Avaliação social como recurso persuasivo em discursos políticos: Clinton vs. Trump

RESUMO Mudanças políticas têm intensificado os debates sobre como certos discursos estabelecem agendas e influenciam eleitores. Nesse escopo, a Análise Crítica do Discurso e a Linguística Sistêmico-Funcional, juntamente com o arcabouço teórico-metodológico da Avaliatividade, demonstraram ser bem-sucedidas na investigação da ideologia em textos, bem como na abordagem de tópicos como a polarização política. Portanto, a presente pesquisa tem como objetivo analisar a avaliação social da Teoria da Avaliatividade como um dispositivo ideológico e persuasivo nos discursos de nomeação de Clinton e Trump, em 2016. Os resultados enfatizam usos de estratégias retóricas que podem levar os eleitores à preferência de certos conjuntos de valores e crenças. A análise revelou que, naqueles discursos, os candidatos usaram estratégias diferentes para persuadir os eleitores, construindo abordagens discursivas ideológicas negativas (Trump) e positivas (Clinton), aplicando a avaliação social como estratégia retórica central.


Introduction
The American presidential election of 2016 occurred on November 8, 2016. The 58th quadrennial American presidential election had the Republican candidate businessman Mr. Donald Trump beat the Democratic candidate former Secretary of State Mrs. Hillary Clinton.
Mr. Trump took office as the 45th President "despite being considerably outspent in the general election; losing the national popular vote by over two percentage points but still winning the Electoral College, something that has happened rarely in American history" (SABATO et al., 2017, p.7).
Following the election's unforeseen outcome, protests broke out across the United States with some street clashes continuing for several days. On the other hand, President Trump also received support for his triumph. News all over the country and from the world over aired clear indications that the country was poles apart. Beyond the election, questions at the core of ideological views of candidates were a sign that polarization in the United States had reached its zenith.
Values and beliefs are at the core of political polarization. As stated by Moreno (2019, p.1), more than the left-right axis of political competition, the "new politics" is defined by "the salience of particular concerns among the mass electorates"; and "new issues", or issues of the 'new politics'such as environmentalism, minority rights, feminist issues, and gay and lesbian expressionshave affected not only the meaning of party competition but also the social basis of party support.
More recently, there is an overall perception that President Trump's election has intensified the division in the US territory (PILKINGTON, 2017;SHI et al., 2019). It is reasonable to say that Clinton-Trump race represented a major ideological rupture in the American society, that later was reflected in other nations. Many factors have to be considered when judging the success or failure of political candidates; fact is that discourse is an important aspect of the equation since it embeds their ideologies. Van Dijk (2002, p.1) posits that politicians usually have at least two ideologies, a professional one to function as All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0 politicians, and "the socio-political ideologies they adhere to". The latter identifies them as members of parties or social groups: "conservative and progressive politicians, socialists and neoliberals, Christian-democrats, greens, nationalists and racists, and so on. And it is likely that these ideologies will show up and combine in the discourse of politicians." In the spectrum of political agendas and discourses, it is not a goal of this study to provide a sense of the rightness or wrongness of any political orientation. I intend to shed some light on some of Trump's and Clinton's discursive strategies by analyzing one speech from each candidate. More specifically, the corpus is composed by the two speeches sustained during their nomination in their respective parties' conventions: The Republican National Committee (hereafter RNC) -Trump, and The Democratic National Committee (henceforth DNC) - Clinton. 1 Conventions have become important instruments for candidates to shape their images through long speeches loaded with ideological meaning. Decker sees the conventions with "a specific purpose: to sell or redefine" (2016, p.2). For that reason, in this study, the focus will be given to social evaluation and how efficient it is for political messages from a rhetoric standpoint. Moreover, this paper aims to analyze the Clinton-Trump nomination speech to address how, due to a cumulative groove of semantic patterning, the target audience is dynamically positioned to interpret each candidate's ideological stance through persuasion deployed to interact with their electors and set ideological agendas.
To achieve the previously mentioned goals, this paper draws on underpinnings of Systemic-Functional Linguistics (hereafter SFL) (HALLIDAY, 1994) together with Critical Discourse Analysis (hereafter CDA). Moreover, I apply Martin's (2000) theoretical and methodological framework of Appraisal to show how interpersonal resources can be strategically deployed in discourse to shape interpersonal relations (power and engagement) and political identities. From that common ground, the then-candidates were able to set their agendas whilst establishing their political views to that specific audience and context. Next, I will discuss the combination of SFL and CDA to evince ideology in text.
Authors from this tradition (e.g. FAIRCLOUGH, 1992;FOWLER, 2013;WHITE, 2006) postulate that SFL is an efficient theoretical-methodological approach to study ideological complexes in texts once it considers language as a social functional phenomenon.

CDA and SFL: A Critical Analysis of Ideology
As part of the political practices when candidates engage in campaigns, political discourses construct (and are constructed by) ideologies. Van Dijk (2006, p.115) understands ideologies "sociocognitively defined as social representations of social groups, and more specifically, as 'axiomatic' principles of such representations". Sustaining social groups selfimages, ideologies organize their "identity, actions, aims, norm and values and resources, as well as its relations to other social groups" (van DIJK, 2006, p.115).
Furthermore, from a socio-cognitive perspective, "ideologies are defined as basic systems of fundamental social cognitions and organizing the attitudes and other social representations shared by members of groups" (van DIJK, 1995, p.1). Sociopolitical or sociocultural ideologies are intertwined with discourse (LI, 2010;van DIJK, 2002). A basic premise of CDA is that language in use implies ideological meanings while restraints the use of language and meanings involved (FOWLER, 2013).
According to Fuoli (2015), the SFL-based CDA has tended to focus on investigating the ideological implications for patterns in discourse. The adoption of SFL is commonly done by the much-adopted CDA in critical text analysis (e.g. DEDAIC; NELSON, 2012), as long as conceived necessary when the issue of ideology is at stake.
In Halliday's SFL (1994), language in use deals with the mechanism of text structure, function and meaning of language. Having a pragmatic approach, the author proposes an analysis of language in social contexts where a particular lexicogrammatical choice is made under the influence of the social and cultural contexts. For SFL, language is structured to build simultaneously three types of meanings (metafunctions): ideational, interpersonal and textual. This semantic simultaneity stems from the fact that language is a semiotic system, a  (HALLIDAY, 1994).
Plenty of studies (e.g. WHITE, 2006;FAIRCLOUGH, 2003) apply SFL's functional view of linguistic choices as indexes of meanings aligned with CDA premises: both are guided by the underlying assumption that linguistic forms and choices express ideological meanings. SFL provides an analytical tool for the systematic examination of power relations in the text, as well as the motivations, purposes, assumptions, and interests of text producers.
With its focus on the selection, categorization, and order of meaning in the microstructure of the sentencerather than only in the macro level of discourse -, SFL is especially useful for a systematic analysis of linguistic traits at the micro-level of discourse and a critical insight about the organization of meanings in a text.
In essence, SFL is a multi-perspectival model, designed to provide analysts with complementary lens for interpreting language in use. SFL can be coupled with CDA for the organization of meanings in metafunctions: (1) textual, related to the organization of both ideational and interpersonal meanings; (2) ideational, meanings associated with the representation of the "state-of-affairs" (HALLIDAY, 1994); and (3) interpersonal, dealing with the negotiation in social relations. In regard of the latter, the development of researches of other systems (e.g. involvement, evidentiality, tense, among others) has expanded the role of the interpersonal metafunction and the attention devoted to Appraisal.
At the core of interpersonal metafunction (SFL), Appraisal is a discursive-semantic system that provides a model to investigate evaluation in texts and its cumulative semantic patterning. Since Martin's groundbreaking research (2000), studies have been associating CDA and Appraisal to unveil ideological nuances in political discourses (WODAK, 1999).
Some other studies and frameworks also approached ideology in text. For example, Van Dijk's ideological square (1998;2002) investigates how underlying ideologiesand the social attitudes and personal opinions influenced by themare generally polarized, and can be organized in a combination of general discursive strategies in all levels of discourse analysis of opposing pairs, emphasizing (positively) and de-emphasizing (negatively), respectively, the ingroup ("Us") versus the outgroup ("Them"); also based on studies of Social Psychology (TAJFEL, 1978). Likewise, De Fina (1995), Fetzer (2014) andKranert (2017) deal with discourses of the ingroup vs. the outgroup through pronominal choices. Wodak (1999, p.33) discuss political discursive polarization and provide a few notions thereof: (1) "outgroup" is conceived as the "enemy", and it is defined in terms of difference, also oftentimes derided or criminalized; (2) "We-discourse" builds a "good-orbad" dichotomy (ingroup vs. outgroup) based on assumptions of moral quality and moral character; (3) recurrent positive self-image of the ingroup; (4) scapegoating; and (5) exaggeration of the negativity when referring to the outgroup sayings and behaviors.
Another recent wave of "we vs. them" type of studies weighs on the flourishing of right-wing and left-wing populism (e.g. BLOCK; NEGRINE, 2017). According to Wodak (2015, pp.1-2), populism is different from other rhetorical devices because "it does not only relate to the form of rhetoric but to its specific contents". She coined two concepts: (1) the "politics of fear", based on the "us vs. them" argument, and (2) the "arrogance of ignorance", based on a common-sense and anti-intellectualism appealing attitude. Although these two rhetoric devices of populist discourses are regularly applied in political contexts, this paper has not the intention of depicting populismconsidered a broad topic within different fields and with no agreement in the literature in this regard. Nonetheless, March (2017, p.282) points out that "host ideology is more important than populism per se", and "what is often called 'thin' or 'mainstream' populism' is not populism but demoticism (closeness to ordinary people)", and advises analysts not to "label parties 'populist' just because their rhetoric is demotic". In this line of thought, as showed latter in analysis, both Trump and Clinton apply different persuasive strategies based on demoticism and aligned with the strategies of "politics of fear" and "arrogance of ignorance".
Li (2010) draws on Van Dijk's (1995;1997) approach to reveal how social practices are linked to ideological meanings in texts. A similar approach will be adopted in this study to evince the use of the rhetoric device of social evaluation to address questions like: does political discourse, regardless of whether it is disguised under "objective" expressions, favor certain values of attitude? Is it true that this persuasive resource coupled with underlying culture may lead to the preference of certain sets of values of attitude? Do the same discursive parameters apply in all cultural contexts, taking into account the class setting of the voters?
All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0 Before investigating the corpus to address the aforementioned questions, the next two sections will discuss Appraisal as a much-adopted system for political discourse analysis.

Appraisal: A Functional Model of Linguistic Analysis
The Appraisal was devised by Martin (2000) and developed by collaborators such as Christie and Martin (2005), Scherer, Schorr and Johnstone (2001), Macken-Horarik (2003), among others. According to Martin (2000, p.145), "appraisal deals with the semantic resources used to negotiate emotions, judgements and appreciations. They are resources used to expand and commit to these evaluations", being the term coined to denominate all evaluative uses of language, including those in which speakers/writers adopt particular views or ideological positions. Martin and White (2005, p.61) conceive realization as the idea that language is a stratified semiotic system involving cycles of coding at different levels of abstraction.
Besides phonology and lexicogrammar, abstraction is a level of coding, adds , to refer to discourse semantics. Abstraction emphasizes the fact that it is concerned with meaning beyond the clause (with text, co-text and context). This level is concerned with various aspects of discourse organization, including the question of how people, places and things are introduced in the text and kept track of once there (identification); how events and states of affairs are linked to one another in terms of time, cause, contrast and similarity; how participants are related as part to whole and sub-class to class (ideation); how turns are organized into exchanges of goods, services and information (negotiation); and how evaluation is established, amplified, targeted and sourced (appraisal). Martin's underpinnings are placed in discourse semantics because it deals with different levels of abstraction beyond the lexicogrammar, thus decoding meanings in the discursive flow in an ample variety of systems. The realization of an attitude tends to emerge from a phase of discourse more prominently, irrespective of grammatical boundaries. Hence, discourse semantic "cannot be seen simply as a list of cohesive ties relating one grammatical unit to another, but as a further level of structure in its own right" (MARTIN, 2014, p.9), expressing the need to move out of lexicogrammar to generalize the evaluative meaning common to this kind of scatter.
All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0 Following Martin (2000), Appraisal can be located as an interpersonal system at the level of discourse semantics where it co-articulates interpersonal meaning in three subsystems for attitudinal positioning: attitude, engagement and graduation, all of which deal with evaluative interpersonal meanings of participants, processes and circumstances (core elements to represent the experience). Attitude relates to evaluations of emotional/affective values of interactants, and is, in turn, divided into subsystems of affect, judgement and appreciation. Martin (2000, p.173) defines them as: (1) (HYLAND, 1998) to construe an epistemic positioning, evidentiality, intensity and other combinations. In this paper, I will focus on the attitude subsystems since they provide more substantial data to prove the argument proposed.
Lastly, other than being positive or negative attitude, two basic types of Appraisal can occur separately or in combination: (1) inscribedexplicit through evaluative lexicon and syntax (epithets, relational attributes or comment adjuncts); and (2) evokedimplicit through lexical enrichment as, for example, in figurative language (MARTIN, 2000).
In political discourses, Appraisal enhances the metadiscourse (HYLAND, 2018)not conceived as the "discourse about discourse", but a notion that embodies the idea that communication is more than just the exchange of information, goods or services. It "involves the personalities, attitudes and assumptions of those who are communicating" (HYLAND, 2018, p.3). Metadiscourse focuses on how "both interactive and textual resources are used to create and maintain relations with readers" (AGUILAR, 2008, p.87), in a way "in which context and linguistic meaning are integrated to allow readers to derive intended interpretations" (HYLAND, 2018, p.437).
According to Aguilar (2008), in her extensive study about metadiscourse, the "allinterpersonal" developments made by functionalists is one of the most important contributions for theories of metadiscourse so far. This tradition interweaves interactive metadiscourse (guiding the reader through the text with transitions, frames and endophoric markers, etc.) with interactional metadiscourse (involving the reader in the text with hedges, boosters, attitude markers, etc). Appraisal is, therefore, an important vector of metadiscourse.

Attitude: Social Evaluation as an Instantiation of Appreciation
Evaluations propagate or ramify through a text which leads to the assumption that evaluation flows through cohesive reference chains (LEMKE, 1998). Sometimes these chains can be long enough that they imply subtle meanings across whole texts, which is particularly frequent in political discourses (COFFIN, 2002). Coffin and O'Halloran (2005, p.148) had a particular interest in these "longer range cohesive propagations" (cf. LEMKE, 1998, p.53). According to the authors, there's ongoing criticism to analysts who approach texts focusing on single pieces of text as a proof of a text's ideological slant, failing to consider textual modification (cf. WIDDOWSON, 2008;O'HALLORAN, 2003). Hence, Appraisal studies have been emphasizing that patterns of evaluation in parts of a text serve to construct a particular evaluative position over the course of a text (COFFIN, 2002;COFFIN;O'HALLORAN, 2005;MACKEN-HORARIK, 2003;ROSE, 2003;DE OLIVEIRA, 2017). Coffin and O'Halloran (2005, p.149) indicated that indirect evaluations (evoked) are really common in political texts, as they exemplify: Consider the following sentence: 'TWO million jobs will be lost if Tony Blair signs the new EU treaty, it was feared last night.'. There is no direct Judgement on Blair's personality here […]. However, this sentence, we would argue, functions as an indirect Judgement, since it is likely to prompt many readers to judge Blair's action as morally irresponsible. Indirect appraisals can only be recovered from contextual meanings in the discursive flow which imposes a standpoint for their classification and make them borderline. Martin (2000, pp.58-62) and Martin and White (2005) discussed borderline appraisals but did not reach out perspectives of shadow areas for affect, judgement and appreciation. 2 One of these shadow areas is the social evaluation, conceived as in the border of judgement and appreciation whereas it refers to the positive or negative of products, activities, processes or Adopting Martin's traditional evaluation model, both expressions in blue are only appreciation appraisals since they evaluate "things." 3 However, readers can reach and individualize the people (politicians) who are the main target of the author's criticism. In this sense, social evaluation seems to be the best classification in both cases. Recovered from the discursive flow, both lexical choices in italics institutionalize or group people to make a comprehensive evaluation of politicians and countries, for this reason social evaluation is a persuasion device that also propitiates a tension between ideational and interpersonal dominance over the interpretation of the categories, that is, appreciation embedding judgement as facts (e.g. "the houses are in Manhattan") in the perspective of social groups.
Joffe sees 'populists' as 'authoritarian,' grouping Marine Le Pen, Geert Wilders, Ukip and Donald Trump with their respective parties. Hence, when affirmed that Britain is a 'great exception'also an appreciation appraisal (cf. MARTIN'S model) -, the author judges people "in command" positively. With this intent, social evaluation is applied as a powerful resource in political discourses to condense meanings in readers' minds by constructing implied associations that would be hardly validated in judgment-type appraisals.
Social evaluations account for syllogisma kind of logical argument that applies deductive reasoning to arrive at a conclusion based on two or more propositions that are asserted or assumed to be trueas in the example: (1)  combines appreciation and judgement to create social evaluation: a subsystem 4 of appreciation that associates persons, mindsets, parties, ideologies as the same "thing".

Methodology
As stated before, the corpus of the present study is composed of two speeches, one from Donald Trump and the other one from Hillary Clinton. Both speeches are particularly relevant since they were held during their respective nominations as candidates for the presidential race at the time (RNC on July 18 -21, 2016 and DNC on July 25 -28, 2016).
As the procedures adopted in analysis, the then-candidates' nomination speeches were subdivided into three sections representing the first, second and third initial moments of the candidates' speeches (three 500-word parts from their beginnings -Sections 1, 2 and 3). This division is justified by the intention to analyze how candidates established their rhetorical strategy and set the tone of their discourses from their beginning on. Next, I mapped the persuasive strategies that each candidate applied in each speech section based on Appraisal and focusing on social evaluation. Appraisals were divided into categories: Affect, Judgement, Appreciation, and Appreciation -Social Evaluation.
In the quantitative part of the research, a radar chart was created based on the appraisal occurrences in each candidate speech (divided into the three aforementioned sections). The radar chart (also known as web chart, spider chart, or star chart) is a graphical method of displaying multivariate data in the form of a two-dimensional chart of three or more quantitative variables represented on axes starting from the same point. According to Kaczynski et al. (2008), radar charts are primarily suited for strikingly showing "outliers" and "commonality", or when one chart is greater in every variable than another, and used for ordinal measurements, wherefore they provided a visual representation of how much each candidate applied positive and negativeappraisals, as well as their proportion and distribution throughout the first segment of the nomination speeches.
For the qualitative part of the research, I focused on appraisals as rhetoric devices for persuasion aiming at preserving and changing electors' attitudes or behaviors toward relevant ideas of political agendas. The analysis demonstrates how each candidate invested on certain values of attitude and approaches to portray their ideological views. They adopted different strategies for appraisal deploying in their speeches' openings combined with persuasion based on the notions of ingroup ("Us") vs. the outgroup ("Them"). In the next sections, I demonstrate how social evaluation was effective in the last chairperson race in USA (2016).

Hillary Clinton: A Discourse of Positivity
Hillary Clinton herself states in her book (CLINTON, 2017) that her campaign motto "Stronger together" bears a lot of information about the image she wanted to construct of herself. In a sense, it summarizes the complex ideological activity marketers may have wanted to associate to the Democrat candidate, in her words in Annex A. As a result, analysis of Clinton's DNC speech showed a purposeful overly positive discourse that invested heavily on positive self-evaluation emulated mainly by appraisals of the social evaluation type.
All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0 Chart 1 (and 2next section, for Trump), in Annex B, represents how appraisal was mapped in Clinton's speech. For the charts reading, speeches were divided into three 500word parts from their beginnings -Sections 1, 2 and 3. The background circles were subdivided into three sections representing the first, second and third initial moments of candidates' speeches, and into types of appraisals (Affect, Judgement, Appreciation, Appreciation -Social Evaluation). The size of the blue and red stain represents, respectively, the proportion of use of positive and negative appraisals while stains' ranges point to the types of appraisals. This proportion is based on appraisals occurrences in numbers. For example, the broader stain area in Chart 1 (Section 3 -Apr -Social Evaluation) represents the use of 29 appraisals of this type in Section 3.
Regarding the particularities of Clinton' evaluations (Chart 1 -Annex B), the candidate starts evaluating more constantly in Sections 2 and 3. The blue area in the Chart 1 illustrates a highly positive stance (tone), generally given by social evaluations in all sections.
For example, in Section 1, Clinton offers a guideline of what will be her discursive strategy 5 : What a remarkable week it's been. We heard the man from Hope (+ve), Bill Clinton. And the man of Hope (+ ve), Barack Obama. America is stronger (+ve) because of President Obama's leadership (+ ve), and I'm better (+ ve) because of his friendship (+ve).
In this excerpt, Clinton's discursive strategy and ideological approach go through the balance of positive judgement and appreciation to persuade. Her reasoning: Bill Clinton is someone who personalizes the Democrats' values ("the man from Hope"positive judgement). Barack Obama succeeds Clinton and also personalizes the Democrats' values ("man of Hope"positive judgement). As a consequence, "America is stronger" (positive appreciation), which comes as a result of Clinton + Obama. Then, Hillary uses positive affect to talk about herself, and, at the same time, modalizes the utterance by giving credit to Obama and Bill Clinton. Clinton's persuasive strategy enhances emotion as an effort to present herself as a humble person. Besides, a depiction of her discourse, in terms of affect, reveals a highly positive stance, mainly toward members of the Democrat party (   (Clinton) Target: Electors, family, friends and Party members (4 occurrences) Example: "I'm also grateful to the rest of my family and the friends of a lifetime. To all of you whose hard work brought us here tonight. And to those of you who joined our campaign this week". Target: HERSELF (1 occurrence) / Obama (1 occurrence) Example: "I'm better because of his (Obama) friendship". Target: Joe Biden (3 occurrences) Examples: "the one-and-only Joe Biden, who spoke from his big heart"/ "We heard from our terrific vice president". Target: Bernie Sanders (4 occurrences) Examples: "Bernie, your campaign inspired millions of Americans," / "Your cause is our cause" / "Our country needs your ideas, energy, and passion". (1) "Tim Kaineyou're soon going to understand why the people of Virginia keep promoting him: from city council and mayor, to Governor, and now Senator. And he'll make our whole country proud as our Vice President".
In Sections 2 and 3, Clinton establishes a kind of "character battle" against Trump.
There are nine (9) positive judgements to herself and thirteen (13)  To shed some light on this interpretation, I named "We-Americans" (ingroup) and "They-Americans" (outgroup) as the two ideological islands (or blocks) that candidates attempted to put themselves in, or reject (cf. van DIJK, 2002). Example 3 shows Clinton using "anyone" (exclusive) and "we" (inclusive) to set apart social groups. These "social islands" share values, beliefs, and a well-established set of ideas that Clinton and her target audience reject (anyone = Republicans) and align to (we = Democrats). Therefore, the ingroup and the outgroup perception for each candidate is different: for Clinton, the ingroup (inclusive) is We-Americans = We-Democrats, while the outgroup (exclusive) is They-Americans = They-Republicans. For Trump, the ingroup (inclusive) is We-Americans = We-Republicans while the outgroup (exclusive) is They-Americans = They-Democrats.
Social evaluation became more frequent and condensed (in the discursive flow) as Clinton's discourse went on. For instance, a longer range of cohesive propagation is built by Clinton in Section 3 (excerpt from Annex C). This excerpt, in Table 2, reveals how social evaluation targets (in italics) Democrats with positive social evaluation and Republicans with negative social evaluation throughout the discursive flow (1 to 5). It is important to notice that Clinton refers to the ingroup as embedded in social evaluation, that is, she is associating Americans with Democrat values and beliefs: "We-Americans" equals "We-Democrats" (ingroup) so to speak. In this example, by entangling evaluations of affection, judgment and appreciation, Clinton applies social evaluation to create an echoic evaluation of "us" vs. "them", which is erected on assumed and projected (hypothetically built) ideological categories of Americans: the ones that are better, the ingroup (We-Americans = We-Democrats) and the worse ones, the outgroup (They-Americans = They-Republicans). This type of evaluation establishes a metadiscourse and leads the audience to set apart social groups. Lexicon items as "troops", "firefighters" and "police officers" are apparently neutral, but bring an implied cumulative interpersonal meaning, for example, "Doctors and nurses who care for us" (= Democrat doctors and nurses) stand against "a nation where one person had all the power" (= the nation of Republicans, the rejected outgroup). Therefore, these appraisals seem to be better classified as social evaluation (and not only appreciation -"nation"), whereas implied meanings of nation (thing), Republicans (social group) and Trump (individual) put them all aligned and grouped as the target. Primary and secondary targets may vary to each person's discretion based on his/her point-of-view, context and ideological background.
Then, in this case, evaluations of affect (e.g. "Mothers who lost children to violence") and judgement (e.g. "Americans don't say: "I alone can fix it") complement implying meanings to discuss a greater ideological theme of "unity and dissention" of Americans (Republicans vs. Democrats), and other co-related ones. 6 Finally, as the most prominent trace in Clinton's persuasive strategy (cf. van DIJK'S ideological square), she is overly positive discussing ideological themes. Clinton opts to emphasize more the ingroup with positive appraisals (9 in Section 1, 34 in Section 2, 29 in Section 3), rather than de-emphasize the outgroup with negative ones (1 in Section 1, 4 in Section 2, 4 in Section 3).

Donald Trump: A Discourse of Negativity
Trump's campaign motto "Make America Great Again" implies the message that America is not great "now". This reasoning anchors the Republican discursive strategy, thus, contrarily to Clintons' discourse of positivity, Trump's speech invests more on negative appraisals of social evaluations. As can be seen from Chart 2, in Annex D, Trump evaluates more than the Democrat (colored area representing 154 appraisals against 136 from Clinton), being highly negative in social evaluations (red stain: 91 negative appraisals against 23 from Clinton). In his discursive strategy throughout the speech, Trump opts to focus on the alleged turnaround caused by the Democrat takeover on The White House.
Diagram 1, from the excerpt of Annex E (Section 3 of Trump's discourse), illustrates the use of a recurrent discursive strategy of his: the positive evaluation of the "pre-Hillary" time/people associated to the negative evaluation "After four years of Hillary Clinton".
Arrows indicate the discursive flow throughout the reading in Diagram 1. In the excerpt: -positive appreciation is in blue (ingroup); -negative appreciation is in red (outgroup); then, -Trump provides a sanction as a result of this process: the "legacy of Hillary Clinton". Trump's strategy of using appreciation rather than affect or judgement strengthened his criticism by constructing a "state-of-things" validation for his argument. Despite the debatable accountability of Clinton to that situation, negative evaluation amount by targeting the Democrat candidate in evoked judgements: Clinton can be only established as the target through lexical enrichment in the discursive flow. A depiction of Trump's speech in terms of affect, judgement and appreciation reveals similar results, implying he is overly negative, preferring to de-emphasize more the outgroup (They-Americans = They-Democrats) with negative evaluations (cf. van DIJK's [1998] ideological square), as I analyze next in detail.
Affect is barely used by Trump. From the only three occurrences in all sections, two of them are related to Sarah Root, a girl that was then-recently murdered ("…an innocent young girl"; "their amazing daughter). Distinctively than Clinton who used 15 affect evaluations, Trump's initial sayings shift away from the emotional appealing.
For judgement, in all Sections, Trump is more evident than Clinton to establish the once referred "character battle". Positive judgement appraisals to himself (18) and negative judgement appraisals to Clinton (17)  Social evaluations are also a big part of Trump's speech. He uses this type of rhetoric device differently than Clinton, though. When defining his "social islands", "We" and "They-Americans", he combines judgement, appreciation and social evaluation in a way the audience loses sight of the real target of his criticism, as can be noted in the following excerpt: "Tonight, I will share with you my plan of action for America. The most important difference between our plan and that of our opponents, is that our plan will put America First. Americanism, not globalism, will be our credo." As detailed in Diagram 2, deixis analysis of the excerpt reveals a combination of appreciation (things), judgement (candidate) and social evaluation (parties). This bridging of evaluation is allowed by an ideological identification of the background (real) target: Clinton.
Diagram 2: Evaluation through combined appraisal types Source: Author Social evaluations are applied by Trump to consolidate ideological complexes. When blending non-humans participants with human processes, the now-President creates a tension of ideational and interpersonal meanings through the use of processes as grammatical metaphors. This persuasive strategy implies a sort of "institutional behavior" that can only be obtained from a certain "group-thinking" standpoint reflecting ideologies that his target audience is aligned with ("we" -inclusive) or reject ("they" -exclusive). See in the following excerpt (social evaluation participants in italics): As long as we are led by politicians who will not put America first, then we can be assured that other nations will not treat America with respect the respect that we deserve. Non-human Participants: America, other nations Human Processes: leadputbe assuredtreatdeserve Therefore, in the excerpt, social evaluations are a particularly strong subtype of Appraisal because, by bridging candidates and parties, one can potentially establish a common ground for evaluations crafted out of shared or rejected ideological sets that puts both candidates and institutions (parties, organizations, social groups, companies, etc.) on the reach of a lesser degree (at the level of lexicogrammar) type of evaluation, but connected to implied ideology scattered throughout the text. This can be observed in the following excerpt, detailed in Diagram 3. "Big business, elite media and major donors are lining up behind the campaign of my opponent because they know she will keep our rigged system in place" Diagram 3: Microdiscursive and macrodiscursive unfolding Source: Author Diagram 3 presents the macro notion of ideology accessed from the micro notions of group members' discourses and social practices (cf. LI, 2010). This rhetoric effect the result of a micro and macrodiscursive unfolding: pronominal variations of the outgroup do not define neither "things" nor "people", that is, "business", "elite", "donors" and "she" are all encapsulated as the same "thing" that "knows" (human process) of a problematic situation: it is the accountability and the ideological identification that establishes this bridging.

Persuasive Strategies of Candidates
After examining the then-candidates' discursive strategies, it is possible to garner some information about the types of strategies they deployed. In Table 3, we can see whether or not, and to which extent, social evaluation was applied by Clinton and Trump as rhetoric devices based on demoticism and variations of "politics of fear" and the "arrogance of All content of Bakhtiniana. Revista de Estudos do Discurso is licensed under a Creative Commons attribution-type CC-BY 4.0 ignorance" (cf. WODAK, 2015). Their discourses did not differ much in terms of the types of rhetorical arsenal applied to persuade voters. Both of them deployed discursive strategies aligned to what Wodak defines as "argumentation strategies used by ingroup members to delineate differences between groups and distance themselves from the outgroup" (1999, p.31). Although all listed devices were identified in both speeches, they were applied in different ways, varying from a positive-negative standpoint to a different proportion of evaluative lexicon. Examples of persuasive strategies in Trump's speech 1) "outgroup" is conceived as the "enemy", and it is defined in terms of difference, then oftentimes ridiculed or criminalized "Trump is arrogant and does not operate with people" "The most important difference between our plan and that of our opponents" 2) "We-discourse" builds a good or bad dichotomy (ingroup vs. outgroup) based on assumptions of moral quality and moral character "America is unity" / "Trump is I can fix it" "A change in leadership is required to produce a change in outcomes. But Hillary Clinton's legacy…" 3) recurrent positive self-image of the ingroup "we are not afraid. We will rise to the challenge, just as we always have. We will not build a wall. Instead, we will build an economy where everyone who wants a good job can get one.
And we'll build […]" "But here, at our convention, there will be no lies. We will honor the American people with the truth, and nothing else." 4) scapegoating "you'll find contractors and small businesses who lost everything because Donald Trump…" "The problems we face now poverty and violence at home, war and destruction abroad" 5) exaggeration of the negativity when referring to the outgroup members' statements and behavior "He wants to divide us" "Another humiliation came when president Obama drew a red line in Syria"

Source: Author
The big picture obtained from Charts 1 and 2 (Annexes B and D) evinces, as I argue, that political speeches of this kind in electoral or polarized scenarios, when based on social evaluations, ought to be assumed from an "antithetical-background" perspective, in other words, ideological meanings are directly connected with background meanings encoded in agendas and/or personified in specific political rivalries. This antithetical-background meaning is scarcely accessed in small corpora of isolated texts (e.g. speech vs. speech), it has to be extracted from political agendas and discourses. Therefore, one's discourse is usually identified by the target audience based on a biased identification of a "nemesis" figure that triggers a "set-of-ideas-we-go-against" stance, especially in polarized scenarios.
In this study, by pairing together RNC and DNC speeches, uses of social evaluation revealed important aspects to be considered in dichotomic political analysis like this, such as: (a) proportion of the evaluative meaning; (b) propagation of evaluation in text; (c) negative meaning (outgroup) versus positive meaning (ingroup); and (d) steadfast negativity or positivity. Also, the identification of ideological or evaluative solidarity is, therefore, a key interpersonal parameter for electors to make meaning of political speeches.

Concluding Remarks
This research studied persuasion through appraisal with a focus on social evaluation in Clinton-Trump nomination speeches. Results indicated that social evaluation proved to be a strong persuasion device because it works on the domain of the belief systems encapsulated in discourses, changing and/or preserving ideologies. In this sense, from a metadiscursive perspective, social evaluation can be particularly effective in political discourses due to its capacity of battling ideological agendas by reaching a higher-ranked target of criticism that is not as effectively accessed by judgement or appreciation appraisals separately.
The analyses revealed that the then-candidates adopted different approaches in tone, proportion and scale to stand ideological views and field evaluations in speeches. In the first part of her speech, Clinton employed fewer evaluations when compared to Trump, who managed to consistently voice them. While Trump adopted an overly negative stance, based on judgement and social evaluation, Clinton rather loaded her speech with positive evaluation, using affect and judgement substantially. Both candidates invested heavily on social evaluation as a persuasive device to single out from each other ideologically.
The We-Americans and They-Americans categories were wisely applied by both candidates. Clinton's strategy was based on distancing herself and her electors from all the negativity she tried to associate with the Republican candidate. On the other hand, Trump evaluated Clinton constantly, mainly by using evoked appraisals to pair a negative state-of-affairs (appreciation) with people in command (judgement), while transferring accountability to Clinton due to the ideological bridging that social evaluation allows.
The Appraisal framework proved to be efficient for mapping ideological meaning in political speeches. For being loaded with social evaluation, these speeches are capable of increasing political polarization since they tend to amplify a group-thinking orientation based on distinguishing bad-or-good values that electors reject or identify with. Therefore, shareable sets of values and beliefs associated to parties, politicians, and electors make this type of audience particularly susceptible to social evaluation.
To sum up, political discourses favor certain presupposed values of attitude that are usually based on underlying cultural orientations of that specific political/ideological group.
Discursive parameters applied in this very specific cultural and situational context set the rules of the game for candidates and voters beyond their encoded beliefs. In the analyzed case, the way then-candidates' speeches influenced voters inside their "allied" and "enemy" ideological groups for the outcome of the election involves a much more complex political analysis, however, it is undeniable that speeches were a pivotal aspect of that election since it provided a cumulative groove of semantic patterning with the potential to lead voters' interpretation and reinforce ideological views.