Search Results

This chapter examines the prospects for explaining away anti-substitution intuitions as due to conversational implicatures. It does this by exploring existing accounts, but also by considering alternative possible accounts. It argues that these accounts all fail for the same reasons as those considered in Chapter 1: The Enlightenment Problem and The Aspect Problem.

This chapter reviews efforts to explain away the problems with Millian theories, and shows that they are unsuccessful. Soames’s solution to Russell’s problem (existence failures) either adopts a radical modal realism or uses an intentional sense of reference. Some attempts to rebut Frege’s problem (substitutivity failures) ignore opaque interpretations. Kripke took the argument against substitutivity to depend on strong disquotation principles, and presented a puzzle to show that they are problematic. But weaker disquotation principles are available, and substitutivity arguments can be presented without relying on disquotation principles. Gricean attempts to explain away substitutivity failures in terms of metalinguistic or mode implicatures do not help with Russell’s problem or Fine terms, and are unsatisfactory even for Frege’s problem.

Grice’s definition of conventional implicatures, employed throughout this book, has four primary components: conventionality, speaker commitment, speaker orientation, and semantic independence. This closing chapter explores the regions around the core cases of conventional implicature by asking, for each of these central properties, what happens if we remove it from the definition. Some of these alterations characterize rich areas of natural language meaning, whereas others seem unattested.

This chapter offers a brief introduction to the core ideas and gives some notation concerning truth conditional semantics. It aims to revive earlier experiences with the field and ease later contact with semantic representations of the items under investigation. Particular emphasis is laid on the explicitness of the paradigm which encourages to specify literal contents of words, the modes of semantic composition, presuppositions, implicatures and default inferences, indexical anchors of utterances, and other conceptual parameters that may influence the use and interpretation of language.

This chapter shows how the expression theory can account for four cases of nondescriptive meaning: interjections, syncategorematic terms, words with conventional implicatures, and non-declarative sentences. The meaning of interrogative and imperative sentences is explained in terms of non-propositional thoughts. Syncategorematic terms express thought-parts that are no more incomplete or dependent than those expressed by other terms. Interjections provide an exception to the rule that meaning consists in the expression of ideas; they express mental event types that are not thought parts (nor emotions). Conventional implicatures are explained in terms of appositional thoughts, complexes with a main thought and a subordinate thought.

Conversational implicatures do not normally fall within the scope of operators because they arise at the speech act level, not at the level of sub-locutionary constituents. Yet in some cases they do, or so it seems. ‘Embedded implicatures’ are seeming implicatures that arise locally, at a sub-locutionary level, without resulting from an inference in the narrow sense. The phenomenon has received a good deal of attention recently and has led many researchers to cast doubt on the traditional division of labour between semantics and pragmatics. This chapter discusses the relation between various possible approaches to the phenomenon, including Truth-Conditional Pragmatics.

This chapter gives a nuanced account of the phenomena which have led moral theorists to explore so-called ‘thick concepts’. It argues that the phenomena are far more deeply embedded in linguistic practice that usually thought, and do nothing to support the idea of specific kinds of concept.

An introduction to the notion of modularity of mind and an argument as to why only formal semantic theories are compatible with the claim that semantic comprehension is the product of a (Fodorian) modular system. This chapter also looks at some initial challenges to formal semantics stemming from the apparent place of pragmatic reasoning in our grasp of meaning. These include arguments concerning the nature of speech acts, the analysis of implicatures, word learning, and ambiguity.

A new analysis of convention and of linguistic function yields a robust description, in naturalistic terms, of the distinction between semantics and pragmatics. For a number of different but instructive reasons, this distinction is intrinsically blurry, depending on statistics over individual psychological processing techniques. Implications for the theory of language interpretation are discussed, with the conclusion drawn that there are many ways of grasping the content that the specific speaker intends to convey without employing a theory of mind. A coordinate claim is that during normal conversation, it is not language that is most directly perceived by the hearer but rather the world that is most directly perceived through language.

The standard model in the pragmatics literature focuses on ordinary conversations, in which the parties are presumed to engage in a cooperative exchange of information. The legal context offers an example of conversation that is strategic in nature. This chapter shows that the pragmatics of strategic conversation has certain features that deviate from the standard model. The first section focuses on two main instances of implied communicative content, namely, implicatures and utterance presuppositions. It argues that in both of these cases, there is an important distinction between implied content that is semantically encoded in the utterance — and therefore forms part of what the law communicatively determines — and implied content that is essentially contextual and thus much more problematic in the legal case. The second section focuses on the idea of pragmatic commitments and their normative foundations. The chapter explores the normative framework of strategic speech and ways in which it differs from ordinary conversations. Finally, it tries to explain in what sense legal speech is strategic, and demonstrate how the pragmatic aspects of strategic speech actually work in the legal context.

This chapter reviews accounts that explain conversational implicatures by appeal to distinctively linguistic knowledge and principles, rather than as consequences of general rationality. The chapter begins with Horn’s approach: Horn assumes that pragmatic principles offer potentially conflicting ways to enrich semantic content, and speakers resolve these conflicts in part using linguistic knowledge, specifically scales of alternative items ordered by strength or markedness. The chapter continues with Levinson’s approach. He argues that pragmatic principles don't just apply to the propositions speakers contribute to conversation; rather, they pervasively disambiguate and enrich linguistic meanings in tandem with compositional semantics. Ultimately, the chapter suggests that such neo-Gricean explanations, while they seem to exhibit much the same form as Grice's, in fact involve very different philosophical commitments.

Grice takes hints and suggestions as paradigm cases of conversational implicatures, but he does not attempt to delimit them precisely. This chapter works to characterize a specific, common-sense practice of hinting: using an utterance to shift the initiative by pressing the addressee to flesh out and push forward a direction to the conversation that the speaker will not or cannot develop further herself. Hints may seem to illustrate cooperative reasoning in action but in fact they differ in key ways from the technical apparatus of familiar pragmatic theories. Most importantly, the interpretation of a hint is not dictated by principles of rationality or general interpretive mechanisms. Hints leave information hidden; their open-endedness shows in the fact that there's no clear answer as to what exactly a hint requires the addressee to appreciate, and in the fact that they are difficult, unreliable and consciously interpreted.

This chapter develops a research program on implicatures and discourse structure with implications for grammar. The program shows the centrality of discourse structure in linguistic interpretation by detailing how discourse structure provides an important source of information to computing so-called scalar implicatures. While scalar implicatures have typically received a Gricean treatment based on reasoning about the content of sentences in isolation since the work of Larry Horn, the Gricean tradition misses an important, indeed decisive, component in the calculation of implicatures—discourse structure. This has general implications for the way discourse structure is treated within grammar. This chapter considers settings where implicatures are generated in the absence of strong Gricean cooperativity, in particular, non-cooperative settings where Gricean principles do not really apply as they are formulated by current proposals. The study proposes a different mechanism for deriving implicatures in such settings.

Chapter 2 characterizes Free Choice Disjunction (FCD), i.e. the behavior of disjunction (and plain indefinites) under modals. This is a well known phenomenon whereby a sentence of the form possible (p or q) winds up unexpectedly being interpreted as possible and possible q. It is argued that such a phenomenon bears the hallmark of Scalar Implicatures (SIs). Gricean theories of SIs and some of their limitations are reviewed. A novel approach is presented, based on the idea that SIs come about via a covert counterpart of only (‘exhaustification’) that operates on grammatically determined alternatives, and is subject to a principle of ‘Maximize Sternght/Informativity’. Then a version of the theory of FCD by D. Fox is articulated, that derives the FC effect through recursive exhaustification of subdomain alternatives. This approach to FCD provides independent motivation for the thesis that polarity phenomena are related to the activation of subdomain alternatives by indefinites, put forth in Chapter 1. It is proposed that covert only (O) targets alternative bearers in its domain, much like wh-words are targeted by interrogative complementizers, an idea that is implemented via a feature checking mechanism. O assign ‘+’ to alternative bearers in its domain, thereby making their alternatives active. In absence of O, alternative bearers get a default value ‘-‘, and their alternatives remain inactive. This sets the stage for an integrated theory of polarity sensitivity and free choice.

Chapter 5 studies indefinites that are restricted to Free Choice environments, sometimes also described as ‘epistemic indefinites’. This includes German irgendein, Italian uno qualunque and Romanian vreun. It is argued that the FC effects that these items give rise to are identical to those that come about with ordinary disjunction and ordinary indefinites in similar contexts. The switch between pure NPIs that disallow FC uses and Polarity Items that allow them is located in the (non) availability of ‘pre exhaustified’ (i.e. recursively exhaustified) domain alternatives: NPIs disallow them (and are thereby restricted to DE contexts), FC indefinites allow them (and thereby are admitted in modal contexts where the FC effect comes about). This arguably constitutes a genuinely uniform theory of Negative Polarity and Free Choice Items. Several further dimensions of variation in epistemic indefinites are identified and a small number of formal parameters that generates such variation are identified. For example, domain size is responsible for total vs. partial variation, obligatory strengthening for ‘antilicensing’ in strongly negative environments. A novel generalization pertaining to the relation between Epistemic Indefinites and numerals is identified and discussed.

It has been noted in the literature that NPIs are degraded if there is an intervening element (typically, a non weakest member of a scale) between them and, in traditional terms, their licensor (in our terms, the exhaustifying operator). This chapter discusses this phenomenon and investigates several arguably related cases of intervention by presupposition triggers. The proposal is that this effect follows as a case of minimality. If. in probing for a Polarity Sentitive item, O finds along its path a closer alternative bearer (i.e. a scalar term), it is forced to activate the alternatives of the scalar term. In negative contexts, this brings about a positive implicature that makes exhaustification fail. Relativized minimality explains why phrasal interveners (e.g. strong quantified NPs) are generally more disruptive than heads (e.g. modals). This approach is extended to presuppositional intervention (diagnosed by V. Homer) in two ways. Some presuppositional triggers (most prominently factives) turn out to be a special case of scalar triggers (as argued by J. Romoli). Others are argued to take on the behavior (and featural make up) of strong NPIs. In all three major cases of intervention considered in this chapter, syntactic locality plays a crucial role exposing the interconnected nature of the syntax and semantics of exhaustification.

Referential descriptions can vary in the quantity and nature of the information they encode. How much and what types are jointly determined by the communicative needs of the addressee and cognitive pressures on the speaker. This chapter explores the interplay of these two forces in cooperative and competitive contexts. Participants were asked to play a communication game in which a speaker directed an addressee to manipulate objects in an array. The perspectives of speaker and addressee were arranged so that one of the objects visible to the speaker was concealed from the addressee. Speakers were adept at adapting their descriptions according to their communicative goals and the needs of their audience. However, they did not do so perfectly in either cooperative or competitive situations. There were indications that speakers may have been more sensitive to their audience’s knowledge and needs in competitive than cooperative situations. This may be because cooperative communication usually allows speakers to rely on addressees to indicate when their needs have not been met. Competitive communication does not allow for such feedback, and thus places greater urgency on keeping independent track of the addressee’s perspective.

This chapter focuses on various forms of implicated content in statutory law, mostly conversational implicatures and presuppositions. The main argument of the chapter is that the strategic nature of legal communication calls into question the reliability of implicated content in the law. The argument shows that both the legislatures and the courts have an interest in maintaining a strategic form of communication between them, and that this strategic communication is enabled by a level of uncertainty about the normative framework that governs their ongoing conversation.