Abstract

Modelling the semantic behaviour of expressives in such a way that one be in a position to predict some, at least, of their distinctive properties is the purpose of any theory aiming at accounting for so-called expressive meaning. In this paper, we review a semantic framework coined by its author the Theory of Bias. We show that the two-tiered account of the meaning of expressives (notably, sentential interjections and pejorative epithets) favored by this framework has the resources to provide a consistent explanation of their semantic behaviour in prima facie awkward statements such as ‘alas, I am unfavorably disposed towards something’ or ‘that stupid Trump isn’t stupid’. On the critical side, we argue that, not being intended to account for the meaning of expressives alone, the theory has hard times explaining the unshiftability of their bias, as opposed to the shiftability of the bias of discourse particles —a further member, according to some authors, of the category of expressives lato sensu— and that it fails to pinpoint unpluggability as the crucial test for expressive content. We conclude that, in spite of its merits, the theory need be theoretically enhanced and methodologically sharpened.