Rohan Sud

Our words are vague, yet the world is precise. This dissertation consists of three essays that jointly attempt to articulate, explore, and defend this claim. In the first essay, I present my positive proposal for a non-metaphysical treatment of vagueness. Inspired by research on the Problem of the Many, I claim that we are speaking many perfectly precise languages simultaneously and that each speech act involves uttering many perfectly precise sentences. This position, which I call supersententialism, is the most plausible alternative to metaphysical vagueness: it provides a fully reductive account of vagueness that preserves the T-schema. The second essay aims to address a rampant, albeit unnoticed, argument schema that purports to show that non-metaphysical accounts of vagueness cannot explain vagueness in fundamental, or perfectly natural, properties. In response, I propose a position that I call ersatz metaphysical vagueness according to which the term ‘perfectly natural’ is vague. Ersatz metaphysical vagueness retains the idea that the world is precise, but replicates the advantages of genuine metaphysical vagueness. I show how, by adopting this hypothesis, there can be vagueness in attributions of perfectly natural properties arising from the vagueness in our notion of perfect naturalness rather than from ‘the world itself’. Ersatz metaphysical vagueness thus provides us with a highly general alternative to genuine metaphysical vagueness. The final essay explores the relationship between metaphysical vagueness and ontological deflationism. Ontological deflationist positions treat the quantifier as having a highly plastic meaning. This plasticity naturally leads to vagueness in the quantifier. Thus, ontological deflationists are committed to vague existence. Is this vague existence ‘metaphysical vagueness’ or ‘vagueness in the world’? The rest of the essay clarifies this question and answers it under various clarifications.

Many accept David Lewis's (1983) claim that, among the candidate meanings for our predicates, some are more natural than others -- they do better or worse at ``carving nature at its joints''. Call this claim predicate naturalism. Disagreement remains over whether the notion of naturalness extends ``beyond the predicate'' (à la Sider, 2011). Are the candidate meanings of logical vocabulary also more or less natural? Call this claim logical naturalism. One motivation for predicate naturalism comes from its supposed ability to help rebut various radical indeterminacy arguments associated with Hilary Putnam (1977, 1980, 1981). It does so in combination with a popular meta-semantic theory called reference magnetism. I argue that the same threats of Putnam-style radical indeterminacy rearise for proponents of predicate naturalism -- threats which logical naturalism helps rebuts. So, if we accept predicate naturalism because it helps fend off threats of indeterminacy, we should also accept logical naturalism.

This paper asks: Is the quantifier variantist committed to metaphysical vagueness? My investigation of this question goes via a study of vague existence. I’ll argue that the quantifier variantist is committed to vague existence and that the vague existence posited by the variantist requires a puzzling sort of metaphysical vagueness. Specifically, I distinguish between (what I call) positive and negative metaphysical vagueness. Positive metaphysical vagueness is (roughly) the claim that there is vagueness in the world; negative metaphysical vagueness is (roughly) the claim that there is vagueness that is not in our language or thought. I’ll argue that the quantifier variantist’s commitment to vague existence comes with positive metaphysical vagueness -- even if it doesn't require negative metaphysical vagueness.

One desideratum for a theory of fundamentality is to give us the conceptual tools to articulate fruitful metaphysical distinctions between the assortment of ‘realist’ and ‘anti-realist’ positions in a given domain such as meta-ethics. The ability to articulate such distinctions gives us a way to assess rival theories of fundamentality, such as Fine’s grounding theory and Sider’s metaphysical semantic theory. Indeed, Sider has argued that metaphysical semantic theories have an edge with respect to this desideratum and takes this as an important reason to prefer those theories over rival grounding theories. This paper takes a closer look at how the rival theories compare with respect to questions of realism and defends the grounding theorist’s ability to meet the desideratum.

Does moral vagueness require ontic vagueness? A central challenge for nonontic treatments of moral vagueness arises from the referential stability of moral terms across small changes in how they are applied: if moral vagueness is not ontic vagueness, it’s hard to explain this referential stability. Pointing to this challenge, Miriam Schoenfield has argued that moral vagueness is ontic vagueness, at least for a moral realist. I disagree. I argue that a moral realist can use a conceptual role semantics for moral terms to give a purely semantic treatment of moral vagueness.

I propose a view that I call "Ersatz Metaphysical Vagueness" according to which the term "perfectly natural" can be semantically vague. As its name suggests, the view mimics traditional metaphysical vagueness without the radical metaphysical underpinnings. In particular, the ersatzer avoids a widely accepted argument schema (advanced by JRG Williams, Ted Sider, Cian Dorr, John Hawthorne and others) according to which, if there is no metaphysical vagueness, F-ness cannot be both perfectly natural and vague.

According to the plurivaluationist, our vague discourse doesn’t have a single meaning. Instead, it has many meanings, each of which is precise—and it is this plurality of meanings that is the source of vagueness. I believe plurivaluationist positions are underdeveloped and for this reason unpopular. This paper attempts to correct this situation by offering a particular development of plurivaluationism that I call supersententialism. The supersententialist leverages lessons from another area of research—the Problem of the Many—in service of the plurivaluationist position. The Problem reveals theoretical reasons to accept that there are many cats where we thought there was one; the supersententialist claims that we are in a similar situation with respect to languages, propositions and sentences. I argue that the parallel suggested by the supersententialist reveals unappreciated advantages and lines of defense for plurivaluationism.

Adam Elga (Philosophers’ Imprint, 10(5), 1–11, 2010) presents a diachronic puzzle to supporters of imprecise credences and argues that no acceptable decision rule for imprecise credences can deliver the intuitively correct result. Elga concludes that agents should not hold imprecise credences. In this paper, I argue for a two-part thesis. First, I show that Elga’s argument is incomplete: there is an acceptable decision rule that delivers the intuitive result. Next, I repair the argument by offering a more elaborate diachronic puzzle that is more difficult for imprecise Bayesians to avoid.