Word and Plan

John MacFarlane (University of California, Berkeley)

Umeå University

The Department of Historical, Philosophical and Religious Studies

THE BURMAN LECTURES IN PHILOSOPHY 2025

Word and Plan

John Macfarlane, University of California, Berkeley

Lecture 1: Felicitous Underspecification

Wednesday October 8, 13:15-15:00, Hörsal HUM.D.220

Abstract: In recent work, Jeffrey King has called our attention to the phenomenon of “felicitous underspecification”: felicitous uses of context-sensitive language in the absence of determinate intentions about the needed contextual supplementation. An example would be talk of a “local shop” in the absence of a determinate intention defining the scope of the locality (local-to-the-neighborhood, local-to-the-city, local-to-the-region?). After discussing the challenge this phenomenon poses for standard theories of communication, I consider King’s own solution and argue that it is inadequate. I then describe the solution I think is needed, which makes use of the basic ideas of Allan Gibbard’s plan expressivism. According to this approach, ordinary descriptive claims like “I went to a local bar” must be understood as expressive of practical plans for the use of words, as well as ordinary beliefs. Indeterminacy amounts to practical indecision.

Lecture 2: Disagreement and Meaning

Thursday October 9, 13:15-15:00, Hörsal HUM.D.210

Abstract: Philosophers often argue from premises about disagreement to conclusions about meaning. For example, from the fact that a fan of brutalist architecture who calls a building “beautiful” thereby disagrees with a traditionalist who calls it “not beautiful,” we may infer that the two parties mean the same thing by “beautiful.” For if they did not, their claims would only have the surface appearance of inconsistency. This form of argument has played a central role in meta-ethics, aesthetics, and discussions of contextualism in epistemology and philosophy of language, but recently its validity has been challenged (most influentially by David Plunkett and Timothy Sundell). It has been argued that the two parties can disagree even while meaning different things by “beautiful” and asserting compatible claims; the locus of disagreement is not what they have asserted, but the competing normative views about how “beautiful” ought to be used they have thereby expressed. This sort of disagreement has been called a “metalinguistic negotiation.” I give reasons for doubting that any interesting cases of disagreement are metalinguistic negotiations in this sense. But I think there is something right about the idea that, in making assertions, we express normative proposals for the use of words. I argue that the expressivist theory sketched in Lecture 1 captures what is plausible in the metalinguistic negotiation account while avoiding its implausible features (and vindicating the argument from disagreement).

Lecture 3: Panvariabilism

Friday October 10, 13:15-15:00, Hörsal HUM.D.220

Abstract: Several philosophers have argued recently that pronouns and proper names should be treated semantically as variables. This treatment allows us to see them both as directly referential (in agreement with Kaplanian orthodoxy) and as potentially shifting their reference in modal and doxastic contexts (against Kaplanian orthodoxy). I show that if the arguments for variabilism are any good, they generalize to all semantic categories and therefore support panvariabilism, the view that all lexical items should be treated as variables. I then give an independent motivation for panvariabilism: it is needed to solve an analogue of the problem of felicitous underspecification that arises, not for the contextual supplementations of lexical items, but for the lexical items themselves.  Panvariabilism is the semantics we need to fit the pragmatics argued for in the first two lectures.

All interested are welcome to these lectures!