Abstract
► I show that some arbitrary items (arbs) are semantically definite, others are variable. ► This means that no uniform semantic analysis for arbs is possible. ► I show that there are fine distinctions in the salience of arbs & non-arbitrary NPs. ► This means that there is no way to derive pragmatics from semantics. ► I conclude that we must separate semantic calculations from pragmatic ones.
This article explores a typology of impersonal constructions [1] (constructions with arbitrary interpretations or arbs) in Russian, German, and English based on their semantic and discourse-pragmatic properties.
(1)They speak English in America.
The goal of this work is twofold. First, I want to introduce a meaning-based typological distinction into the realm of impersonals. I will argue that some arbs are semantically definite, while others are variable, drawing attention to previously unobserved behaviour of different arbs with respect to adverbial quantification. This divergent behaviour points to the inadequacy of any semantic analysis that treats the two types of arbs in the same fashion. I then point out further differences between the arbs (and non-arbitrary NPs) that emerge in their interaction with topic structure and discourse anaphora, which further corroborate the separation between the two types of arbs. Secondly, I want to use the study of arbs to explore the architecture of the semantics–pragmatics interface. I conclude that the behaviour of arbs forces a division of labour between semantics and pragmatics in the calculations pertaining to reference and referent-tracking.