Word-Order Change and Grammaticalization in the History of Chinese (review)

Reviews 561 Chaofen Sun. Word-Order Change and Grammaticalization in the History ofChinese. Stanford: Stanford University Press, 1996. xiii, 207 pp. Hardcover $39.50, isbn 0-8047-2418-0. Grammaticalization is concerned widi how language change occurs; its focus is the evolution ofindependent lexical items into more constrained morphosyntactic elements. More broadly, it concerns the constant movement oflanguage toward structure, both synchronically and diachronically (Heine et al. 1991; Lehmann 1982). Grammaticalization theory is crucially linked to pragmatics: to structural and semantic development within specific discourse contexts. In this way, theories of grammaticalization are diametrically opposed to Saussurean and generativist models, in which language is an idealized, all-inclusive rule system with fixed boundaries. In die generativist model, die communicative discourse of speakers is the end product of the operation ofgrammatical rules; in the grammaticalization model, however, it is discourse itselfthat motivates the emergence of grammatical structure (Hopper 1988). The current work on Chinese focuses on the grammaticalization of free lexical verbs into morphosyntactic elements, showing how in modern Mandarin the free lexical verbs often coexist with isographemic grammatical markers. Such gradient change and co-occurring forms are problematic for the structuralist/generativist model, since it strictly separates synchrony from diachrony, and hence change can only be dealt witii in discrete stages. Grammaticalist approaches to language, however, consider change to be an ongoing process. Co-existing grammaticalized and lexical forms exemplify layeringand divergence, the co-occurrence offorms that are either functionally distinct but derived from the same etymon or functionally identical and derived from different etyma (Hopper 1991). This book is actually a history ofmodern Mandarin, and not one ofall "Chinese " as die tide implies. Data from dialects odier dian Mandarin have not been considered, and "Chinese" is treated as a monolithic entity with a linear development from Old Chinese to modern Mandarin. It also examines the development ofcertain prepositional phrases (marked witii grammaticalized verbs) in an effort to challenge claims ofconstituent-order change ("SOV > SVO") in Chinese. In this book it is claimed that the Chinese were the first to discover grammaticalization: "since the time ofthe Song Dynasty (960-1279 a.d.) Chinese scholars have been using the term xuhua EHb meaning 'grammaticalization'-----© 1998 by University The notion of'grammaticalization' was not widely used in die West until . . . ofHawai'i PressMeillet" (p. 11). Xuhua reflects a philological distinction (much earlier, in fact, than the Song dynasty) between xuzi JU^ 'empty (- function) words' and shizi If^ 'full words' (i.e., a closed class oflexical items). Xuhua implies both that 562 China Review International: Vol. 5, No. 2, Fall 1998 grammaticalization involves semantic and syntactic changes and that all empty words were formerly full words. Meillet, however, introduced two other facets: (1) that the process of grammaticalization constituted both a continuum and two discrete phases and (2) that there is a correlation between a decrease in expressive value of elements and an increase in their frequency (Heine et al. 1991, p. 9). Since the focus is on discourse and the communicative function oflanguage, grammaticalization studies tend to analyze quantitative data from colloquial texts and spoken-language corpora. In this study, the author attempted to include as far as possible written texts that approximate the spoken language of a given period : literary and semi-colloquial Tang texts, Yuan dramas, Ming and Qing colloquial novels, bilingual language primers (e.g., the fifteentii-century Laoqida), and excerpts from Early (1000-1900) and Modern Mandarin novels. Before these periods , so-called "colloquial texts" are harder if not impossible to find. For Old Chinese (500 B.c.-A.D. 200), Sun examined excerpts from the Zuo zhuan and Shiji (100 b.c.), and for Early Middle Chinese, excerpts from the Shishuo xinyu (500 a.D.). The author surveyed selections (often unspecified) from the texts above without the aid of a computer. (Sun, for example, searched the first one thousand characters in seven works for adpositions, and the first forty pages of one version of the Zutangji for de \%—p. 126.) This methodology is understandable given that there are very few digitized colloquial Chinese texts at present. Still, one would like to know how much of each text, and what exactly, was surveyed. Short selections oflarge works may...