Winds of War: Epistemic and effective control in political

use of these resources in the discourse of war and presents a case study on their use by two politicians, President George W. Bush and Prime Minister Tony Blair, in political speeches and statements during the build-up to the second Iraq war. Results indicate significant qualitative and quantitative differences in the preferred stance strategies in the discourse of the two politicians.

. El marco analítico se inspira en la distinción de Langacker (2009de Langacker ( , 2013 entre el nivel epistémico y el efectivo en la gramática, y en la oposición que se deriva entre pretender el control de las concepciones de la realidad y el control de las relaciones existentes a nivel de la realidad. Las estrategias epistémicas conciernen la legitimación epistémica de las aserciones, al aportar soporte epistémico y justificación epistémica para la proposición (Boye 2012). El control efectivo tiene por objeto la legitimación de acciones y planes de acción. El despliegue estratégico conjunto de actos de posicionamiento epistémico y efectivo realiza una estrategia combinada de control sobre la aceptación por parte de oyente/lector de concepciones de la realidad y de planes de acción. Este artículo estudia el uso estratégico de estos recursos en el discurso de la guerra, y presenta un estudio de caso sobre su uso por dos politicos, el Presidente George W. Bush y el Primer Ministro Tony Blair, en discursos politicos y declaraciones durante los preparativos para la segunda guerra de Irak. Los resultados indican diferencias significativas cualitativas y cuantitativas en las estrategías de posicionamiento preferidas por los dos politicos.

INTRODUCTION
This paper explores epistemic and effective legitimisation strategies in the discourse of war. The framework is based on a model for the analysis of stancetaking which posits two macro categories of stance, the epistemic and the effective (Marín Arrese 2011Arrese , 2015Arrese , 2021. In characterising these two macro-categories, the model draws on Langacker's (2009;2013) distinction between the epistemic and the effective level, and the systematic opposition thereof between striving for control of conceptions of reality and control of relations at the level of reality. Epistemic stance pertains to speaker/writer's striving for control of conceptions of reality, which involves their estimation of the veracity of the event designated and the likelihood of its realization, and/or their specification of the sources of information whereby they feel entitled to make a factual claim. The category of effective stance pertains to the positioning of the speaker/writer with respect to the realization of events, to the ways in which the speaker/writer carries out a stance act aimed at determining or influencing the course of reality itself.
Epistemic stance strategies may serve ideological purposes in the striving for control of hearers' acceptance of information, by overcoming the epistemic safeguards of their audience and offering 'guarantees' for the truth of their assertions (Chilton 2004(Chilton , 2011Hart 2011). Speakers have a vested interest in justifying their assertions, since they may be held politically accountable for the veracity of their claims and for their decisions and actions based on those claims. Effective stance strategies are directly aimed at the legitimisation of actions, and the coercion of hearers/readers' acceptance of those actions and plans of action, by discursively constructing them as necessary, desirable, righteous, or feasible, and constructing the controllers as fully committed to those actions and plans. It will be argued that the joint deployment of epistemic and effective stance acts effects a strategy of combined control over hearers/readers' acceptance of particular conceptions of reality and of plans of actions and constitutes the basis for the legitimisation of specific actions and decisions.
This paper presents a case study on the use of epistemic and effective stance in discourses of war, and more specifically on the second Iraq War. Studies on the Gulf war and the second Iraq war have mainly focused on the use of metaphor or the use of topoi in the representation of social actors and events (Lakoff 1991;van Dijk 2005;Cap 2006; Mohammed Alrubaian 2019). The present case study examines the discourse of US President George W. Bush and UK Prime Minister Tony Blair, in the build-up to the Iraq War, from 11 September 2001 to the capture of Saddam Hussein in 2003. The aim is to characterize and to make visible their preferences in the deployment of epistemic and effective stance resources as legitimising strategies in the justification of the Iraq invasion, which may relate to differences in their perceived political involvement and interests, and thus in their rhetorical goals, or differences derived from the interactional identities of the two politicians.
The paper explores the following research questions: (a) the degree to which there are similarities or significant differences in the use of expressions of epistemic and effective stance in the discourse of the two politicians. (b) the degree to which there are similarities or significant differences in the pattern of distribution in the categories of epistemic and effective stance resources in the discourse of the two politicians, that is, whether they display a preference for particular categories of epistemic and effective resources.
It is hypothesized that there will be significant differences in the deployment of epistemic and effective stance strategies by Bush and Blair, and their preferred legitimising strategies, which may derive from: (a) interactional identities and intercultural discourse practices; (b) politicians' perceived political involvement and responsibility for the war.
In order to address these aims and hypotheses, the paper presents a quantitative study on the discourse of Bush and Blair during the build-up to the second Iraq war with the following research objectives: (i) to identify, annotate and quantify epistemic stance markers in order to compare the categories and the variation in the deployment of these resources; (ii) to identify, annotate and quantify effective stance markers in order to compare the categories and the variation in the deployment of these resources.
The paper is organized as follows. Section 2 discusses the domains of epistemic and effective stance. The case study is presented in Section 3. The results are discussed in section 4. The final section is devoted to the conclusions.

STANCE & CONTROL IN DISCOURSE
The multifaceted nature of stance has been linked to the expression of beliefs, attitudes or emotions, and positioning in the discourse, and has been addressed from diverse and often overlapping perspectives and frameworks, such as the work on stance (Biber and Finegan 1989;Biber 2015), affect and emotion (Ochs & Schieffelin 1989, DuBois & Kärkkäinen 2012), evaluation (Thompson and Hunston 2000Thompson and Alba-Juez 2014), stancetaking and dialogicality (DuBois 2007;Englebretson 2007), and the sociolinguistics of stance (Jaffe 2009). Stancetaking strategies involve the use of a variety of lexico-grammatical elements which "overtly express an evaluative frame for some other proposition" (Biber at al. 1999: 967).

EPISTEMIC & EFFECTIVE CONTROL
Epistemic stance pertains to speaker's striving for control of conceptions of reality, for 'epistemic control' in the discourse (Langacker 2013). Speakers make strategic use of epistemic stance expressions in order to provide justificatory support for the proposition (cf. Boye 2012), with the aim of controlling hearers' acceptance of "the validity of a proposition" as "part of a conception of reality" (Langacker 2009: 303), and thus persuading them to accept the veracity of their claims.
Effective stance pertains to speaker's striving for control of relations at the level of reality (Langacker 2013), to their positioning with respect to the realization of actions and of plans of action. The relations profiled by effective control elements "represent actions or assessments aimed at determining what actually happens" (Langacker 2009: 303). As Langacker (2009: 167) notes, effective control elements are "force-dynamic at the social level, involving the exercise of authority because I am trying to make the hearer actually do something (not just think of it or acknowledge it)". Speakers make strategic use of effective stance expressions with the aim of urging for action and persuading or coercing hearers to accept those actions and decisions.
By exploiting 'epistemic legitimising strategies', speakers may indirectly manipulate hearers' acceptance of the veracity of their assertions, as a prior condition for the discursive legitimisation of actions, in the latter case through more direct coercive 'effective legitimising strategies' (Marín-Arrese 2011, 2021).

PERSUASION, MANIPULATION & COERCION
Persuasion and persuasive intent have been addressed from various perspectives in the literature. From the field of evolutionary Psychology, Sperber et al. (2010: 359) have argued that "humans have a suite of cognitive mechanisms for epistemic vigilance, targeted at the risk of being misinformed by others", that is, the hearer's ability to detect deceptive language use ('cheater detector', Cosmides 1989), and their ability to assess the reliability of the source and the validity of the information ('source tagging', Cosmides and Tooby 2000). Chilton (2011: 772) notes that "human language could not have evolved without a cooperative assumption that predisposed humans to believe other communicators". But this natural tendency to accept as true and valid the communicated information is coupled with the need "to be epistemically vigilant and check for cheaters". Cheaters will then need to find ways "to circumvent the epistemic defences of their audiences", which would account for the development of the ability both "for persuasion" and "for detection of persuasive intent". Chilton (2011: 774), however, makes the caveat that "the tendency to accept is variable within and across situations, societies, polities and cultures", which raises the issue of why some claims go unchallenged or why some people fail to adopt a critical stance.
Critical discourse analysts focus on the use of deceptive tactics by the dominant groups (cf., Fairclough 1989;Van Dijk 2006). Pragma-cognitivists have studied the cognitive processes involved in manipulative language use (cf., Saussure 2005). Sorlin (2017: 133) argues for an approach which focuses on "the informational/beliefbased/cognitive impact of the phenomenon", that is, she defines manipulation in terms of "the (a) ulterior motives behind a communicative effort" and "the (b) course of actions the discourse generates". Sorlin (2017: 135) has argued that there is no clear dividing line between persuasion and manipulation, and that "what manipulators tamper with, cognitively speaking, is the Hearers' interpretative possibilities, limiting their 'freedom of interpretation' so to speak". The manipulative spectrum, according to Sorlin (2017), may be conceived as a continuum involving interpretative freedom, with 'persuasion' at one end and 'coercion' at the other, as in the following diagram: Figure 1. The manipulative spectrum (Sorlin, 2017: 136) Persuasion presumably tends to be more "belief-based", involving the manipulator's intent to persuade Hearers "into adopting the Speaker's beliefs" (cf., 'epistemic control') through covert deception (Sorlin, 2017: 135). But in coercive manipulation, the manipulator would seek to impose some form of pressure on Hearers, "to coerce the victims into acting in a certain way (through linguistic/pragmatic --not physical --means)" (Sorlin, 2017. 135) (cf., 'effective control'). In sum, making others believe and do things, and thereby satisfying the interest of speakers/writers, is one of the main components of manipulative language use (van Dijk 2006).
An additional dimension of manipulation is the play on the Hearers' emotions and emotional needs (cf., Marín-Arrese 2021). This dimension is undoubtedly present in the discourse of the two politicians, though it is beyond the scope of this paper.

EPISTEMIC STANCE
Categories of epistemic stance discursively construct (the reality status of) events as certain/probable or possible (epistemic modality), and the representation of events as verifiable (evidentiality), as fact (factives), or as beliefs (cognitive attitude) (Marín- Arrese 2015Arrese , 2021. The focus in this study is on those expressions which reflect a higher degree of commitment to the validity of the communicated information. Epistemic stance (EP) resources include the following categories: (a) Epistemic modals (EM): These express the speaker's assessment concerning the existence of the event designated, its actual or potential occurrence. Expressions of high or medium degree of epistemic support include epistemic necessity (must, certainly, of course) and probability modals (will, probably, likely).
(1) I recall a few weeks ago talking to an Iraqi exile and saying to her that I understood how grim it must <EP, EM> be under the lash of Saddam. (Blair-T13) (2) Your response and probably <EP, EM> that of most people would have been very similar to the response of some of you yesterday on Iraq. There would have been few takers for dealing with it and probably <EP, EM> none for taking military action of any description. (Blair-T6) (b) Indirect Inferential Evidentiality (IIE): Evidentials primarily indicate the source of information (Aikhenvald 2004), and the evidence on the basis of which the speaker feels entitled to make a factual claim (Anderson 1986). A broader conception of evidentiality also includes an estimation of source reliability, as Chafe (1986) posited in his seminal publication.
The focus here is on the subtype of Indirect Inferential Evidentiality (IIE). These expressions may indicate both perception-based and conception-based inferences (cf. Diewald and Smirnova 2010), as well as inferential meaning based on information acquired through oral and written texts (communication-based or report-based inferences) (cf. Marín-Arrese 2015a), as the following examples show. It is interesting to comment that Bush sometimes uses the perception predicate see, albeit with an indirect inferential value to refer to intentions of the agents described, but which strategically evokes direct perception of resulting actions. Expressions found with an indirect inferential value include the following: see, seem/s, clearly, obviously, plainly, it is clear/plain.  (Kiparsky and Kiparsky 1970). These expressions indicate the speaker's assignment of knowledge status to a representation, that is, full support for the proposition, so that they reflect the speaker's high degree of epistemic support and speaker commitment. Personal cognitive factives (CFV) include expressions such as: I/we know, I/we recall, I/we remember. Impersonal factives (IFV) are expressions such as: It is true, The fact is, The truth is.
(5) We know <EP, CFV> from prior weapons inspections that Saddam has failed to account for vast quantities of biological and chemical agents, including mustard agent, botulinum toxin and sarin, capable of killing millions of people. We know <EP, CFV> the Iraqi regime finances and sponsors terror. And we know <EP, CFV> the regime has plans to place innocent people around military installations to act as human shields.  In the discourse of both Blair and Bush we often find instances of the rhetorical strategy of repetition of the same stance expression at the beginning of successive clauses or sentences, in order to emphasise the veracity of their claims. Metadiscourse is defined by Ädel (2010: 75) as "reflexive linguistic expressions referring to the evolving discourse itself or its linguistic form, including references to the writerspeaker qua writer-speaker and the (imagined or actual) audience qua audience of the current discourse". In these cases of metadiscourse framing, the validity assigned by hearers to the communicated information would normally depend on the "evidential standing" of the speaker as source (White 2006: 64). In this case study, an obvious feature of authority and credibility which may be exploited by the speakers is the fact that the two politicians are supposedly 'experts' in matters of national security.

EFFECTIVE STANCE
Effective stance strategies discursively construct proposed actions and decisions as necessary (Deonticity: "We must act"), desirable (Normativity: "We abide by social norms and social morality"), and possible (Potentiality: "We can"), and construct the controllers as fully committed to those actions and decisions (Intentionality: "I/We have the will", "I/We am/are committed") and to their enforcement (Directivity: "I urge you to").
This paper focuses on those effective expressions which indicate a higher degree of speaker's striving for control, that is, a higher degree of involvement in achieving their aims and in their commitment to some course of action.
Effective stance (EF) categories and expressions are the following: (a) Deonticity (DM): Within this category we include expressions of deontic modality and participant-external necessity (must, should, cannot) (cf. van der Auwera and Plungian 1998). Similar meanings are expressed by so-called semi-modals or quasimodals (have to, need to) (cf. Leech 2003). Examples found include: must, should, cannot, can't, has/have to, need/s to.
As a rhetorical device, in the discourse of Blair we often find reiteration of the same stance expression for emphasis and to stress the need for some course of action, as in the following example. (9) Mr Speaker, in the face of this evidence, our immediate objectives are clear. We must <EF, DM> bring Bin Laden and other Al Qaida leaders to justice and eliminate the terrorist threat they pose. And we must <EF, DM> ensure that Afghanistan ceases to harbour and sustain international terrorism. If the Taleban regime will not comply with that objective, we must <EF, DM> bring about change in that regime to ensure that Afghanistan's links to international terrorism are broken.   (Krug 2000). The expression of resolve or strong intention is found in some 'effective matrix predicates' (we are determined, we are resolved) (cf. Langacker 2009).
(12) We will <EF, INT> starve terrorists of funding, turn them one against another, drive them from place to place, until there is no refuge or no rest. And we will <EF, INT> pursue nations that provide aid or safe haven to terrorism. Every nation, in every region, now has a decision to make. Either you are with us, or you are with the terrorists.  Rhetorical effects involving repetition are often found in the discourse of Bush. It is quite common to find a cumulative effect through the use of various expressions of the same category involving the effect of an increase in force of commitment.
(13) We will <EF, INT> not wait to see what terrorists or terror states could do with weapons of mass destruction. We are determined <EF, INT> to confront threats wherever they arise. And, as a last resort, we must <EF, DM> be willing to use military force.  This example provides an interesting progression from intention to determination, with a final show of deontic force involving all the American citizens. Also interesting is the strategic play on intersubjectivity (evoking inclusive collective commitment). The invasion of Iraq was particularly controversial in the UK, as it attracted widespread public opposition, including that of a substantial number of Blair's MPs, as well as the Cabinet resignations of Guy Cook and Clare Short. At the time of the invasion and subsequently in the Iraq War Inquiry, Blair faced severe criticism over the policy itself and the circumstances in which it was decided upon. From 11 September 2001 to the 2003 invasion of Iraq, both Bush and Blair delivered a substantial number of public and institutional addresses centred on the issue of the 'War on Terror'. This paper examines a variety of texts within that time frame and up the capture of Saddam Hussein. (i) BUSH: Address to the Nation, address to Congress, radio addresses, and key public speeches (22 texts, 27,917 words). (ii) BLAIR: Parliamentary statements, and key public statements and speeches (16 texts; 27,988 words).

THE TEXTS
The variety of text types inevitably means that there will be certain genre-related variation in the use of epistemic and effective stance resources, which is beyond the scope of this study. Also, since most of the texts by Blair are parliamentary statements, there might be both cultural and normative differences in the way the speaker displays rhetorical strategies of persuasion and legitimisation, in comparison to the various public addresses by Bush.

RESEARCH QUESTIONS, HYPOTHESES AND RESEARCH OBJECTIVES
The corpus study looks at the presence and patterning of stance markers in each sub-corpus, as expressions of epistemic or effective stance. The case study explores the following issues: (i) Whether there is variation in the use of epistemic and effective stance resources in the discourse of the two politicians; (ii) Whether there is variation in the pattern and distribution of the categories of effective and epistemic stance in the discourse of the two politicians.
The assumption is that the particular discourse domain (political discourse), and the specific topic (war) and rhetorical goals (persuasion/coercion, justification of the legitimacy of war) will in great part determine the type and frequency of epistemic and effective stance expressions. It is hypothesized that there will be significant differences in the deployment of epistemic and effective stance strategies by Bush and Blair, and their preferred categories of epistemic and effective legitimising strategies. Variation in the use of stance resources between speakers may derive from: (a) differences in interactional identities and intercultural discourse practices in argumentative and persuasive styles, (b) differences in the politicians' perceived political involvement and degree of responsibility and accountability for the war. The paper posits the following null hypotheses: (i) H01 (Dependent variable: DV/Independent variable: IV): There is no difference between the frequency of epistemic stance expressions and effective stance expressions (DV) in the discourse of Bush and Blair (IV). The null hypothesis is expressed as H01: π1 = π2 (ii) H02 (Dependent variable: DV/Independent variable: IV): There is no difference between the pattern of distribution of categories of epistemic stance (DV) in the discourse of Bush and Blair (IV). The null hypothesis is expressed as H02: π1 = π2 (iii) H03 (Dependent variable: DV/Independent variable: IV): There is no difference between the pattern of distribution of categories of effective stance (DV) in the discourse of Bush and Blair (IV). The null hypothesis is expressed as H03: π1 = π2 The following research objectives are set in relation to the above hypotheses: (a) Identification of stance expressions in the texts: epistemic and effective. (b) Analysis of the texts, identifying the categories of epistemic and effective stance, and manual annotation of data.

RESEARCH PROCEDURE
The texts were annotated manually, and searches were made using Monoconc to ensure that all instances of stance expressions were identified. Frequency comparison of stance expressions was carried out using UCREL's Log-likelihood (LL) test, which indicates whether there is relative overuse or underuse of the observed frequency (O1) in Corpus 1 in comparison to the observed frequency (O2) in Corpus 2: overuse in corpus 1 relative to corpus 2 is indicated by '+', and underuse is coded with a '-' sign (differences in scores are significant at > 3.84 score). The log-likelihood test was applied to each category of stance in the two corpora.
Relative frequencies and possible associations between the variables was tested using the SIGIL online utilities for statistical inference (Corpus Frequency Test Wizard), which provide frequency comparisons between corpora, and give relative frequencies of tokens for each category (per thousand words). A Chi-square test was run to establish whether there was association between the use of stance resources by the two politicians (Bush vs. Blair) and the frequency of tokens in each of the categories of epistemicity and effectivity. The value of significance was established at p<0.05. Results for FCT values of X 2 are marked as: ***significant at p < .001, **significant at p < .01, *significant at p < .05.
The criteria for the inclusion/exclusion of expressions found in the texts were the following: (i) The focus is on those expressions which reflect a higher degree of epistemic support and epistemic justification for the assertion or claim. Excluded: modals of "neutral support", possibility and uncertainty (may, perhaps).
(ii) Expressions having propositional scope where the profiled occurrence is immediate to the ground (I know); all the expressions which are cases of non-immediacy (I was aware) have been excluded (cf. Langacker 2009). (iii) Included only high and medium commitment expressions. Modal will with a predictive or a predictability meaning was included. Excluded: modal will designating future and planned events. Excluded: modal will under the scope of a cognitive attitude or cognitive factive predicate.
(iv) Excluded: modals will and would in an irrealis context, such as if-clauses and other expressions within if-clauses (protasis), and non-assertive modalities or contexts (questions).
(v) Excluded: modal expressions under the scope of matrix predicates not indexing the speaker as conceptualizer. (vi) Excluded: discourse-pragmatic markers functioning as management markers or style disjuncts (you know, you see, I mean), and metadiscourse expressions involving references to the audience.

RESULTS AND DISCUSSION
The results for effective and epistemic stance resources, Table 1, show that the preferred type of legitimisation in the discourse of the two politicians is the use of effective strategies. This preference is particularly marked in the case of Bush, totalling 76.6% of the stance resources used. Blair also shows a clear tendency to use effective expressions, 57.8%, but balanced by a considerable number of epistemic stance expressions, 42.2%. There are also differences between the two politicians in the overall figures for the ratios: Blair makes a significantly higher use of stance expressions in general (R= 23.22), which is patent in the higher use of epistemic expressions (R= 9.79), which practically doubles the ratio of use by Bush (R= 4.98). With respect to Hypothesis 1, a chi-square test of independence was performed to examine the relation of the distribution of the categories of effective and epistemic stance (dependent variable) and the speaker variable (independent variable). Results (chi-square statistic with Yates correction= 47.3061, p-value is < .00001; significant at p < .05) indicate that the relation between these variables is significant so that the null hypothesis, H01, may be discarded.
Blair seems to effectively combine epistemic and effective resources in striving for control of both conceptions of reality and control of relations at the level of reality in urging for action. In contrast, Bush appears to be focused on the more direct strategy of control of relations at the level of reality. Table 2 shows the results for the categories of Epistemic Stance for both politicians. With respect to Hypothesis 2, there is a significant relationship between the pattern of distribution of the subcategories (dependent variable) and the two speakers (chi-square statistic= 18.5799, p-value= .00095; significant at p < .05). Results indicate that the null hypothesis may be discarded and we may adopt the alternative hypothesis, Ha2, which claims that variation is due to association between the variables. The frequency comparison test (FCT) for each of the subcategories shows that differences between both politicians are significant for the categories of cognitive attitude (CGA), impersonal factives (IFV) and marked enunciational positioning (MEP). Overuse of expressions in these categories is very significant in the discourse of Tony Blair: the subjective expression of beliefs (CGA), the claims to what is true or fair (IFV), and the overt marking of the act of enunciation (MEP) appear to be distinct features of his rhetoric of legitimisation (cf., results of his oral evidence in the Iraq War Inquiry hearings, November 2009 -February 2011, Marín-Arrese 2015b). His discourse stance seems to reflect Blair's concern with his image and personal credibility, and with the acceptance of his claims by the audience. The Log-likelihood (LL) score shows significant overuse in these categories and in the total of expressions of epistemicity (-49.14) in the discourse of Blair.

EPISTEMIC STANCE: RESULTS
In striving for epistemic control, both speakers show a preference for expressions of epistemic modality, which involve a lower degree of commitment to the validity of the communicated information. As Bybee et al. (1994: 179) observe concerning epistemicity and truth, «markers of epistemic modality indicate something about less than a total commitment by the speaker to the truth of the proposition». It would seem that both speakers are reluctant to fully commit to the conceptions of reality they propound. Table 3 shows the results for the categories of Effective stance. With respect to Hypothesis 3, the pattern of distribution of the categories of effective stance is also significant (chi-square statistic= 36.8401, p-value= < 0.00001; significant at p < .05), so that we may discard the null hypothesis and adopt the alternative hypothesis, Ha3, that variation is due to association between the dependent variable of the categories of effective stance and the independent variable of the speakers. These differences might have a contextual basis, in that most of the speeches and statements by Bush are addressed to the American people, to whom he makes pledges regarding their victory over Sadam, guaranteeing the future safety of the USA. Blair, on the other hand, is mostly addressing the House in his statements, so his main concern is that of urging the MPs to back his policy of support for the invasion of Iraq. In the discourse of Blair, the higher ratio and overuse of expressions of Deontic modality (DM) and Directivity (DIR), in urging the House for action is also congruent with the contextual basis, as is the overuse of Normativity (NRM) with Blair's predilection for 'ethical discourse' (cf., Charteris-Black 2005, for Blair's use of metaphor in political discourse; Charteris-Black 2011, for use of metaphor by Bush).

CONCLUSION
This paper has explored the strategic use of epistemic and effective stance resources in the discourse of war, and more specifically, in a series of key speeches, statements and addresses by Bush and by Blair, on the 'War on Terror' and the involvement of Iraq, from 11 September 2001 to the capture of Saddam Hussein in 2003. It has been argued that the joint deployment of epistemic and effective stance resources conforms a strategy of striving for combined control in the discourse over hearers/readers' acceptance of conceptions of reality and belief states about the 'War on Terror', which constitute the basis for control of relations at the level of reality, for the legitimisation of particular actions and decisions taken thereof. Epistemic stance resources realize the indirect legitimisation strategy of providing epistemic justificatory support, knowledge and information, for the speaker's proposed conception of reality, while effective stance resources provide a more direct strategy of urging for action by claiming the necessity, desirability, righteousness or feasibility of the proposed plans of action, or the speaker's commitment towards those proposed actions.
Regarding the categories of epistemic stance, both speakers show a preference for expressions of epistemic modality (EM), which reflect a lower commitment to the validity of the communicated information. Overall, results for epistemic stance are low for Bush in comparison with Blair, who shows significant figures for expressions of cognitive attitude (CGA) and impersonal expressions of factivity (IFV). Also notable is Blair's use metadiscoursal comments involving marked enunciational positioning (MEP), which seems to play a distinct role in his goal of epistemic control in the discourse. One may surmise that in having to answer to Parliament, Blair is under greater pressure regarding responsibility for his assertions and communicated information, as well as accountability for actions and decisions thereby legitimised (cf., the Iraq War Inquiry, 2009Inquiry, -2011. Results for effective stance expressions point to a significant preference for effectivity over epistemicity in the discourse of Bush, whose preferred strategy is the use of expressions of intentionality (INT). Since Bush is addressing the American people in most of his speeches and statements, he is primarily concerned with expressing his commitment to the victory over Sadam, which involves both retribution for the 9/11 attacks and the guarantee of future safety in the USA. On the other hand, results for Blair show a marked preference for Deontic modality (DM) and Directivity (DIR), which reflect his main concern in urging Parliament to support the invasion of Iraq, and side with the USA in international relations.
Further research is necessary regarding the extent to which these stancetaking patterns may reflect speakers' interactional identities, in speeches on non-war issues, and the possible variation in different genres of political discourse. Some studies seem to point to a tendency towards mitigation and decreased commitment in the discourse of Blair, a result of his preference for expressions of epistemic modality (EM), as well as the use of agnostic stance (Brandt 2004) markers reflecting reluctance to commit to his assertions in his role as witness in the IWI (Marín- Arrese, 2015). A study on epistemic and effective stance in the farewell address by Bush shows a preferred tendency for the use of effective stance, as well as the use of expressions of epistemic modality (EM) as his main choice in the domain of epistemicity (Hidalgo-Downing, 2021).