How to distinguish hypothetical from actual speech in fiction: Testing the typicality hypothesis

This paper investigates the connection between counterfactuality and stereotypicality in direct speech representation. In Monika Fludernik’s theory of schematic language representation, quotations typify rather than reproduce, and typicality coincides with stereotypical expressivity in the form discourse particles, among other features. By distinguishing hypothetical speech proper from the more general concept of typifying direct speech, we can see that in fiction hypothetical speech is not always stereotypically expressive. In conversational storytelling, discourse markers serve the functions of source-tracking, emplotment, and expressing the quoter’s emotions and evaluation. I discuss reasons why fiction differs from conversational storytelling in this respect. Fludernik’s treatment of discourse markers or ‘typicality markers’ in direct speech representation is here complemented with Bakhtinian notions of dual expressiveness, speech genres, and the responsive quality of utterances. The arguments presented are illustrated by passages from the fiction of Carol Shields, Peter Bichsel, and Junot Díaz.