BOOK REVIEWS
157
forms should be included or the student is going to be left in the dark on numerous items which are often heard in everyday speech. K. P. K. Whitaker ("A Study of the Modified Tones in Spoken Cantonese", Asia Major, New Series, Vol. V, Parts 1 and 2) has treated this subject intensively and a glance at her long lists of words normally appearing in changed tone will convince anyone that a student of Cantonese will certainly need some way to handle unknown items showing this phenomenon.
Admittedly, as Rev. Cowles points out in defending his decision to ignore the changed tones, they vary considerably from area to area; it would indeed be impractical to attempt to record all the local variants. The point here should be that there is no practical way to design a dictionary to cover all the great multitude of regional varieties of the Cantonese dialects. A choice will have to be made concerning just which dialect form will be treated and the most likely selection would seem to be Standard Cantonese. I believe that this choice should have been made and that this dictionary should have included as many as possible of the common changed tone forms used by the speakers in Hong Kong and Canton. Furthermore, these forms should not be listed under the basic tone of the character but in such a way that the student can look them up in the dictionary on the basis of what he hears. Thus, since the high rising changed tone is often confused with basic tone of similar contour, it might be best to list these under the high rising basic tone and indicate in the symbolization that historically such forms are members of other basic tone categories.
Rev. Cowles has indeed made a very important contribution and I do not mean to detract from this by quibbling over minor points. Nevertheless, in striving for totality in a single dictionary the compiler necessarily takes on an impossible task. Obviously decisions to include and exclude face him at every turn, and no two compilers could be expected to make the same decisions. A lexicographer should define his area and depth of concentration then be as thorough as possible within these limitations. One should not in one paragraph (p. vii) defend the size of a dictionary on the grounds that the forms included ‘are in the language, and being there, call for a record and interpretation into English' then three paragraphs later argue against inclusion of the changed tone forms because they "are simply multitudinous, and usage differs widely in many localities". It would seem wise to skip local