In lieu of an abstract, here is a brief excerpt of the content:

  • Weighing in on scales: A reply to Goldberg and Jackendoff*
  • Stephen Wechsler

1. Introduction

Goldberg & Jackendoff 2004 (henceforth G&J) includes an analysis of corpus data on the English resultative construction presented in Boas 2000, Appendix A.1G&J fail to mention that Wechsler 2001—a paper G&J cite in other connections—analyzes the same data set (Boas 2000, Appendix A), investigating some of the same questions as G&J. Surprisingly, given that they worked from the same data set, G&J and Wechsler 2001 arrive at strikingly different descriptive generalizations— almost direct opposites. The purpose of this brief note is to compare the two accounts. I argue that the analysis offered by G&J is wrong, and that the approach proposed in Wechsler 2001 and developed in subsequent work (Beavers 2002, 2005, Wechsler 2003, 2005) remains the more promising one.2

2. Scalar adjectives

At issue are constraints on the adjective appearing in the English resultative construction. The following puzzling contrasts were noted by Green (1972, ex. 6b/7b).

(1) He wiped it clean/dry/smooth/*damp/*dirty/*stained/*wet.

Such contrasts have led some researchers to conclude that the allowable verb-adjective pairs are listed as idiosyncratic facts, for example, as special lexicalized compounds (Dowty 1979:303).

Instead, both Wechsler 2001 and G&J attempt more general semantic accounts. G&J (p. 560), following earlier work by Goldberg (1995: 195ff.), offer this semantic characterization of the allowable adjectives: they tend to be nongradable. They explain with reference to the contrast between wet and dry, illustrated in 1 above (G&J, p. 560):

There are some generalizations about which APs are more productive. For example, more productive APs tend to be nongradable and when used as RPs [Resultative Phrases—SW] strongly tend to encode a clearly delimited state. Consider the contrast between dry, which is quite productive as an RP, and wet, which is not. Things are either dry or not dry—it is odd to refer to them as ‘a little dry’; on the other hand, things can be more or less wet.

But G&J’s claim that dry is nongradable is clearly wrong. Both wet and dry are gradable. Throughout the considerable literature on (non)gradable (also called (non)scalar) adjectives, the key test for gradability has been felicity of comparative forms and degree modifiers (Sapir 1944, Klein 1980, Kennedy 1999, inter alia). For example, gradable fine and ominous contrast with nongradable prime (as in prime numbers) and unanimous: [End Page 465] finer, finest, extremely fine; more/most/very ominous; but *primer/*primest/ *more prime/*most prime/*extremely prime numbers; *more/*most/*very unanimous.3 Turning to wet and dry, comparative and degree modifiers are unexceptionable with both of them: wet, wetter, wettest, very wet, dry, drier, driest, very dry. Despite G&J’s comment in the above quote, there is nothing ‘odd’ about characterizing a sponge, for example, as ‘a little dry’. Indeed, my clothes drying machine has a knob that can be turned in either direction, to any degree, with arrows labeled more dry and less dry. There is hardly a clearer case of a gradable adjective than dry.4

If gradability is not the key to the resultative adjective puzzle, then what is? Through an examination of the Boas 2000 corpus data, Wechsler 2001 arrived at a different generalization—roughly the opposite of what G&J claim. Durative verbs such as wipe in 1 actually favor GRADABLE, not nongradable, adjectives, in the resultative construction.5 What is the crucial property distinguishing, for example, wet from dry?

Wechsler 2001 approached this question by drawing on relatively recent work on the semantics of gradable adjectives (Kennedy 1999, inter alia).6 For more recent treatments see Rotstein & Winter 2004 and Kennedy & McNally 2005.7Gradable adjectives have long been modeled with scales, that is, ordered sets with a measure function. The adjective is interpreted relative to a standard degree on that scale: Michael Jordan is tall means that Jordan’s height is greater than some contextually determined standard (Kennedy 1999, inter alia).

Gradable adjectives can be subclassified according to the structure of the scale. Three types of gradables, plus the nongradables, concern...


Additional Information

Print ISSN
pp. 465-493
Launched on MUSE
Open Access
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.