Michael Glanzberg

Semanticists will often casually remark that Noam Chomsky rejects semantics. Chomsky has frequently noted how poorly understood some aspects of semantics are, and has shown little inclination to grant the status of reasonably well‐developed science to many parts of semantics. One specific reason Chomsky has often voiced skepticism about semantics is that he saw the wrong kinds of appeals to semantics in the wrong places. The arguments for the autonomy of syntax in Chomsky's early writing have been described as part of an extended argument for the “existence of syntax”. Semantics for Chomsky must be part of I‐language: the part that provides useful information from the Language Faculty to the Conceptual‐Intentional system. There are many traditions in semantics, and many current research programs in sematics are embedded in firmly anti‐Chomskyan views of language in general. Chomsky's skepticism about truth‐conditional semantics has been, and remains strong.

This chapter reviews Davidson's main work on truth. It focuses on the connections between truth, meaning, and interpretation that form the core of Davidson's views, and on the relations of his views to traditional theories of truth. It highlights several key ideas that comprise Davidson's approach to truth: Tarski's work on truth is fundamental to understanding the concept, as is the relation of truth to meaning, and we fail to understand that connection adequately unless we take into account the constraints of radical interpretation. Furthermore, in understanding those connections, we see that truth is not a redundant or trivial concept, but equally, it is not a locus for metaphysical commitments.

Gary Ostertag: Quine and Russell: The chapter provides a selective overview of themes common to Russell and Quine, focusing on Russell's theory of descriptions and the notion of contextual definition. It begins by discussing Russell and Quine on modality, along the way highlighting the following topics: how C.I. Lewis's metalinguistic understanding of the modal operators shaped the subsequent debate about modality – in particular, how it rendered the very idea of de re modality unintelligible; how Quine's inattention to matters of scope led him to conclude, somewhat misleadingly, that he had derived a modal paradox; how the failure of the reductionist program provided the basis for a challenge to a cornerstone of the understanding of modality – the idea of analytic truth. I then turn to Russell and Quine on the semantics of attitude reports, where the issue of scope takes center stage again. A final section discusses Russell's influence in the shaping of Quine's ontological criterion.

This paper revisits a challenge for contextualist approaches to paradoxes such as the Liar paradox and Russell’s paradox. Contextualists argue that these paradoxes are to be resolved by appeal to context dependence. This can offer some nice and effective ways to avoid paradox. But there is a problem. Context dependence is, at least to begin with, a phenomenon in natural language. Is there really such context dependence as the solutions to paradoxes require, and is it really just a familiar linguistic phenomenon at work? Not so clearly. In earlier work, I argued that the required form of context dependence does not look like our most familiar instances of context dependence in natural language. I called this extraordinary context dependence. In this paper, I shall explore, somewhat tentatively, a way that we can see the context dependence needed to address paradoxes as not so extraordinary. Doing so will also allow us to connect thinking about the context dependence of quantifier domains with some interesting ideas about the distinctive semantic properties of certain quantifiers.

This paper reviews the claim that certain predicates, including what are called predicates of personal taste, have a sometimes-hidden element for a judge or experiencer. This claim was advanced in my own earlier work, as well as a number of other papers. My main goal here is to review some of the arguments in favor of this claim, and along the way, to present some of my earlier unpublished work on the matter. In much of the earlier literature, this claim was part of a debate between relativists, contextualists, and others about the semantics of ‘subjective’ or ‘perspectival’ predicates. I shall argue here that these issues are independent. Whether we opt for experiencer or judge parameters is independent of whether we prefer relativist semantics to any other kind.

This chapter introduces readers to the empirical questions at issue in debates over gendered pronouns and assesses the plausibility of various possible answers to these questions. It has two parts. The first is a general introduction to the linguistics and psychology of grammatical gender. The second focuses on the meanings of gendered pronouns in English. It begins with a discussion of some methodological limitations of empirical approaches to the topic and the normative implications of those limitations. It then argues against three simple theories of the semantics of gendered pronouns in English and proposes an alternative that fares better: the Gender-First View. Finally, it discusses the singular use of 'they' and its connection to nonbinary gender identities.

This paper revisits a challenge for contextualist approaches to paradoxes such as the Liar paradox and Russell’s paradox. Contextualists argue that these paradoxes are to be resolved by appeal to context dependence. This can offer some nice and effective ways to avoid paradox. But there is a problem. Context dependence is, at least to begin with, a phenomenon in natural language. Is there really such context dependence as the solutions to paradoxes require, and is it really just a familiar linguistic phenomenon at work? Not so clearly. In earlier work, I argued that the required form of context dependence does not look like our most familiar instances of context dependence in natural language. I called this extraordinary context dependence. In this paper, I shall explore, somewhat tentatively, a way that we can see the context dependence needed to address paradoxes as not so extraordinary. Doing so will also allow us to connect thinking about the context dependence of quantifier domains with some interesting ideas about the distinctive semantic properties of certain quantifiers.

In this paper, we defend a traditional approach to semantics, that holds that the outputs of compositional semantics are propositional, i.e. truth conditions. Though traditional, this view has been challenged on a number of fronts over the years. Since classic work of Lewis, arguments have been offered which purport to show that semantic composition requires values that are relativized, e.g. to times, or other parameters that render them no longer propositional. Focusing in recent variants of these arguments involving quantification and binding, we argue that a correct understanding of how composition works gives no reason to relativize semantic values, and that propositional semantic values are in fact the preferred option. We take our argument to be mainly empirical, but along the way, we defend some more general theses. Simple propositional semantic values are viable in composition, we maintain, because composition is itself a complex phenomenon, involving multiple modes of composition. Furthermore, some composition principles make adjustments to the meanings of constituents in the course of composition. These adjustments are by triggered syntactic environments. We argue such small contributions of meaning from syntactic structure are acceptable.

One of the difficulties surrounding the ancient Paradox of the Liar is that it is unclear what philosophical problem it poses. In this thesis, I offer a diagnosis of the problem, and defend a solution to it. In Chapter I, I argue that the problem posed is one about the relation between a sentence uttered in a context and the proposition expressed. Chapter II supports this conclusion, by arguing against most notions of truth-value gap, and by elaborating on the notion of failing to express a proposition. ;To develop a solution to the problem, I argue in Chapter III that the notion of context must be expanded to take into account structural features of the discourse in which an utterance is made. So understood, a context determines a domain from which the truth conditions available for the construction of propositions expressible in the context must be drawn. ;This theory is applied to the Liar Paradox in Chapter IV. Under my enhanced notion of context, we can see certain inferences as changing a context just by introducing novel words into a conversation. In the Liar Paradox case, the term for the semantic relation of expressing a proposition is introduced. Because such relations have a highly complex local character, this context-change induces a substantial change in the domain of available truth conditions. The result is, I show, that there are effectively more propositions that can be expressed in the new context than there were in the old. Among the new propositions that can be identified in the new context is the one expressed by the Liar Sentence. ;One of the most striking features of a hierarchical account of this kind is the result that any quantification, whether syntactically unrestricted or not, is interpreted as being over some restricted domain. I go on to argue in Chapter V that in fact there are independent reasons for thinking that quantification must behave in this way. ;Finally, I elaborate some of the more general consequences of the theory I develop. First, I claim that the notion of expression I propose undermines any sharp and substantial distinction between semantics and pragmatics. Second, in Chapter VI, I argue that the theory has some consequences for contemporary debates over realism and truth

∗Thanks to J. C. Beall, Alex Byrne, Jason Decker, Tyler Doggett, Paul Elbourne, Adam Elga, Warren Goldfarb, Delia Graff, Richard Heck, Charles Parsons, Mark Richard, Susanna Siegel, Jason Stanley, Judith Thomson, Carol Voeller, Brian Weatherson, Ralph Wedgwood, Steve Yablo, Cheryl Zoll, and an anonymous referee for valuable comments and discussions. Versions of this material were presented in my seminar at MIT in the Fall of 2000, and at the University of Maryland, Baltimore County. Parts of this paper also derive from my comments on a paper of Scott Soames at the ‘Liars and Heaps’ conference at the University of Connecticut in the Fall of 2002. I am grateful for the help of these audiences, and especially to Prof. Soames.

The Taming of the True continues the project Neil Tennant began in Anti-realism and Logic of investigating and defending anti-realism. Tennant’s earlier book anticipated a second volume, in which issues related to empirical discourse would be addressed in greater detail. The Taming of the True provides this sequel. It also attempts a ground-clearing project, by addressing challenges to some of the presuppositions and implications of Tennant’s anti-realist position. Finally, it takes an opportunity to revisit some of the issues examined in the earlier book, in light of the evolving debate over realism and anti-realism.

The discussion of supervenience is replete with the use of in?nitary logical operations. For instance, one may often ?nd a supervenient property that corresponds to an in?nite collection of supervenience-base properties, and then ask about the in?nite disjunction of all those base properties. This is crucial to a well-known argument of Kim (1984) that supervenience comes nearer to reduction than many non-reductive physicalists suppose. It also appears in recent discussions such as Jackson (1998).

One of the great successes in the study of language has been the application of formal methods, including those of formal logic. Even so, this chapter argues against one way of accounting for this success, by arguing that the study of natural language semantics and of logical consequence relations are not the same. There is indeed a lot we can glean about logic from looking at our languages, and at our inferential practices, but the semantic properties of natural languages do not determine genuine logical consequence relations. We can get from natural language semantics to logical consequence, but only by a significant process of identification of logical constants, abstraction, and idealization. The chapter also discusses different approaches to the nature of logical consequence, and examines which allow logic and natural language to come closer together.

Quantification is haunted by the specter of paradoxes. Since Russell, it has been a persistent idea that the paradoxes show what might have appeared to be absolutely unrestricted quantification to be somehow restricted. In the contemporary literature, this theme is taken up by Dummett (1973, 1993) and Parsons (1974a,b). Parsons, in particular, argues that both the Liar and Russell’s paradoxes are to be resolved by construing apparently absolutely unrestricted quantifiers as appropriately restricted.

The story goes that Epimenides, a Cretan, used to claim that all Cretans are always liars. Whether he knew it or not, this claim is odd. It is easy to see it is odd by asking if it is true or false. If it is true, then all Cretans, including Epimenides, are always liars, in which case what he said must be false. Thus, if what he says is true, it is false. Conversely, suppose what Epimenides said is false. Then some Cretan at some time speaks truly. This might not tell us anything about Epimenides. But if, to make the story simple, he were the only Cretan ever to speak, and this was the only thing he ever said, then indeed, he would have to speak truly. And we would then have shown that if what he said was false, it must be true.