Hi, I'm Mike Deigan, a philosophy PhD student at Yale. I have broad interests, but mostly work on language and ethics, trying to answer questions like “What does the word ‘a’ mean?” and “How should we live?”.
Comments/questions are welcome!
We sometimes lead double lives in the consequents of counterfactuals. In counteridenticals like 'If I were you, I would like me.', 'I' and 'me' seem to pick out different individuals. This, I argue, is a problem for the orthodox Kripke-Kaplan view of indexicals as rigid designators, a view which requires 'I' and 'me' to have the same referent in all worlds of evaluation.
I show this problem is not limited to counteridenticals, but also appears in 'ordinary' counterfactuals like "If I were a policeman, I would arrest me." and modal subordination, as in "I could have been a policeman. I would have arrested me for what I just did." So the problem cannot be solved by a special treatment of counteridenticals. Nor, I argue, can it be analyzed as involving descriptive indexicals. Instead, we should make a more dramatic fix, such as moving to counterpart theory for de re modal ascriptions.
I present data that suggest the universal entailments of counterfactual donkey sentences aren't as universal as some (Van Rooij (2005) and Walker and Romero (2015)) have claimed. I argue that this favors the strategy of attributing the universal entailments to a special property of the similarity ordering on worlds provided by some contexts, rather than to a semantically encoded sensitivity to assignment.
Sometimes when someone says something, we don't understand what it means, but nevertheless go along with it and signal acceptance. I argue that this phenomenon, which I call ‘stupefying’, is both practically important and theoretically interesting. Unlike the backdoor, non-at-issue content-based mechanisms emphasized by Langton, Stanley, and others, stupefying is a means of conversationally short-circuiting rational deliberation which can rely entirely on at-issue meaning.
Modelling and understanding stupefying, then, is a worthwhile endeavor. I try to make some headway on this in the remainder of the paper by extending some tools developed by Yalcin (2016 and elsewhere) and Bledin and Rawlins (2016).