Published Work

According to traditional versions of the mental file theory, we should posit mental files—that is, mental representations with containment structure—to explain both rational relations between the attitudes, and the persistence of the attitudes across time. However, Goodman and Gray (2020) offer a revisionary interpretation of the file framework, according to which its explanatory commitments are better presented by positing mental filing, as a process, but not mental files, as mental representations with file structure. Goodman and Gray focus on a certain class of synchronic explanations but files have been thought to play an essential diachronic role in the maintenance and

update of object-directed thought. This paper has two aims. First, we clarify the aims and commitments of Goodman and Gray (2020). Second, we extend their mental filing view to show how it can account for continued belief, change of mind and persistence of the attitudes in general.

Plausibly, two subjects' thoughts about the same object are only coordinated—that is, only allow for genuine communication, disagreement, etc.—if they are part of a representational system that makes it no accident that they are thinking about the same thing. Traditionally, coordination has been understood as an equivalence relation. On that model, coordination induces equivalence classes of subjects who are capable of `thinking the same'. But there is reason to think that coordination is intransitive. In this paper, I explore what an intransitive conception of coordination would look like. I argue that to model it we need to employ a non-standard, because intransitive, notion of logical consequence. It is an upshot of this approach that there is no simple answer to the question of whether two subjects are `thinking the same'. 

Among the virtues of Relationist approaches to Frege's puzzle is that they put us in a position to outline structural features of the puzzle that were only implicit in earlier work. In particular, they allow us to frame questions about the relation between the explanatory roles of sense and sameness-of-sense. In this paper, I distinguish a number of positions about that relation which have not been clearly distinguished. This has a few payoffs. It allows us to shed light on recent controversies about the `essential indexical'. It also allows us to see what's at stake between Relationists and their opponents. When we see what's at stake, we can see that we have reason to adopt an account of cognitive significance that incorporates elements of both Relationist and Fregean approaches. 

We draw attention to a series of implicit assumptions that have structured the debate about Frege’s Puzzle. Once these assumptions are made explicit, we rely on them to show that if one focuses exclusively on the issues raised by Frege cases, then one obtains a powerful consideration against a fine-grained conception of propositional attitude content. In light of this consideration, a form of Russellianism about content becomes viable.

The goal of this paper is to offer an interpretation of the mental files framework. In particular, we aim to characterize the basic explanatory commitments of the approach in non-metaphorical terms. We show how to replace the file-metaphor, and talk of thoughts being ‘contained’ within files, with theses framed in semantic and metasemantic terms. In brief, we will argue that the metaphor of mental files can be cashed out in terms of relational representational facts (‘de jure coreference’) and a metasemantic thesis about the role that information-relations to objects play in grounding those relational representational facts.

Frege's Puzzle is a founding problem in analytic philosophy. It lies at the intersection of central topics in the philosophy of language and mind: the theory of reference, the nature of propositional attitudes, the nature of semantic theorizing, the relation between semantics and pragmatics, etc.  This chapter is an overview of the puzzle and of the space of contemporary approaches to it.

Fregeanism and Relationism are competing families of solutions to Frege's Puzzle, and by extension, competing theories of propositional representation. My aim is to clarify what is at stake between them by characterizing and evaluating a Relationist argument. Relationists claim that it is cognitively possible for distinct token propositional attitudes to be, in a sense, qualitatively indistinguishable - to differ in no intrinsic representational features. The idea of an `intrinsic representational feature' is not, however, made especially clear in the argument. I clarify it here and, having done so, offer reason to doubt the argument. This will put us in a position to draw some lessons about how Relationism and Fregeanism ought to be pursued.

[This is the descendant of a paper called "Name-bearing and Response-dependence"]. Metalinguistic approaches to names hold that proper names are semantically associated with name-bearing properties. I argue that metalinguistic theorists owe us an account of the metaphysics of those properties. The unique structure of the debate about names gives an issue which might look to be narrowly linguistic an important metaphysical dimension. The only plausible account of name-bearing treats name-bearing properties as a species of response-dependent property. I outline how such an account should look, drawing on forms of response-dependence identified in the literature on colour, moral properties, etc. Having done that, I show how the account can illuminate a feature of the communicative function of names which would otherwise be puzzling from the perspective of metalinguistic accounts.

Frege's Puzzle challenges us to explain what, in addition to how it portrays the distribution of properties and relations over objects, determines the cognitive significance of a representation. Fregean solutions hold that whenever an object is represented, it is presented in a particular way. And the fact that the same object can be presented in different ways explains how the cognitive significance of a representation can exceed what object it presents. In this article, I characterize a class of alternative approaches to Frege's puzzle: relational approaches. Relational approaches can be characterized by two theses: 1) the cognitive difference between referentially-equivalent terms is explained in terms of how those terms are `coordinated' with other terms, 2) the fact two terms are coordinated is not explained in terms of a match between the intrinsic representational features of those terms, but rather in terms of irreducibly relational representational features. I then characterize a range of arguments that have been offered for relational approaches, and distinguish between two ways of developing the approach: Formal Relationists posit rationally relevant relations that hold between vehicles of propositional content; Semantic Relationists posit relations that hold between between propositional constituents.

Predicativists hold that proper names have predicate-type semantic values. They face an obvious challenge: in many languages (English among them) names normally occur as, what appear to be, grammatical arguments (call these bare occurrences). The standard version of predicativism answers this challenge by positing an unpronounced determiner in bare occurrences. I argue that this is a mistake. Predicativists should draw a distinction between two kinds of semantic type - underived semantic type and derived semantic type. The predicativist thesis concerns the underived semantic type of proper names and underdetermines a view about the semantic type of bare occurrences. I'll argue that predicativists should hold that bare names are derived individual-denoting expressions. I end by considering what this result means for the relationship between predicativism and other metalinguistic theories of names.

This paper is about how to evaluate evidence for predicativism about names. I aim to point out some under-appreciated thorny issues and to offer both predicativists and non-predicativists some advice about how best to pursue their respective projects. I argue for three related claims: (1) That non-predicativists have to posit relatively exotic, though not entirely implausible, polysemic mechanisms to capture the range of data that predicativists have introduced; (2) that neither referentialism nor extant versions of predicativism can offer a very plausible account of  the interpretive possibilities for singular unmodified definite descriptions containing names; and (3) that the most plausible version of predicativism would treat bare names as non-anaphoric definite descriptions.

[This the final form of a paper that used to have slightly different titles] Despite its increasing popularity, Predicativism about names - the view that names are metalinguistic predicates - has yet to confront a foundational issue: how are names individuated in the lexicon? In this paper I argue that the most straightforward answer to this question, the one implicitly assumed by those who defend the view, is untenable. Predicativists hold the meaning of a name N can be given by the following schema: x satisfies N iff x is called `N'. This entails that for every sign s there is exactly one name-predicate articulated with s. I show that this approach will not work - it founders on cases in which individuals have names which are spelled in the same way but pronounced differently (or vice versa). I consider two possible modifications to the Predicativist picture, and argue that neither will work. I close by suggesting that this challenge is the expression of a fundamental problem for Predicativism.

Call an account of names satisfactionalist if it holds that object o is the referent of name a in virtue of o’s satisfaction of a descriptive condition associated with a. Call an account of names minimally descriptivist if it holds that if a competent speaker finds ‘a=b’ to be informative, then she must associate some information with ‘a’ which she does not associate with ‘b’. The rejection of both positions is part of the Kripkean orthodoxy, and is also built into extant versions of the file-picture of reference. In this paper, I argue that the rejection of minimal descriptivism only follows from the rejection of satisfactionalism given certain implausible assumptions about the nature of competence with a proper name. I do this by showing that considerations internal to the file-picture - in particular the idea that competence with a proper name constitutes an ‘epistemically rewarding’ relation to its bearer - motivate an acceptance of minimal descriptivism.

Proponents of the predicate view of names explain the reference of an occurrence of a name N by invoking the property of bearing N. They avoid the charge that this view involves a vicious circularity by claiming that bearing N is not itself to be understood in terms of the reference of actual or possible occurrences of N. I argue that this approach is fundamentally mistaken. The phenomenon of ‘reference transfer’ shows that an individual can come to bear a name in virtue of the referential practices of a group of speakers. I develop a picture of name-bearing which captures this fact by treating the extension of name as a function of the way that extension is represented in the presuppositions of groups of speakers. I show that though there is a form of circularity inherent in this approach, it is not vicious circularity.