An Alleged Ambiguity and the Dynamics of Context-Change

The observation that demonstrative expressions allow for both bound and referential readings, and can be bound across sentence boundaries, provides independent motivation for a shifty account of context. Dynamic semantics offers an elegant model of shiftiness, in treating the context as a running record of potential interpretive dependencies, and utterances as instructions to update and possibly change extant dependencies. Such an account advances over the static Kaplanean model insofar as it allows for the interpretation to be dynamically affected by the linguistic elements in the preceding discourse. However, due to the way it represents linguistic dependencies, and due to its reliance on both linguistic and non-linguistic effects of context to determine interpretation, the account still makes demonstrative pronouns indefinitely ambiguous at the level of logical form, thus inheriting some of the theoretical problems of the static account.

The Grammar of Prominence

This chapter draws theoretical conclusions and outlines directions for future developments. It summarizes the key theoretical and philosophical upshots of the account developed in the book and discusses further extensions of this framework. It discusses how the account can be applied to model context-sensitivity of situated utterances, in a way that can offer insights into puzzles concerning disagreement in discourse and communication under ignorance, which have plagued standard accounts of context and content. Further, it outlines the way the account is to be extended and applied to various types of context-sensitive items, including relational expressions, gradable adjectives, and domain restriction.

Prominence Semantics for Modality

This chapter develops a formal model of context-sensitivity of modal discourse. Much like demonstrative pronouns, modals are prominence-sensitive, selecting the most prominent candidate interpretation. The prominence ranking of candidate interpretations is recorded in the conversational record, and is maintained through the effects of discourse conventions represented in the logical form of a discourse. In this way arguments are individuated as structured discourses that underwrite a particular propositional pattern. It is shown that such account provably preserves classical logic. Further, this chapter argues that its model offers a more satisfactory account of the individuation of argument patterns in natural language discourse then the competing alternatives. Any adequate account, it is here argued, will have to take into account not just the contribution of individual sentences, but also of discourse conventions. Indeed, the contribution of discourse conventions is crucial for determining the contribution of individual sentences in the first place.

Introduction

The chapter provides an introduction to the key themes of the book. It introduces the problem of context-sensitivity, and its theoretical significance. It then outlines the key elements of the account the book develops—the notion of context, of content, and of context-content interaction—situating them with respect to the dominant tradition in theorizing about context-sensitivity. The chapter, finally, outlines some of the philosophical ramifications of this account and of its criticism of the traditional model for the nature of context, content, and their interaction. The book argues that the traditional model of context-sensitivity underlies the linguistic arguments for non-propositional accounts of content and non-classical semantics that have gained support in recent literature. But it is the traditional model that should be abandoned instead, and replaced with a more adequate linguistic model of context this book develops.

The Challenge: Non-propositionalism

Recent literature has presented a serious challenge for propositional accounts of content. It has been argued that certain bits of natural language discourse, in particular, modal claims, pose a fundamental challenge for propositional accounts, as they fail to express propositional content even relative to a context. The puzzling linguistic behavior of modal discourse suggests that context simply cannot determine propositional content for such claims. This appears to call for a re-thinking of the interaction between context and content, and their role in communication. Indeed, various non-propositionalist accounts that have been proposed to capture this puzzling behavior call for such re-thinking. Such accounts have received various implementations, for instance, in various expressivist and dynamic update semantics. These accounts deny that modals express ordinary propositional content, and they also deliver a non-classical logic. This chapter introduces the challenge, and the main features of non-propositional accounts that have been proposed as a solution.

Content in Context

On the traditional picture, sentences express content relative to context. This content then is, or determines, truth-conditional, propositional content, which is what we assert and believe, and which can guide our action. If I have a thought about the world, and I want to convey it to you, I should utter a sentence which, in this context, expresses that thought. You can then understand it, and come to believe it, and it might guide your action. But on the current proposal the context is constantly changing, even mid-utterance, and utterances are interpreted as instructions to update the context. What of our simple account of thought, communication, and action? This chapter shows our dynamic account still delivers propositional content. While utterances are semantically assigned dynamic meaning, this meaning serves as an instruction to build ordinary propositional content.

Pointing Things Out: Prominence and the Attentional State of a Discourse

This chapter provides the key elements of the account of context and context-sensitivity. Like the shifty, dynamic account of context in Chapter 3, this account models context as a running record of candidate interpretations for context-sensitive items, for exmaple, demonstrative pronouns. However, by contrast, the chapter argues that context organizes candidate interpretations by prominence. Further, it argues that prominence is fully linguistically determined: the context is updated exclusively by linguistic rules, through effects triggered by elements in the logical form of a discourse. These elements are richer than standardly thought, including, for example, contributions of demonstrative gestures. A demonstrative, then, receives a simple unambiguous meaning: as a matter of its meaning, it simply selects the most prominent interpretation that satisfies its lexically encoded constraints. This account is argued to be empirically more adequate, and avoids the theoretical challenges facing the alternatives discussed in earlier chapters.

Context and Discourse Conventions

This chapter further develops the idea that discourse conventions govern the dynamics of prominence, and determine the state of the conversational record, fixing the interpretation of an occurrence of a prominence-sensitive expression, such as a demonstrative pronoun. The chapter identifies a range of linguistic mechanisms—discourse conventions—that affect prominence as a matter of their grammatical contribution reflected in the logical form of a discourse. Specifically, it is argued that mechanisms of discourse coherence—the inferential connections between individual utterances that signal how they are organized into a coherent discourse—affect the contextual prominence ranking of candidate interpretations for demonstrative pronouns as a matter of their grammatically encoded contribution. The meaning of a demonstrative is then determined linguistically through and through. While demonstratives are prominence-sensitive, they are not sensitive to non-linguistic features of utterance situation.

Interlude: Context and Common Ground

An influential alternative account of context that likewise models context as a body of information that changes with an evolving discourse is Stalnakerian common ground model. On this model, however, the context is projected from a body of information mutually accepted by the interlocutors for the purposes of a conversation—a common ground. While the context constantly changes, these changes simply reflect the agents’ rational and cooperative response to manifest evidence. Might one attempt to assimilate the kinds of effects on prominence simply to such rational responses to manifest evidence? Might we then do without the rich discourse structure posited in this chapter? It is argued here that this account would be empirically inadequate, failing to capture the special status linguistic conventions have when weighed against our overall evidence.

Dynamic Propositionalism

This chapter tackles the challenge of non-propositionalism. It argues that the source of the puzzle motivating non-propositionalism is the implicit assumption of the traditional, extra-linguistic account of context-sensitivity resolution. The problem is not in the idea that modal claims express truth-conditional content, but in the underlying assumption of how a context operates to determine this content. With a more nuanced understanding of the linguistic mechanisms driving context-sensitivity resolution, which captures the effects of discourse conventions, the apparent non-propositionality of modal discourse turns out to be an illusion. The account delivers ordinary propositional content even for discourses that prima facie evade propositionalist treatment. More importantly, a broader range of data suggests that such propositional content is required to properly account for the range of interpretations modal discourses allow. Thus, any adequate account has to take into account how discourse conventions identified in this chapter interact with the interpretation of modality.