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

178 LETTERS IN CANADA 2000 various aspects and tenses); at this stage, no examples are given. The section on the verbal system concludes with an alphabetical list of auxiliaries (mainly second members of >combined verbs=), which may still be improved. The grammar is rounded off with a short introduction to Hindi syntax, demonstrating, with a few examples, how one has to set about analysing and understanding more complex syntactical structures. This will prove especially useful for students who are not used to looking at a language analytically. The book has an attractive outward appearance and is impaired by only a few (printing) mistakes. If it is reprinted, the author might consider providing it with a list of abbreviations (they become clear somewhere in the book, but not necessarily at their first appearance), and perhaps an index and select bibliography. (RENATE SÖHNEN-THIEME) Tadao Miyamoto. The Light Verb Construction in Japanese: The Role of the Verbal Noun John Benjamins 1999. xiv, 234. US $87.00 A light verb is a type of verb whose semantic content is >light= (or has little lexical meaning), as opposed to >heavy= (or lexically more specified), and much of the semantic content is obtained from its arguments. English verbs give and take are good examples of light verbs, as in give a kiss and take a nap. The Japanese verb suru, often translated as do, is such a verb, and the V(erbal)N(oun)-o suru construction (-o is the accusative case marker in Japanese) is highly productive in the language. For example, Japanese has an enormous number of words which originate from Chinese, and they must be accompanied with suru to have a verbal function in a clause. Thus, ryokoo >travel= expresses its verbal sense by combining with suru, as in Johnga Tokyo-ni ryokoo-o suru >John travels to Tokyo.= Tadao Miyamoto=s work presents a new approach to this construction. There has been much debate regarding the treatment of suru, owing to the fact that it imposes semantic restrictions and patterns in some ways as if it were a heavy verb. Miyamoto argues that none of the previous analyses of suru are entirely satisfactory, since they focus on the verb per se, ignoring the status of the verbal nouns (henceforth, VN). His claim is novel in that he puts little emphasis on suru, but more on the role of the VNs. He claims that VNs can be classified into two groups, >simple nominals,= which do not have argument structure and are referential, and >complex nominals,= which have argument structure and are not referential. He argues that a number of Japanese VNs can fall in either category. Miyamoto then proposes that the puzzles for previous researchers are in fact a consequence of this dichotomy. Hence, suru acts as a heavy verb in a construction with a simple HUMANITIES 179 nominal (or >mono-predicational= construction), whereas it acts as a light verb in a construction with a complex noun (or >bi-predicational= construction). Miyamoto=s claim is certainly a preferred one, as two ad hoc lexical representations for light suru and heavy suru are unnecessary, and it is obvious that the VN, rather than suru, decides the lexical meaning of the VNo suru construction. Miyamoto further exhibits the differences between the VNs by examining the issue from different theoretical perspectives. He employs Jackendoff=s Conceptual Semantics to depict the lexical semantic representations of the two types of VNs in an explicit manner. He then gives a syntactic analysis of the two types based on Chomsky=s minimalism, arguing that there are three kinds of accusative cases (structural, partitive, and inherent), and, owing to the difference in semantics of referentiality and specificity, they are associated with different types of VNs B structual to mono-predicational telic, partitive to mono-predicational atelic, and inherent to bi-predicational. Miyamoto=s work has both breadth and depth. In terms of depth, his argument for the dichotomy between the VNs is detailed and largely based on empirical evidence. In terms of breadth, he even discusses the peripheral issue of the >unaccusative hypothesis=; that is unaccusative VNs (i.e. the subject of the event somehow has attributes that are...

pdf

Share