Hilary Chappell, ed. Sinitic Grammar: Synchronic and Diachronic Perspectives. Oxford University Press, 2001. xxvii, 397 pages.

I must confess that I was powerfully attracted to this book by its title. Having long been an advocate of "Sinitic" as the proper designation for the group of languages known in Mandarin as Hanyu, I was curious to see whether the editor of this book had made a principled choice when she decided to use this word rather than the customary "Chinese." After I received the book and had a chance to peek inside it, I could see that the author had very good reasons for choosing "Sinitic" to describe her frame of reference.

In her "Introduction," the editor lays out the rationale for the whole book, which is to take a synchronic and diachronic look at Sinitic languages, with the emphasis definitely on the plural. Her aim is not to examine only Mandarin grammar, or some abstrusely homogenized "Chinese." Rather, she wants to analyze a whole series of different languages. After one reads the chapters on Xiang, Shang, Cantonese, and Taiwanese, anyone who has an open mind will have to recognize that these are separate languages, not merely dialects of some disembodied, yet all-inclusive "Chinese." Furthermore, her focus is not merely on modem forms of these languages, but on earlier stages of their development as well, which leads to questions of relatedness within and beyond the languages of this (the Sinitic) group.

Hilary Chappell clearly made an intelligent decision when she chose to characterize her book as dealing with "Sinitic." On July 7, 2003, I wrote an e-mail to Dr. Chappell telling her how pleased I was with the title of her book. I was dismayed, however, when she wrote back from Paris the same day, saying that Oxford University Press was asking her to change the title of the book to the nondescript, imprecise Chinese Languages ...! Clearly, the forces of thoughtless conformity were going to work on Hilary Chappell.

I find it surpassingly strange that even the most eminent historical linguists can speak of "Sino- Tibetan" in one breath and of "Chinese" in the next breath, without being able to recognize the gross contradiction of the latter name. Surely, if Sino-Tibetan is a family, then the two main subdivisions of the family should be Sinitic and Tibetan (assuming that one admits the existence of a Sino-Tibetan family [and nobody calls it "Chinese-Tibetan"!], which most historical linguists studying the languages of China do). Even if one does not accept the existence of a Sino-Tibetan language family, there are dire problems with "Chinese" as an umbrella term for all the Han languages of China.

What, after all, do we mean when we say "Chinese" in the linguistic sense? Do we mean just Modem Standard Mandarin (MSM; Putonghua) as the national language of China? Fair enough, if that's what one wants to call Mandarin. By extension, then, we could also speak of "Middle Chinese" and "Old Chinese," but, logically and strictly speaking, we should do so only when we are referring to the actual forerunners of Mandarin per se (cf. "Middle English," "Old French," etc.). If we accept "Chinese" as the designation of Mandarin, a distinctly vernacular language, we should technically not also apply it to the dead book language wenyan(wen) commonly called "Classical Chinese," but which I prefer, for the sake of clarity and accuracy, to call "Literary Sinitic." After all, we make an unambiguous distinction between Italian, Middle Italian, and Old Italian on the one hand, and Latin (with its Classical, Late, and Vulgar varieties) on the other. Ditto for Hindi, Middle Hindi, and Old Hindi on the one hand, and Sanskrit (with its Vedic, Classical, and Buddhist Hybrid varieties) on the other -- not to mention Pali, Prakrit, Ardha- Magadhi and other vernaculars that came before Hindi. We do not speak of some amorphous "Italian" or "Hindi" that covers all stages of development and all styles of vernacular and literary related to the current, modern, nationally approved exemplars designated by the quoted terms. Even supposing that one does accept "Chinese" as the proper designation for MSM and its vernacular forerunners, it is inappropriate to refer to prior, non-Mandarin stages with the same term. This is especially the case because, both historically and contemporaneously, Hanyu is such a large, elaborate, complicated group of distinct languages: It is not a small, closely coherent, tightly homogeneous group with minimal variety .All the more, it has NOT been a single, invariant, monolithic language for all time.

So much for "Chinese" in the context of Mandarin versus Literary Sinitic (wenyan[wen]). It is even more questionable to apply this single designation to all the varieties of languages that are now being spoken or have ever been spoken by the Han people. And, by any impartial linguistic standards, these are separate languages, not merely dialects. In the first place, as I have pointed out over and over, "dialect" is a mistranslation of the Mandarin term fangyan. "Topolect" is a far more accurate, neutral rendering for this linguistically vague term. Conversely, the English word "dialect" should be translated into Mandarin as tong1yan2 通言 (a form of "speech" [-lect] that goes "across" or "through" [dia-]). Compare "dialog(ue)" (a conversation between two or more people), which is formed from the same Greek roots (dia- + legein ["to speak"]) as "dialect" and is suitably translated into Mandarin as dui4hua4. Thus, the word "dialect" etymologically implies mutual intelligibility. And, by any unbiased linguistic standards or tests, dialects of a given language are patently and unmistakably mutually intelligible. To make exceptions for the many varieties of mutually unintelligible "Chinese" alone by insisting that they are "dialects" of a single language is to destroy the viability of any linguistically rigorous definition of the word dialect.

What about the argument that there is only one "Chinese" language because all forms of it are written with characters? That would be like saying that all the languages of the world that are written with Roman letters are a single language (!), or --looked at from another angle closer to home --that Old Japanese (written with Chinese characters) is the same linguistic entity as "Chinese." As for all the spoken varieties of "Chinese" having the same written form, I have long ago and repeatedly demolished this totally specious argument, so I will not waste time on it again here, except to summarize ever so briefly. In China, what gets written down in Chinese characters is almost always Mandarin (or, in the past, Literary Sinitic). The other Sinitic languages have customarily not been written down. When they are written down in their full, unadulterated (by Mandarin) forms, they are very different from Mandarin and unintelligible to monolingual speakers and readers of Mandarin.

The last resort for those who support the notion that there is only a single "Chinese" language is also the feeblest: "the Chinese people feel that they have only a single language." But if linguistics is ever to become a full-fledged science (a goal toward which it continually strives), we cannot rely on feeling to decide such fundamental questions as whether "Chinese" is a single language or a whole group of languages. We must try to be precise and rigorous, and we must have rational grounds for our analyses. It would be conceivable for a Nordic supremacist to say "I feel that there is only one great Germanic language," but we certainly would not be obliged to accept his emotional claim as fact. When all the linguistic evidence (phonological, lexical, grammatical, syntactical, morphological, historical) points to the existence of separate Sinitic languages in China (each with a considerable time-depth of its own), it is being willfully obscurantist and anti scientific to continue to insist that there is now, always has been, and always shall be only a single "Chinese" language. And what about all the non-Sinitic languages of China? Referring to the Sinitic languages of China as "Chinese" leaves us without a way to speak unambiguously of "the Chinese languages" or "the languages of China" (the title of S. Robert Ramsey's widely read book, which badly confused me when I first encountered it) inclusively (i.e., comprising both the Sinitic and the non-Sinitic languages of the political entity known as "China").

In linguistic classification, "Sinitic" is quite comparable to "Indic." Both consist of a number of separate languages (Mandarin, Wu, Southern Min, Yue...; Hindi, Bengali, Marathi, Gujarati... ), each with its own historical development. And both Sinitic and Indic are part of a larger language family (Sino-Tibetan [if one accepts the existence of that family] and Indo-European).

I have been using the word "Sinitic" for decades as an umbrella term to refer to all the languages spoken by the people who style themselves "Han," both in their modern forms and in their earlier stages. Nearly all of my linguist friends profess not to comprehend any problem with "Chinese" or why I should devote so much energy and time trying to show that "Sinitic" is a far more suitable designation. (When I start to talk about this subject with them, their eyes glaze over and their minds go shut.) Therefore, it was with an enormous sense of relief and gratification when I discovered that Søren Egerod, the great Danish linguist, used the designation "Sinitic" for exactly the same reasons I outlined above. Now, everybody respects Egerod, but --without any explanation whatsoever --they completely ignore his principled views about the problem (and it is a serious issue) of "Chinese" versus "Sinitic."

So, when I happened upon Hilary Chappell's edited book, it was as though I had met a kindred spirit. Apart from the fact that she chose a brilliant title for her book, she has also filled it with a wonderful collection of first-class papers (delivered at the First International Symposium on Synchronic and Diachronic Perspectives on the Grammar of Sinitic Languages, which was held at Melbourne, Australia in July, 1996) by a galaxy of top-flight scholars from Australia, France, America, Taiwan, Hong Kong, and Japan. The papers are grouped into four parts: typological and comparative grammar, historical and diachronic grammar, Yue (Cantonese) grammar, and Southern Min (Taiwanese) grammar. To provide an idea of the range of interesting topics covered, here is a list of the papers in the volume and their authors:

In terms of its breadth of coverage and depth of investigation, this is an unprecedented volume. The organizers of the conference where the papers were first presented and the editor of Sinitic Grammar, together with the individual authors, are to be warmly congratulated for making a major contribution to our understanding of the nature and history of the Sinitic language group.

All language examples in the book are provided with characters, romanization, glosses, and translations, making them accessible to specialists and non-specialists alike.

There is a companion volume, written in Mandarin, and edited by the symposium co- organizer (with Hilary Chappell), Yunji Wu. It is entitled Hanyu Fangyan Gongshi Yu Lishi Yufa Yantao Lunwenji (Synchronic and Diachronic Perspectives on the Grammar of Sinitic Languages) (Guangzhou [Canton]: Jinan Daxue Chubanshe, 1999). This book is composed of three studies on different Wu dialects, one on Hui, two on Xiang, one on Xiamen (Amoy) Southern Min, and five others on historical, typological, and comparative topics, including reduplication, copular verbs, verbs of having, comparative constructions, disposal constructions, modals, interrogatives, and affixal morphology.

Sinitic Grammar bristles with technical terms such as. epistemic, deontic, and radial structure / category .Even when one looks these terms up in the index, it is very difficult to find a page on which they are succinctly defined. As a result, it would have been helpful if a glossary of the key concepts employed in the volume had been provided.

There is a common Bibliography for all the papers in the volume, necessitating constant turning to the back of the book to check references, plus the impossibility of photocopying the references for an individual chapter (doubtless intended on the part of the publisher!). An odd feature of the Bibliography is that some of the entries for Chinese and Japanese items are given with characters only for the titles and the publishers (the authors and place of publication are given in romanization) but no romanization or translation, while most of the titles are given in romanization but with no characters or translation. The entry for "Hafu Yanjing Xueshe" should be "Hafo Yanjing Xueshe" or, better yet, "Harvard-Yenching Institute."

The endpaper maps showing the distribution of Sinitic languages within China are spectacularly revealing and, considering that they employ only gray-scale tones, ingeniously informative. A question before closing: there is an assumption in this volume (as in most other treatises on Sinitic languages) that it is improper -- but not impossible -- to write Cantonese, Taiwanese, etc. Why hasn't anyone discussed the linguistic, cultural, social, psychological, and political grounds for that assumption?

Despite all of its virtues and the valuable data it supplies, reading through this book from cover to cover is like an exercise in schizophrenia. From the very first page of the text, where we encounter the theme of "studies of grammar of Sinitic languages (or Chinese dialects." to the last chapter, which -- in the same breath -- discusses "the linguistic features of Medieval Chinese and Mandarin" and Min "positioned in the hierarchy as a great-great uncle on the genealogical tree of Sinitic languages, with Mandarin, a direct descendant of Medieval Chinese, and some other Chinese languages at a lower section on this tree [!!]," the authors waver unsteadily between Sinitic and Chinese (as well as between dialect and language), with no idea of the difference between the two. It is obvious from reading between the lines that the editor, being exceptionally bright and possessed of uncommon analytical acumen, realized clearly that "Sinitic" is far superior to "Chinese" as a category for comprehending the large group of disparate languages grouped under the umbrella term Hanyu, but her co-authors were intellectually and emotionally incapable of following her lead. The best defense they can muster for persisting in their references to "Chinese dialects" rather than "Sinitic languages" is the pathetic admission of the author of chapter 4 (p. 85n1): "In this study I use the traditional term 'Chinese dialects' for the sake of convenience. This does not in any way imply a rejection of the concept of 'Sinitic languages'."

My conclusion? "Chinese" is not a convenient term. It is an unworkable, impossible lintuistic category, both synchronically and diachronically. The only reason people continue to use it is out of the sheer force of habit and the sentimental desire to please non-linguist cultural homogenizers who are devoted to the unrealistic ideal of a monolithic CHINESE lasting through five millennia and occupying the territory from Manchuria to Taiwan to Tibet to Uyghurstan (Eastern Central Asia).