La notion de pertinence au défi des effets émotionnels

Dans cet article, nous proposons d’intégrer les effets émotionnels (à la suite de Wharton, 2003), à savoir l’expressivité, donc les aspects affectifs de la communication linguistique, dans une conception pragmatique cognitive générale, en l’occurrence la Théorie de la pertinence de Sperber & Wilson (1986/95). Nous partons du constat que cette dimension cruciale de la communication linguistique a été pratiquement entièrement laissée de côté par le courant formel et propositionnaliste en sémantique et pragmatique. En nous appuyant sur quelques cas empiriques, et notamment sur celui d’un « malentendu » émotionnel, nous proposons de lier ces effets à la notion d’ineffabilité descriptive. Nous défendons l’idée que les effets émotionnels permettent de dépasser les limitations de la communication propositionnelle et interagit étroitement avec elle ; nous proposons d’envisager ces effets comme atteignant leur pertinence par leur résonnance expérientielle, à savoir leur capacité d’évoquer des émotions similaires par l’accès à des traces mémorielles ou par l’imagination, ce qui ouvre à une forme particulière de créativité langagière. Since the communication of information about emotional states clearly plays a central role in human interaction, it might be presumed that pragmatic accounts of linguistic communication would include well developed views on how these states are communicated. However, for a range of reasons, aspects of linguistic communication which feel as if they go beyond the strictly propositional dimension have long been dismissed by scholars interested in meaning: as a result, there is a conspicuous void in theories of pragmatics where the emotional dimension to communication should be. Although speech-act philosophers found ways to incorporate aspects of non-truth conditional meaning in terms of propositional attitudes, the direct expression of emotional states, as opposed to the description of such states is largely ignored. Indeed, in most modern theories of pragmatics, the domain of research is limited to those cases that fall within the category of non-natural meaning. This excludes expressive meaning. This paper synthesizes an account of emotions and emotion-reading that fits with work in a cognitive model of pragmatics – relevance theory (Sperber & Wilson, 1986/1995) – and with work on emotions in cognitive and affective science (Cosmides & Tooby, 2000; Deonna & Teroni, 2012). Turning first to pragmatics, there are two main ways in which the relevance theory account of utterance interpretation diverges from both traditional Gricean and Neo-Gricean ones and we believe that these two divergences underpin our ultimate claim: that relevance theory is capable of rising to the challenges we raise in this chapter. The first of these is that within relevance theory the informative intention need not always be described as an intention to communicate a single proposition and propositional attitude. In relevance theory the informative intention is construed more broadly than merely an intention to communicate a proposition p: as an intention ‘to make manifest or more manifest to the audience a set of assumptions I’ (Sperber and Wilson, 1986/1995 : 58 - our italics). When what is communicated is quite vague, it typically involves a marginal increase in the manifestness of a very wide range of weakly manifest assumptions, resulting in an increased similarity between the cognitive environments of communicator and audience. The second difference concerns the line Grice (1957) famously drew between showing and non-natural meaning (meaningNN). It has often been remarked that this line has had a huge influence on the development of pragmatics. Many pragmatists continue to focus on the notion of meaningNN and abstract away from cases of showing. So where, in fact, should the line be drawn? According to relevance theory, it should not be drawn at all. Cases of both showing and meaningNN qualify as instances of ostensive-inferential communication and instead of there being a cut-off between the two notions, there is a continuum of cases in between. Turning to cognitive and affective sciences, we adopt two ideas. Firstly, from cognitive science, we view emotions as one type of evolved superordinate cognitive mechanism, the function of which is to mobilize cognitive processes responsible for perception and attention, physiological changes etc. Secondly, we the key concept from affective science that emotions are attitudes bearing on evaluations. So, rather than treating all stimuli as equal while scanning the environment, (which would, presumably, result in some kind of cognitive overload), appraisal theorists claim that people scan for inputs as a function of particular criteria, among which goal relevance. Attention is therefore paid to stimuli when once appraised as being relevant and, in certain conditions, an emotional episode may result. Our account builds on these two ideas observations using relevance-theoretic pragmatics. The kind of information conveyed during emotional communication puts the user into a state in which emotional procedures are highly activated, and are therefore much more likely to be recognised and selected by an audience (Wharton, 2009 ; de Saussure, 2013). Central to this thinking is the idea that the notion of cognitive effect needs to be complemented by a new notion of emotional effect, typically activated by emotion-reading procedures, which trigger immediate experiential responses either in memory or in imagination, which are straight away relevant – potentially very much so – in one’s mind.