Automatic Detection of Irony

2Toward Automatic Detection of Figurative Language

2.1. Introduction

As we saw in Chapter 1, irony is a complex linguistic phenomenon that has been studied in detail in both philosophy and linguistics (Grice et al. 1975, Sperber and Wilson 1981, Utsumi 1996). Although authors differ in their definition of irony, all agree that it implies a mismatch between what is said and the reality. Taking account of the differences between approaches, irony can be defined as a mismatch between the literal and intended senses of an utterance. The search for non-literal meaning begins when a listener becomes aware that the utterance in question, when taken literally, makes no sense in the context (Grice et al. 1975, Searle 1979, Attardo 2000a). In most studies, irony is considered in conjunction with other forms of figurative language, such as humor, satire, parody and sarcasm (Clark and Gerrig 1984, Gibbs 2000). The distinction between these different forms of figurative language, particularly that between irony and sarcasm, is highly complex. This is the result of a blurred distinction between the notions in question at a linguistic level, and of the complexity involved in differentiating between notions within a text at a computational level.

The theories described in Chapter 1 form the basis for most of the cues used for automatic detection. A study of the state of the art in this area shows irony and sarcasm to be among the most widely studied forms of figurative language, unlike metaphor ...