A corpus-based investigation into lexicogrammatical incongruity and its relation to irony

Callaghan, David James (2023). A corpus-based investigation into lexicogrammatical incongruity and its relation to irony. University of Birmingham. Ph.D.

Preview

Callaghan2023PhD.pdf
Text - Accepted Version
Available under License All rights reserved.
Download (2MB) | Preview

Abstract

Presented is a corpus-based investigation into the lexicogrammatical features of irony. A common understanding of irony is of a trope in which the dictum and the implicatum are seen as incongruous. I argue that patterns of lexicogrammatical incongruity can reflect this incongruity at the pragmatic level. Additionally, a bottom-up examination of authentic examples of ironic utterances can reveal common lexicogrammatical patterns. This study attempts to readdress the paucity of linguistic studies into irony by focusing on real-world examples of irony as a source of data.

Examples of irony were taken from two irony-rich discourse environments and ironic examples were extracted using an independent framework of irony. Commonalities of patterning were first identified, and then interrogated across the two DIY corpora, as well as two general corpora, in order to measure both frequency (raw/t-score) and fixedness. Finally, a deeper examination of the concordance lines revealed whether such patterns carry an ironic force.

Three significant findings are presented. Firstly, the study explores lexicogrammatical patterns of collocation concerning multiple hedging: that is, two or more lexical items which ostensibly have a hedging function, yet often frame strong evaluative or rhetorical statements. Secondly, I present patterns of collostruction in which the progressive aspect colligates with cognition verbs. It is the lexicogrammatical incongruity within these patterns that is often a source of irony. Usage of these phrases does not, however, guarantee that the statement will always be ironic. Yet, when compared within larger general corpora, these patterns demonstrate high tendencies of pragmatic characteristics related to irony. Therefore, the final results chapter argues that such patterns can be considered as having ironic priming.

Identification and awareness of such patterns may help audiences in accurately reaching ironic interpretations. More practically, these patterns may also help NLP methodology by building upon previous attempts of automated irony detection to create more robust algorithms. Furthermore, there are wider implications to what corpus linguistic methodology can explore in regard to connections between pragmatics and lexicogrammar.

Type of Work:

Thesis (Doctorates > Ph.D.)

Award Type:

Doctorates > Ph.D.

Supervisor(s):