Textually Constructed Meaning Shifts in Jokes: A New Stylistic Theory of Jokes

This thesis takes a stylistic approach to joke analysis, in order to answer the question of ‘what makes a joke text funny?’. According to Simpson et al (2019), humour research is often neglected in stylistics, and research into humour from outside of this discipline lacks a consensus or joined up approach. The three main families of humour research (Attardo, 1994) are introduced, with the suggestion that they can form a unified approach as they are all aspects of foregrounding occurring through deviation from norms at either linguistic, ideational or interpersonal levels of meaning. This leads to the conclusion that foregrounding is a necessary feature of humour, but not sufficient to define a text as humorous.

Raskin’s (1985) ‘Semantic Script Theory of Humour’ (SSTH) has become a mainstay of humour research due to the claims that this theory does provide the necessary and sufficient conditions to define a text as joke carrying. The SSTH hypothesis that jokes are constructed using a pair of overlapping and opposing scripts remains unfalsified, though this thesis makes the argument that this is due to an unfalsifiable methodology, and suggests revisions to the SSTH from a stylistic approach. Using Jeffries’ (2010b) and Davies’ (2012;2013) work on constructed opposition a sample of 80 jokes is analysed, finding that a majority of these joke texts are not based on constructed oppositions which does not offer support to the SSTH approach.

The remainder of the thesis then details the text-based discovery process which was taken to develop a new theoretical framework for joke analysis. In a three-stage process, quantitative and qualitative textual analysis is applied to a total of 645 jokes from the Edinburgh Fringe Festival, to determine what patterns are present in the humorous textual meanings. This leads to the proposal of a new theoretical framework of Textually Constructed Meaning Shifts in Jokes, with a testable hypothesis that joke texts will contain at least one of five foregrounded Textually Constructed Meaning Shifts: bisociation, reinterpretation, asymmetrical comparison, contradiction and performative reinforcement.