Norrick (2004: 1738) stated that ‘hyperbole in general, and ECF in particular, suggest endless avenues for future research’ – but books are not endless and this one has now reached its end. In it, I have walked down some of the avenues, not all of them, and I have not always followed those I have chosen through to their very ends.

I have decided to treat hyperbole as a unified phenomenon, combining hyperbole, overstatement and extreme case formulation under one big umbrella, although separated by Gibbs and Norrick. I think this is justified by the likely fact that non-extreme I've been waiting for hours and extreme you're always late will be perceived as functionally and stylistically equivalent, and by the impossibility of pinning down intentionality in specific cases. Thus, all the above labels are covered by the definition given in Chapter 2, which contrasts the surface meaning and contextual fit of the hyperbolic and possible literal expressions and derives the transferred, emotively coloured meaning from the nature and degree of this contrast. Expressions falling within this definition have been found to come in various guises. Basic types, i.e., those where only hyperbole is at work, are joined by composite realisations, which combine hyperbole with metaphor or metonymy and are less frequent on the whole. These types have been treated together in the study (except for with regard to comprehension, cf. Chapter 5), because in many everyday cases the distinction in effect might be minimal (cf. hyperboles based on die).