CHAPTER 1. The concept of irony in the study of literature
1.1. Irony as a stylistic device in mass-media discourse
With the term irony we refer to many different types of verbal expressions, but also to situations that are not necessarily expressed verbally. Irony is therefore not a unitary phenomenon. For instance, in a narrative the protagonist can utter a sentence that is ironic; this is called verbal irony. However, the sequence of events that unfold in the narrative can itself be called ironic relative to the protagonist. In the first case, the protagonist utters a sentence (performs a speech act) which is insincere; for example, it might mean the opposite of what she considers as the veridical situation. In the latter case, the protagonist acts in view of a goal, but in the situation there are circumstances which, unknown to the protagonist, cause her acts to have the opposite effect. There is a schematic similarity between the two cases, since in both of them, the act of the protagonist is negated by the veridical situation; but there is also a big difference, since the discrepancy between the act and the veridical situation is intended in the verbal case, but not in the non-verbal case. In fact, in the verbal case it is exactly the intended mismatch between the veridical situation (as seen by the speaker) and the meaning of the utterance that qualifies it as ironic, whereas in the ironic event it is the unintended mismatch that makes up the irony.
This amounts to saying that, in accordance with the general theory of categorization developed in cognitive linguistics, cf. Lakoff (1987), the term irony denotes a category with a heterogeneous structure. The difference depicted above holds true at least as regards ironic events versus ironic discourse, but even if we restrict ourselves to verbal irony, there is no reason to believe that the category of ironic utterances has a uniform structure, as assumed in many analyses of irony, such as those offered by Brown and Levinson (1978: 226) and Searle (1979), who suggest that irony is understood by assuming the opposite of the sentence's literal meaning. In Sperber and Wilson (1981), it is claimed that an ironic statement is like an echo reminding the listener of a similar statement which on a previous occasion, has been or could have been uttered; through the echoic form, the speaker expresses her attitude towards the situation. In Clark and Gerrig (1984), verbal irony involves pretense, meaning that the speaker of an ironic sentence pretends to be some other person proclaiming the utterance to an unknown audience. KumonNakamura et al. (1995) make the claim that ironic remarks have an effect by alluding to a failed expectation. Many theories on irony argue that the phenomenon can be defined through necessary and sufficient conditions, but in so far as we are talking about semantic conditions, this seems not to be a fruitful approach.
A cognitive approach to irony will proceed from the following assumptions:
1) Irony is not primarily a rhetorical device or a literary technique; it is found in ordinary language with a relatively high frequency. In the corpus studied in Tannen (1984), irony was found in 8% of all turns; in Gibbs' (2000) study, it was found in 7% of all turns. In many cases, a sentence is understood as ironic without this being the intention (Gibbs and O'Brien 1991). Conversely, the meaning of a sentence is often understood without the sentence being recognized as ironic. All of this points to irony as an integrated part of human communication which works implicitly and automatically. The studies mentioned above show that irony is used in talk among friends to establish group membership by commenting on individuals who are not group members. The case of irony is a bit like what Lakoff and Johnson (1980, 1999) found concerning metaphor. Metaphor is not (just) a rhetorical figure found in literary texts; it is used in ordinary language all the time and it is processed without necessarily being recognized as a metaphor. However, there is also a crucial difference, because the cognitive ability to think in metaphors (and metonymy) is a necessary condition for language use; without these two operations, the lexicon would explode and put too big a load on the semantic memory. Irony, on the other hand, is a pragmatic operator with no influence on the structure of language.
2) This then raises the question, what is the pragmatics of irony? Again, it is probably not possible to give a unique definition of the pragmatic effects of irony. In the literature, we find different – sometimes contradictory – claims about irony which, however, can all be true. For instance, some argue that irony makes a negative meaning less rude (Dews et al. 1995), whereas others suggest that ironic statements are more rude than literal statements (Colston 1997). Interestingly, Ivanko and Pexman (2003) hypothesize that in a strongly negative context, an ironic statement is considered more mocking and less polite than a literal statement, whereas in a slightly negative context the opposite is true. We will return to this study in more detail below.
There are also competing theories about the social function of irony, considering it to be either a source of affiliation or of conflict between individuals. Such competing view do not constitute a problem for a cognitive approach to irony, since there is no reason to assume that irony is a unitary phenomenon that can be described using a single set of necessary and sufficient conditions. In fact, it might even be unclear whether a given sentence is meant to be ironic or not.