Pragmatics of power On the intricacies of metadiscourse in the Labour leadership debates 2015 and 2016 Johannes Scherling Metadiscourse, the use of language to refer to any aspect of discourse, such as by clarifying one’s own stance or by criticizing other participant’s discourse conduct, is a common feature in debates in the political domain. By pointing to perceived inadequacies of discourse tone and conduct as well as by - more or less explicitly - stating agreement or disagreement or by explicating one’s arguments, metadiscourse here functions as a means to position politicians in relation to other politicians, but also to deflect from questions of content by diverting to the nature of the discourse. Arguably, heavy use of metadiscourse suggests that the person employing it feels the need to distance themselves from others, so as not to be seen as being too similar, as well as to justify their own positions, while a more moderate use of metadiscursive devices may indicate a higher degree of confidence. In this paper, I intend to show how this is relevant by analyzing metadiscourse in several instances of the UK Labour leadership debates of 2015 and 2016 between Jeremy Corbyn and his respective contenders, thereby also shedding light on the power struggles between Corbyn - who was largely identified as ‘anti-establishment’ - and the party establishment. 1. Introduction Metadiscourse can be a signifier of power. The fact that participants in a conversation feel it is necessary to revisit previous parts of discourse instead of progressing linearly, to clarify or justify their positions instead of asserting them, to criticize their opponents instead of proving them wrong, may on occasion be revelatory of a certain self-perception. It suggests that the participants in question feel the need for clarification or AAA - Arbeiten aus Anglistik und Amerikanistik Band 42 (2017) · Heft 2 Gunter Narr Verlag Tübingen Johannes Scherling 282 justification, feel a need for criticism which merely delegitimizes the other rather than providing alternatives, and thus arguably feel insecure and lacking the power to define the narrative of the discourse. On the other hand, metadiscursive elements can also be indicative of the wielding of power in that they may imply a certain hierarchy, projecting somebody as being in control of the narrative by employing such elements to criticize the opponents’ conduct or weak points, or to entangle them in contradictions. This is highly relevant in political debates, as such impressions of power - or its projection - based on the use, non-use, or type of metadiscourse may well affect the image a candidate is trying to bring across. It affects whether they will be perceived as a self-confident person who can assert their views and avoid becoming defensive following criticism, or an insecure one with the constant desire to emphasize and justify their standpoints. A leadership debate is a prime example where this unseen metadiscursive conflict takes place, as various candidates compete to be seen as the most confident, the most professional and the most decisive person in the room. The way that these candidates employ metadiscourse to organize, explicate and interpret their own discourse as well as that of their competitors can therefore be seen as an important factor in both how they perceive themselves as well as how the audience will perceive them. The Labour leadership debates in the UK in 2015 and again in 2016 are a good case in point and allow for an interesting juxtaposition of how (or whether) a candidate - Jeremy Corbyn in this case - acts differently when he is an outsider or when he is an incumbent, on the level of metadiscourse. In this paper, debates from the two time frames will be analyzed with regard to what kind of metadiscourse is employed, to what effect, and what this can tell the audience about the relative power a candidate wields. In Chapter 2, a few theoretical concepts will be clarified before Chapter 3 provides some background information about the leadership debates of 2015/ 6. Chapter 4 briefly outlines research methodology and corpus data, while Chapter 5 is dedicated to the analysis and juxtaposition of the respective candidates’ use of metadiscursive devices, which will then be discussed in Chapter 6. The purpose of this analysis lies not in a quantitative account of the use of metadiscourse (although some quantitative terminology is inevitable), but rather in a qualitative analysis of certain strategies employed by the candidates, their differences, and their impact. Pragmatics of power 283 2. The intricacies of metadiscourse 2.1. Metalanguage When people use language, they do so not only to refer to their mindinternal and -external worlds, but to a considerable extent also to discuss, clarify or object to the language code itself, a function that is known as ‘metalanguage’. As such, it stands in contrast to what is called object language with its focus on language-external entities. It is an important feature of human communication, and one of Charles Hockett’s (1963) design features of language, referred to as reflexivity, meaning that human language can be used to draw attention to the code itself. As such, metalanguage is a broader concept, and differs from metadiscourse in that the latter should be concerned only with the current, rather than any other text (cf. Mauranen 1993). The purpose of metadiscourse is not to add propositional content to a text (and is hence also not part of its truth conditions), but to signal the author’s presence and interaction with the text (cf. Vande Kopple 1985). This means that the two sentences X is the case and Let me be very clear that X is the case would have the same proposition (namely “X IS THE CASE”), but the latter would be employing metadiscursive elements to comment on the proposition. Such metadiscourse, i.e. the act of referring to the language code self or to language use, may be called reflexive as “it reflects on what is being, or has been said” (Mey 2001: 174). Schiffrin (1980: 231, cited in Hyland/ Tse 2004: 157) more broadly defines metadiscourse as the author’s linguistic and rhetorical manifestation in a text in order to “bracket the discourse organisation and the expressive implications of what is being said.” 2.2. Approaches to metadiscourse A very famous approach to the metafunctions of language was put forth by Halliday (cf. 1996), who distinguished between ideational, interpersonal and textual meanings. The first, ideational meaning, is concerned with the representation of the world; the second, interpersonal meaning, denotes the additional meaning that may be explicit or implied and can be infered about the speaker’s attitudes and feelings about an issue; the third one, textual meaning, is concerned with the organization of information (both on a macroand micro-level), its relation to other texts or text passages. Among scholars specializing on metadiscourse, there is disagreement on whether metadiscourse encompasses both the interpersonal and textual function, or only the latter, i.e. whether metadiscourse also includes notions such as attitude or perspective. According to Hyland and Tse (2004: 158), metadiscourse is an “open category” where “the same items can function as metadiscourse in some parts and not in others.” This means that the same words or phrases can function as meta- Johannes Scherling 284 discourse in some cases, but have no such function in others. Thus not their form, but their function in a text is important. Approaches to metadiscourse can roughly be categorized along the lines of the aforementioned disagreement, and consequently into what has been called the broad approach and the narrow approach, respectively. The broad approach, as adopted by linguists such as Markkanen et al. (1993) or Hyland (1998), incorporates both the interpersonal and textual functions of language; it also takes into account related categories such as stance, which express “personal feelings, attitudes, value judgements, or assessments” (Biber et al. 1999: 966). Within this approach, metadiscourse is seen as modifying and/ or highlighting certain aspects of the text and providing the writer’s attitude to it. For proponents of the narrow approach, on the other hand, interpersonal meaning is mostly considered to be outside of metadiscourse proper, as it conveys information about the participants rather than the text itself. This approach, represented by, among others, Mauranen (1993) or Dahl (2004), particularly explores all aspects of text organization (textual meaning), and thus focuses on the notion of reflexivity on text, addresser and addressee (socalled textual metadiscourse), where “reflexivity excludes modal expressions and stance from metadiscourse, because they reflect the writer’s attitude and perform the truth value of the proposition, they are textexternal states of affairs of the writer, and they do not show awareness of the text” (Tuomi 2009: 67). Based on this notion and drawing on Jakobson’s (1960) functional model, Ädel (2006) distinguishes between metadiscourse that is directed at the addresser (expressive), at the addressee (directive) and at the text itself (metalinguistic). Even within the narrow approach, however, there is no clear agreement on the exact extent of metadiscourse. Mauranen (1993), for example, has a strong focus on reflexivity, proposing only a distinction between highly explicit reflexivity (as in, e.g., In the following section, I will…) and reflexivity of low explicitness (particularly internal connectors such as moreover, firstly or it follows). Dahl (2004), on the other hand, maintains that rhetorical metatext - “meta-elements which assist the reader in the processing of the text by making explicit the rhetorical acts performed by the writer in the argumentation process” (1812), such as argue, believe or claim - should also be included. Schiffrin (1980), by contrast, makes a distinction between metalinguistic referents (so-called discourse deixis), metalinguistic operators (including evaluative words such as true, false, right, wrong) and metalinguistic verbs (such as say, tell, assert). This helps to illustrate that there is no single definition of metadiscourse and that the borders between what should be considered textual versus interpersonal meaning are blurred at best. For the present paper, I have adopted a hybrid approach in which I include rhetorical metatext to some extent, but mainly focus on textual metadiscourse, i.e. on linguistic items that are used to refer to the nature Pragmatics of power 285 of the conversation or back to previous parts of discourse, or that are used to structure the current discourse. I have, however, also chosen to include here references to past discourses in cases where the purpose was to show consistency (e.g., I have always said that) or, conversely, to delegitimize the opponent by illustrating supposed contradictions or antagonisms in a candidate’s discourse history. I consider these to be metadiscursive rather than intertextual (i.e. referencing to any other than the current text) as the purpose is not to borrow authority, not even primarily to point toward the existence of other texts, but to indicate that the current discourse is either in line with previous discourses or antagonistic to them, and hence they can be seen as a metadiscursive commentary, a strategy for validating or devalidating a text. 3. The Labour leadership debates 2015/ 2016 3.1. Nature of discourse and functions of metadiscourse in leadership debates According to Ilie (2003), public political debates fundamentally presuppose two somewhat contradictory conditions - a spirit of adversariality and a spirit of cooperativeness: adversariality because candidates must distinguish themselves and thus provide a clear choice to eligible voters and because they must establish preferability through acclaiming, attacking and defending (cf. Dudek/ Partacz 2009), i.e. by discursively constructing themselves in contrast to the other(s); cooperativeness because, as future party leaders, they need to present themselves as individuals who can also embrace and work with their opponents at the end of the day. Additionally, challengers to the leadership have a general need to be more aggressive than incumbent leaders, as they first need to establish their credibility and potentially undermine the incumbent’s as they cannot draw on previous experience (cf. Dudek/ Partacz 2009). Metadiscourse in such debates is both of a high value as well as highly frequent. As arguably the main concern in political debates lies in reinforcing one’s own credibility - often at the expense of one’s opponent(s) - candidates need to show professional competence, political stamina and consistency between statements (cf. Ilie 2003). For this purpose, metadiscursive elements are used to underline and emphasize each of those points: by employing explanatory metalanguage, professional competence can be highlighted; by using metadiscourse indicating strong opinions in the face of opposition, a candidate can show political stamina; and by referring back to previous texts and statements, a candidate can illustrate how s/ he remained consistent in their statements. Metadiscourse, hence, tends to be used frequently to highlight and promote the speaker’s image rather than to further expand on their ideas. Johannes Scherling 286 3.2 The Labour leadership debates After a devastating defeat in the British general elections, Labour leader Ed Miliband resigned in May 2015, resulting in a debate over the next Labour leader over the following summer months. The candidates running for leadership were three members of the party establishment - Liz Kendall, Yvette Cooper and Andy Burnham - and one backbencher - Jeremy Corbyn. Against all expectations, the vote led to a landslide victory of Corbyn (59.48%). This caused an uproar in the party and led to infighting between Corbyn supporters and the party establishment, which climaxed in a vote of no confidence by Labour’s parliamentary club in May 2016. Corbyn refused to step down, assured that he had the majority of party members behind him. The vote of no confidence did, however, result in a new leadership vote in September 2016, this time between the now incumbent Jeremy Corbyn and Owen Smith, which Corbyn won by an even greater margin (61.8%) (cf. Mason 2015, Eaton 2016). 4. Research question, methodology and data In light of the theory that challengers to leadership generally need to be more aggressive, this begs the question as to whether this also applies in the case of Jeremy Corbyn, who - within a year - moved from challenger to incumbent leader. Since metadiscourse is an important tool in establishing power and credibility in such debates, the aim is to establish whether and how the use of metadiscourse is different and whether and how metadiscourse is an instrument in establishing and emphasizing a position of power in the Labour leadership debates of 2015/ 6. The analysis of metadiscourse is conducted according to Ädel’s (2006) reflexivity triangle and thus distinguishes between text-oriented, addresser-oriented and addressee-oriented metadiscourse while making further distinctions on the level of function adapted from Hyland and Tse (2004) as outlined in Table 1. LEVEL OF ORIENTATION text-oriented addresser-oriented addressee-oriented LEVEL OF FUNCTION antagonizing/ criticizing antagonizing/ criticizing antagonizing/ criticizing taking/ holding the floor taking/ holding the floor taking/ holdingthe floor agreeing/ claiming common ground agreeing/ claiming common ground agreeing/ claiming common ground clarifying/ emphasizing/ justifying message clarifying/ emphasizing/ justifying message clarifying/ emphasizing/ justifying message Table 1: Analytical categories. Pragmatics of power 287 The individual functional categories were obtained through an inductive analysis of several leadership debates, in which metadiscursive elements were marked and then categorized for their function. Table 1 includes only the major categories that were identified. The boundaries are obviously not clear-cut, as for example metadiscourse functioning as clarification may also imply criticism of the opponent for misinterpreting an utterance. For this reason, the classification was done based on the dominant function. The label antagonizing/ criticizing applies to metadiscourse whose function it is emphasize differences between the candidates by pointing out contradictions or disagreements. The label taking/ holding the floor refers to metadiscursive acts that aim at giving a candidate the opportunity to speak or to retain the right to speak. Agreeing/ claiming common ground is concerned with metadiscursive attempts at showing that the candidates are on the same page, while clarifying/ emphasizing/ justifying message denotes any use of metadiscourse to return to a previous utterance in order to elaborate, repeat or justify its content. The data for the 2015 debates were taken from the Labour hustings hosted by Sky News (length: 1: 22: 42), Channel 4 (length: 47: 33), The Guardian (length: 1: 25: 22) and the BBC (length: 58: 24); those for the 2016 debates were obtained from the hustings hosted by Sky News (length: 56: 03) and the BBC (length: 59: 20), as well as the hustings in Birmingham (length: 1: 32: 31) and Glasgow (length: 1: 30: 55). As there were no scripts available for any of these, they were then broadly transcribed by the author for linguistic content. 5. Analysis The analysis that follows has a clear qualitative focus, though some quantitative information will also be provided. The reason for giving greater weight to a qualitative analysis is that the same metadiscursive element may have different functions which only reveal themselves through an analysis of context. Additionally some of the categories do not feature that often across the various texts. Moreover, the main purpose of the paper is to identify certain strategies and explore whether or not they were successful in projecting and/ or gaining power. 5.1 2015 Debates In the 2015 debates, four candidates competed for leadership: Liz Kendall (LK), Yvette Cooper (YC), Andy Burnham (AB) and Jeremy Corbyn (JC), with an interviewer (Int) asking the questions. At that time, a victory by Corbyn was not anticipated and hence he was debating from an underdog position. Johannes Scherling 288 Clearly, the genre being a debate for primacy, criticism and antagonism may be expected to feature largely, not merely with regard to content, but also with regard to the function of metadiscursive elements. Most metadiscourse identified with this function was textand/ or addressee-oriented, as engaging in extensive self-criticism is certainly not the aim of such debates. ADDRESSER-ORIENTED (1) YC: So we set out that alternative, oppose the 40% cuts, but don’t think we can do it just by printing money… and that’s my disagreement with Jeremy [Corbyn] on that because I think you have to do this in a way that is actually credible. TEXT-/ ADDRESSEE-ORIENTED (2) LK: You actually said that it was NATO that had, you know, provoked Russia... JC: No, I didn’t. I said that one thing provokes another. LK: You said they had not been unprovoked. JC: I said that NATO’s excessive and obsessive expansion during the 1990s has been a problem. ADDRESSEE-ORIENTED (3) LK: You don’t have to tell me about people who are suffering. Table 2: Metadiscourse functioning as criticism/ antagonism. In example (1) in Table 2, Yvette Cooper mentions that’s my disagreement with Jeremy, thereby explaining the nature of her previous comment, which is why the metadiscourse here is addresser-oriented. In pragmatic terms, however, it is implied that she is criticizing Corbyn’s position. In this sense, the metadiscourse may be argued to be directed towards the text as well as towards Corbyn, too. In example (2), which is part of a longer dispute between Jeremy Corbyn and Liz Kendall, both Kendall and Corbyn are using metadiscursive elements to refer back to a previous utterance regarding NATO by Corbyn. It is metadiscursive in that it qualifies the following utterance as an utterance that was (or was not) made in the past, and hence is meant to associate Corbyn with a particular ideology. The purpose of this, however, is not merely text-related, in that they want to revisit and discuss Corbyn’s standpoint, but rather criticism by Kendall and thus an attempt to discredit Corbyn with her interpretative reproduction of his utterance (i.e. addressee-oriented). Corbyn’s use of metadiscourse in his reply, conversely, is meant to clarify what he actually said, and in turn to criticize/ discredit Kendall. Example (3), on the other hand, is addressee-oriented and functions as antagonism to Corbyn’s implicit claim that Kendall does not care about the troubles of ordinary people. Pragmatics of power 289 TEXT-/ ADDRESSEE-ORIENTED (4) LK: For some people like Jeremy and others, the Iraq War does wipe out for them all the good that we did as a government. But for me... but for ME… JC: I just said a whole lot of other things. LK: You said you so profoundly disagreed with him including on economic policy ADDRESSER-ORIENTED (5) JC: I don’t say it’s all the fault of NATO. (6) YC: Let’s be clear, I am not criticizing Jeremy’s values in any way or anything that Jeremy thinks. Table 3: Metadiscourse functioning as clarification/ justification. Example (4) in Table 3 is taken from a stretch of discussion in which Liz Kendall attacks Jeremy Corbyn for what is perceived as damning criticism of the Blair era, right after Corbyn points out what he perceives to have been negative as well as positive aspects. Again, the exchange employs references to previous parts of discourse (using a representative speech act), whose purpose is not to quote and reiterate the issues, but to clarify what is seen as a purposeful misrepresentation by the other. In this sense, it is - similar to (2) - directed at the text, but also implicitly addressee-oriented, as it is employed as a means to criticize. Examples (5) and (6) are both worth mentioning in that they involve a negation rather than an assertion, i.e. explaining the nature of an utterance in terms of what it is not. In both cases, the metadiscourse is addresser-oriented, as its function is to justify and clarify the purpose of a previous utterance. However, in each of these examples, arguably, the inference that is invited - drawing on Lakoff’s (2004) notion of frame-negation in which by negating a proposition, it is exactly this proposition that is being evoked 1 - is exactly the opposite, namely that Corbyn is entirely blaming NATO and that Cooper is criticizing Corbyn’s values. ADDRESSEE-ORIENTED (7) YC: Can I just finish? (8) JC: If I can finish, Andy... No, but I wanna finish if I may (9) AB: I’ve said I... Well, let me answer your question, Liz. Let me answer your question. It’d be nice if you’d let me answer your question 1 An example by Lakoff (2004) is US-President Richard Nixon, who famously stated, I am not a crook, arguably therefore associating himself with the ‘crook’-frame in the audience’s mind. Johannes Scherling 290 (10) LK: I don’t think Jeremy’s and my politics is anything, like, the same. I think there’s a challenge for centre-left parties across Europe, on the one hand, avoiding the kind of austerity that sees people left behind, though I have to say the kind of fantasy politics of Syriza, which would see us out of government for a generation... JC: Liz, can I help you on this? Table 4: Metadiscourse functioning as an attempt to hold/ take the floor. Taking and holding the floor are undoubtedly standard practices in any discussion, but even more so in a leadership debate where participants are competing in impressions of self-representation. Consequently, all of the participants employ metadiscourse to this effect. From its very nature, such metadiscourse is addressee-oriented, as the purpose is to either get the opponent to yield the floor or make it clear to them that one does not intend a speaker change. In examples (7) and (8) in Table 4, Yvette Cooper and Jeremy Corbyn, respectively, are struggling to hold the floor, and by using the metadiscursive phrase can I finish (thereby commenting on the ongoing discourse), not only imply that they still have something to say, but also implicitly criticize their opponent(s) because they were interrupted, which hence portrays their opponents as ignoring basic rules of politeness. Conversely, examples (9) and (10) show attempts by Andy Burnham and Jeremy Corbyn to take the floor from the current speaker - Liz Kendall in both cases. In (9), Burnham uses let me answer your question to imply that he considers the speech act of asking the question as completed and hence that it is his turn now. His repetition of the same metadiscourse serves to emphasize this. In (10), Corbyn is employing linguistic politeness (can I help you on this) to frame his attempt at taking the floor as a friendly act of assistance, when in fact he means to correct what Kendall just outlined. This also has a certain patronizing quality as both are knowledgeable professionals, but serves to position him above her in the hierarchy of power. It can be seen as metadiscursive because it is an implicit evaluation of her previous utterance and thus a comment on the text. ADDRESSEE-ORIENTED (11) JC: …and Yvette’s right on the exploitation of foreign workers. (12) YC: I think we have agreement that this is a serious issue, I just think that these policy proposals are the same that we had at the time of the general election... (13) JC: I don’t see how it could make it any better because you end up with destruction, of course, you end up with civilian deaths, of course, what’s that message across the region? Does this make to some people ISIL a more attractive operation or do we try to contain them and cut them off, as I think is what Liz was saying a moment ago. Pragmatics of power 291 (14) JC: The crash of 2008/ 9 was brought about by deregulated banks by subprime mortgage crisis, and it was brought about by the ineffectiveness of control of what the financial markets did. I think about that everyone would be agreed. Bringing the banks back into public ownership, again we’d probably all be agreed that was the right thing to do. Table 5: Metadiscourse functioning as agreement/ establishing of common ground. Table 5 illustrates metadiscourse used to express agreement, an important aspect in leadership debates as outlined above. The potential new leader must show that s/ he will be able to embrace his/ her competitors and cooperate with them in the future. In the data analyzed, this type of metadiscourse was predominantly used by Jeremy Corbyn, whose attempts to claim common ground may be revealing of either a character trait or his increased need as an underdog to show he can lead, or both. Alternatively, and seeing that he was largely perceived as being on the extreme fringes of the party, it might be an attempt to show he is not an outsider entirely. In examples (11) and (13), we can see that Corbyn - rather than merely explicating his standpoint - is referring to previous utterances by his opponents in order to indicate commonalities between their points of view, by evaluating Cooper’s utterance as right (11), while framing his opinion as, essentially, a paraphrase of Kendall’s statement before (13). Example (12) shows Cooper establishing a tentative common ground (metadiscursively stating that the evaluation of the issue under discussion as serious is an evaluation they both share) only to contrast this (I just think) with an utterance attempting to construct her opinion as different from the rest. Example (14) can be seen as a prime example of the ‘leadership posture’ that Corbyn is attempting to present himself as having. By presupposing everybody’s agreement on the issue discussed, by categorizing what was said in terms of a representative speech act (agree), he aims to preclude any disagreement by inviting the others to join his argument, a strategy similar to that employed in (11) and (13). The 2015 debates are, in short, characterized by fights for the floor, by criticism of opponents and by justifications of the participants’ own views. What stands out somewhat are Corbyn’s attempts to create common ground and using metadiscourse for conciliatory purposes, which the others struggle to reject or at least not openly embrace. The quite frequent references to previous discourse - both current and past - show attempts to deconstruct opponents’ views, while their counterattempts to disclaim the individual interpretations of such discourse by merely negating the previous assertion creates the odd impression that they are actually confirming its content. The overall impression of the debate as capitalizing on conflict and demarcation is expected; the only surprising aspect Johannes Scherling 292 is represented by Corbyn’s repeated attempts to create common ground, in order to highlight commonalities, rather than differences. 5.2 2016 Debates By the time the 2016 leadership vote took place, Jeremy Corbyn had been Labour leader for almost a year, so his position had changed from challenger to incumbent. As such, he faced resistance from Labour’s parliamentary club which forced him to call for a new leadership election in the autumn of 2016. His challenger, this time, was Owen Smith, previously a member of his shadow cabinet who had - in May 2016 - resigned and co-triggered the new debate. The 2016 debate was characterized by a high degree of emotional involvement (both by the competitors and the audience), as the internal quarrels had created a considerable amount of turmoil, which is also reflected in the more aggressive tone and adhominem accusations that dominated them. Against this background, it is perhaps not entirely surprising that by far the most frequent type of metadiscourse was related to criticizing and antagonizing the opponent. This function is realized in various ways, such as by referring to perceived inconsistencies in past and present discourse to discredit the other, by criticizing claims and the demeanor of the other or, by far the most striking case, by criticizing the nature of the debate itself. TEXT-/ ADDRESSEE-ORIENTED (15) OS: He used to say he wanted a contest for the Labour leadership every year. He gave an interview a couple of years ago in which he said that he didn’t want to be Labour leader because he wanted to be able to disagree with his party. (16) Int: Just in terms of the conduct of this leadership campaign: when people say they hate the former leader and prime minister, when there’s booing and hissing, do you... You said you wanted a kinder politics. JC: I don’t do booing and I don’t do hissing. Int: But some of your supporters clearly do. JC: Well, I ask people to be kind. In some cases it takes longer than others. OS: You see, Jeremy laughs it off again and he said earlier in the debate... JC: You’ve done plenty of it. OS: He said he laughs it off... and I haven’t, Jeremy, I haven’t. That’s not true. (17) OS: We shouldn’t have booing here tonight, because it’s just true. TEXT-/ ADDRESSER-/ ADDRESSEE-ORIENTED (18) JC: I think some of the language used by some of my colleagues in parliament has been more than unfortunate. I think we should treat each other Pragmatics of power 293 with respect. Where there is a political difference, say so, it doesn’t have to be personal; it never should be abusive. (19) OS: The reason I resigned is because I do not think you are serious [...] someone who says, frankly, a lot of platitudes [...] (20) OS: I agree with you that there is no room for personal abuse in politics, but you know, you’re not the only socialist in the village... [laughter] I don’t know why you’re laughing! [to audience] Jeremy’s been pretty [undecipherable] with his language on occasion. JC: Owen, I’ve never used bad language on people, you never find me using bad language on individuals. OS: I got attacked a bit this week for using the word ‘lunatic’, for which I have apologized JC: Good, thank you. OS: For which I have apologized. Jeremy Corbyn called a Tory MP a lunatic in 1992 and used the word again in parliament in 1993, so... [laughter] Interviewer: I’ll give you a few moments to respond to that and then we move on... JC: I’d rather we go on with discussing politics. Table 6: Metadiscourse functioning as criticism/ antagonism. As example (15) in Table 6 illustrates, Owen Smith is referring to past statements by Jeremy Corbyn (thus making this text-oriented), in which he called for yearly leadership elections and maintained he would never want to be leader. The purpose, clearly, is to delegitimize his current conduct (addressee-oriented) as he disapproved of the call for an early election after only one year. Though the reference undoubtedly has an intertextual function, it is also metadiscursive in that in makes clear that what follows is a past statement of Corbyn’s and thus attempts to position it in contrast to Corbyn’s recent statements. Examples (16), (17), (18) and (20) are vivid illustrations of metadiscourse related to the heated nature of the debate (text-oriented), in which the audience, in particular, intervened with paralinguistic signals of disapproval such as laughter, hissing and booing - specifically directed against Smith. In (16), Smith is attacking what he perceives to be Corbyn’s dismissive attitude regarding audience responses (addressee-oriented), by saying Jeremy laughs it off again, which also implies that Corbyn not only does not mind the behavior of some of his supporters, but that he is actually complicit in it by belittling Smith’s concerns. The metadiscursive nature of the phrase concerns the fact that in saying so, Smith is attempting to reframe and classify Corbyn’s preceding statement as ridiculing rather than asserting. The rising insecurity of Smith can be seen in that he also accuses the audience directly of not taking what he says seriously, as in (20), rather than implic- Johannes Scherling 294 itly, as in (17). The ad-personam nature of the debate is most apparent in (20), where Smith first asserts that he apologized (addresser-oriented) for using the word lunatic to describe Corbyn (without performing an apology there and then), only to then attempt to turn the tables by accusing Corbyn of the same language use. However, as the incident lies more than 20 years in the past, his metadiscursive - as well as intertextual - comment is ridiculed by the audience, thus defeating the purpose of his previous criticism directed at the audience. Corbyn, picking up on the audience’s mood, responds to this comment by metadiscursively disqualifying the content of Smith’s criticism and asserting that he would rather go on with discussing politics, i.e. that he would like to shift from the level of metadiscourse to that of content, which is both textand addresseeoriented as it is a statement that he considers the discussion tedious, but at the same time a criticism of Smith, who is bringing up the topic. In (18), Corbyn also shows a different approach to Smith in dealing with such ad-hominem metadiscourse by saying that such criticism never should be abusive, which directly relates to Smith’s rather open and personal attacks on Corbyn, as in (19), in which he delegitimizes Corbyn’s political ideas as platitudes (addressee-oriented). Overall, Corbyn gives the impression of being more in control of the argument, as he resists getting involved in Smith’s direct attacks on the behavior of the audience, which in turn earn Smith only more negative reactions. TEXT-/ ADDRESSEE-ORIENTED (21) JC: I’m sure we can agree on two things: firstly, all of us together are going to make sure we defeat any aspects of antisemitism in our party and within our society. That I’m sure we’re absolutely agreed. OS: We ARE agreed, but I’m not sure you’re entirely committed to it, Jeremy. (22) JC: Owen, why don’t we try and discuss how we can make sure that the party policies are effective and that we win the next election by after this election unifying and coming together around the issues, which we DO agree on. (23) JC: As you know I received reports, I was concerned and asked Shami Chakrabarti to undertake an inquiry into this, which she did. She’s proposed a series of rule changes as well as an information and education program in the party, which will help. [...] We’ll review the process next year, to make sure that we are getting rid of any vestige of antisemitism whatsoever within our party and I’m sure you will agree with me on that. Table 7: Metadiscourse functioning as agreement/ claim to common ground. By contrast, metadiscursive elements delineating agreement or attempts to establish common ground are much rarer, and if they occur, they are almost exclusively used by Corbyn. Table 7 shows some of the few in- Pragmatics of power 295 stances where such metadiscourse is employed. In (21), (22) and (23), we see Corbyn attempting to force Smith’s hand, by almost presupposing agreement as in I’m sure you will agree with me on that (23), thus using rhetorical metalanguage with a clear orientation towards the addressee, while also being text-oriented as it refers to and qualifies the previous discourse. As (21) illustrates, this strategy is occasionally successful, as Smith is required to take a stand. It is interesting to note how Smith struggles between agreeing to an assertion that he cannot reject and not giving his opponent the privilege of embracing his claim. He does this by commenting that they are agreed, but by immediately attempting to question Corbyn’s commitment, hence not denying the accuracy of the propositional content, but Corbyn’s actual intentions. In this sense, they both are in line with their respective roles as challenger and incumbent, as it is Smith’s only chance to establish himself in contrast to Corbyn rather than being ‘suffocated’ by too much common ground. This is similarly the case in examples (24) and (25) (see Table 8), in which Smith overtly agrees, but implicitly criticizes Corbyn, as indicated by the use of but immediately following the explicit agreement, which is achieved through rhetorical metalanguage. AGREEMENT (explicit) - CRITICISM (implicit) (24) OS: I’m pleased of course that Jeremy condemns anti-Semitism [...], but it is very worrying that [...] many in the Jewish community do not feel welcomed in the Labour movement right now. (25) OS: I agree we need to come back together whoever wins this. But you talk about trying to unite the party. I find that quite hard to reconcile with something your campaign did just this evening, which was to publish a list - a deselection list if you like Table 8: Metadiscourse functioning as explicit agreement and implicit criticism. The second largest area in which metadiscourse is employed is for the purpose of clarifying or justifying one’s own position. A great part of this is realized by references to the text, in the sense that the competitors attempt to explicate their intentions or to cast doubt on the other’s intentions. In some cases, and particularly with Jeremy Corbyn, this is also done by directly addressing the audience and/ or Owen Smith (i.e. addressee-oriented). In example (26) in Table 9, Corbyn is defending himself against criticism regarding low ratings. By responding with a metadiscursive element (we are putting the message out) relating to the manner in which Labour tries to communicate with the public (addresserand text-oriented, as it both serves as a statement on the discourse, i.e. that they are doing enough, and a justification, i.e. that there’s nothing more to improve), he attempts to shift the focus from the messenger to the Johannes Scherling 296 medium and the audience, i.e. to render it a problem not of the message and its creator, but of its transmission and interpretation. He immediately follows this up by directly addressing the audience (likewise in (30) and (31)) using rhetorical metalanguage (I tell you this) and asking them for patience until they have won the elections and implemented their policies. In (27), Smith is asked to provide justification for why he feels he is more electable than Corbyn. His reply features a metadiscursive tactic of indicating a longer list of characteristics to follow (for starters), when in fact only one aspect is mentioned, but the implication that there is actually more remains, creating friction in the argument. In (28) and (29), respectively, Smith uses two different tactics to deal with criticism/ clarification request directed at him. While in (28), he negates a claim by clarifying what he did not say (thus arguably reinforcing what he is attempting to deny, see reference to frame negation above), in (29) he merely states that everything has been said already, i.e. that he does not feel it is necessary to clarify his position. One of the effects of Corbyn’s addressee-oriented metadiscourse is to show he is in a higher position from which he can explain his ideas and ‘educate’ the audience, especially by indicating the kind of speech act to follow (e.g. the use of tell as a representative speech act, but implicitly also a directive as it seems to express a command to listen). The reason this announcement may be seen as metadiscursive is that it qualifies what follows as something of import, as something that needs emphasis, and that it instructs the audience to pay increased attention to what is to follow. The impression of his clarifications is therefore not so much defensive as it is offensive, in that such metadiscourse seems to imply the issue of comprehension lies with the addressee(s), rather than in his own explanations. TEXT-/ ADDRESSER-/ ADDRESSEE-ORIENTED (26) Int: Why are your poll ratings the worst of any opposition Labour leader a year on, that we’ve seen in history? JC: We’re putting the message out in the best way that we can. We’re putting it out through the media as best as we can. We’re putting it out through the social media as best as we can. And I tell you this: when the leadership debate is over, and the party comes together, to defeat the Tories [...] people will see what the alternative is. (27) Int: Your case seems to be that you’re more electable than JC. What’s your evidence? OS: Well, for starters... ADDRESSER-ORIENTED (28) OS: I didn’t say we should be having talks with Islamic State, I absolutely didn’t. What I said was that chances are we would never be able to negotiate with Islamic State. Pragmatics of power 297 (29) OS: I think I’ve made my position clear on that. ADDRESSEE-ORIENTED (30) JC: I tell you this now: there has been a re-energization of politics over the past year. (31) JC: We can all learn from what we do, but I tell you this: parliament has to understand that we’re there on the sufferance of the people who’ve put us there. The questions we should put, should be the questions that come from the people. (32) OS: Well, let me be clear that I will never play any part in splitting the Labour party. Table 9: Metadiscourse functioning as clarification/ justification. It becomes quite clear that the 2016 debate was led in a much more direct and pugnacious way, seeing that by far the most frequently identified categories revolved around criticizing the other and justifying oneself. The very fact that so much of the discourse involved references to the purportedly aggressive tone of contributions helps to illustrate that the debate took place in an emotionally (over)charged atmosphere, unlike the 2015 debates. 6. Discussion and conclusion Direct comparison of the findings in the usage of metadiscourse in the Labour leadership debates of 2015 and 2016 show that, while both include an extensive amount of metadiscursive language, their functions vary quite noticeably between the two sets of events. The 2015 debates were characterized by a great effort to hold or take the floor as well as by attempts at clarification/ justification and the establishment of common ground, whereas the 2016 debates showed a foregrounding of direct criticism, often inviting dismissive reactions from the audience, expressed through paralinguistic elements such as booing or hissing. This can be explained, at least in part, due to the difference in the setting (debate between four people vs between two people) and in the preconditions (heated atmosphere and deep ideological trenches between the parliamentary club and party members). In 2015, there were four candidates, who were all challengers to the party leadership; by 2016, the situation had changed. There was an incumbent leader (Corbyn) who was challenged by an establishment candidate (Smith). This necessitates a different style in the debates, as the challenger needs to act more aggressively in order to overcome the general advantage of the incumbent candidate (cf. Dudek/ Partacz 2009), while the latter needs to show that s/ he can balance out the various opinions in Johannes Scherling 298 a party and bring people together. In this respect, however, it is quite striking that Corbyn stands out as the person with by far the highest usage of metadiscourse functioning to claim common ground, even though in the first debate, he, too, was a challenger. While his opponents, in both 2015 and 2016, demonstrably use metadiscourse to this effect as well, they virtually never do it for the sake of agreement, but merely use it as a stepping stone to distance themselves from the stances of the other candidates. Another difference is that Corbyn uses such claims for common ground proactively and for topics that are of a very general nature and difficult to disagree with, such as fighting anti-Semitism, and thus forces his competitors to put their cards on the table; however, at the same time he in part appropriates their purpose as competitors by getting them to agree with his positions and then creating the impression that these are all originally his ideas. In the 2016 debates, Smith shows a general tendency to try and discredit Corbyn by referring to past statements or by highlighting what he sees as contradictions in Corbyn’s persona; through the use of such metalanguage, Smith attempts to avoid having to agree with his opponent, not on the basis of content, but on the basis of sincerity. But in doing so, he also occasionally discredits himself, such as when he calls Corbyn out for the use of a derogatory reference to an MP from 20 years ago, which evokes outward derision from the audience. Looking at the data in its totality, it becomes clear that Corbyn makes effective use of metadiscourse, while his opponents, especially Owen Smith, more frequently misfire regarding the underlying intention of their utterances, such as when Smith is attempting to criticize Corbyn, but creates a comical effect or when he criticizes the audience only to be taken less seriously. This may, in part, be attributable also to an antiestablishment mood that was prevalent during the debates. Overall, the data shows a greater amount of metadiscourse functioning as clarification or justification by Corbyn in the 2015 debates, which has a defensive purpose (e.g. what I meant was or what I said was). On the other hand, in 2016, a decrease of defensive metadiscourse becomes apparent, while offensive clarification metadiscourse increases (e.g. I am very clear that, I simply say to them or I tell you this now). This suggests a difference in interpersonal power relations between the debate participants in 2015 and 2016, which can clearly also be traced to the different roles they have with Corbyn taking on the role of the incumbent leader in the 2016 debates, thus requiring and allowing him to speak from a position of power. Smith, on the other hand, occasionally uses metadiscourse to refer to the nature of the discourse, by e.g. addressing what he perceives to be an aggressive and personal tone. Such discourse, which at times evolves into extended sequences, represents a shift from the level of content to the level of code, and thus marks a change in tactics to criticize and thus discredit Corbyn on the basis of the discourse conduct of his supporters Pragmatics of power 299 rather than on content issues; a tactic which clearly backfires, as the audience reactions show. Analyzing the metadiscursive devices, especially with regard to the 2016 debates, paints a certain picture: metadiscourse is employed to the effect of revealing particular self-perceptions of the competitors. While in 2015, there seem to have been rather balanced power relations, by 2016 this had changed. A lot more space was dedicated to criticism, in particular by Owen Smith, while Jeremy Corbyn attempted to create common ground, not so much to convince Smith, but rather to preemptively corner him and force his hand. It is clear that Corbyn makes different use of metadiscourse in the second set of debates and that this is one of the factors that help to establish his power to dominate the discussion, not only through the specific use of discourse, but also by its absence. Not using metadiscourse here functions as a signifier of certainty, of conviction, as it helps frame an utterance not in terms of overt explication and subjective interpretation, but in terms of an assertion, which suggests both conviction and professionalism. The same applies to his rhetoric of addressee-oriented metadiscourse when clarifying his position, which suggests a knowledgeable - if at times slightly patronizing - position. It would be rather bold to claim that metadiscourse decided the outcome of the Labour leadership debates; it is, however, reasonable to suggest that it can serve at least as a revealing indication of how the participants position themselves in the debate and what they perceive their power status to be. In this sense, as was put forth at the beginning of this paper, metadiscourse can indeed be argued to serve as a signifier for power - or at the very least as a signifier for the perception of that power. References Ädel, Annelie (2006). Metadiscourse in L1 and L2 English. Amsterdam: John Benjamins. Biber, Douglas/ Stig Johansson/ Geoffrey Leech/ Susan Conrad/ Edward Finegan (1999). Longman grammar of spoken and written English. Essex: Longman. Dahl, Trine (2004). “Textual metadiscourse in research articles: a marker of national culture or of academic discipline? ” Journal of Pragmatics 36. 1807- 1825. Dudek, Patrycja/ Slawomir Partacz (2009). “Functional theory of political discourse. Televised debates during the parliamentary campaign in 2007 in Poland.” Central European Journal of Communication 2. 367-379. Eaton, George (2016, September 24). “How Jeremy Corbyn won the Labour leadership election.” New Statesman. http: / / www.newstatesman.com/ politics/ uk/ 2016/ 09/ how-jeremy-corbyn-won-labour-leadership-election [accessed 29 May 2017]. Johannes Scherling 300 Halliday, M.A.K. (1996). “Systemic Functional Grammar.” In: Keith Brown/ Jim Miller (eds.) (1996). Concise Encyclopedia of Syntactic Theories. Oxford: Pergamon. 321-325. Hockett, Charles (1963). “The problems of universals in language.” In: Joseph Greenberg (ed.) (1963). Universals in Language. Cambridge: MIT Press. 1-29. Hyland, Ken (1998). “Persuasion and context: The pragmatics of academic metadiscourse.” Journal of Pragmatics 30. 437-455. Hyland, Ken/ Polly Tse (2004). “Metadiscourse in Academic Writing: A Reappraisal.” Applied Linguistics 25(2). 156-177. Ilie, Cornelia (2003). “Discourse and metadiscourse in parliamentary debates.” Journal of Language and Politics 2(1). 71-92. Jakobson, Roman (1960). “Linguistics and Poetics.” In: Thomas Sebeok (ed.) (1960). Style in Language. Cambridge, MA: MIT Press. 350-377. Lakoff, George (2004). Don’t think of an elephant! - Know your values and frame the debate. White River Junction: Chelsea Green. Markannen, Raija/ Margaret S. Steffensen/ Avon Crismore (1993). “Quantitative contrastive study of metadiscourse problems in design and analysis of data.” Papers and Studies in Contrastive Linguistics 28. 137-151. Mason, Rowina (2015, September 12). “Labour leadership: Jeremy Corbyn elected with huge mandate.” The Guardian. https: / / www.theguardian.com/ politics/ 2015/ sep/ 12/ jeremy-corbyn-wins-labour-party-leadership-election [accessed 29 May 2017]. Mauranen, Anna (1993). Cultural Differences in Academic Rhetoric: A Text-Linguistic Study. Frankfurt a.M.: Peter Lang. Mey, Jacob (2001). Pragmatics: An Introduction. 2 nd ed. Malden, MA: Blackwell. Schiffrin, Deborah (1980). “Metatalk: Organisational and evaluative brackets in discourse.” Socio-logical Inquiry: Language and social interaction 50. 199-236. Toumi, Naouel (2009). “A Model for the Investigation of Reflexive Metadiscourse in Research Articles.” Language Studies Working Papers 1. 64-73. Vande Kopple, William J. (1985). “Some Exploratory Discourse on Metadiscourse.” College Composition and Communication 36(1). 82-93. Corpus data Labour debate 2015 hosted by Sky News: https: / / youtu.be/ FdT1Mz1cYqA Labour debate 2015 hosted by Channel 4: https: / / youtu.be/ 996Lnep7Hfk Labour debate 2015 hosted by The Guardian: https: / / youtu.be/ UuCRIuHrKAo Labour debate 2015 hosted by the BBC: https: / / youtu.be/ jrNfhEfowlM Labour debate 2016 hosted by Sky News: https: / / youtu.be/ h-kpFHSxvJI Labour debate 2016 hosted by the BBC: https: / / youtu.be/ 1KVrN9F6F6U Labour debate 2016 in Birmingham: https: / / youtu.be/ J7W1NA0rMJE Labour debate 2016 in Glasgow: https: / / youtu.be/ 8qJb75V90p8 Johannes Scherling Department of English Studies Karl-Franzens-University Graz Austria