| Co-Investigator(Kenkyū-buntansha) |
TOMIZAWA Naoto Yamagata University, Faculty of Humanities and Social Sciences, Associate Professor, 人文学部, 助教授 (40227616)
MARUTA Tadao Yamagata University, Faculty of Humanities and Social Sciences, Professor, 人文学部, 教授 (10115074)
|
|---|
| Research Abstract |
This research focuses on the peculiar syntactic and semantic behaviors of "adjuncts," which have been traditionally characterized in default as "not being arguments." We advance an alternative theoretical view from two interrelated perspectives, namely, lexical semantics and minimalist syntax, in which adjuncts are given more explicit and proper treatment.
In a lexical semantic approach to the English tough-construction, Maruta presents a type shift analysis based on morphological operations, explaining the relevant alternation in the construction. The analysis is extended to the two types of Japanese tough-constructions, which are argued to have two-layered thematic domains mutually related via a type shift operation called successive predicate formation, and it is proposed that in both constructions the adjectives serve as experiencer predicates being predicated of the domain-external arguments. The necessity of the type shift operation is argued to follow from the "do-ability" of the
… More
things described.
Identifying the fundamental function of the resultative construction as a description of idiosyncratic events with a significant transition on a unique scale, Suzuki proposes a unified analysis of selectional restrictions on the result phrases in both adjectival and prepositional resultatives, employing the notion of complementary opposition on a scale. It is also argued that the boundedness constraint on the construction can be derived from interacting requirements of event composition and complementary opposition.
Based on examination of the syntactic behavior of wh-phrases and anaphors, Tomizawa presents evidence to adopt a "derivational model," in which various features are processed derivationally, instead of a "representational model," and proposes a single fundamental principle of feature fixation in terms of the computational efficiency. The principle not only gives a unified account to the effects of the Subject Condition and the Adjunct Condition, but also explains why extractions from A'-raised wh-phrases are possible while extractions out of extraposed phrases or shifted objects are not. Less
|
