How affect modulates conversational meanings: a review of experimental research: invited review

ABSTRACT

Affect has been found to play important role in word and sentence processing. What is less understood is the role it plays in the process by which interlocutors arrive at what speakers mean. In the present review, the way affect modulates how we comprehend what others mean is examined. This is done by reviewing studies that have employed experimental methods using both written materials and spoken utterances. The goal of the present review is to better understand how the inferential process is framed by affective factors and to propose ways of integrating affectivity into pragmatics. In Part 1, the motivation for the present review is explicated. In Part 2, experimental evidence is presented and suggestions are offered about how to best operationalise variables crucial for disentangling affect from other mechanisms that contribute to utterance interpretation. In Part 3, central notions of pragmatics are discussed with regard to affectivity, which it is proposed that, if construed as a goal-oriented activity that serves interpersonal configuration, can account for several pragmatic phenomena.

KEYWORDS:

• Affect
• conversational inference
• pragmatics
• speech acts
• prosody

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 These are not the only conversational phenomena that require inference for their interpretation. For example, pragmatic inference is at work even when listeners attempt to connect a pronoun with the right antecedent.

2 I use the phrase what is said in a non-technical sense. In Gricean pragmatics, what is said refers to “the conventional meaning of a sentence uttered with the exclusion of any conventional implicature” (Huang, 2010, p. 489). For a detailed discussion of the concept and its development, see Huang (2010).

3 There is recent evidence from behavioral and neurological measures (Shamay-Tsoory et al., 2005; 2007; Shamay-Tsoory & Aharon-Peretz, 2007) that, except for a cognitive Theory of Mind (ToM), i.e. our ability to reason about others’ beliefs, humans also have an affective ToM, the ability to detect and reason about others’ emotional states, which is based on different neural mechanisms from cognitive ToM, and, unlike the latter, is highly correlated with empathic abilities (Shamay-Tsoory et al., 2003; 2002) as measured by a battery of diagnostic questionnaires (Mehrabian & Epstein, 1972).

4 Note however that emotion can also serve as a trigger that leads to impolite behavior such as when indignation results in (linguistic) insults (Vergis 2015).

5 Despite the fact that Janney and colleagues’ proposal was supported by a wealth of empirical evidence and was theoretically motivated, it never became part of mainstream pragmatic theories. Possibly the main reasons for this exclusion are that emotive meanings have a clearly interpersonal orientation and the nature of multimodal signs (prosody, gestures) is not discreet but gradient, which takes us again to one of the reasons why linguistics did not integrate affect (see van Berkum’s comment in Section 1).