Rethinking the semantics of attitude reports




Rausch, Alexander Peter

Journal Title

Journal ISSN

Volume Title



In response to John Locke’s death in 1704, Pierre Coste (1824) penned a letter containing the following passage: [H]e was naturally somewhat choleric. But his anger never lasted long...I remember, that two or three weeks before his death, as he was sitting in a garden, taking the air in a bright sunshine, whose warmth afforded him a great deal of pleasure...; we happened to speak of Horace, I know not on what occasion, and having repeated to him these verses, where that Poet says of himself, that he was Solibus aptum; Irasci celerem, tamen ut placabilis essem: “That he loved the warmth of the sun, and that tho’ he was naturally choleric, his anger was easily appeased.” Mr. Locke replied, that if he durst presume to compare himself with Horace in any thing, he thought he was perfectly like him in those two respects. (170, bold emphasis mine) This passage loosely contains a sentence of the form ⌜x says of y that p⌝ and ends with an anaphoric reference to “two respects” that, at face value, are designated by clausal complements, viz. ‘that he loved the warmth of the sun’ and ‘that tho he was naturally choleric, his anger was easily appeased.’ This dissertation concerns the semantics of these and related constructions, with an emphasis placed on those of the form ⌜x believes about y that p.⌝ These belief-about reports have received less attention than reports of the form ⌜x believes that p,⌝ but the former – along with their syntactically derived forms – are just as much a part of present-day, natural language English as the latter. Indeed, once I started studying ‘about’-constructions systematically, I immediately began to notice their regular appearance “in the wild.” This dissertation represents my attempts at taming this small corner, replete with philosophical significance, of the English language. There are three chapters, and they all concern the semantics of attitude verbs with clausal complements; I focus on ‘believes’ for ease of exposition. Chapters 1 and 3 can be read as standalone works. Chapter 2 should be read after Chapter 1. In Chapter 1, “A Puzzle about Belief-about,” I argue that certain valid inferences involving belief-about reports are prima facie inconsistent with orthodox views of the belief relation as binary and propositional. In response, I propose a conservative departure from orthodoxy according to which certain ‘that’-clauses designate novel devices of semantic type <e, t> called open propositions; the view of belief as binary and propositional is retained. I give some reasons for thinking that open propositions are properties of a certain kind, give a bridge principle between belief-about and belief simpliciter, and formally implement the resulting view in accordance with contemporary theories of syntax and compositional semantics. The upshot is that theorists committed to orthodoxy must complicate their account of certain ‘that’-clauses in surprising ways. In Chapter 2, “Belief is a Ternary Relation,” I object to the semantic complexity required by the proposal advanced in Chapter 1 and investigate a more radical departure from orthodoxy, viz. that belief is a ternary relation between subjects, objects (“targets”), and properties (“contents”). After showing how the resulting Target and Content View can be formally implemented, I respond to a variety of objections that fall roughly into one of two categories: semantic and metaphysical. Responding to the semantic objections requires developing accounts of truth, assertion, and related notions, while responding to the metaphysical objections requires defending a particular view on the nature of propositions. The upshot is that if theorists are unwilling to countenance the semantic complexity required to save orthodoxy in the way proposed in Chapter 1, then the Target and Content View is an attractive alternative with theoretical benefits that are significant in their own right. The formal implementation of the view proposed in Chapter 2 presupposes an intensional semantic framework attributable to the linguistic development of variable-based theories of intensionality, which purport to explain the transparency of determiner phrases in the context of attitude reports. In Chapter 3, “Variable-based Intensionality for Structured Propositions,” I argue that these theories in their simple, traditional forms are not available to advocates of structured propositions, and that the only attempt so far to unify these approaches is unsuccessful. So, I develop an improved variable-based theory of intensionality for structured propositions. Due to the underappreciated generality of the intensional phenomena at issue, however, it turns out that all theories face further challenges still. The upshot is that advocates of structured propositions might even be in a stronger position than other theorists when it comes to tackling these challenges purely semantically, but the complications required also suggest that non-semantic explanations of transparency are worth investigating now more than ever. I’ll end with a note on methodology. This dissertation is couched within the theoretical framework of generative grammar and compositional semantics. One of its guiding principles is, accordingly, that truth-conditions for English sentences in context must be compositionally derived on the basis of syntactically respectable logical forms. This dissertation is also strongly influenced by the methods and tools of contemporary analytic metaphysics. So, another one of its guiding principle is that we ought to take seriously the nature of the entities appealed to by our best scientific theories, including the science of language. A recurring theme is therefore the dovetailing of considerations from formal semantics, on the one hand, and philosophical argumentation pertaining to the metaphysical natures of compositional semantic values (in context), on the other. I’ve tried to treat linguistic and philosophical considerations as on a par with one another throughout, conceiving of their distinction ultimately as an arbitrary matter – or, at least, as one that can be safely bracketed for present purposes.



LCSH Subject Headings