On the syntactic composition of manner and motion (review)

One of the most actively studied topics in the area of the lexicon-syntax interface is the analysis of verb classes that display crosslinguistic regularity of aspectual properties and mapping of the predicate and its arguments. At issue in this area of study is whether semantic generalizations underlie the syntactic structure, or whether structural generalizations give rise to certain facets of meaning, thereby producing aspectual and other semantic generalizations. In On the syntactic composition of manner and motion, Zubizarreta and Oh take up this topic, arguing that structural generalizations can account for crosslinguistic generalizations of mapping and interpretation. Their argument is based on a detailed crosslinguistic study of manner-of-motion verbs like walk, crawl, fly, which are particularly interesting because in some languages they can behave like accomplishments or activities, as in 1, while other verbs, especially transitives, are not variable in the same way.

1. 1.

   1. a. Jan walked to the park (in an hour).

   2. b. Jan walked in the park (for an hour).

2. 2.

   1. a. Lee ate the sandwich in an hour.

   2. b. *Lee ate the sandwich for an hour.

Z&O show that manner-of-motion verbs do not have the same variable behavior in all languages. They argue that differences between Germanic, Korean, and Romance can be explained by independently motivated syntactic properties of functional categories in these languages. This in turn points to the underlying source of the generalization as a syntactic one. In broad terms then, Z&O defend a syntactic account of manner-of-motion predicates, and by extension, of regularities of lexical meaning in general. They take as a starting point the observation that: 'The most noteworthy development in the area of lexicon-syntax interface since the 1980's is the realization that there are "constructional" meanings, which are independent of the particular lexical items that make up the sentence' (1). In developing their account of constructional meanings, Z&O adopt and extend the framework of Hale and Keyser (1993, 2002), which represents regularities of lexical meaning in terms of structural relationships (Spec-head, head complement) between the constituents of decomposed lexical items in 'l(exical)-syntax'. The centerpiece of Z&O's analysis is the hypothesis that certain specific components of meaning are structural, and in a given language, a closed-class item (such as light verbs come, go, become) can in effect 'spell out' a particular constructional environment, so that its presence reflects meaning, rather than producing it. They argue that the range of interpretations available in a particular language follows from (i) generalizations of constructional meaning made available universally by l-syntax, and (ii) limitations imposed by independent syntactic differences occurring from language to language.

The main argument advanced by Z&O in support of their view of the lexicon-syntax interface is that it explains the nature of language-particular differences in the behavior of manner-of-motion verbs. They develop this theme over the course of three chapters. Ch. 1 provides an excellent introduction and overview of the entire work. It first introduces the empirical issues presented by manner-of-motion verbs that are addressed by the study, then sketches two major types of approaches: lexicalist approaches, exemplified in Levin & Rappaport Hovav 1995 and Rappaport Hovav & Levin 1998, and syntactic approaches (Borer 1994, 2005, Ritter & Rosen 1998, Folli 2001). Z&O's discussion concentrates on the extent to which language-particular differences in the variability of manner-of-motion verbs has been accounted for. They conclude that crosslinguistic differences either have not been addressed, or have not been explained in terms of independent properties of functional items in the languages under discussion. It is their goal to develop such an account. They proceed to introduce the framework of Hale and Keyser (1993, 2002), on which their approach builds, and then summarize the main ingredients of their proposal. With Hale and Keyser they assume that fundamental attributes of event interpretation are derived from the...