The analysis of evaluative language: words and their

In this dissertation, the analysis of evaluative language is used to explore arete and eunoia as components of ethos. According to Jim R. Martin and Peter R. R.

White (2005: 1), the appraisal theory of language concerns ‘the construction by texts of communities of shared feelings and values, and with the linguistic mechanisms for the sharing of emotions, tastes and normative assessments’. The language of evaluation, moreover, has been of ‘longstanding interest for functionally and semiotically oriented approaches and for those whose concern is with discourse, rhetoric and communicative effect’ (Martin & White 2005: 1). I found this approach particularly applicable when considering the idea of emotions as relations to objects (see Ahmed 2004) and as evaluative expressions that indicate what is seen or not seen as valuable (see Sayer 2006: 457).

In the analysis of evaluative language, emotions are treated as socially meaningful sentiments. According to Martin and White (2005), evaluative expressions are meanings of attitude that can be categorized into three groups: affect (positive or negative feelings in general), judgement (attitudes towards the behaviour and moral character of people) and appreciation (evaluation of things and phenomena). The starting point of the analysis of evaluative language is the idea that evaluative expressions always have a target. Thus evaluative expressions are ways of positioning oneself in relation to something or someone. When using the analysis of evaluative language, I took into account those expressions in which online gossip participants described human beings as either moral agents or

‘things’ whose moral worth was denied. Note that for Martin and White (2005),

‘affect’ is a general group of evaluative expressions and does not focus on bodily reactions in particular (cf. Paasonen 2011: 22–23, 54–55, 232–236; also Graefer 2013). From the viewpoint of this linguistic categorisation, I characterized reactions to ‘things’ (such as voyeuristic pleasure as a reaction to a celebrity’s picture) as ‘appreciation’. From the viewpoint of previous research into online

celebrity gossip, ‘affect’ is close to what Martin and White call appreciation. As Graefer (2013: 60) suggests, affect is aesthetic evaluation that it is freely attached to objects and therefore may change the meaning of what is seen as ‘good’ and

‘bad’ over time. In online celebrity gossip, affect as a freely-moving content finally becomes a social phenomenon – a shared way to react – which makes it part of collective norms, particularly those relating to consumption and commercialism (Graefer 2013). This is line with Martin’s and White’s (2005: 45) notion that there is no such thing as ‘pure’ affect: it is always related to social systems of evaluation.

In the analysis of arete, I focused on celebrity gossip participants’

autobiographical moralizing in which they described themselves and private people around them as moral beings. Of the three categories of attitude suggested by Martin and White (2005), the analysis of autobiographical moralizing concerned the category of judgement. Since the objective to present oneself in a better light often goes hand in hand with the evaluation of other people as immoral actors (Bergmann 1993: 128), I also took into account the ways that online gossip participants evaluated people in their everyday lives. Moreover, presenting oneself in a better light may also take place through critical self-reflection in which language users narrate their learning process, which indicates that ‘now I know better’. In rhetorical terms, rhetors who tell autobiographical experiences may position the ‘self’ in the place of other people and thereby aim at constructing a virtuous impression of their own moral character. Bearing this in mind, I made a distinction between confessions and testimonies as types of autobiographical telling. While confessions are autobiographical stories (or fragments of stories) targeting the moral character of the ‘self’ (see Foucault 1990 [1978]: 53–73), testimonies deal with the moral character of private individuals whose acts the rhetors describe as personally witnessed (see Foucault 1990 [1978]: 59; Felman & Laub 1992).

The analysis of evaluative language can be seen as an applicable method to explore emotions (eunoia) as a rhetorical issue. It concerns ‘shared feelings and values’, as Martin and White (2005: 1) describe. The combination of emotion and its target indicates values. Values, moreover, represent goals serving as guiding principles in life, both individually and socially (Schwartz 1992). Values are often more abstract and hidden than concrete evaluations (Sayer 2011: 25–28). This is because their meaning is based on the combination of the evaluative expression and its target (referent). In other words, evaluative language has a rhetorical function beyond a literal understanding, since it reflects the goals of evaluators and their community (Thompson & Hunston 2000: 6, 13, 21). These goals as values of communities can be, moreover, seen as specific (special) topoi. In the

theory of rhetoric, specific topoi are often regarded as material contents of argumentation that indicate time, place, circumstances and emotional involvement (see Grimaldi 1972: 124–133; also Miller 1987). Specific topoi refer to what matters to rhetors and their audience(s) in a particular kairos.

Of the categories of evaluative language suggested by Martin and White (2005), affect (such as ‘I like her’) is at the heart of evaluative language and it can be seen as central in the language of children who are learning to express what they want and desire (see Painter 2003). In everyday language of adults and young people, however, affect is transformed into ‘institutionalized feelings’ of judgement or those of appreciation, as Martin and White (2005: 45) argue. Thus ‘pure’ affect, as they seem to suggest, is rare in the language use of adults or young people (see ibid.; see also Martin 2000; Painter 2003). The categories of judgement and appreciation are particularly interesting in terms of values and power because they take place on the level of communities as ‘institutions’ of evaluation in which desires are transformed into social meanings.

I see the idea of ‘institutionalized feelings’ suggested by Martin and White (2005:

45) to fit in with Lanham’s (2006: 166–176) ‘motive spectrum’. Judgement as the linguistic realisation of moral norms resonates with the idea of a socially shared moral purpose, whereas appreciation as the evaluation of things, such as a picture of a celebrity, can be seen as the sign of play in which evaluation is motivated by the pleasure of things. The third category, namely game (see Lanham 2006: 166–

176), may be involved in both purpose and play because both of them can be used in persuasion. Bearing these ambiguities in mind, this study holds that there are two basic ways of using evaluative language that contribute to emotivist morality.

Firstly, those persons considered ‘others’ are described by choosing words from the category of judgement to disguise game (power) with purpose. Common to sexism, ageism, racism, anti-Semitism, classism, etc. is that game comes in the guise of moral purpose expressed in the form of categorical, present-tense claims in which moral ‘badness’ is associated with people seen as representatives of a certain group (see Dijk 1993; Reisigl & Wodak 2001; Pälli 2003: 218, 220;

Enberg 2011).

Because sexist domination is prominent in celebrity gossip discourses online (see Fairclough 2008; Meyers 2010; 2013), I take it here as an example of emotivist morality. For instance, saying that ‘women provoke men to violence’ is an example of evaluative use of language in which the idea that women should be obedient to moral norms (purpose) is utilized to reinforce sexist preferences (game) in the guise of the vocabulary of judgement. On the contrary, a self-expression according to which ‘she looks like an ugly, fat pig’ is a way of

contributing to sexist domination (game) according to which women are aesthetic objects to be judged in the guise of a playful, appreciation-based evaluation (play). In appreciation, people are considered entities, not agents who behave (see White 1998: 36). Figure 3 illustrates the categories of evaluative language (affect, judgement, appreciation, see Martin & White 2005) and their relation to values classified in three rhetorical motives (for more on game, play, purpose, see Lanham 2006: 166–176).

Figure 3. Evaluative language (see Martin & White 2005: 45).

As Figure 3 shows, evaluation of human beings is transformed into either meanings of judgement or appreciation. Although the category of judgement may seem more natural than appreciation for the making of moral meanings, this study holds that the evaluation of human beings as material things is emotivist morality in its most extreme form because it explicitly aims at denying the functional role of certain human beings (for more on evaluating human beings as mere objects, see MacIntyre 2003 [1985]: 58–59; Ahmed 2004: 195; Bauman & Donskis 2013:

37–40). Especially where celebrities are concerned, their worth is often equated with their physical appearance. This is evident when treating young female celebrities belonging to certain ethnic groups as stereotypical ‘others’ to be sexualized. For instance, the pop singer Shakira is treated as the cultural object of sexual desire in various media contents through which the stereotype of Latina identity as overly sexual is supported (Orgad 2012: 112–113, 115). In such media contents, Shakira is dehumanized by using the vocabulary of appreciation in which game comes in the guise of play. In rhetorical terms, the evaluation of Shakira as a sexualized object is an example of eunoia connecting the rhetor and audience on the basis of shared sentiments which contribute to a sexist and racist characterization of the ‘third persona’ (see Wander 1984).

•evaluation of human

human beings