- The semantics of evaluativity by Jessica Rett
Jessica Rett's book investigates the phenomenon called 'evaluativity' (also referred to as 'norm-relatedness' by Bierwisch 1989) in degree and other constructions. A construction is evaluative if it requires that a degree exceed some contextual standard. The author argues in favor of [End Page 723] an approach that treats evaluativity as an implicature. The self-defined goal of the book is to apply the Gricean theory of conversational implicature (CI) to R's earlier work (Rett 2007, 2008) in terms of markedness and semantic competition. Knowledge of R's earlier work is advantageous but not obligatory when approaching the present book. Besides offering a new analysis of evaluativity as implicature, this book is an excellent overview of evaluativity in degree constructions, with extensions to other domains. It presupposes knowledge in generative grammar, more specifically familiarity with degree semantics (see e.g. Cresswell 1976, von Stechow 1984, Beck 2011 for an overview) and CI (Grice 1975). The theoretical orientation of this work is formal compositional semantics in the spirit of Heim & Kratzer 1998.
The book is structured as follows: Chs. 2–4 provide the background by introducing the null morpheme POS in Ch. 2, the null morpheme EVAL (cf. Rett 2007, 2008) in Ch. 3, and a brief review of a theory of implicature in Ch. 4 along with a discussion of markedness. Ch. 5 is central in that it introduces the new account of evaluativity as an implicature and applies it to some core phenomena. Possible extensions to other phenomena are presented in Ch. 6, and Ch. 7 contains overall conclusions.
In the introductory chapter, the author addresses two problems of previous approaches to evaluativity in positive constructions such as 1a: the extra argument and the evaluativity problem. She points out that it is wrong to conflate these two issues. The covert operator originally dubbed POS, used to explain positive constructions in a framework working with degrees, is very powerful in that it solves both the extra argument problem (by existential closure) and the question of evaluativity. R discusses examples like 1a–c, in which POS cannot account for the presence of evaluativity, and 2a–b, in which it cannot account for its absence.
(1) a. Adam is tall.
b. How short is Adam?
c. Adam is as short as Doug.
(2) a. Adam is shorter than Doug.
b. Adam is as tall as Doug.
R also discusses the wide distribution of the phenomenon to other constructions besides comparatives rejecting POS.
In Ch. 2, R introduces the basics of degree semantics, according to which degrees are semantic primitives. She addresses the semantics of gradable adjectives and elaborates on the nature of scales. She provides concise overviews on numerals, measure phrases (MPs), and degree quantifiers. The last part of Ch. 2 is dedicated to a more profound discussion of evaluativity in positive constructions and to POS. R discusses the different formulations of POS by authors over the course of time and finally possible extensions of POS to the verbal domain where evaluativity occurs optionally. This is the strongest argument in favor of divorcing the evaluativity component of the original POS from its type-shifting component (separation of evaluativity and existential quantification).
In Ch. 3, R explains why the covert EVAL-operator (Rett 2007, 2008), a modifier version of POS, is not sufficient and why it needs to be supplemented with the account of its distribution outside of positive constructions and comparatives. She critically assesses her own EVAL theory, outlines its strengths and weaknesses, and anticipates the possible objection that we do not see an overt manifestation of this operator in the grammar crosslinguistically. Among the strengths of her old account, she highlights that it correctly predicts that evaluativity can exist outside of positive constructions, since it treats evaluativity as a property independent of whether the adjective's degree argument is overtly manipulated. As to the weaknesses, the internal argument asymmetry in equatives needs a stipulation, MPs are also a problem, and the explanation of evaluativity in the...