Skip to main content
Skip to main content

Philosophy Colloquium - John MacFarlane / Panvariabilism

Close portrait of the face older man in glasses and a cap, with a short grey beard and mustache, standing on a California beach at dusk, the silvery Pacific behind him as the sky glows with peachy warmth.

Philosophy Colloquium - John MacFarlane / Panvariabilism

Linguistics | Philosophy Friday, April 26, 2024 3:00 pm - 5:00 pm Skinner Building,

Friday April 26, John MacFarlane joins us from Berkeley to give a colloquium talk titled "Panvariabilism," arguing that every lexical item should be modeled semantically as a variable. The abstract is below.


Sam Cumming, Paolo Santorio, and Dilip Ninan have argued that singular referring terms (pronouns and proper names) should be treated semantically as variables, and that epistemic expressions (including epistemic modals and attitude verbs) should be treated as assignment-shifters. Building on their work, I argue for a more radical view: all lexical items should be treated as variables, and the contents of attitudes and assertions should be modeled as sets of world/assignment pairs. I motivate this view through two distinct lines of argument. The first argues that the considerations invoked by Cumming, Santorio, and Ninan extend to all lexical categories. The second argues that assignment-sensitive contents and semantic values are needed to make sense of felicitous underspecification and metalinguistic negotiation. On the resulting picture, many questions we intuitively regard as questions about meaning are outside the scope of semantic theory.

Add to Calendar 04/26/24 3:00 PM 04/26/24 5:00 PM America/New_York Philosophy Colloquium - John MacFarlane / Panvariabilism

Friday April 26, John MacFarlane joins us from Berkeley to give a colloquium talk titled "Panvariabilism," arguing that every lexical item should be modeled semantically as a variable. The abstract is below.


Sam Cumming, Paolo Santorio, and Dilip Ninan have argued that singular referring terms (pronouns and proper names) should be treated semantically as variables, and that epistemic expressions (including epistemic modals and attitude verbs) should be treated as assignment-shifters. Building on their work, I argue for a more radical view: all lexical items should be treated as variables, and the contents of attitudes and assertions should be modeled as sets of world/assignment pairs. I motivate this view through two distinct lines of argument. The first argues that the considerations invoked by Cumming, Santorio, and Ninan extend to all lexical categories. The second argues that assignment-sensitive contents and semantic values are needed to make sense of felicitous underspecification and metalinguistic negotiation. On the resulting picture, many questions we intuitively regard as questions about meaning are outside the scope of semantic theory.

Skinner Building