The Japanese-Korean connection : A contrastive study of the inventory and properties of their IHRCs
A. Grosu,Daeyoung Sohn
Abstract:The earlier literature on the Internally Headed Relative Constructions (henceforth: IHRCs) of Japanese and Korean has generally assumed that these constructions exhibit the same characteristic properties in the two languages, and in particular, those that distinguish them from the IHRCs of other languages. However, the earlier literature has sometimes attributed incorrect properties to the IHRCs of each of these two languages, so that, depending on the studies that one reads, one may get the impression that significant differences exist between the IHRCs of the two languages. To take just one example, Shimoyama (2001, Chapter 3) proposes that the IHRCs of Japanese disallow proper names as internal heads (henceforth: IHs), and Kim (2007) provides numerous data in which putative IHRCs have precisely such IHs, thereby giving rise to the possible impression that the two languages differ in this respect. Grosu & Hoshi (this volume; henceforth: GH) undertake a detailed examination of various earlier proposals concerning the properties of Japanese IHRCs, and argue for specific conclusions, which we propose to view as correct. The central goal of this article is to check the extent to which the properties of Japanese IHRCs are also found in Korean IHRCs, and more generally, whether the two languages have the same inventory of subtypes of IHRCs. In so doing, we will take GH as point of departure and standard for comparison, and we will also assume, as background, the typology of IHRCs outlined in Grosu (2012), the IHRCs of Japanese being one of three types described in that paper. This article thus constitutes a follow-up to GH, and may in fact be viewed as a logical appendix to the latter. It will thus be helpful for the reader to take a look at GH before reading this paper. By and large, GH argue that a number of earlier studies have incorrectly characterized some of the properties of Japanese IHRCs, in most, although not all, cases due to a failure to appreciate the full implications of the fact that some IHRCs are string-wise homophonous with adverbial clauses. We believe that a comparable, even if not entirely identical, state of affairs exists with respect to the earlier literature on Korean IHRCs, and we will attempt to rectify it in what follows. The conclusion that we will eventually reach is that, as far as we can tell at the moment, there are no significant differences in the inventory and properties of IHRCs in the two languages. The judgments of the Japanese and Korean examples in this paper are primarily those of the authors whose names are listed second and third respectively, and they have so far been counter-checked with only a small number of additional speakers. In view of a certain amount of idiolectal variation that exists in both languages with respect to IHRCs, additional checking with greater numbers of consultants seems highly desirable.
Linguistics