Irony and reversal of evaluation

This paper outlines a corpus-assisted investigation into the nature and functions of irony in both spoken interaction and written texts. We begin with a review of some of the principal current debates in irony studies, which have until recently often been conducted with little recourse to authentic examples of use in interactive discourse types. We consider, then, from an examination of corpus-based real-life data, firstly, how explicit irony operates and then from this whether there might be a more objective way of identifying and defining episodes of implicit irony than simple reliance on the researcher’s unsupported intuition. Potential sites of implicit irony are then examined in the data to see how and why speakers and writers employ it and how hearers and audiences respond. The data analyses afford strong evidence that the principal mechanism driving all irony is an implied reversal of the evaluative meaning of the utterance (rather than of the propositional / ideational meaning, as argued in many traditional theories of irony). In addition, they reveal how irony in discourse always has a strategic argumentative point. Three corpora were employed, the first of semi-spontaneous interactive spoken discourse (WHB: circa six million words of White House press briefings in transcription [1998-2004]), the second of spoken interviews (Ints: 250,000 words of transcribed televised UK political interviews) and the third of written texts (Papers:100 million words of UK broadsheet texts). Techniques from Corpus Linguistics, principally concordancing, were employed. Most of the corpus interrogation was conducted using the WordSmith Tools suite of programs. Corpora have only rarely been used to investigate participants’ interaction in discourse and this paper is intended as a contribution to the nascent interdisciplinary field of Corpus-Assisted Discourse Studies (CADS).