This is an open access article distributed under the terms of the Creative Commons
Attribution License (Politeness is a social norm but, in certain contexts, so too is impoliteness. In many everyday social contexts (e.g., dinner parties, first-time meetings), it is typically normative to be polite (Kádár & Haugh, 2013). However, there are also certain social situations where it is normative to be impolite (Culpeper, 1996, 2011). One notable example is that of PMQs – the premier political debate held weekly in the UK House of Commons. In PMQs, the Prime Minister (PM) responds to questions, which may be posed by any Member of Parliament (MP). In terms of exchanges between the PM and opposition MPs, the event has been likened to Punch and Judy, a traditional British knockabout comic puppet show that features domestic strife and violence between the two central characters: Mr Punch and his wife Judy. PMQs is not violent but it is certainly characterised by strife, which takes the form of sustained and intense adversarial verbal conflict. Notably, the word adversarial is derived from the Latin adversarius, meaning an opponent, a rival, or an enemy. The principal aim of this paper is to consider to what extent adversarialism is the norm in PMQs, by identifying what are the normative discursive features of PMQs, and to analyse the extent to which these features may be seen to reflect adversarialism.
The concept of social norm has been defined in many ways; there is no single canonical definition. For example, according to Fiske et al. (2010), social norms are powerful underlying forces guiding how we should behave, which often go unnoticed. For Bicchieri (2006), social norms are the unwritten social codes of conduct and the foundation of the world by which we live. According to Cialdini and Trost (1998, p. 152), social norms are “rules and standards that are understood by members of a group, and that guide and/or constrain social behaviour without the force of laws”, whereas Oni (2021) conceptualises norms in relation to a variety of socially dependent behaviours (based on approval or disapproval of others) with accompanying expectations. Our focus is on discursive norms in PMQs, which are regarded as forms of discourse that are both customary and expected, hence may be regarded as normative. In this paper, analyses of PMQs are reported, intended to illustrate how social evaluations and norms underlie the interactions that take place between the protagonists in this highly distinctive setting.
The paper is based exclusively on interactions between the PM and the Leader of the Opposition (LO). The focus can be amply justified in terms of the way in which their interactions have become increasingly central in PMQs, as documented in an analysis of the initial sessions of five PMs: Margaret Thatcher1, John Major2, Tony Blair3, Gordon Brown4, and David Cameron5 (Bates et al., 2014). Over a 31-year period (1979 to 2010), results showed that the proportion of time taken up both by LO questions and PM responses had increased, that the LO tended to ask longer and increasing numbers of questions, and that the PM typically gave lengthier responses.
We begin with an overview of the multidisciplinary and methodological approaches adopted in this paper. There follows a brief account of adversarialism with particular reference to parliamentary discourse in general and PMQs in particular. Then five numbered sections featuring empirical studies by the authors and colleagues are reported, each identifying a distinctive aspect of PMQs discourse: (1) face-threats (Bull & Wells, 2012); (2) personal attacks (Waddle & Bull, 2020a; Waddle et al., 2019); (3) the rhetorical use of quotations (Fetzer & Bull, 2019; Bull & Waddle, 2019); (4) equivocation by the PM (Bull & Strawson, 2020; Bull & Waddle, 2019) (5) traditional forms of address to the Speaker6 and third-person references (Bull et al., 2020). Finally, a further study is reported (Bull, 2013), intended to assess whether PMQs discourse is little more than mere point-scoring or if it plays a more substantive role in political opposition.
Overview
This paper also draws on relevant research on adversarialism conducted in a variety of academic disciplines: notably, sociology (Goffman, 1955, 1967), linguistics (e.g., Brown & Levinson, 1987; Culpeper, 1996, 2011; Harris, 2001), political science (e.g., Bates et al., 2014), as well as social psychology (e.g., Bavelas et al., 1990). Given the highly relevant research that has emerged from such a range of related academic disciplines, it would be inappropriate not to take a multidisciplinary approach to an analysis of PMQs.
The methodological approach taken throughout all six empirical studies was based on the detailed analysis of transcripts from Hansard, so named after Thomas Curson Hansard (1776–1833), a London printer and publisher, who was the first official printer to the Parliament at Westminster. It should be noted that Hansard is an edited not a verbatim version of parliamentary proceedings, so video recordings of PMQs are also used in these studies in order to enhance analytical accuracy.7 Transcript analysis was based either on existing coding systems (e.g., for equivocation, Bull, 2003; Bull & Mayer, 1993), or new systems devised especially for a particular study (e.g., for face-threats in questions, Bull & Wells, 2012, and for personal attacks, Waddle et al., 2019). Analyses were both qualitative and quantitative. Where quantitative analyses were reported, these were primarily in the form of descriptive statistics; where inferential statistics were employed, details are provided in the relevant sections.
Adversarialism
The adversarial discourse of PMQs was the focus of a highly influential study by Harris (2001), entitled “Being politically impolite”. With analysis based on politeness theory (Brown & Levinson, 1987), Harris argued that much of PMQs discourse is composed of intentional and explicit face-threatening acts (FTAs), which can either threaten positive face (i.e., making the targeted person look bad) or negative face (i.e., in some way impeding the targeted person’s future freedom of action). Thereby, adversarialism may be seen as a form of systematic impoliteness, which is central to those interactions. A useful distinction concerning forms of impoliteness can be drawn from that which Haugh (2003) makes in relation to politeness: anticipated or inferred. Anticipated politeness is typically based on expectations/anticipations, whereas inferred politeness is based on implicature. In the context of PMQs, given that adversarialism is very much the norm, impoliteness would typically be expected/anticipated rather than inferred.
Much of the inspiration for politeness theory (Brown & Levinson, 1987) came from Goffman’s (1955) seminal essay “On face-work”. Goffman (1967) highlighted three kinds of facework: an avoidance process (avoiding potential FTAs), a corrective process (performing various redressive acts), and making points (using aggressive facework). Interestingly, within the politeness theory framework, rudeness is essentially viewed as a deviation from the rules of polite/cooperative communication. However, rudeness can be motivated and deliberate (Kienpointner, 1997). Furthermore, Lakoff (1989) proposed a three-way categorisation of linguistic behaviour (polite, non-polite, and rude), that politeness is associated primarily with regular conversation, and that linguistic politeness theories be extended to institutional/professional contexts where rude discourse can be more prevalent. Kasper (1990) proposed a further distinction: unmotivated and motivated rudeness. Whereas the former relates to rudeness due to ignorance of certain behavioural norms, motivated rudeness relates to deliberate norm violation (i.e., the speaking person’s rudeness is wilful and intentionally hurtful).
Notably, a model of impoliteness has been proposed (Culpeper, 1996, 2011; Culpeper et al., 2003), which is opposite but parallel to politeness theory. According to Culpeper (1996, 2011), in certain contexts (e.g., military training) impoliteness is not a marginal activity – it is central to the interaction. Another such context is PMQs, which has been figuratively likened to a form of verbal pugilism (Bull & Wells, 2012). Just as professional boxers are evaluated in terms of their boxing ability, so too the PM and LO are evaluated on their ability to deliver and counter verbal punches. Boxing is conducted under a strict code known as the Queensberry rules.8 Similarly, verbal clashes between politicians in parliament are not a free-for-all, they are subject to a strict code of conduct. In PMQs, MPs’ dialogue should follow a question-response (Q/R) pattern; and they are expected to observe certain conventions regarding what is termed unparliamentary language. Specifically, they should refrain from abuse and insults – including direct accusations of lying – nor should they misrepresent one another. These conventions are enforced by the Speaker, who may ask members to withdraw objectionable utterances. Historically, Speakers have raised objections to many abusive terms, including coward, traitor, hooligan, stoolpigeon, blackguard, guttersnipe, swine, rat, and git (House of Commons Information Office, 2010). An MP who refuses to comply may be suspended from the House (referred to in parliamentary procedure as naming). A notable example occurred during a debate9 on financial affairs (11 April 2016): Labour MP Dennis Skinner was suspended for the remainder of the day’s sitting for not withdrawing his invective directed at PM Cameron – namely, referring to him as “Dodgy Dave”.
1. Face-Threats
Harris (2001) identified techniques whereby FTAs may be performed in PMQs and provided several illustrative examples (discussed below). The study by Bull and Wells (2012) followed up on the work of Harris by devising a typology of FTAs in PMQs. Eighteen sessions from 2007 were analysed. During that period of Labour government, Blair resigned as PM (27 June) and was succeeded by Brown, hence it was possible to sample PMQs performance from two different PMs. Specifically, the study was based on Blair’s final nine sessions (18 April - 27 June) and the first nine of Brown’s premiership (4 July - 21 November).
Method
The focus of analysis was on Q/R sequences between PM and LO. However, in practice it is not just questions and responses that form the interactional units of PMQs but speaking turns (Harris, 2001). Characteristically, a question may be formed by a series of propositions followed by an interrogative (e.g., “will the PM assure the House that…”, “is the PM aware that…”), with a final action- or information-seeking summarising proposition.
Prefaces may, for example, be used to make political points, attack other politicians, or make jokes or asides (Harris, 2001). Thus, FTAs may be performed that are part of the wider speaking turn, but do not necessarily constitute interrogatives as such. Accordingly, the analysis was focused not just on questions and responses but on what were termed questioning and response turns. In addition, attention was given to Harris’s concept of mitigating techniques, whereby participants may soften the force of FTAs.
Results
The results are presented across three sections: questioning turns, response turns, and mitigating techniques (Bull & Wells, 2012).
(i) Questioning Turns
Six different techniques were identified whereby a questioner may perform FTAs:
1. Questions may follow a preface, which can be used to perform FTAs. For example, on 18 April 2007, Cameron10 (then LO) prefaced a question to the PM about the leadership ambitions of Brown (then Chancellor of the Exchequer) with an extended personal attack:
Cameron: We know why we do not want the Chancellor – he has complicated the tax system and virtually bankrupted the pensions system. He is impossible to work with and he never says sorry. That is why we don’t want the Chancellor. What does the Prime Minister think is wrong with him? (Hansard HC Deb, 18 April 2007, col. 296)
2. A detailed question (Harris, 2001) may contain a request for highly specific information, which the PM may not have available or wish to publicise, for example:
Cameron [to Blair]: This week we have the scandal of the Prime Minister, in his last few days in office, opening the prison gates and releasing 25,000 prisoners on to our streets. Can he tell us when he was first warned that the prison population would go over 80,000? (HC Deb, 20 June 2007, col. 1368)
In their follow-up, the LO, in continuation of their attack, may provide those details if the PM declines or fails to reply – as was the case in this instance, when Cameron stated, “The truth is that the Prime Minister was told by the Home Office in 2002, five years ago”.
3. The PM may be presented with an invitation to perform a face-damaging response (e.g., apologising, criticising a fellow party member, or admitting that a particular policy has failed or a government department has been incompetent). For example:
Cameron: Ten years ago, he told us that he would be tough on crime; now he is releasing 25,000 criminals on to our streets. Shouldn’t he, just this once, apologise for what can only be described as an abject failure to deliver? (HC Deb, 20 June 2007, col. 1371)
4. Questions in PMQs may be based on contentious presuppositions that are in themselves highly face-threatening (Harris, 2001). For example, in the foregoing extract from 20 June, Cameron’s use of the phrase “just this once” presupposes that it is typical for PM Blair to never apologise.
5. According to Bavelas et al. (1990), there are certain questions where all possible responses have the potential for negative consequences, but a response is expected. These have been labelled communicative conflicts, and re-conceptualised in terms of threats to face (Bull et al., 1996); so, whichever way a politician responds to such conflictual questions, they run the risk of losing face. For example, a question from LO Cameron related to a recent loss of supposedly secure personal data: “Does the Prime Minister accept systemic failure in this department?” (HC Deb, 21 November 2007, col. 1180). If the PM were to respond with affirmation, this would indicate government incompetence; conversely, a denial, given the publicity already generated by the data loss, would lack credibility. Brown avoided both options – instead responding in an equivocal manner with proposals for a procedural review.
6. Politicians may deviate from a question format by making asides, which, for example, can take the form of a quip intended to berate or mock opponents.
(ii) Response Turns
The following five ways whereby the PM can respond to FTAs in question turns were identified: ignore; attack; talk up positive face; rebut; self-justify. These different tactics are not necessarily individual alternatives, they may be used in combination (as in the example below), nor do they necessarily involve the performance of FTAs.
The first four of these tactics are identifiable in the PM’s response to the question from LO Cameron highlighted above (“What does the Prime Minister think is wrong with him?):
Blair: Let me tell the right honourable Gentleman what is right with the Chancellor [ignores question – says what he thinks is right with Brown]. The right honourable Gentleman has some experience of the economy, has he not? He’s had something to do with the British economy once, hasn’t he? Back in 1992, did he not? He [Cameron] was the special adviser to the Chancellor of the time – we remember Black Wednesday11 [attack on Cameron]. The Chancellor has delivered the strongest economic growth that this country has ever seen – interest rates half what they were under the previous Conservative Government, the highest employment, the lowest unemployment for years, and rising living standards [talks up positive face of Brown, thereby rebuts Cameron]. What’s he [Cameron] delivered for the British economy? A bit part on Black Wednesday [attack]. (HC Deb, 18 April 2007, col. 296-297)
The fifth strategy, self-justification, relates to responses where the PM offers explanation, reason, or excuses for their actions. The following example is in response to a question from LO Cameron challenging the PM over his refusal to hold a referendum on the UK’s position in the European Union. The whole of Brown’s response can be seen as a self-justification for this position:
Brown: Mr Speaker, if we were deciding to join the Euro, we would have a referendum. If it was the old constitutional treaty, we would have a referendum. Because it is an amending treaty that is not fundamental change, we have managed to negotiate red lines in Europe which mean that the national interest is protected. And Britain will decide on justice and home affairs. Britain will decide on foreign policy where it’s multilateral. Britain will decide on social security, and Britain will decide on national security. And we will, at all times, stand up for the British national interest. (HC Deb, 17 October 2007, col. 819)
(iii) Mitigating Techniques
Mitigating techniques (Harris, 2001) – ways in which the full force of an FTA may be somewhat softened or controlled – can occur in questions or responses. Three such techniques are described below.
1. MPs should refer to one another by formal titles (e.g., the Prime Minister, the Foreign Secretary, the right honourable Gentleman), and address their remarks to the Speaker. Such third-person language functions to distance the user, can soften the force of an attack, and is expected in accordance with parliamentary convention (Harris, 2001). The use of direct, second-person pronouns (e.g., you) is regarded as unparliamentary, thereby may be challenged by the Speaker.
2. Humour can soften the full force of an attack on another politician. In the following example, the PM mocked LO Cameron’s pretensions by using one of his own quotes against him, in which Cameron had compared himself to Arnold Schwarzenegger12:
Brown: I know that he’s good at PR, but didn’t he go too far last weekend when he went to California and he said in an interview in the newspaper, he said, “Look at me… and think of Arnold Schwarzenegger”? That’s the last thing on anybody’s mind. (HC Deb, 17 October 2007, col. 819)
3. In the above example, Brown referred to a quotation from a newspaper interview, thereby using Cameron’s own words to ridicule him. In this way, Brown could imply that Cameron had ridiculous delusions of grandeur without explicitly saying so. The use of quotations in PMQs as a rhetorical device is discussed in greater detail in Section 3.
Discussion
The foregoing typology clarifies how questions can pose FTAs and how PMs may respond to such attacks (Bull & Wells, 2012). Notably, every question posed by LO Cameron to PMs Blair and Brown in these eighteen sessions (108 questions) included one or more face-threats in accordance with the typology. This finding graphically illustrates how FTAs in LO questions are customary practice, thereby very much a social norm in interactions between the leaders. FTAs in questions are characteristically attacks on the PM’s competence, hence are highly adversarial. Notably, asking face-threatening questions might be seen as highly impolite in many other social contexts.
In PMQs, Harris (2001) has argued that systematic impoliteness is not only sanctioned but rewarded in accordance with members’ expectations, through an adversarial and confrontational political process. Even the most serious FTAs rarely, if ever, result in a breakdown in interpersonal relationships, nor is that the intention. MPs clearly perceive that the main role of political opposition is to oppose, namely, to criticise, challenge, subvert, or ridicule the policies and positions of the government. Nowhere is this more evident than in these weekly exchanges between PM and LO. Indeed, the latter will likely want to promote their own reputation as a skilful and effective adversary and highlight their leadership credentials. Via such adversarial behaviour, the face of an LO can be enhanced whilst undermining that of the PM. This would suggest that MPs are fully aware of these social norms, and that this influences both how they evaluate the main protagonists and how the protagonists respond to and interact with one another.
2. Personal Attacks
Arguably, there can be no discourse more adversarial than personal attacks, as it effectively highlights those on the receiving end as opponents, rivals, even enemies. In this section, two studies of personal attacks are reported (Waddle et al., 2019; Waddle & Bull, 2020a).
In the first study (Waddle et al., 2019), the period of analysis followed that of Bates et al. (2014), namely, the premierships of Thatcher, Major, Blair, Brown, and Cameron. However, whereas Bates et al. analysed only the opening PMQ sessions for each PM, we analysed both their early and latter periods. Thereby, it was possible to assess changes in personal attacks over a period spanning five decades (1979-2016) and within the tenure of each PM. This same data set was also utilised for the second study (Waddle & Bull, 2020a).
In the study by Waddle et al. (2019), two hypotheses were tested. Bates et al. (2014) had found an increase in rowdiness across their period of analysis as well as a greater likelihood for MPs to become more troublesome in their questioning the longer their tenure in Parliament. Accordingly, the following predictions were made: (1) an increase in personal disrespect over time; (2) more personal attacks in the latter periods of PMs’ time in office.
The aim of the second study (Waddle & Bull, 2020a) was to analyse personal disrespect in relation to question topic, comparing questions and responses focused on domestic policy with those on foreign policy. This analysis followed a previous study of aggressive questioning in US presidential news conferences (Clayman et al., 2007), in which questions on foreign policy were found to be significantly less aggressive than those on domestic policy. A short-term boost in presidential popularity following the nation’s involvement in an international crisis (especially one involving military action) is known as the rally ‘round the flag effect (RE) (Mueller, 1970). Considering the foregoing findings from the US, the aim of this second study was to test whether any related form of reduced hostility was observable in the exchanges between party leaders at PMQs.
Method
In both studies (Waddle et al., 2019; Waddle & Bull, 2020a), the first and last 6013 Q/R exchanges were analysed for each PM.
Leaders’ exchanges were separately analysed for instances of personal attacks, defined as personal references considered disrespectful (Waddle et al., 2019). These included negative personality statements, implied enduring negative character traits, aspersions/disparaging insinuations, patronising/condescending remarks, mockery, badgering (comments regarded as personal harassment), and negative names/labels (for example, on 24 November 2010, PM Cameron called Ed Miliband14 “the nowhere man of British politics").
In the second study (Waddle & Bull, 2020a), the analysis of question topic was based on the UK policy agenda codes (John et al., 2013), which comprises 19 major topics. These were then categorised as either of a foreign or domestic agenda – three matching the former, the remaining 16 all identifiable as domestic (e.g., economy, health, education, crime, government operations). The three befitting foreign policy (foreign affairs, foreign trade, and defence) accounted for 15% of the questions in the analysed exchanges (notably, the topic foreign affairs included policy issues related to the European Union).
Results
The results of the first study (Waddle et al., 2019) showed that, overall, 31.4% of the leaders’ speaking turns included personal attacks on their opponent. Data were analysed using a generalised linear model, and statistically significant results refer to that model. Of the five PMs, the highest proportion of personal attacks came from Cameron (39.2%), then Brown (37.5%) – both of whom were significantly higher than Blair and Thatcher (both at 24.2%) – with Major at 30%.
Notably, when asking questions as LO, Cameron also made the highest proportion of personal attacks (45.6%). The period of his highest level of personal antagonism came towards the end of Brown’s premiership, when 61.7% of his questions contained disrespect – a level significantly higher than that observed by any other LO. The lowest level of personal disrespect in questions to the PM came from LO Jeremy Corbyn15, at just 8.3%. (Corbyn’s questioning style in PMQs is considered at greater length in Section 3).
An additional analysis assessed personal attacks by PMs over time. Overall, it was found that PMs made significantly more personal attacks in the latter than in the early periods. In terms of individual premierships, the increase in personal attacks was significant for Thatcher (from 8.3% to 40%), Major (11.7% to 48.3%), and Brown (25% to 50%); Blair’s increase was non-significant (18.3% to 30%). The one exception, however, was Cameron, who was found to make significantly fewer personal attacks in his latter period than at the beginning of his premiership (reducing from 58.3% to 20%).
In the second study (Waddle & Bull, 2020a), exchanges on foreign and domestic policy were compared. LO questions based on domestic policy issues were over twice as likely to be personally antagonistic (36% containing disrespect) than those concerning foreign policy matters (just 17.2%). For PM responses, this difference (also highly significant at p < .001) was even more pronounced, with those on domestic issues almost three times more likely to be couched in personal disrespect (37.4% compared to just 13.1% for foreign policy-based responses).
Discussion
Overall, the results of both studies (Waddle et al., 2019; Waddle & Bull, 2020a) showed that personal attacks are very much a feature of PMQs discourse, and that there were significant increases in personal antagonism over PMs’ time in office. However, the effect was mitigated by question topic, being far less pronounced (and thereby, generally, more polite) for questions on foreign affairs than on domestic affairs (Waddle & Bull, 2020a).
From the beginning of Thatcher’s premiership in 1979 to the end of Cameron’s in 2016, the results also pointed towards a general trend for PMs to be more personally offensive than their predecessors. When comparing a combination of their early and latter periods, all PMs except Blair were found to make more personal attacks in their speaking turns than the PM they succeeded, while Cameron was shown to use more personal attacks both as PM and LO than any of the other leaders.
Personal attacks are highly adversarial and in many other social contexts would be seen as downright offensive. On average, nearly one third of the LO-PM exchanges included personal attacks and they were significantly more frequent for PMs the longer they were in office. From these data (Waddle et al., 2019), personal attacks are common practice in LO-PM interactions, hence may be regarded as a normative form of discourse in PMQs.
3. Quotations
Quotations are very common in PMQs discourse. Politicians may quote themselves, other politicians (allies or opponents), or expert sources (e.g., official reports). An analysis of word counts (Fetzer & Bull, 2019), based on 40 sessions during Cameron’s premiership (20 each with LOs Miliband and Corbyn – overall, 240 Q/R sequences), showed quotations accounted for 9% of the Miliband-Cameron exchanges, 10% of Corbyn-Cameron. Quotations are typically used in an adversarial manner, either to deconstruct the argumentation, credibility, or leadership of political opponents, to enhance such qualities in themselves, or occasionally both (Bull & Waddle, 2019).
However, quotations were used in a very different fashion by LO Corbyn. During his successful leadership campaign, he called for a “new kind of politics” (ITV, 2015). One feature of his apparent “new” approach was when speaking at PMQs to include questions sourced from members of the public. Consequently, we conducted a study focused specifically on the impact of such questions (referred to as public questions, those not so sourced as non-public questions) (Bull & Waddle, 2019).
Method
Analyses in this study were based on exchanges between Corbyn and Cameron – the first 20 sessions for Corbyn as LO. Our aims were to assess whether this innovation had any mitigating effect on the adversarial discourse of PMQs: specifically, to test whether public questions were associated with an increased reply-rate (the proportion of questions to which the PM gave an explicit answer) and whether they led to a decrease in personal attacks.
An example of a public question was evident in his opening turn at his first PMQs as LO on 16 September 2015. Here, Corbyn quoted a member of the public; the PM’s response follows:
Corbyn: […]16 Many told me that they thought Prime Minister’s Question Time was too theatrical, that Parliament was out of touch and too theatrical, and that they wanted things done differently, but above all they wanted their voice heard in Parliament. So, I thought, in my first Prime Minister’s Question Time, I would do it in a slightly different way. […] So, I sent out an email to thousands of people and asked them what questions they would like to put to the Prime Minister, and I received 40,000 replies. There is not time to ask 40,000 questions today – our rules limit us to six – so I would like to start with the first one, which is about housing. Two-and-a-half thousand people emailed me about the housing crisis in this country. I ask one from a woman called Marie, who says, “What does the government intend to do about the chronic lack of affordable housing and the extortionate rents charged by some private sector landlords in this country?”
Cameron: […] Let me now answer, very directly, Marie’s question. We do need to see more affordable housing in our country. We delivered 260,000 affordable housing units during the last Parliament, and we built more council houses in our country than had been managed in the previous 13 years, but I recognise that much more needs to be done. […] (HC Deb, 16 September 2015, col. 1037)
Results
In Corbyn’s first PMQs as LO, all six of his allotted questions were sourced from members of the public. This reduced to three then four at subsequent sessions, then only one or two over the next 16 sessions (it was not until his twentieth session that no such questions were asked).
Results showed that Cameron gave explicit answers to 23% of Corbyn’s public questions and 20% of his non-public questions – a difference that was not statistically significant (Wilcoxon matched-pairs test). However, a notable difference did emerge from our statistical analysis of personal attacks, which was based on a generalised linear model. Overall, of the 120 Q/R sequences, Cameron made attacks on 31 occasions (25.8%), Corbyn on 18 (15%) – a statistically significant difference (p = .039). Similarly, for non-public questions (n = 89), the difference was also significant (p = .031): Cameron included a personal attack in 25 (28.1%), Corbyn in 13 (14.6%). Yet, interestingly, for the public questions (n = 31), the difference was not significant (p = .740): Cameron made personal attacks on six occasions (19.4%), Corbyn on five (16.1%).
Discussion
It was argued above that quotations in PMQs are typically used in an adversarial manner – either to deconstruct the argumentation, credibility, or leadership of opponents, or to enhance these qualities in themselves. In many other social contexts, quoting someone else’s words against them might be seen as very impolite, while quoting oneself might be seen as boastful.
However, the results here suggested that the use of quotations in PMQs is not intrinsically adversarial, but what matters is the way in which they are used. Although Corbyn’s public questions had no effect on the level of equivocation, they did somewhat curtail Cameron’s personal attacks, which was notable given the previous finding of Cameron’s high level of personal attacks compared to his predecessors (Waddle et al., 2019). Furthermore, the results of both this study (Bull & Waddle, 2019) and the data reported by Fetzer and Bull (2019) show that the use of quotations in PMQs is a customary practice, and hence may be regarded as normative.
4. Equivocation
An interesting feature of the study by Bull and Waddle (2019) was the noticeably low reply-rate for PM Cameron in response to Corbyn’s questions (21% overall). This contrasts with reply-rates found in studies of televised political interviews with UK party leaders. One analysis of 33 interviews (broadcast between 1987 and 1992) showed a mean reply-rate of 46% (Bull, 1994). A more recent study of 26 interviews from the 2015 and 2017 general election campaigns (Waddle & Bull, 2020b) showed a comparable mean reply-rate of 38%. Of course, PMQs is a markedly different setting from political interviews but the foregoing low reply-rate for Cameron highlighted the need for further investigation of equivocation at PMQs.
Accordingly, an analysis was conducted of PMQs featuring Cameron’s successor as PM, Theresa May17 (Bull & Strawson, 2020). The aims were twofold: to analyse May’s reply-rate in response to questions from LO Corbyn and to test whether she had a distinctive equivocation style.
Method
The analysis was based on 23 sessions (all those from May’s first term of office: July 2016 - April 2017). The LO-PM exchanges were analysed in terms of typologies for identifying questions and responses (Bull, 1994, 2003, 2009), and for coding different forms of equivocation (Bull, 2003; Bull & Mayer, 1993).
Results
May’s overall reply-rate to the 138 questions posed over the 23 sessions was just 11%, significantly lower (Mann Whitney U = 117, p = .006) than Cameron’s 21% reply-rate from the foregoing Bull & Waddle (2019) study. In terms of equivocation type (Bull, 2003; Bull & Mayer, 1993), May’s most commonly used forms were: making political points (92%), ignoring the question (43%), giving a non-specific response to a specific question (26%), making personal attacks (23%), stating or implying the question has already been answered (19%), and acknowledging the question without answering (16%).18 Two of the said categories are unremarkable in identifying a distinctive equivocation style for May. Namely, that used most frequently (making political points) was also identified as the most common in an analysis of equivocation in televised interviews by three former party leaders (Thatcher, Major, and Labour’s Neil Kinnock) (Bull, 2003). Although interviews differ from PMQs in many ways, it is unsurprising that making political points is the most common form of equivocation in both situations. Also unremarkable, in consideration of the findings of Waddle et al. (2019) concerning her five predecessors, was PM May’s frequent use of personal attacks.
However, May’s four remaining recurrent equivocation categories are interesting in the context of the overt-covert distinction, as proposed in political interview research (Clayman, 2001). Thus, overt equivocation is typically quite explicit (e.g., the politician states their unwillingness to answer a question), whereas covert equivocation is not acknowledged explicitly by the user, who may even attempt concealment. Notably, all four of these categories can be considered covert. Two of them (acknowledging the question without answering and ignoring the question) may be understood in relation to each other. In the former, the politician responds with an acknowledgement of the asked question but does not provide an answer; whereas in the latter, they not only make no attempt to answer but even fail to acknowledge that the actual question was asked. An example of the former is evident in May’s following response:
Corbyn: […] Is it not the case that her cuts to universal credit19 will leave millions worse off?
May: […] On the point that the right honourable Gentleman raised in relation to universal credit, the introduction of universal credit was an important reform that was brought about in our welfare system. It is a simpler system, so people can see much more easily where they stand in relation to benefits. Crucially, the point about universal credit is making sure that work always pays. As people work more, they earn more. It is right that we do not want to see people just being written off to a life on benefits and that we are encouraging people to get into the workplace. (HC Deb, 2 November 2016, col. 880-881)
Here, it is apparent that May includes an acknowledgement to the question (i.e., “On the point that the right honourable Gentleman raised in relation to universal credit”) but does not provide an answer (she neither affirms nor denies whether people will be worse off due to these cuts). This can be contrasted with the following example of ignoring the question, when Corbyn challenged the PM about her appointment of Boris Johnson (who eventually became PM himself) as Foreign Secretary. Corbyn’s question here relates to the contents of two newspaper articles: a Daily Telegraph article (10 January 2002), in which Johnson referred to Commonwealth members on the African continent as “flag-waving piccaninnies”, and a more recent article in The Sun (22 April 2016), where the future PM had seemingly questioned the motives of US President Barack Obama because of his part-Kenyan heritage. Noticeably, May’s response here relates only to her previous remarks on racial discrimination in the criminal justice system, entirely ignoring Corbyn’s question concerning Johnson’s appointment to the Foreign Office.
Corbyn: The Prime Minister is rightly concerned that: “If you’re black, you’re treated more harshly than if you’re white”. Before appointing her new Foreign Secretary, did she discuss with him his description of black people as “piccaninnies” and ask why he had questioned the motives of US President Obama on the basis of his “part-Kenyan” heritage?
May: […] The right honourable Gentleman referred to the remarks I made [concerning racial discrimination in the criminal justice system]. It is correct that if you are black, you will be treated more harshly in the criminal justice system. That is exactly why, as Home Secretary, I dealt with the issue of stop and search. I was concerned to make sure that nobody should be stopped and searched on the streets of this country because of the colour of their skin. I did that as a Conservative. In 13 years, Labour did nothing on it. (HC Deb, 20 July 2016, col. 818)
Both forms of equivocation used by May in these examples can be considered covert. In the first, in acknowledging the question, she gives the misleading impression that an answer is forthcoming. In ignoring the question in the second example, she does not even acknowledge the question asked. Furthermore, in instances where May states (or implies) that she has already answered a question, this can conceal the fact that the question was not answered. In giving a non-specific response to a specific question, Bull and Strawson (2020) argued that May, in effect, modifies the question, and it is to her own modified version that she responds. This may be viewed as the most covert of all four of these equivocation techniques because thereby May seemingly creates the impression of providing an answer but not to the actual question that was asked. In the first example above, May responded by talking about generalities of universal credit. That was not the focus of Corbyn’s question, which was modified by May, and it is to that modified version that she responds.
Discussion
From the data presented above on reply-rates (Cameron 21%, May 11%), equivocation by PMs is common practice in PMQs and hence very much the norm. In many other social contexts, repeatedly not answering questions might be seen as impolite. The detailed analysis of May’s equivocation style also illustrates specific covert ways in which she equivocated in response to LO Corbyn’s questions.
In political interviews, it has been shown that politicians typically equivocate in response to adversarial questions which create what is termed a communicative conflict, where all the principal possible responses to a question are potentially face-threatening and equivocation is arguably the least face-threatening response (Bull, 2008; Bull et al., 1996). It is highly possible that equivocation also occurs in response to such conflictual adversarial questions in PMQs, however this hypothesis has yet to be tested, therefore would require further empirical analysis.
5. Traditional Forms of Address
In PMQs, MPs are expected to direct their remarks through the Speaker – addressing her/him as “Madam Speaker/Mr Speaker” – and this form of address occurs commonly throughout parliamentary debates (Bull et al., 2020), hence may be seen as another normative feature of parliamentary discourse. From this perspective, PMQs may be considered a form of mediated address, as participants should not interact directly without the involvement of the Speaker, who plays a go-between role, maintaining order in the debates, determining who may speak. If, as argued above, PMQs is seen as a form of verbal pugilism, then the Speaker might be compared to a boxing referee, there to ensure fair play and that, figuratively speaking, there are no blows below the belt (in the parliamentary context, this might mean instances of unparliamentary language or excessive interruptions).
From the perspective of Harris (2001), addressing remarks to the Speaker and the use of third-person language are mitigating factors, which soften the force of FTAs in parliamentary discourse. Whether this is the case, or whether addressing the Speaker as “Mr Speaker” may work as an FTA booster, was the focus of this study (Bull et al., 2020).
Method
The study was based on 40 PMQ sessions during Cameron’s premiership: 20 with LO Miliband (21 March 2012 - 10 July 2013) and 20 with LO Corbyn (16 September 2015 - 20 April 2016).
In line with the foregoing research, analyses were based on Hansard transcripts. Interestingly, when these transcripts were checked against video-recordings of PMQs, it was found that, for the most part, references to the Speaker were edited out (i.e., they did do not appear in the written record), presumably because they were not regarded of particular significance.
A preliminary analysis of interactions between PM Cameron and the LOs suggested there were two main functions served by explicit references to Mr Speaker: signalling conflictual situations and interaction ritual/discourse organisation. The first of these relates to instances where the LO challenged the PM’s policies and attempted to deconstruct their argumentation. The second typically fits one of two forms: (1) accepting a speaking turn from the Speaker, which is ceremonial and recurrent, and part of the ritual dynamics of PMQs, thereby, can be seen to play a role in discourse organisation; and (2) disambiguating an utterance, making clear who is the intended recipient. Overall, it was found that all instances of addressing the Speaker could be coded as either conflictual or interaction ritual/discourse organisation.
Results and Discussion
In the case of the interaction ritual/discourse organisation function, the concept of addressing remarks to the Speaker as a mitigating factor (Harris, 2001) would not be relevant since these features would not require any mitigation. However, in the context of conflictual situations, using the form of address “Mr Speaker” (or “Madam Speaker”) might be understood as a type of mitigation, especially given that the expected role of the LO is to attack the government (thereby performing more FTAs than the PM).
An alternative explanation concerns the relative power difference between LO and PM. An LO’s primary role is to attack the government, and typically they have far less power than the PM. In these terms, the LO might be seen as calling on the Speaker to support their viewpoint. Conversely, PMs have far less need for this, as they already have authority as head of government. Similar to a litigant in a court case appealing to the judge, the LO may be seen as calling the PM to account, deriding their performance to a third party (e.g., for failing to answer a question or not meeting legitimate public concerns). From this perspective, addressing the Speaker need not be a mitigating factor but may be a signal that a stronger FTA is forthcoming. Thereby, via such ritualistic discourse, an LO may highlight government failings. In short, addressing the Speaker may well be a potent means of conducting political opposition.
By the same token, the concept of third-person language as a mitigating factor (Harris, 2001) can also be challenged. In the following example (Bull et al., 2020), Cameron – then LO – launched a wholesale attack on the leadership of Brown:20
Cameron: Mr Speaker, for 10 years he has plotted and schemed to have this job – and for what? No conviction, just calculation; no vision, just a vacuum. Last week he lost his political authority, this week he is losing his moral authority. How long are we going to have to wait before the past makes way for the future? (HC Deb, 10 October 2007, col. 289)
A conceivable, more personalised version is:
For 10 years you have plotted and schemed to have this job – and for what? No conviction, just calculation; no vision, just a vacuum. Last week you lost your political authority, this week you are losing your moral authority. How long are we going to have to wait before the past makes way for the future?
This more personalised version would be regarded as unacceptable parliamentary language and, undoubtedly, the Speaker would admonish the LO for failing to address his remarks accordingly. However, an alternative viewpoint is that, through such third-person language, Cameron, by not addressing Brown personally – effectively, treating him as an object of talk – was able to make a more damning and brutal indictment.
Overall, these foregoing modes of address are conventional forms of PMQs discourse, hence may be seen as normative. However, whether adversarialism is mitigated by these ritual practices is open to question, given that they may be weaponised by politicians to bolster their attacks on opponents. From this perspective, these ritual practices might also be seen as adversarial. In many other social contexts, this kind of covert aggression may well be considered impolite.
Functions of Adversarial Opposition
PMQs is characterised by highly adversarial discourse between its two main protagonists, and five distinctive features of that adversarialism have been identified above. However, if PMQs discourse is essentially adversarial, does it serve any useful purpose beyond political point-scoring? In this final section, it is proposed that PMQs can be an important form of political opposition, taking as a case study its role in the British phone-hacking scandal (Bull, 2013).
The phone-hacking scandal was an ongoing controversy involving the News of the World and other British tabloid newspapers published by News International (now trading as News UK). The newspaper’s employees were accused of engaging in phone-hacking, police bribery, and exercising improper influence in the pursuit of publishing stories. Of particular concern was the revelation that journalists had accessed the voicemails of a missing 13-year-old schoolgirl.21 These allegations prompted not only a huge public outcry but the subsequent closure of the News of the World; according to its final editorial (10 July 2011), “Phones were hacked, and for that this newspaper is truly sorry... there is no justification for this appalling wrongdoing”.
Results and Discussion
In the context of this furore, the then LO (Miliband), across two sessions of PMQs (6 & 13 July 2011), launched a wholesale attack on both News International and PM Cameron. There were four principal themes to Miliband’s attack:
1. A demand for a public inquiry into the culture and practices of UK newspapers. This was subsequently granted by Cameron.
2. Opposition to News International’s takeover bid for broadcast organisation BSkyB.22 Were the bid successful, it would have created a UK media giant, whose revenues would dwarf all competition, including the BBC. The intended takeover aroused huge opposition, both within and outside Parliament, and was withdrawn on 13 July 2011.
3. Call for the resignation of Rebekah Brooks (chief executive of News International and former News of the World editor). Notably, the phone-hacking of Milly Dowler’s voicemail took place during her editorship of the newspaper. Brooks resigned as chief executive of News International on 15 July 2011.
4. Criticism of the PM for his appointment of another former News of the World editor – Andy Coulson – as his Director of Communications. Coulson was subsequently charged with phone-hacking (24 July 2012). He was found guilty of a charge of conspiracy to intercept voicemails (phone-hacking) in June 2014 and received an 18-month prison sentence.
Thus, although PMQs is often castigated as no more than a pointless exercise in political point-scoring, Miliband’s questions concerning the phone-hacking scandal related to substantive political issues – and arguably, considering subsequent events, achieved a certain degree of political success.
Conclusions
PMQs is notorious for its adversarial discourse, especially for the gladiatorial encounters between PM and LO. To what extent is this reputation justified? The main aim of this article has been to identify discursive features that are normative to PMQs and to consider if they can be conceptualised in terms of adversarialism. Through this analysis of discursive practices, it has been possible to specify what is meant by adversarialism in PMQs and to argue that it is the underlying norm of these leader encounters.
Thus, five specific discursive features of PMQs have been identified: face-threats in questions, personal attacks, the rhetorical use of quotations, equivocation by the PM, and traditional forms of address. Given that each of these features has been identified as customary practice in PMQs, we argue that they may all be regarded as normative.
Further consideration was given to whether these discursive features can be regarded as adversarial. In the study of face-threats (Bull & Wells, 2012), it was found that every question posed by LO Cameron to PMs Blair and Brown across eighteen PMQ sessions (108 questions) included at least one face-threat – in the sense that they might make the PM look bad (e.g., by questioning their competence) or by limiting their future freedom of action. Hence, LO questions to the PM may be considered highly adversarial. Personal attacks by their very nature are intrinsically adversarial. On average, almost one third of the LO-PM exchanges included personal antagonism (Waddle et al., 2019); these were also significantly more frequent for PMs the longer they were in office. It is clear from these data that personal attacks are common practice between the leaders, hence may be regarded as normative in PMQs.
The use of quotations is customary in PMQs discourse, hence may be regarded as normative. Politicians typically use quotations adversarially – either to deconstruct the argumentation, credibility, or leadership of opponents, or to enhance such qualities in themselves. Thus, they may quote themselves, other politicians, or expert sources. However, findings from the study reported herein (Bull & Waddle, 2019) suggest that quotations in PMQs are not intrinsically adversarial but what matters is their manner of use. It was argued that LO Corbyn’s use of questions sourced from members of the public – which was associated with a significantly reduced number of personal attacks by PM Cameron – might be seen in this case not as adversarial but an attempt to mitigate the ritualistic and customary verbal aggression of PMQs.
Equivocation by PMs in response to questions in PMQs is normative, as shown by the foregoing reported low reply-rates for both Cameron and May. It has been hypothesised that such equivocation may be a response to adversarial questioning, particularly for questions that create communicative conflicts. In broadcast political interviews, there is ample evidence to support this hypothesis, but it remains untested in the context of PMQs.
Addressing remarks to the Speaker and third-person references are conventional forms of PMQs discourse, hence may be seen as normative. However, whether adversarialism is mitigated by these ritual practices is open to question, given that they may be weaponised to bolster attacks on opponents. From this perspective, these ritual practices may also be seen as adversarial.
If PMQs discourse is essentially adversarial, does it serve any useful purpose beyond political point-scoring? In the final study above (Bull, 2013), it is proposed that PMQs can facilitate highly effective political opposition, taking as an illustrative case study its role in the British phone-hacking scandal.
Overall, five distinctive features characteristic of PMQs discourse have been highlighted. Given their frequent occurrence, we argue that all may be considered normative; we also propose that they may all be seen to reflect adversarialism in PMQs. However, it should be noted that these discursive features are by no means confined to PMQs. There is now an extensive research literature on political interviews – especially on face and facework (e.g., Bull et al., 1996) and equivocation (e.g., Bull & Mayer, 1993; Bull, 2003; Waddle & Bull, 2020b). Similarly, both personal attacks (Waddle & Bull, 2016, 2020b) and the strategic use of forms of address (Bull & Fetzer, 2006) have been analysed in the context of political interviews.
It is also noteworthy that the tradition of questioning heads of government in parliamentary settings is by no means confined to the UK. Bull and Waddle (2022, p. 84) note that:
In Canada, this convention is known as Question Period, in Australia and New Zealand as Question Time, in India as Question Hour. A fixed period for questions – Question Time – is a feature of many European parliaments. In some countries, such as Austria and Finland, the procedure for asking questions is even enshrined in the constitution. In the European Parliament, a Question Time – allowing members to ask questions in plenary sessions – was introduced in 1973 (Norton, 1993).
Hence, there are plenty of opportunities for further research analysing these different discursive techniques in different cultural settings. For example, one such study (Rasiah, 2010) was based on Question Time in the Australian House of Representatives, where questions without notice may be posed by MPs. The questions are usually directed at government ministers regarding their portfolios, and all ministers are expected to be present in the House to answer these questions. Rasiah used an equivocation typology devised by Bull and Mayer (1993) to classify what were termed “agenda shifts” (instances where politicians shifted the agenda of the question).
In the Introduction above, it was proposed that PMQs can resemble a form of verbal pugilism. Thus, just as professional boxers are evaluated on their boxing ability, so too the PM and LO are evaluated on their ability to deliver and counter verbal punches. As such, awareness of these social norms and evaluations undoubtedly influences the behaviour of party leaders, who thereby will seek to gain esteem from their fellow MPs. Adversarial questioning is the norm at PMQs; it is the expected role of LOs to scrutinise government policies and actions, and to call the government to account.
However, adversarial discourse is not to everyone’s taste. PMQs has been described as a kind of political “Marmite” (Allen et al., 2014) – people either love it or hate it. Hence, there are some who favour its abolition, some who want reform, and some who are fully supportive just as it is! Notably, however, PMQs remains the most famous parliamentary event worldwide. As such, in one form or another, its future as a regular key feature of UK politics looks assured.