The Zhuang character script which David Holm discusses in this monumental book is a system for writing the Zhuang language of Guangxi in southern China using Chinese characters and invented Chinese-like characters. It is not a standardized writing system, and has never been used as a general purpose script for writing the Zhuang language by the general Zhuang-speaking population. Rather, it has mostly been used by a small class of ritual practitioners and Taoist priests to record ritual texts for recitation. This restricted pattern of usage has meant that manuscripts written by different practitioners in different locations at different times tend to use different repertoires of characters to represent the same Zhuang words. In this book Holm systematically analyses the character usage in a range of texts from known locations across the Zhuang-speaking areas of Guangxi and neighbouring parts of Eastern Yunnan, as well as some texts written using the same writing system to represent the related Bouyei language of Guizhou, and the Thổ, Nùng and Tày languages of northern Vietnam. In an attempt to tie down Zhuang character usage geographically, Holm defines fifty-one survey points covering much of Guangxi, as well as neighbouring parts of Yunnan, Guizhou, and Vietnam. These survey points represent forty-five Zhuang, Bouyei or Tày source texts (some original manuscripts, some based on modern editions) and six eighteenth-century or modern vocabularies.

The main part of the book, some 660 pages in length, provides a detailed analysis of the characters used to represent a sample of sixty high-frequency Zhuang words. For each of these words Holm tabulates all of the characters used in the various sources, indicating in which locations each character is found. Most words are represented by about twenty different characters, and Holm groups the characters used for each word into a number of semantic or phonetic series for discussion. He attempts to explain the usage and distribution of each series of characters on the basis of historical phonetics and the modern pronunciation in different Chinese dialects (Pinghua, Cantonese and Southwest Mandarin). My only complaint is that the total number of occurrences of each character in each source is omitted, and without this information it is difficult to evaluate the significance of some characters that only occur in a single manuscript. For example, Word 3 (bae “to go”) is represented by the character 丕 (Chinese pī) in twelve sources and by the character 不 (Chinese bù) in just one source. If the anomalous form occurred multiple times in the same source then it could be considered to be a deliberate scribal choice, but if it only occurs a single time then it would be best explained as a transcription error for 丕, but the reader cannot tell without the figures. In this particular case Holm treats 不 as a “graphic approximation” of 丕 (p. 114), a term which he uses frequently to explain characters which look similar to the expected character form but are inexplicable in their actual form. The simple explanation, which Holm seems unwilling to state explicitly, is that such characters may be the result of misreading and transcription errors during the process of copying. As the ritual manuscripts that form the main body of Zhuang character script literature belong to individual ritual practitioners, and are burnt on the death of their owner, texts are copied and recopied from one generation to the next, and so there is much scope for the introduction of scribal errors.

Each of the sixty words is accompanied by two maps, one showing the distribution of the most frequent characters for the word, and one that shows the pronunciation of the corresponding Zhuang word in the local dialect. For a few words the maps do show some correlation between local pronunciation and manuscript character forms, but in most cases there is no obvious correlation. However, from Holm's analysis of the sixty key words it seems that the choice of how to represent a particular Zhuang word in Chinese or Chinese-like characters depends more on the local Chinese dialect than on the local Zhuang dialect, so the second map is less useful than the first.

One obvious problem with attempting to map the geographic distribution of Zhuang character forms is that although a particular manuscript may have been collected from a certain location, there is no guarantee that its text was written at the same place by a local resident. Indeed, Holm describes patterns of migration by Zhuang ritual practitioners, and he points out that at least one of the source texts was not written at its nominal location. Nevertheless, Holm argues persuasively that on the whole the mapping of source texts does help explain how and why different characters are used to represent the same word in different manuscripts.

This book will be an essential tool for anyone studying the Zhuang character script, and will be particularly useful for researchers studying Zhuang manuscripts, as the word list (in conjunction with the valuable Zhuang character index) should help them to identify the Zhuang character script tradition behind the manuscript and even the general location where the manuscript was produced. Important as this book is, it is only the first stage in an ambitious project by Holm and his colleagues to document systematically the Zhuang character script. He plans to follow this up by publishing the Zhuang source texts used in this book as a Traditional Zhuang Texts series, with photo-reprints of the original manuscripts, transcriptions, glosses, translations and textual notes, as well as sound recordings of the texts. At a later stage Holm plans to produce a set of Zhuang character dictionaries for key localities, based on a selection of texts from each area. This approach should revolutionize the study of the Zhuang character script, and I eagerly await the publication of future stages of this project.