Magdalena Kaufmann

Stojnić (2021) argues that the content of linguistic utterances is determined by the rules of natural language grammar more stringently than what is generally assumed. She proposes specifically that coherence relations are encoded by the linguistic structures and determine what individuals count as most prominent, thereby serving as the referents of free (“demonstrative”) pronouns. In this paper, I take a close look at the empirical evidence from English and Serbian that she offers in support of this position. Considering these data points in connection with additional linguistic data (also from German and Japanese), I argue that there is no compelling evidence for the assumption that coherence relations directly determine the resolution of pronouns. Instead, grammatical restrictions imposed by different types of pronouns and tenses have a larger impact on the meaning conventionally expressed by complex utterances than what is generally assumed in the literature on coherence relations.

This paper refutes the widespread view that disjunctions of imperatives invariably grant free choice between the actions named by their disjuncts. Like other disjunctions they can also express a correlation with some factual distinction, but as with modalized declaratives used for non-assertive speech acts this needs to be indicated explicitly. A compositional analysis of one such indicator, depending on, constitutes the point of departure for a uniform analysis of disjunctions across clause types. Disjunctions are analyzed as sets of propositional alternatives that correlate with a partition that the speaker may or may not be able to indicate explicitly. Free choice arises as the specific case where the partition is induced by the preferences of the addressee.

We discuss the semantic significance of a puzzle concerning ‘ought’ and conditionals recently discussed by Kolodny and MacFarlane. We argue that the puzzle is problematic for the standard Kratzer-style analysis of modality. In Kratzer’s semantics, modals are evaluated relative to a pair of conversational backgrounds. We show that there is no sensible way of assigning values to these conversational backgrounds so as to derive all of the intuitions in Kolodny and MacFarlane’s case. We show that the appropriate verdicts can be derived by extending Kratzer’s framework to feature a third conversational background and claiming that the relevant reading of ‘ought’ is sensitive to this parameter