354
BOOK REVIEWS
sources. For such purposes we could use dozens of studies like Sagart's in all the Chinese dialects.
Of course there is much more that one can do with materials such as these. The synchronic description of this particular subdialect at this particular time is useful in many ways. For example, Sagart's lexicon leads us into the interesting area of borrowed words in Hakka, loans from both Cantonese and English. We might hope for a future study of the phonology and semantics of loans in this subdialect along the lines of Samuel Cheung's chapter on loan words in Cantonese (Zhang Hóngnián 香港粵語語法的研究, Hong Kong 1972).
The few references in Sagart's study to syntactic details are intriguing and suggest the possibility of a fruitful expansion in that area. Although syntax and phrase construction are treated only cursorily in a section entitled Grammaire in the lexicon, we see some interesting details of usage that call for elaboration, hopefully at an early date. Page 20, entry 475 has a locative coverb phrase after the main verb in a construction that would require special explanation in other dialects. (cf. Cantonese phak gà chè hài nī douh 泊喺呢度 and also hài nī douh pāak chè, ‘park here' with a difference of nuance that needs fuller explanation). Also, I am fascinated by a dialect that uses throughout (Mandarin zhī) as the classifier for humans, monsters, deer, and other creatures. In some parts of China the use of this classifier is an insult when applied to people, but in this subdialect it seems to be the standard form for human beings. Divergent usages of this kind could constitute the base for an interesting study in its own right.
We also find Sagart's teu: kjius ‘les chiens', suggesting a plural form alternating with ais kjius 'le chien'; one wonders if teu, is equivalent to the Cantonese form dī in post-verbal position.
It is just in these areas of syntax and semantic shifts that one would like to see an expansion of Sagart's work. For too long we have taken it for granted that syntactic features are so similar among Chinese dialects that they are seldom worth separate study. In detailed studies of the kind Sagart has done we begin to see