequency. The present study only takes interviews with Fox News into consideration to see if their political affiliation affects their way of questioning. The present study selected interviews by Fox News as this is a new channel that has been accused of having biased media coverage in the past. This study aims to find out whether Fox News asks different types of questions depending on who they are interviewing: a fellow Republican (Trump), or a Democrat (Obama). By analyzing both the questions and the replies, this study aims to show whether a relationship exists between the threatening questions the Obama and Trump receive, and the replies they provide to these questions. This study will also look at interviews from before and after their elections and compare the equivocal responses in these interviews to see if the equivocal behavior of Obama and Trump changes after the elections. No previous studies were found that compare questions and replies in interviews before and after the elections. Additionally, this study has not found any previous research that compares one news channel that interviews politicians of opposing parties.

(6)

This study first categorized all the questions according to their 3 levels of threat. This study has mostly used Clayman & Heritage (2002) and Brown & Levinson (1987) to determine the levels of threat of a question. This also study categorized all the different types of threat questions can pose and distinguished between 7 different types of threat. This study mostly used categories that were developed by Bull et al. (1996) (section 3.4). For the analysis of the questions, this study made use of categories that were developed by Bull & Mayer (1993), and developed 9 different types of equivocal responses(section 3.5).

Lastly, Montgomery (2008) and Fairclough (1989) have shown that political news interviews are usually accountability interviews in which interviewer and interviewee adhere to an institutionalized turn-taking system with pre-allocated roles and turns. This makes the interviewer in news interviews the more powerful participant, and it allows him to control the topics. Section 3.3 will therefore show how the present study also investigated the different threats and replies in the addressed topics.

The next chapter of this study will discuss findings of previous studies that are relevant to the present study. Chapter 3 will explain the method behind the present study, and chapter 4 will present the findings of this study. The last chapter will discuss the findings of the current study and conclude with recommendations for future research.

(7)

2. Literature Review

2.1 Introduction

This chapter will provide an overview of previous studies that are relevant for the present study. The first section of this chapter will discuss the notion of “face” and politeness strategies, and is necessary background information to understand reasons for equivocation. The second section deals with political news interviews, with a specific focus on stance of the interviewers, potentially threatening questions and required responses of interviewees. This second section will provide more insight into the various ways in which questions can be threatening. The third section of this chapter will look at equivocation theory in detail and discuss reasons for equivocation and different ways in which people can equivocate.

2.2 Face

2.2.1 Explanations of Face

The term equivocation is usually broadly defined as an evasive or ambiguous response to questions that have the potential to threaten someone's social self-image, often referred to as “face” (Bavelas et al, 1988; Bavelas, 1983; Wodak et al. 1999/2009; Bull & Mayer 1993; Bull et al., (1996); Bull, 2008; Bull & Simon-Vandenbergen, 2014; Clementson, 2016). Two key studies on the notion of face describe it as the mutually negotiated social self-image that all individuals have prior to entering a conversation; participants bring their self-image to the conversation which in turn needs to be accepted by their interlocutor (Goffman, 1967; Brown & Levinson, 1987). Additionally, participants aim to maintain their own face as well as protect the face of their interlocutor (which also contributes to the concept of face as mutually-negotiated) (Goffman 1967).

Brown & Levinson (1987) and Goffman (1967) show that face is something that can be lost or gained in a situation. According to Brown & Levinson (1987) people employ politeness strategies to avoid threatening the face of their interlocutors or their own face. At the same time they can also attempt to gain face in a conversation. A more recent study on the construction of face shows that damage to face, or losing face has a greater impact on individuals than gaining face (Zhang, Cao & Grigoriou, 2011). Although face can be gained or regained after a loss of face, participants in a conversation are more preoccupied with avoiding loss of face, rather than possibly gaining face.

Brown and Levinson (1987) distinguish between two types of face. On the one hand there is positive face, which can be defined as a desire for inclusion. This involves an

(8)

individual’s want to feel equal to other individuals, and their want to be appreciated and respected by others. Positive face is often what is at stake for politicians in news interviews. Politicians want to be elected, or they want their ideas to be accepted by the people which means that they want their positive face to be accepted by others (Bull et al., 1996). On the other hand, there is negative face, which is the desire for independence: people’s want to be unimpeded in their actions (Brown & Levinson, 1987). Negative face is less relevant to the present study. The pre-allocated turn system of news interviews forces interviewees to reply, which impedes them in their actions (Bull et al., 1996). The interviewee’s want to be unimpeded in their actions is irrelevant in news interviews, as they have already agreed to adhere to this interview structure beforehand (Bull et al., 1996) (see section 2.3.1). Negative face is therefore less relevant in the study of political news interviews, and this study will mostly focus on the interviewee’s want for inclusion: positive face.

Separate attention should be given to the notion of face in politics. Although face is a concept everyone has to deal with, the concept of face is different for politicians. Various studies have shown that politicians have a higher risk of losing face since they are public figures who are often watched by an audience (Bull et al., 1996; Jucker, 1986). Another reason politicians are more at risk of losing face is because political news interviews questions are designed to hold them accountable, which increases the risk of loss of face (Greatbatch, 1988; Clayman 1988; Montgomery, 2008; Jullian, 2011; Gnisci et al., 2013). Instead of employing politeness strategies, interviewers often drive politicians into a corner by asking them face-threatening questions, and politicians also risk losing more than one type of face (Bull & Mayer, 1993; Bull et al., 1996; Jucker, 1986; Tracy, 2008; Clayman & Heritage, 2002; Elliott & McDonald, 2007; Harris, 2001 etc.).

Politicians are not only at risk of losing their personal face, but they could also lose the face of their party, or the face of their political allies. (Bull et al., 1996; Montgomery, 2008). Not only do politicians have a high risk of losing face, but they also need to make an effort to gain more face than their opponents (Bull, et al., 1996; Catellani & Covelli, 2013; Clementson, 2016). Section 2.3 will elaborate on face-threats in political news interviews.

2.2.2 Face Threatening Acts

The previous section dealt with the notion of face. This section will elaborate on ways in which face can be threatened. Brown & Levinson (1987) define face-threatening acts (henceforth FTAs) as “acts that by their nature run contrary to the face wants of the addressee and/or of the speaker” (65). To ensure the protection of face of all participants in a conversation, conversationalists usually try to minimize the use of FTAs, and when doing

(9)

FTAs, they aim to minimize the threat (Brown & Levinson, 1987). Brown & Levinson (1987) show that the level of threat of an FTA depends on the social distance between the speaker and the addressee, the power relations between the speaker and the addressee, and the weight of imposition of a particular statement in their culture. The more significant the social distance, the more polite the request will be to avoid imposition on the addressee. Similarly, the more power one’s interlocutor has, the more polite the speaker will be. Strategies for minimizing the level of threat of an FTA to save face are called politeness strategies. Further details on how these are employed and about their significance will be provided in section 2.2.3.

A study on FTAs in romantic relationships found that the perceived levels of FTAs can be subjective (Zhang & Stafford, 2008). They found that the perceived level of threat often depends on the esteem of the person on the receiving end. People with lower self-esteem were far more likely to perceive a message as threatening than people with a high self-esteem. Although FTAs in romantic relationships are different from FTAs in political news interviews, it is possible that interviewees in political news interviews will interpret threats differently depending on their respective self-esteem. This could also cause differences in the types of replies they provide.

Another important aspect to consider regarding FTAs is power relations. Although people often attempt to save their own face and their interlocutors face, FTAs can also be used to exert power over someone and to make them feel uncomfortable for one’s own gain (Goffman, 1967; Brown & Levinson, 1987; Montgomery, 2008; Chang & Haugh, 2011). Chang & Haugh (2011) investigated strategic embarrassment in business situations and show that speakers sometimes embarrass the addressee purposefully to make them feel guilty. Speakers sometimes embarrass the addressee to indicate that they do not feel that the addressee has met their expectations in the relationship, and that they do not feel that the addressee values their relationship. Chang & Haugh (2011) show that if a speaker claims dissatisfaction about the relationship, the addressee is likely to feel guilty and pressured into resolving the problem. In such a situation, the speaker is exerting power over the addressee by making the addressee feel as if they should have made a bigger effort in their relationship. Although the nature of the relationship between business people is different from the relationship between interviewer and interviewee, this study is relevant to the present study as it shows how participants can do FTA’s on purpose rather than protecting someone’s face.

Montgomery (2008) distinguishes between 4 types of news interviews, of which the accountability interview is one. In these interviews, interviewers hold their interviewees accountable for current events, and, similar to the findings of Chang & Haugh (2011),

(10)

interviewers often do FTA’s on purpose. In accountability interviews the interviewees are mostly politicians who have to account for their actions in broadcasted news interviews (Montgomery, 2008). Other studies have also shown that in political debates and news interviews politicians are often put on the spot by other politicians or by interviewers (Clementson, 2016; Ekman, 2009; Tracy, 2008; Montgomery, 2008). Clementson (2016) shows how politicians in political debates do FTA’s when they accuse other politicians of lying or providing an evasive response, even when they are actually responding to the question. Clementson (2016) also states that because of accusations like these, politicians are often perceived as more evasive than they actually are. Clementson (2016) shows people are generally distrusting towards politicians. Because of this distrust, people who watch political debates are often inclined to believe an accusation of evasiveness regardless of whether the accusation is true or not (Clementson, 2016). The study also shows that these accusations enhance the face of the politician who makes the accusation, as the people are more likely to believe him than the accused politician. These accusations are therefore often employed strategically to enhance one’s own face (Clementson, 2016).

Tracy (2008) investigated communicative conduct in local administration meetings. Like Clementon (2016), Tracy (2008) shows that people do not just do FTAs, but they sometimes go even further and deliberately insult their conversation partners (Clementson, 2016; Chang & Haugh, 2011). According to Tracy (2008), most of these attacks occur on a personal level, and most of them are implications. Tracy (2008) shows that one of the participants in a school board meeting states “I am surprised that school board members including one with a PhD in education would misunderstand the purpose of this test” (180). The participant here clearly shows dissatisfaction with the test by implicitly insulting the intelligence of the rest of the board. Similar to other studies, Tracy (2008) here shows that speakers sometimes do not aim to protect the face of their interlocutor, but that they do FTAs on purpose and attack their opponents either to enhance their own face, or as in this case to express their dissatisfaction on a particular matter.

While Brown & Levinson (1987) show that it is expected that people in a conversation aim to save their own as well as their interlocutor’s face, this section illustrated that people sometimes violate this expectancy and threaten someone’s face on purpose and that they are able to do FTAs strategically. Chang & Haugh (2011) have shown how people do FTAs to make people feel guilty about their efforts in a relationship, and Montgomery (2008) has shown that FTAs are embedded in questions and that these FTAs can be done to hold the interviewee accountable for his actions, or to express dissatisfaction. Since studies on equivocation show a relationship between FTAs and equivocal responses (Bavelas, 1988;

(11)

Bavelas, 1990; Bull et al, 1996) this study will analyze the possible threats questions can pose to gain a more precise understanding of the relationship between questions and answers.

2.2.3 Face and Politeness

This section will elaborate on how threats to face can be minimized. As discussed above, Brown & Levinson (1987) distinguish two kinds of face; positive and negative face. A similar distinction exists between positive and negative politeness. Speakers often try to soften the blow of an FTA. To minimize the impact of an FTA speakers can use strategies of politeness to mitigate the FTA. Speakers can use a strategy of positive politeness in which speakers communicate that their own wants are similar to the addressee’s wants, or in which they try to satisfy the addressee’s positive face (Brown & Levinson, 1987). They can also use a strategy of negative politeness in which they try to impose on the hearer as little as possible. Strategies of negative politeness are mostly relevant to the present study. Brown & Levinson (1987) show that FTAs can be mitigated by being indirect, and by hedging the question. Indirectness reduces the weight the of imposition that the request can have on a speaker and allow the speaker to decline the request without losing face (Brown & Levinson, 1987). Hedges are elements in a sentence that weaken the blow of the question or statement. Words like sort of, maybe, perhaps, in a way, might, believe, think, for instance, show that the speaker does not assume that the hearer would be willing to comply with the request. These hedges can also create indirectness as a strategy of politeness (Brown & Levinson, 1987). In interviews this would involve questions like “Would you perhaps be willing to elaborate?”. Both “would you be willing to” and “perhaps” weaken the message of the question’s content: “Can you tell me more?”.

A few more of Brown & Levinson (1987) strategies of negative politeness to mitigate an FTA are important to the present study:

 Use hedges to create vagueness or indirectness in a request or reply  Use of hedges: Don’t impose on hearer’s negative face

 Minimizing the weight of imposition (“I would just like to ask…”)

 Approaching the request from a pessimistic angle (present request as improbable to be granted: “could you” or “would you be willing to” instead of “can you?”)

 Apologizing for the imposition (“I’m sorry to bother you” or “I know you are really busy”)

 Stating that the speaker does not want to impose on the hearer (“I know I am asking a lot”)

(12)

 Using passives instead of actives to distance the hearer from the situation

Other studies add to Brown & Levinson (1987) and show that FTAs are often approached with indirect questions by the speaker to avoid adding to the threat. This indirectness in questions serves as a strategy to mitigate the FTA, since asking questions in an indirect manner imposes less on the negative face of the hearer, and thus minimizes the weight of the imposition (Holtgraves, 1991; Edwards & Bello, 2001). This is further illustrated by a more recent study (Holtgraves & Perdew, 2016), in which participants show higher rates of politeness in more threatening situations. In this study, participants had to show how they would communicate potentially threatening information in a number of situations that differed in levels of threat. The results of this study show that the level of threat in the information greatly affected the way in which this information was conveyed (Holtgraves & Perdew, 2016). Participants who had to give negative information were more careful in conveying the information than participants who had to provide positive information. This shows that the participants of this study employed politeness strategies to mitigate the FTAs they had to make. Since political news interviews are usually accountability interviews (Montgomery, 2008), FTAs are frequently embedded in the question. It is possible that Fox News would want to mitigate some FTAs for Trump as they are both Republicans and have similar political views whereas this would be unlikely for the FTAs Obama receives as he is a Democrat and has opposing political views.

2.3 Political News Interviews

2.3.1 Design of political news interviews

The turn-taking system of institutionalized conversations such as news interviews and the turn-taking systems of natural conversations are very different (Greatbatch, 1988; Fairclough, 1989; Montgomery, 2008). In naturally occurring conversations turn-taking is locally managed, whereas in institutional settings like news interviews, turns are pre-allocated and fixed (Greatbatch, 1988). All turns within the format of political news interviews have to be minimally recognizable as questions and replies (Greatbatch, 1988). Additionally, interviewees usually do not take a turn until the interviewer is clearly done speaking, and they refrain from initiating turns (Greatbatch, 1988). Within this pre-allocated turn-system interviewers often have their questions ready before the interviewee has even entered the room. Both interviewer and interviewee enter the conversation with the standard question-answer format in mind and they are likely to adhere to its structure of pre-allocated turns (Greatbatch, 1988; Montgomery, 2008). Although the interviewer can diverge from the

(13)

questions on his list, the fact that an answer will follow has already been pre-determined. This shows that interviews are conducted in an institutionalized framework of pre-allocated questions and answers. Because of this institutionalized format of pre-allocated turns, interviewer and interviewee have to enact their prespecified roles, which means that an interviewee has to answer the question to avoid looking incompetent (Greatbatch, 1988). Greatbatch (1988) also shows that interviewers can accomplish a number of things within each turn, such as putting pressure on the interviewee, or challenging their opinions. This is always done, however, within the framework of the question-answer format.

Another important aspect in political news interviews is the notion of “stance”. The position of the interviewer is often referred to as stance. Interviewers are required to have a neutralistic stance in their interviews as the interviewers themselves are not the intended audience (Greatbatch, 1988; Rendle-Short, 2007; Bull, 2008). It is the role of the interviewers to obtain information for an audience and to make them feel like the primary addressee (Greatbatch, 1988; Rendle-Short, 2007). Interviewers accomplish this by adhering to their pre-allocated turns and roles as described above. This involves withholding continuation tokens like expressions of agreement or surprise, for instance (Greatbatch, 1988; Montgomery, 2008). Although they are required to have neutralistic stance, interviewers have developed ways of covertly showing their support or disapproval while conducting news interviews (Jullian, 2011; Greatbatch, 1988; Clayman, 1988; Gnisci et al., 2013; Strömbäck & Shehata, 2007; Rendle-Short, 2007; Montgomery, 2008). Jullian (2011) shows that news reporters can covertly voice their opinions by using quotes of authority figures. First, they can use a quote to voice their opinion. They cannot be held accountable for the statement, however, as it is a quote by someone else and they can deny any association with the statement. Second, quotes from authority figures give the interviewers the ability to highlight a specific side of a story (Jullian, 2011). Third, interviewers can quote an authority figure before asking a question. This can give extra weight to the questions as it shows that someone who might have more expertise on the matter has expressed an opinion about it. The interviewer can thus select quotes that put current events in a positive or negative light to covertly voice an opinion or they can quote statements by others that also covertly carry their own opinion and deny association with it.

Clayman (1988) shows a variety of ways in which news interviewers can be covertly critiquing the interviewee such as creating puzzling questions that make the interviewee look like a liar, or incompetent. For example, interviewers can embed several smaller questions into the main question. When the interviewees reply to several of these questions, but not all, the interviewer can already accuse them of being evasive (Clementson, 2016; Rendle-Short,

(14)

2007). Clayman (1988) shows that interviewers also often ask questions within a statement. These statements often carry the interviewer’s opinion, and direct the interviewee towards a preferred answer. For instance, a question like “But it was a horrible situation, wouldn’t you agree?” points the interviewee towards the desired agreement with the statement (Clayman, 1988; Rendle-Short, 2007). Montgomery (2008) has shown that similar agreement from an interview with Blair resulted in headlines stating that Blair claimed involvement in Iraq was a disaster. This shows that these embedded statements are an effective way of covertly taking stance and directing the interviewee towards a desired answer. If, as with Blair, the interviewee is indeed directed towards a desired answer, the interviewers can hold them accountable for the elicited reply (Montgomery, 2008). Lastly, political interviewers sometimes use a politician’s own words against them and aim to make them contradict themselves (Rendle-Short, 2007; Clayman & Heritage, 2002).

A last aspect that needs to be taken into account regarding the design of political news interviews is topic control. Fairclough (1989) shows that the more powerful participant in a conversation is the one who determines the topics in an institutionalized setting of conversation, such as political news interviews. Interviewer and interviewee have prespecified roles in the conversation, and only the interviewer can address new topics. Contributions of the interviewee that are considered irrelevant to the topic are disallowed by the participant in power (the interviewer) (Fairclough, 1989). This suggests that an interviewee who introduces a new topic in political news interviews is off-topic since the institutionalized frame of conversation does not allow the interviewee to switch between topics. The present study will analyze the different topics that occur in the interviews which will show if certain topics are addressed more frequently, and whether certain topics contain more FTAs.

The following section will look at previous studies who have investigated FTAs in questions in political news interviews and it will discuss how interviewers can do FTAs in these questions.

2.3.2 Threats in Political News Interviews

The previous section shows that the adopted neutral stance of interviewers in political interviews often carries covert opinions. This section will elaborate on how interviewers adopt a neutralistic stance during interviews, while covertly expressing their political affiliation in their questions. This seemingly neutral stance of interviewers can present potentially face-damaging situations for the politicians in the interview. Although politicians often come across as evasive and perhaps untruthful, several studies claim that politicians

(15)

often have no other option than to equivocate because interviewers or other politicians often ask threatening questions, thus making equivocation almost unavoidable (Bavelas et al., 1990; Bull, 2008; Clementson, 2016). Various studies have aimed to map out the possible ways in which a question can be threatening to the interviewee. Bull et al. (1996) developed a way to categorize all the possible ways in which a politician’s face could be damaged. Each FTA can be categorized as either a threat to the politician’s personal face, to his party’s face, or to the face of his political allies. Within these categories Bull et al. (1996) developed as many as 19 sub-categories for all the types of threats an FTA could pose. The present study used various categories from Bull et al. (1996) however, some of these FTAs did not occur in the interviews in this study, and some others could be fused together. This will be further explained in section 3.4. The following categories were developed by Bull et al. (1996) and were also some of the most commonly found categories in this study:

 Questions that create or confirm a negative image of the politician  Questions that make it difficult to give an opinion and make the politician seem clueless



Questions that create or confirm a negative image of the party and its policies

 Questions that suggest the politician does not support the electorate  Questions about support for a colleague or opponent

It is important to note that none of these subcategories are mutually exclusive, and one question can thus pose multiple threats (Bull et al., 1996).

Similarly, Clayman & Heritage (2002) made a division of four different categories of threat in political interviews, namely initiative, directness, assertiveness and hostility. Within these categories Clayman & Heritage (2002) distinguish between the various ways in which questions can be threatening and face-damaging. A very long and complex question is often a greater threat than a short question, and a direct question can be more threatening than a question in which politeness strategies are used. An overview of these divisions is provided below:

Initiative:

 The length and complexity of a question. In many cases interviewers ask multiple questions in one turn:

(16)

funeral? If you won’t be going, how come? And also, aside from your personal hopes for peace, do you have reasons to believe that the next coming months might see the new Soviet leadership flex its muscle a bit and a period of increased tension come about?” (Clayman & Heritage, 2002)

 Follow-up questions. These are questions in which the interviewee is directly followed up on his answer. This type of threat occurs in press conferences that adhere to a one-turn-per-journalist norm. These are threatening because they follow the interviewee up on his reply, while this is not expected.

Directness:

 Other-referencing questions. These questions always ask whether the interviewee is willing to answer a question. Although this can be done without other-referencing, these questions mostly start with “could you/would you/ can you/ would you be willing to” etc.

 Presence of the more direct self-references: These are questions in which the interviewer actively puts himself in the question. In these cases interviewers ask questions such as “I would like to know/I wonder/ I want to ask” etc.

Assertiveness:

Questions that direct the interviewee towards a preferred yes/no answer. This is often done by prefatory statements:

“Mr. President, last week the Senate passed a measure enabling both Hawaii and Alaska to achieve statehood. If the House should pass that measure, would you veto the bill?” (Clayman & Heritage, 2002).

 Tag questions & negatively formulated questions. Like the previous example, these are questions that direct the interviewee into answering in a desired way:

“Isn’t it time for some strong action by you to get interest rates down? (Clayman & Heritage, 2002).

Hostility:

 Accountability questions. These are questions that hold the interviewee accountable for certain actions:

“Mr. President, according to yesterday’s report, the administration does not now include the minimum wage in its top measures for passage this year. Would you explain Sir, why this change in signals on the part of the administration?” (Clayman & Heritage, 2002).

 Negative preface questions. Questions that have negative preliminary information. “Mr. President, for months you said you wouldn’t modify your tax cut plan and then

(17)

you did. And when the business community vociferously complained you changed your plan again. I just wondered whether Congress and other special interest groups might get the idea that if they yelled and screamed loud enough, you might modify you tax plan again.” (Clayman & Heritage, 2002). An FTA can be done by adding negative preliminary information, however, an FTA can also be mitigated by adding positive preliminary information.

This study will use the categories of threat as developed by Clayman & Heritage (2002) to develop similar categories that will help determine the different levels of threat questions can pose, as will be explained in section 3.4. Gnisci et al. (2013) argue that questions are face-threatening if they put the interviewee in a situation in which a reply to the question can only have negative outcomes. Especially this last one compares well to Clayman & Heritage (2002) as it often involves attempts at coercing politicians into providing a specific answer. This study will look at how the levels of threat of questions are related to the response that follows, with a specific focus on the different types of equivocal responses that can follow from FTAs.

One last point that needs to be made on the subject of threats in political interviews is that Bull (2008) shows that interviewers who are known to favor certain political parties will ask fewer threatening questions to politicians who are affiliated with said party: “Interviewers identified as favouring Labour all asked the Labour Party leader (Neil Kinnock) fewer conflictual questions than the Conservative Party leader” (340). This is in line with previous studies that state that the personal opinion of the interviewer may influence the ways in which questions are asked. However, this study appears to the only study that shows a relationship between the political affiliation of the interviewers and the questions they ask. Where previous studies have shown possible overall personal preferences, this study suggests a more systematic left-right division in questioning during political interviews.

This section has dealt with the categorization of the types of threats in political interviews as a continuation of issues surrounding interviewer objectivity. The next section of the paper will deal with equivocation theory, and how interviewees deal with threat-questions in interviews.

2.4 Equivocation Theory 2.4.1 Explanation equivocation

Equivocation theory, originally developed by Bavelas et al. (1988), defines equivocation as evasive or ambiguous replies to questions that are potentially face threatening (Bavelas et al,

(18)

1988; Bavelas, 1983; Wodak et al., 1999/2009; Bull & Mayer 1993; Bull et al., 1996; Bull, 2008; Bull & Simon-Vandenbergen, 2014; Clementson, 2016; Clementson & Eveland, 2016). People can choose to equivocate for a variety of reasons, ranging from politeness to protection of the self or other (Edwards & Bello, 2001; Bavelas, 1983; Bavelas et al. 1988; Bull, 2008; Bull & Simon-Vandenbergen, 2014). Studies on equivocation in political debates differ in their conclusions on what causes politicians to equivocate. Ekman (2009), for instance, argues that politicians lie consciously, skillfully and on purpose. Ekman (2009) here paints a negative picture of politicians as inherently deceitful members of society. Bavelas et al. (1988), Bull (2008) and Clementson (2016) go against this claim and show that politicians equivocate because of avoidance-avoidance conflicts.

Avoidance-avoidance conflicts are communicative conflicts in which someone is expected to answer a question, but all the possible direct answers have potentially negative consequences (Bavelas, 1988; Bavelas, 1990). Avoidance-avoidance conflicts are particularly common in political news interviews, as questions here are often meant to put politicians on the spot. Politicians who are presented with an avoidance-avoidance conflict have to answer a question to which a truthful response will result in a loss of face (Bavelas, 1988; Bavelas, 1990; Bull et al, 1996). The institutionalized roles of the accountability interview require the politician to answer (Montgomery, 2008). This means that they are faced with a two-way choice in which they either have to lie or tell the truth and lose face. Politicians thus have to avoid lying and avoid losing face, which shows a double need to avoid providing any reply. This is why they are called avoidance-avoidance conflicts. These avoidance-avoidance conflicts usually result in an equivocal response, as the interviewee attempts to avoid both other options (Bavelas, 1988; Bavelas, 1990; Bull et al, 1996).

There are various ways in which someone can provide an equivocal response. One of these types of equivocal responses is implicature (Riley, 1993; Bull et al., 1996). Implied answers suggest an answer to the question, however, interviewees cannot be held accountable for a statement as they merely implied the answer rather than literally stated it (Riley, 1993). This gives interviewees the opportunity to voice controversial opinions. To the question “Do you think global warming is real?” a politician could answer that he is still wearing his winter coat. The listener then has to infer that the politician does not believe in global warming. When the politician is asked why he does not believe in climate change, he can say he has never said such a thing. Politicians can also put themselves in a positive light without having to actually keep promises they hinted at. An interviewer could ask: “Many people are very worried about tax raises next year, what would you do about this?” to which a politician could reply: “Well their worrying ends where my term starts.” This suggests that there will be

(19)

no tax raises, however, the politician cannot be held accountable as he has only hinted that nothing will change, he has not literally promised anything. A politician can thus put himself in a positive light without having to actually keep promises they hinted at.

Similarly, Wodak et al. (1999/2009) and Bull & Simon-Vandenbergen (2014) show how calculated ambivalence is a form of equivocation used to avoid self-incrimination. In news interviews politicians are sometimes asked questions to which the answer could have legal ramifications (Wodak et al., 1999/2009; Bull & Simon-Vandenbergen, 2014). If politicians have opinions that are deeply racist, for instance, they might risk legal problems if they voiced these opinions, as racism is illegal in many countries. At the same time, however, they need their racist supporters to remain loyal, and they do not want to let them down. A question to which the answer could have legal implications is then likely to be answered equivocally: the politician provides a vague answer that keeps their voters satisfied while at the same time avoiding prosecution. Although most politicians equivocate to some extent, be it because they are skillful liars or because they try to avoid prosecution, calculated ambivalence is a specific type of equivocation that is only seen with extreme right-wing and populist politicians as other politicians usually do not face legal issues with their opinions (Wodak et al., 1999/2009). Both implicature and calculated ambivalence offer politicians a safe way of answering a question without being held accountable, while at the same time giving preferred responses for their audience (Riley, 1993; Wodak, 1999/2009).

2.4.2 Equivocation in Political News Interviews

Politicians have to pay very close attention to their replies in political interviews, as their answers could not only result in loss of face, but also a loss of voters, or even a loss of career (Wodak et al., 1999/2009; Montgomery, 2008). Because politicians are constantly in the media this risk is even higher, and they have to answer carefully when they give interviews. (Montgomery, 2008). Various studies have attempted to categorize the different types of equivocal replies that are given in political interviews.

Bull & Simon-Vandenbergen (2014) have analyzed far right-wing politician Nick Griffin, and found that he employed three different types of equivocation, denials, personal attacks, and implicit meaning. Implicit meaning can then be subdivided between claims of victimization, parallels, disconnection and association. In all his answers Griffin aims to avoid answering questions that could land him in a legally difficult situation. For instance, when he is accused of talking about the white people of Britain when he talks about the indigenous people of Britain, he states “The colour is irrelevant. It’s the people who have been here overwhelmingly, for the last 17 thousand years. We are the aborigines here” (15).

(20)

Griffin here makes use of a few equivocal strategies. He starts by denying that he means white people and states “color is irrelevant”. Then he proceeds to both draw a parallel and claim victimization in the rest of the phrase. The indigenous people of Britain are, according to Griffin, comparable to the persecuted Aboriginals. This parallel is also victimization, as it insinuates that the indigenous people of Britain are currently suffering similarly. This shows that equivocation is not just employed to get through a difficult question, but that it can also be used skillfully to change the focal point, and to show the supporters of the political party the true vision of the party while avoiding prosecution (Ekman, 2009; Bull, 2008; Wodak et al., 1999/2009).

Clementson (2016) created three categories of replies for political news interviews: Refusal to Answer, Same Topic Response and Different Topic Response. The category Refusal to answer contained replies in which interviewees clearly stated they would not respond to the question. Same topic responses were direct replies to the question at hand. Different topic response replies were considered to be equivocal responses. In these cases, the interviewee would answer by going off topic and introducing a new topic to avoid providing an answer to the question. Clementson (2016) argues that politicians only equivocate when they are forced to do so because of an avoidance-avoidance conflict. Clementson (2016) looked at press conferences and political debates with Clinton, Bush Jr., and Obama and found that most replies in the interviews were non-equivocal responses. This opposes other studies that found that most replies in political news interviews were equivocal replies (Bull & Mayer, 1993; Bull & Simon-Vandenbergen, 2014).

Bull and Mayer (1993) analyzed interviews with Margaret Thatcher and Neil Kinnock and categorized the types of responses in interviews. This study developed 11 overarching categories of which some were made up of additional subcategories. This amounted to a total of 30 categories. The present study will now discuss the most important categories:

 Ignores the question

 Acknowledges the question without answering  Attacks the question

 Because it is based on a false premise  Because it is factually inaccurate  Makes political point

 Presents policy

 Appeals to nationalism  Self-justification

(21)

 Incomplete Answer  Partial answer

 Starts to answer but doesn't finish  Decline to answer

This last category, declines to answer, is the same as refusal to answer by Clementson & Eveland (2016). Bull & Mayer (1993) found that Thatcher and Kinnock provided far more attacks and political points than any of the other categories.

Clementson & Eveland (2016) conclude that politicians provided far more direct responses than equivocal responses. However, Bull & Mayer (1993) and Bull et al. (1996) found that most replies were equivocal. As many studies claim that politicians are often forced to equivocate because of avoidance-avoidance conflicts, a possible explanation for this difference in results is the fact that neither of these studies controlled for the potential threat the questions contained. If more threatening questions were asked in Bull & Mayer (1993), this may have resulted in more equivocal responses. The present study therefore aims to analyze the relationship between the threats and the replies.

2.5 Conclusion

Although the studies discussed above have provided an extensive background of information for the present study, various questions remain unanswered. Little to no research has been done on differences between equivocal behavior of American presidents before and after their respective elections. This author of this study is also unaware of any previous research that compares interviews by one news channel with politicians of opposing political parties to see if the political affiliation of the news channel affects the interviews. Many studies have investigated the possible ways in which questions can be threatening, and a wide variety of other studies have investigated all the possible equivocal responses. The present study has not yet found any previous research that combines an analysis of threat in questions with an analysis of equivocal responses in political interviews with two different politicians from opposing political backgrounds before and after their respective elections. The current study aims to begin to fill this gap.

(22)

3. Methodology

This chapter will first introduce in detail the aims of the present study, and the research questions that drove this study. Based on these questions, this chapter will also formulate 4 main hypotheses. The second section will provide information on the data this study used for analysis and the motivation behind using this data. The third section will explain the categorization of all the topics. The fourth section will provide detailed information on the categorization of the questions in terms of level of threat as well as type of threat. The last section will outline how all the replies were categorized.

3.1 Aims of the Study

This study aims to find out how Trump and Obama equivocate to questions in political news interviews that are conducted by Fox News. This study will not only investigate the types of equivocal responses the interviewees provide, but this study will also analyze the questions to see whether certain questions can be more threatening than others. This will show if one interviewee receives more threatening questions that the other, which will indicate whether the political affiliation of the interviewers could indeed affect the types of questions they ask (Bull, 2008). The present study will also look at the potential effects of topics. It is possible that some topics are more threatening than others, and that some may trigger more equivocal responses than others.

A number of research questions have been developed to guide this study as outlined below. In the interviews that were analyzed for this study:

 Does Trump receive more threatening questions before or after the elections?  Does Obama receive more threatening questions before or after the elections?

 Do Obama and Trump receive similar levels of threat in the questions in the “before” interviews?

 Do Obama and Trump receive similar levels of threat in the questions in the “after” interviews?

 What types of threatening questions does Obama receive before and after the elections?

(23)

 Are there any differences between the types of questions Obama and Trump receive before and after the elections?

 Are there any differences in the extent to which Trump equivocates in interviews before and after the election?

 Are there any differences in the extent to which Obama equivocates in interviews before and after the election?

 Are there any differences between the number of equivocal responses between the two politicians, before as well as after the elections?

 What types of equivocal responses are most commonly used by Obama and Trump before the elections?

 What types of equivocal responses occur most frequently after the elections?  Do more threatening questions result in more equivocal responses?

 Are certain topics more threatening than others?

 Do certain topics result in more equivocal replies than others?

Following Bull (2008) it is hypothesized that Obama will receive more threatening questions than Trump, since Fox News is a Republican news channel and Obama is a Democrat. Since other studies have shown that politicians equivocate as a response to a threatening questions (Bull et al., 1996; Bull, 2008; Bull & Mayer 1993; ; Bull & Simon-Vandenbergen, 2014; Clementson, 2016; Clementson & Eveland, 2016; Wodak et al., 1999/2009) it is also expected that Obama will provide more equivocal responses as he is also more likely to receive more threatening questions. A third hypothesis holds that both Trump and Obama will provide fewer equivocal responses after the elections as there is no more risk of losing voters. It is expected, however, that this difference is relatively small as they still suffer the risk of losing face and the loss of support from the electorate.

A fourth hypothesis expects a relationship between the levels of threat questions can pose, and the type of reply; questions that pose higher threats will result in fewer direct replies and more equivocal responses, and questions that pose lower threats or no threats will result in more direct replies than equivocal responses. This study distinguishes 3 levels of threat, and 7 different types of threats that will be further explained in section 3.4. It is expected that certain types of threats may cause more or less equivocation, or perhaps even different kinds

(24)

of equivocation. Since no previous studies were found on the relationship between the types of threat questions and the possible equivocal responses, this part of the study is exploratory and therefore no hypothesis is made.

3.2 Materials

For this study, a total of 9 transcripts of political news interviews were analyzed, all of which were accountability interviews (Montgomery, 2008). The interviews were transcribed and posted on the channels of the Republican Fox News, RealClearPolitics and SBNation. The interviews were all broadcasted by Fox News. For Obama, two of the interviews were conducted before the elections, and three were conducted after. All the transcripts were transcribed by Fox News. For Trump there were two interviews before and two interviews after the elections. One of the “before” interviews was transcribed by Fox News, the other interview was transcribed by RealClearPolitics. Of the “after” interviews one was transcribed by Fox News and one by SBNation. Note that this study used one more “after” interview with Obama since his other two interviews were shorter than Trump’s interviews. To make sure there was an equal quantity of data for analysis for both interviewees this third interview was added. Both Trump and Obama had one interview with Bill O’Reilly and one interview with Chris Wallace before the elections. After the elections Obama had two interviews with Chris Wallace and one with Major Garrett, while Trump had one with Bill O’Reilly and Wallace each. Although not all interviews were consistently conducted by the same interviewee, all the interviewers were employed by Fox News at the time of the interviews. Since this is an overtly Republican News Channel that has been accused of having a biased news coverage on several occasions, this study did not analyze any differences between the interviewers, but treated them all as employees of Fox News.

The “before” interviews with Obama were conducted in April 2008, and in September 2008, the year leading up to his first election. The April interview was conducted during the primaries, while the September interview was conducted after the primaries, when Obama was the Presidential nominee. These interviews were selected since both interviews were conducted in the middle of the presidential campaign, and were therefore expected to contain critical questions, and possibly equivocal responses. The chosen transcripts only involve interviews from Obama’s first run for President to exclude possible confounding variables when comparing the interviews with Trump’s interviews. Obama would have been in office for 4 years and it is likely that people will therefore have formed a more in-depth opinion of him. Since accountability interviews are conducted for an audience (Greatbatch, 1988; Rendle-Short, 2007; Bull, 2008), it is possible that the popular opinion of the people would

(25)

affect the way in which the interviewer conducts the interview. It is also possible that previous experience in running a campaign would affect the types of answers Obama would provide.

The interviews with Obama after the elections were conducted in February 2009, November 2009, and September 2013. This means that the first two interviews were conducted in the first year of his first term, and the third interview was conducted in the first year of his second term. This last interview was conducted later as this study found no other interviews with Obama by Fox News after his first election. This study did not perceive this as an issue since this study aims to make a comparison between interviews before and after the election. Although this last interview was conducted long after the elections, it would still contribute to the before and after picture.

The interviews with Trump before the elections were conducted in October 2015 and in July 2016. The interview in October was conducted shortly before the start of the primaries, and the interview of July was conducted during the primaries. These transcript were chosen as they are expected to contain critical questions and possibly equivocal responses as they are conducted early on in the campaign and Trump still has to convince people to vote for him.

The “after” interviews occurred in December 2016, shortly after the elections, and in February 2017. Although Trump’s interviews after the elections occurred sooner than Obama’s interviews after the election, no issues are foreseen for their comparison since the aim is to make a before and after comparison of the elections. Although a comparison over time would likely be interesting for a follow-up on the subject, it is outside the scope of this study. Additionally, Trump has not been in office long enough to be able to make any comparisons over time.

Because the main aim of the present study is investigating equivocation and the relationship between threatening questions and equivocal responses in political news interviews, a purely textual analysis was deemed best fit for the purpose of this study.

3.3 Topics

Some topics are more controversial than others and they could therefore be perceived as more threatening. For instance, Trump has stated he does not believe in man-made climate change. Since this has resulted in outrage among the electorate, it is possible that Trump will reply more evasively to questions about climate change. Fairclough (1989) states it is also possible that the interviewer, who controls the topics, addresses more controversial topics on purpose. The present study has therefore categorized all the topics in each of the interviews to see if

(26)

certain topics resulted in more equivocal responses, and to see if certain topics in the interviews contained more threatening questions than others. Additionally, section 4.2 provides an overview that shows if questions in certain topics result in more equivocal replies than others. An overview of all 12 different topics that were found across all interviews are provided below.

 Domestic Affairs.

“GARRETT”: Dick Durbin said the new deadline for signing legislation is now State of the Union. Why is that delay acceptable to you? And how upset are you about it?” This question was categorized as the topic Domestic Affairs since it is about the delay of signing new domestic legislation.

 Foreign Affairs:

“O’REILLY: Let’s talk about Mexico. There’s a report that you talked to President Nieto and you told him — this was the report, I want to know if it’s true or not — that if his army couldn’t handle the drug cartels, that U.S. army soldiers would. Did you say that?”

This question was categorized as Foreign Affairs, because it requires Trump to talk about a foreign country. He is asked to comment on how a foreign country handles drug cartels, and if he would intervene if he deemed the country's own policy insufficient. This could put the relationship between Mexico and the U.S.A. under pressure.

 Defense:

“WALLACE: Senator, this week President Bush named David Petraeus, the com-mander of U.S. forces in Iraq, to be the head of Central Command, which controls — oversees military operations across the Middle East and Central Asia. Will you vote to confirm his nomination?”

This question was categorized as the topic Defense since Obama is asked about his opin-ion on the electopin-ion of a new commander of military operatopin-ion

 Personal Question:

(27)

This question occurred in an interview with Obama as well as one with Trump. The question was categorized as Personal since it is a question about their personalities.  Campaign/Race:

“WALLACE: Senator, why are you having such trouble convincing white working class voters that you're their guy?”

This question fell under Campaign as it addresses problems Obama encounters during his campaign. This question was also categorized as race as it also emphasizes that Obama has trouble convincing white working class voters.

 Race/Campaign:

“WALLACE: Senator, for all your efforts to run a post-racial campaign, isn't there still a racial divide in this country that is going to make it very hard for you to get elected president?”

Similar to the previous example, this question was categorized as race and as campaign as it addresses both topics in the question. These two topics were the only topics that were not mutually exclusive.

 Environment/Climate:

“O’REILLY: Yes. And they can if they want. Climate change, they said that you called climate change a hoax. Is that true?”

This question was categorized as Environment/Climate since Trump is asked about previous statements on climate change that state his opinion on climate change.

● Social Media:

“WALLACE: First of all, what’s the deal with the tweets? Why are you sitting there at night watching TV and tweeting out?”

Wallace here asks Trump about his social media behavior, which is why this was categorized as Social Media.

(28)

 Sports:

“O’REILLY: Football. How do you see this game?”

Trump is asked how he feels about the upcoming football game. It was therefore categorized as Sports.

 Economy:

“GARRETT: Good, a couple on the economy, you have a jobs summit next month. You want a jobs bill in 2010. Will that jobs bill raise the deficit or will you demand that it be deficit neutral?”

First of all, the topic is already introduced on being about the economy. Secondly, the question is about what his jobs bill will do for the economy. It was therefore categorized as Economy.

 Healthcare:

“GARRETT: Will you sign legislation on health care that includes the Stupak language?”

Garrett here asks about new legislation on healthcare, and whether he would include anti-abortion laws. It was therefore categorized as Healthcare.

It is important to note that not every interview had the same topics. The topic Campaign, for instance, was only addressed in the interviews before the elections, as the campaigns were important before the elections but not so much after. Other trending issues can also have caused certain topics to be addressed only in one interview. The topic Healthcare is not addressed in any of the interviews with Trump, and it is only addressed in the “after” interview with Obama and Garrett from November 2009. The reason Obama did receive the topic Healthcare can be attributed to the fact that he stated his intent to reform healthcare earlier that year, and that he had since been developing what would later become Obamacare. Additionally, Trump was asked a few questions that fell under the topic Sports since the interview occurred right before the Super Bowl. It is also important to note, that almost all topics were mutually exclusive with the exception of the topic race and campaign. These two topics were often addressed within the same question, and questions that had both

(29)

race and campaign as topic were categorized as both.

The next section will explain the categorization of all the questions in the interviews.

3.4 Assessing Questions

In some cases the interviewer and the interviewee are taking turns, but no specific questions are asked. If no specific question is asked, it is not possible to classify the answer in terms of equivocation. This study therefore only considered turns in which yes/no questions, wh-questions, tag wh-questions, true/false wh-questions, statements with a questioning addition like “right?”, or statements with a rising intonation occurred as questions. Additionally, statements that are made in the imperative are also interpreted as questions. For instance, when Wallace tells Trump: “But give me a number. Give me a number”. Wallace here clearly demands that Trump gives him an answer.

After the present study determined which turns were counted as questions and which were not, this study categorized the questions according to the levels of threat they posed. None of the studies that have investigated threats in questions have distinguished between the different levels of threat a question can pose. Bull et al. (1996) distinguish between threat questions and no-threat questions, however, they do not specify if certain threats are more threatening than other threats. Additionally, Clayman & Heritage (2002) have described the ways in which a question can be perceived as threatening, but no distinction was made between higher levels of threat and lower levels of threat. Most studies only contain a distinction between threat questions and no-threat questions. The present study distinguishes between High-threat, Low-threat and No-threat categories. This decision was made after analyzing the transcripts and finding questions that came across as more hostile than other questions that had already been categorized as threatening questions. The following two examples show two questions that are similar in content, and would both fall into the threat-category as developed by Bull et al. (1996). The examples also show, however, that one question is clearly more threatening than the other, which is why this study distinguishes between high-threat and low-threat.

1. “WALLACE: You are still leading. Why do you think you have gone down a bit?”

Although this question slightly threatens Trump's face since it states there is a decrease in his success, Wallace here softens the blow of the question, making it a Low-Threat question. A similar question for Obama, however, shows a higher level of threat.

(30)

2.

“WALLACE: Anyway, your defeat in Pennsylvania raises new questions about your

candidacy and especially about some of the pillars of the Democratic base. Let's take a look at the numbers. Among white union households, Clinton beat you 72 percent to 28 percent. Among white Catholics, again, same margin — 72 percent to 28 percent. Senator, why are you having such trouble convincing white working class voters that you're their guy?”

Similar to Trump’s question, this is about the polls during the presidential campaign, and their respective positions in the polls. The questions are different, however, in how they are asked. Both questions ask the same thing: why have they become less popular among the electorate. Although both questions are similar in their content, Wallace reduces the level of threat in Trump’s question, while he increases the level of threat in Obama’s question. The question for Trump is mitigated: Wallace uses a strategy of negative politeness by using hedges that increase indirectness of the question: “why do you think…” and “gone down a bit” (Brown & Levinson, 1987). Wallace also minimizes the weight of imposition by adding the positive preliminary information Trump is still leading (Brown & Levinson, 1987; Clayman & Heritage, 2002). And lastly, Wallace reduces directness and association by using a passive form instead of an active form (Brown & Levinson, 1987).

Wallace uses an active form in his question for Obama, however, thereby creating more association and more accountability for the situation (Brown & Levinson, 1987). Additionally he illustrates the large gap between Clinton and Obama by using percentages, and he adds negative preliminary information (Clayman & Heritage, 2002). The wording in this question also pushes Obama towards a desired answer as it already presupposes that he is indeed having trouble convincing white working class voters to support him (Montgomery, 2008). Lastly, instead of just stating “working class voters” Wallace introduces race into the question by adding “white”, which creates a racial contrast between the voters and Obama, and can thus increase the level of threat. These examples show that distinguishing between only two levels of threat was insufficient for the present paper, as threatening questions can differ in their levels of threat. This study therefore distinguishes between three levels of threat: No-threat, Low-threat and High-threat questions. Note that what Clayman & Heritage (2002) defined as threatening is mostly the high-threat category in this study.

The present study has developed the following model for categorization of questions High-threat questions were categorized as such if:

 The interviewer complicates the question by asking multiple questions in one turn  The interviewer includes a negative preliminary statement in the question

Threat & Equivocation in Political News Interviews: An Analysis of Interviews by Fox News with Obama & Trump