Work in progress
A pragmatic account of cross-sentential anaphora on embedded indefinites
I argue that we can maintain a static semantics and account for cross-sentential anaphora using the tools of dynamic semantics – namely updates to the context involving discourse referents – while relegating the tools of dynamic semantics to the realm of pragmatics. In this paper, I look particularlyat the cases of cross-sentential anaphora that involve antecedents that are indefinite descriptions scoped under negation, universal quantifiers, modals, and conditionals. I argue that there’s no clear cut account of the data using a dynamic conception of meaning, i.e., taking semantic values to be context change potentials, and that the pragmatic account fairs just as well or better than the dynamic semantic one in these cases. You can get a draft of the paper here.
Do we need dynamic semantics?
Forthcoming in: Metasemantics, edited by Alexis Burgess and Brett Sherman, OUP
In the first part of the paper, I argue that there is no reason to think that dynamic semantics is better equipped to explain what gets communicated in terms of discourse dynamics than a static semantics plus pragmatics. I argue that a lot of the prima facie evidence for dynamic semantics is based on a narrow conception of both static semantics and pragmatics. In the second half of the paper, I offer some considerations in favor of a static semantic/pragmatic explanation over a dynamic semantic one. You can get a (definitely not final) draft of the paper here. Comments very welcome!
Speakers Reference and Anaphoric Pronouns
What role do a speaker's referential intentions play in unbound pronouns anaphoric on indefinites? I consider discourses such as:
(1a) A woman walked in.
(1b) She ordered some wine.
Some have argued (e.g. Stalnaker 1998, van Rooy 2001, Dekker 2004, Hawthorne and Manley 2012) that the right account of discourses such as (1) involve appealing to the referential intentions of speakers.Specifically, when a speaker utters (1a) she has a particular woman in mind; this woman is the referent of the pronoun 'she' in (1b), or figures in the description on a d-type analysis (as in, the woman who is identical to o, where o refers to the woman the speaker has in mind). I argue that there is little evidence for such a view. First, I argue that the appeal to referential intentions has little explanatory value. Second, I argue the view makes bad predictions about the truth conditions of sentences containing anaphoric pronouns. Finally, I undermine several arguments offered in favor of the view, the most notorious of which is based on cases of pronominal contradiction cases. I'll post a draft soon, but in the meantime, please email me for a draft as I am eager for comments on this paper.
A known problem for counterfactual conditionals, most recently pressed by Alan Hàjek, but also discussed by David Lewis and John Hawthorne, among others, is the following. Take any contingently true counterfactual, the one you are the most certain is true, such as (1) If you had dropped this extremely fragile vase, it would have broken, and you can find a might-counterfactual that seems to conflict with it. For example, if certain intepretations of quantum mechanics are true, then there’s a (close-by) possibility in which the vase lands safely on the couch, supporting the truth of (2) If you had dropped this extremely fragile vase, it might not have broken. (2) for all the world sounds like it contradicts (1). I argue that this is not a feature of counterfactuals brought about by esoteric cases, but a central feature of the way we use counterfactuals in conversation. That is, while counterfactuals are generally thought to ignore far away worlds, I argue that they also often ignore close worlds when these represent possible outcomes that are irrelevant for conversational purposes. I then present a semantic and pragmatic theory that captures this fact, involving a new strategy for ordering worlds. Finally, I suggest that while this new theory of counterfactuals saves many of our everyday counterfactuals from falsehood, there remains a problem for counterfactuals in some of their theoretical contexts, since these are not the sort of contexts in which one can legitimately ignore close possible worlds. Email me for a draft.
A new puzzle about epistemic modals
(Joint work with Ben Lennertz)
In this paper we present a problem for most views of epistemic modals. The problem stems from the interaction of epistemic modals and the term 'because' in sentences like 'Joe might be the thief because he has crumbs in his pocket'. I'll post more about this project soon!