Belief, Rational and Justified

It is clear that beliefs can be assessed both as to their justification and as to their rationality. What is not as clear, however, is how the rationality and justification of belief relate to one another. Stewart Cohen has stumped for the popular proposal that rationality and justification come to the same thing, that rational beliefs just are justified beliefs, supporting his view by arguing that ‘justified belief’ and ‘rational belief’ are synonymous. In this paper, I will give reason to think that Cohen’s argument is spurious. I will show that ‘rational’ and ‘justified’ occupy two distinct semantic categories—‘rational’ is an absolute gradable adjective and ‘justified’ is a relative gradable adjective—telling against the thought that ‘rational belief’ and ‘justified belief’ are synonymous. I will then argue that the burden of proof is on those who would equate rationality and justification, making the case that those who hold this prominent position face the difficulty of explaining how rationality and justification come to the same thing even though ‘rational’ and ‘justified’ are not synonymous.

Informational goals, sentence structure, and comparison class inference

Understanding a gradable adjective (e.g., big) requires making reference to a comparison class, a set of objects or entities against which the referent is implicitly compared (e.g., big for a Great Dane), but how do listeners decide upon a comparison class? Simple models of semantic composition stipulate that the adjective combines with a noun, which necessarily be- comes the comparison class (e.g., “That Great Dane is big” means big for a Great Dane). We investigate an alternative hypothesis built on the idea that the utility of a noun in an adjectival utterance can be either for reference (getting the listener to attend to the right object) or predication (describing a property of the referent). Therefore, we hypothesize that when the presence of a noun N can be explained away by its utility in reference (e.g., being in the subject position: “That N is big”), it is less likely to set the comparison class. Across three pre-registered experiments, we find evidence that listeners use the noun as a cue to infer comparison classes consistent with a trade-off between reference and predication. This work highlights the complexity of the relation between the form of an utterance and its meaning.

Manner implicatures and how to spot them

The goal of this paper is to help develop a general picture of conversational implicature (Grice, 1975) by looking beyond scalar implicature to see how the phenomenon behaves in a general sense. I focus on non-scalar Quantity implicatures and Manner implicatures. I review canonical examples of Manner implicature, as well as a more recent, productive one involving gradable adjective antonym pairs (Rett, 2015). Based on these data, I argue that Manner implicatures—and conversational implicatures generally—are distinguishable primarily by their calculability; their reinforceability; their discourse sensitivity (to the Question Under Discussion; Roberts, 1990; van Kuppevelt, 1995; Simons et al., 2011); and their embeddability (under negation, propositional attitude verbs, quantifiers, etc.). I use these data to draw conclusions about the usefulness of implicature-specific operators and about ways to compositionally represent conversational implicatures.

Adjectival Nominalization

Chapter 6 focuses on the semantics and metaphysical implications of the semantics of the adjectival nominalization “goodness.” Adjectival nominalizations of the form “F-ness” are almost always mass nouns. The mass noun “goodness” derives gradability of a kind from the gradable adjective that it incorporates. So “goodness” is a gradable adjectival nominalization. Mass nouns are distinguished from count nouns on the basis of two semantic properties, called “semantic cumulativity” and “semantic divisibility.” The denotations of mass nouns are then interpreted in terms of the mereological structure of a join semi-lattice. The denotation of gradable mass nouns incorporate scalar as well as mereological structure. In the case of “goodness,” the elements at the base of the lattice structure are instances of goodness. An instance of goodness is a so-called qua quantitative trope, precisely one degree of purpose serving qua exceeding a second degree of purpose serving, where the latter is a standard of comparison.

Gradability

The sense of “good” that has been of principal interest to philosophers and that is the focus of chapters 3, 4, and 5 is evaluative “good.” Hereafter, the modifier “evaluative” is dropped. “Good” is a gradable adjective. Accordingly, chapter 3 examines the semantics of gradable adjectives. The chapter argues that “good” is the unmarked member of an antonym pair of relative gradable adjectives, the marked member being (evaluative) “bad.” The lexical meaning of “good” is associated with a non-significant degree on an open scale of unspecified value. In tokenings of sentences of the form “x is good,” the degree associated with “good” is modulated to a significant degree. Significance of degree is a quantity that exceeds the upper bound of a range of numeric values based on a contextually determined comparison class.

Deriving the distributivity potential of adjectives via measurement theory

The boxes are heavy can convey that each box is heavy (distributive), or that some individually light boxes qualify as heavy when lifted together (nondistributive; Schwarzschild 1996, Schwarzschild 2011). In contrast, the boxes are fragile generally requires each box to be fragile (distributive). Which adjectives behave like heavy or like fragile, and why? This paper proposes a measurement-theoretic account. For a gradable adjective to be understood nondistributively, I argue that a⊕b must exceed a and b along the scale associated with the adjective. That way, the contextual standard θ for what ‘counts as’ (adjective) in the context can be set in such a way that the composite object a⊕b surpasses the contextual standard θ while a and b individually fall short of it – a nondistributive understanding, in that the adjective is true of a⊕b together but not of a or b individually. This ordering is possible for heavy but not fragile, deriving their differences. More generally, researchers agree that an adjective’s potential for distributivity depends on what we know about the property it describes. Making that idea more explanatory, this paper articulates which features of the property described by the adjective matter for distributivity and why.

Cognitive processing of scalar implicatures with Chinese gradable adjectives

In previous research comparing the Context-driven Model with the Default Model of meaning processing, the former was preferred. It predicts that contexts play an exclusively decisive role in meaning processing, whereas the latter holds that the inference of literal meaning generally goes through, unless it is subsequently defaulted or cancelled by the context it is associated with. The Standardization Model, which we added to our experiments, highlights that implicatures are figured out from standardized forms typically based on the mutual background belief and speaker’s intention. We tested whether Chinese people’s processing of the gradable adjective scale <hot, burning> conformed more to the Context-driven Model, the Default Model, or the Standardization Model. The results demonstrated that the Standardization Model is the most acceptable among the three. The findings of this study, which is the first study using the experimental paradigm on Chinese gradable adjectives, highlighted a need for further studies to investigate the same questions with different languages and cultures.

Certainty and possibility

This chapter discusses several more epistemic adjectives. Certain and its near-synonym sure are maximum adjectives that combine with proportional and percentage modifiers. A comparison with non-modal adjectives suggests a ratio-scale classification with at least an upper bound. Several lines of evidence indicate that certainty and likelihood are formally closely related. However, there are puzzles around the interpretation of uncertain that indicate that the relation may not be one of identity. I consider three possible analyses, all of which have certain advantages and drawbacks. I then turn to possible, which is often claimed to be non-gradable. Naturalistic data indicate that possibility is a graded concept (e.g., increase the possibility of), and that possible is gradable (e.g., too/very/n% possible). While an analysis in terms of some kind of scalar coercion is technically feasible, the most natural explanation is that possible is a gradable adjective whose scale is closely related to likely’s scale.