How salient are sarcastic questions?

Abstract

This article explores the connection between salience and sarcasm through focusing on three interrogative constructions that can convey both sarcastic and literal meanings – WH-questions followed by a self-answer, alternative questions, and disjunctive polar questions. A large-scale corpus research was conducted throughout three distinct corpora of American-English to determine how these interrogative constructions use salience to encode sarcasm, and how salience contributes to the preference of their sarcastic meaning over the literal one. Firstly, we examine the salience of sarcasm by comparing the distribution of sarcastic questions between a spoken, non-scripted TV-corpus (COCA), a spoken telephone-corpus (Switchboard), and a written web-corpus (NLDS). The main goal of this comparison was to explore if sarcasm is significantly more salient in one corpus than the others. Results show that there are significantly more sarcastic questions in the written-web corpus than in the non-scripted spoken corpora, suggesting that sarcasm is significantly more salient in this corpus than the others. We attribute these results to situational and emotional salience stemming from the distinct discursive arenas. Secondly, we examine the distribution of the different question types to see if the sarcastic meaning is significantly more salient in one of the three interrogative constructions than the others. Results show that WH-questions followed by a self-answer are the most sarcastic construction out of the three examined, and that its sarcastic meaning is more salient than that of alternative questions and of disjunctive polar questions. We attribute these results to intrinsic syntactic properties and to the notion of predictability and (un)expectedness. We conclude the paper with thoughts regarding future work necessary to further our understanding of salient sarcastic questions.

Keywords

• Sarcasm
• Salience
• Questions
• Corpus linguistics
• Discourse Analysis