-
Preface
- The MIT Press
- Chapter
- Additional Information
Preface When one initially examines some area of the syntax of a natural language , it often seems to manifest what might be called descriptive chaos. For instance, expressions that seem to have parallel meanings and standard syntactic structures nonetheless manifest contrastive syntactic behavior in a variety of ways. An example: (1) a. It impressed me/all the committee members that Quentin was deranged. b. It struck me/all the committee members that Quentin was deranged. c. Quentin impressed me/all the committee members as being deranged. d. Quentin struck me/all the committee members as being deranged. The two main verbs appear to have parallel psychological meanings and to occur in distinct parallel syntactic frames, (1a,b) and (1c,d), with each frame requiring some type of object. One is thus naturally puzzled to observe that those objects behave contrastively—for example, with respect to complex DP shift or right node raising.1 (2) a. Quentin impressed as (being) deranged — all the committee members.2 b. *Quentin struck as (being) deranged — all the committee members. c. Quentin may have impressed as (being) deranged and he certainly impressed as (being) strange — all the committee members. d. *Quentin may have struck as (being) deranged and he certainly struck as (being) strange — all the committee members. There might be some impulse to dismiss these contrasts, to set them aside as probably unreal or as merely some lexical fact of little or no significance. In general, I believe any such impulses contribute negatively to the project of uncovering the grammatical truth. Rather than ceding to them, one can and I will systematically consider a distinct possibility : the facts are as they seem to be because despite the real superficial similarities in the structural features of cases like (1a–d), they also manifest less than glaring but perfectly genuine di¤erences in grammatical structure. The thought that seeming grammatical chaos like that in (2) is in fact a symptom of some ungrasped structural regularity is supported because there are other contrasts between the two verbs, equally unexpected a priori.3 (3) a. It was me that Quentin immediately impressed as (being) deranged. b. *It was me that Quentin immediately struck as (being) deranged. c. the one who Quentin impressed several friends of as (being) deranged4 d. *the one who Quentin struck several friends of as (being) deranged (4) a. Jerome, impressing whom as clever is not easy, . . . b. *Jerome, striking whom as clever is not easy, . . . Here, too, the verb impress has greater constructional freedom than strike, for no recognized reasons. In what follows, I argue inter alia that various ideas that are well supported by facts about English objects, many distinct from those in (1)–(4), provide simple and justified solutions to the question of how to characterize the manifold di¤erences between the uses of impress and strike in (1)–(4), di¤erences that I have only partially documented here. These ideas will involve rejecting the notion that an English phrase of the form [V þ DP] invariably involves a grammatical relation properly characterized as direct object. Rather, I will claim that at least three distinct relations occur in such a structure. Hence, what appear from one point of view to be anomalies in the behavior of direct objects will emerge as nonanomalous , as regularities stated on the distinct relational types. And I will argue that these relational types as well as others play a fundamental role in explicating the notorious problems of characterizing English double object (ditransitive) constructions, as well as in illuminating a host of others from possible passivization targets to Visser’s Generalization. xvi Preface ...