Reviewed by Marjorie K. M. Chan, The Ohio State University. To appear in: Cahiers de Linguistique Asie Orientale (1999) 28.1:101-112.

(This online version includes a few GIF's for displaying IPA and Chinese characters; hence, adjust font size as needed for best viewing.)

[ back to publications page ]

Modern Cantonese Phonology

By Robert S. Bauer and Paul K. Benedict [Trends in Linguistics: Studies and Monographs 102] Berlin and New York: Mouton de Gruyter. 1997. xlvii + 559 pages. Hardcover. ISBN 3-11-014893-5.

Modern Cantonese Phonology by Robert S. Bauer and Paul K. Benedict is the first comprehensive, Englishlanguage publication on Cantonese (Yue dialect) phonology since [Yue-] Hashimoto's Phonology of Cantonese (1972), published a quarter of a century ago and long out of print. Bauer and Benedict (hereafter B&B unless specific reference is to the first author) present a very detailed description of the phonetics and phonology of modern Cantonese, with "modern" to be broadly interpreted to cover the language over a fiftyyear span; that is, changes since the early 1940's are chronicled in this book. Furthermore, fully aware that today's Hong Kong and Guangzhou (Canton City) varieties of Cantonese are not identical, B&B are careful to note differences between Hong Kong and Guangzhou Cantonese throughout the book. They also draw on other subdialectal differences among the Yue dialects in tracing historical changes in Cantonese, as in the case of the search for the historical origin of the high-rising changed tone in Cantonese by analyzing corresponding phenomena in other Yue dialects. The book (p.xii) states that its intended audience includes "beginning students of Cantonese, Cantonese native-speakers, teachers of English to Cantonese-speaking students, professional linguists, and laymen curious about one of Southeast Asia's major languages" (where Cantonese is treated as being spoken in the same part of the world as languages such as Burmese, Thai, and Vietnamese). Despite the stated readership, however, a beginning student of Cantonese may be quite overwhelmed by the amount of detail, especially those pertaining to other Yue subvarieties. At the same time, learners of Cantonese, native speakers, and English teachers with no background in articulatory phonetics before opening the book are bound to find the linguistic terms and phonetic transcriptions that use the International Phonetic Alphabet (IPA) quite daunting and intimidating. The book's steep price would also discourage the casual reader from purchasing it. Given the rich details, informative endnotes, and careful documenting and referencing of sources throughout the book, those who would benefit the most from this work are researchers, linguists who are especially interested in Cantonese and Chinese dialectology. The study does not purport to be based on any particular phonological theory, opting instead for a more general descriptive approach, and making use of phonetic details and acoustic research at various times as part of the study in determining phonological contrasts. Even then, the authors prefer broad phonetic transcriptions to strict phonemicization. For phonologists and dialectologists (which include the reviewer), the book is a treasure trove - a rich reference source for information that draws from numerous publications spanning half a century or more and brings them together into a single volume. This book is very much data-driven, containing systematic and detailed articulatory and acoustic descriptions of different aspects of the phonology of Cantonese. The information is based on both published sources and the authors'own research and data-collection and recording. Scattered throughout the book are tables that summarize and list much of the data in a format that is convenient for the reader to consult. Modern Cantonese Phonology is surely a labor of love, begun by Paul K. Benedict in 1941-1942 as a small, sixty-odd page, typewritten manuscript. Descriptions of the dialect in that manuscript thus captures an earlier stage of the dialect, from which one can then chart changes that have taken place over the past half century. The manuscript was handed to Robert S. Bauer in 1988 for a collaborative volume. The younger author revised and vastly expanded upon the manuscript to produce the present, monumental work that will serve as a lasting tribute to his co-author. Given the amount of new research and publications that were

consulted, the data on modern Cantonese summarized and tabulated therein, the additional data that was collected from informants, including recordings, and so forth, the current book is almost single-handedly the work of the younger scholar, who well merits recognition as the book's first author. From its modest beginnings half a century ago, the final product is a hefty, 559-page volume. The book contains a seventeen-page introduction followed by three main chapters, each over a hundred pages in length, and ending with a short, six-page postscript. Chapter 1 is on "Cantonese consonants and vowels," Chapter 2 on "Cantonese tones," and Chapter 3 on "Cantonese syllables and words." The postscript discusses "Cantonese in the 21st century," and looks at phonetic changes, lexical changes, and the relationship between Cantonese and Putonghua in the next century. In the paragraphs below, some comments will be made on various parts of the three chapters. Due to the length of the book and the details therein, remarks given in this review can only be quite superficial. Chapter 1 is the shortest chapter, although still over a hundred pages in length. It provides a brief introduction to articulatory phonetics, and includes such details as duration measurements of syllables (from Kao 1971). Further details on duration measurements are given later in the chapter concerning whether or not vowel length should be treated as a contrastive feature in the dialect. Although most phonological studies have generally not treated vowel length as contrastive (except with respect to the low vowels, B&B (p.39) make use of duration measurements from Kao (1971) and Li (1985) to draw the conclusion that vowel duration should be considered a distinctive characteristic of the vowel system of Cantonese. The chapter also contains a fairly detailed phonetic description of the segments in modern Cantonese. Changes in the language over the past half century are observed. The authors document and discuss the loss of contrast between plain and labialized velars before the back, rounded vowel, , among many speakers in Guangzhou, Macao, and Hong Kong. The loss of lip-rounding, resulting in the merger of [k] and [kw] (and the corresponding aspirated series) before seems to be more prevalent in Hong Kong and Macao speech than in Guangzhou speech, based on the subjects in Bauer's study. A systematic, larger scale study would be needed in the future for a more in-depth study of this sound change that involves not only sociolinguistic factors such as age and register, but also subdialectal differences. For instance, modern Zhongshan Yue dialect does not exhibit this contrast (Chan 1980); nor did it exhibit that contrast even as far back as the mideighteenth century in Macao (Chan 1982). Among the interesting topics dealt with briefly in section 1.4 is that on the "Alternation of homorganic final consonants: -m/-p, -n/-t, and ." As noted by B&B, Cantonese has a number of word pairs with alternation between homorganic stop and nasal coda in which the members are semantically related. These pairs form so-called 'word families,' a term first used by Bernhard Karlgren (1934) to refer to sets of words with similar (but not identical sound) in Archaic Chinese that were related in meaning, representing relics of morphological processes. (See also Pulleyblank 1973.) Similarly, Bauer (p.93) notes that the phenomenon in Cantonese is believed to be a remnant of an ancient word-derivation process that is no longer productive, but one in which different types of suffixes are attached to lexical roots. Table 1.90 (p.94) presents twenty-five pairs of such words that can be found in modern Cantonese, some of which have standard Chinese characters and others do not. One pair of examples is 'to expand, spread' and 'to enlarge, extend.' The following section 1.5 contains extensive literary versus colloquial pairs that have their source in historical mergers versus preservation of earlier rhyme categories ([Yue-]Hashimoto 1972, Chan 1980). Bauer's study of some Cantonese dictionaries of the past half century suggests that one or the other member of some of these pairs have become obsolete over time. Chapter 2 is the longest chapter, containing about 170 pages. It focuses on the tones in Cantonese, with the chapter divided into two main parts. The first part is a description of the tones based on aural perception and a transcription of the tonal inventory using Yuan Ren Chao's system of tone numbers and tone letters. The second part is an acoustic study using the Summer Institute of Linguistics' speech analysis freeware, WinCECIL (Computerized Extraction of Components of Intonation in Language -- for Windows). The program tracks fundamental frequency for visual display of the tones. Measurements of peaks, endpoints, and dips of tones are also made. Duration measurements are made for checked, or Rusheng, syllables (i.e., those ending in a coda, -p, -t, or -k, referred to in the book as 'dead syllables') and non-checked syllables.

Differences between Hong Kong and Guangzhou tones are noted, such as the observation that most Hong Kong speakers appear to have lost the High Falling tone contour on Pingsheng syllables, pronouncing them as simply High Level and not making a distinction between the two, as Guangzhou speakers do. Macao speakers (Zhan and Cheung 1987:10) are similar to Hong Kong speakers in having only the High Level form of Pingsheng. In addition, they do not have a distinction between Low Rising /13/ and High Rising /35/. This is identical to Zhongshan Cantonese (Chan 1980). However, similar to standard Cantonese and in contradistinction to Zhongshan, Macao Cantonese retains a distinction between Yinqu and Yangqu (/33/ versus /22/), whereas Zhongshan has merged the two. Currently, very little has been published on intonation in Cantonese; in section 2.8 on this subject, only two acoustic studies are cited in the book, namely, one article by Vance (1976 - also see Vance and Walker 1976), and another one by Guo and Luke (1986). Given the limited research on Cantonese intonation, Bauer conducted a small study, reported in section 2.8.2, to investigate the falling intonation pattern that has been perceived as being associated with simple declarative or imperative sentences. Several native Cantonese speakers were recorded using WinCECIL as they read a series of declarative sentences with neutral expression. The results were digitized and the displays of fundamental frequency on the computer monitor were measured. The displays are included in the book as Figures 2.24 through 2.31. Studying the figures, Bauer observes that "the F0 values of tone contours are higher for any type of tone at the beginning of the utterance than at the end" and that "these lower F0 values produce the impression of the falling intonation pattern" (p.151). Bauer also addresses the question of downdrift, the progressive lowering of pitch over a sentence or some other unit; he observes that "tone contours tend to become progressively lower as the speaker arrives at the end of the utterance" (p.154). While the study provides some preliminary observations, a much more controlled experimental study of intonation focussing on declination in Cantonese is conducted by Johnson (1987), using utterances consisting of all high level or low level tones. For a tone language such as Cantonese, with pitch contours and three contrastive level tones, sentence-final particles often serve functions similar to that of intonation in a non-tone language such as English. These particles are treated later in Chapter 3 under "Cantonese morphology" (pages 291-295). It is noteworthy that whereas only seven sentence-final particles are cited for standard Chinese (Matthews and Yip 1994:338), for Cantonese, some sources list thirty or more basic forms (Kwok 1984:8, Ouyang 1993:63), and still others have listed as many as seventy-seven (Ball 1924:122-125). These do not include concatenation of two or more particles. There is, thus, fertile ground here for future research in Cantonese, with respect to the study of intonation both in the absence of sentence-final particles, and with sentence-final particles present. In contrast to the limited research on, and information about, Cantonese intonation, two-thirds of Chapter 2 is devoted the so-called 'changed tone' (bianyin) phenomenon in the dialect. As a result of the detailed study of the topic, Chapter 2 contains a staggering number of tables and figures. To be precise, there are 75 tables and 51 figures in the chapter, leaving the reader feeling rather overwhelmed. The tables are unquestionably useful; they give systematic sets of examples: reduplicated forms; onomatopoeic phrases; derivation (of nouns from verbs) by tone change; names (including surnames and nicknames); address terms; reduplicated and non-reduplicated kinship terms; names of animals, fruits, plants, and vegetables; names of kinds of people; names of body parts; and so forth for many more categories. This is probably, to date, the most comprehensive compilation of forms exhibiting this phenomenon in Cantonese. Complementing the tables are figures display the F0-tracking of the lexical tones followed by numerous 'changed tone' examples. Chapter 3, containing about 150 pages, deals with "Syllables and words." This topic is included in the book, since in Cantonese, the syllable tends to correspond to the word, such that the syllable functions as both a phonological unit and a morphological unit of analysis. This chapter covers topics that are often neglected or ignored, and includes data that are not readily found elsewhere. In the first half of the chapter, they examine homophonous words involving taboo and obscene words, as well as auspicious and inauspicious words. Also studied in that portion of the chapter is the topic of Cantonese sound symbolism and meaning. Many examples are given that involve semi-reduplicated patterns and onomatopoeic forms. That first half of the chapter also investigates contraction in Cantonese, and the occurrence in casual speech of consonant clusters containing the lateral, /l/. As noted by the authors, this is not limited to standard Cantonese; it is also found in the Yangjiang dialect (Huang 1986). There are also implications for historical phonology, as some of

these forms provide evidence - other than xiesheng series - for the reconstruction of consonant clusters in Old Chinese (Chan (1984). The bulk of the remaining half of Chapter 3 is an extremely detailed and systematic study of English loanwords in Cantonese. It deals with loan translation and phonetic borrowing. Here, as elsewhere in the book, data are compiled and listed in some 20 or so tables. The chapter closes with a section on the Cantonese syllabary. It goes systematically through a study of syllables that correspond to reading pronunciation of standard Chinese characters, colloquial Cantonese syllables, as well phonotactics and labial dissimilation. While the book per se is divided into three chapters, thumbing through it, one finds roughly just over half of the space in the book is actually occupied by text. The other half consists of endnotes, lists of maps, figures, tables, appendices, references, glossary of linguistic terms, and author and language/subject indices. References alone occupy 21 pages, not to mention 9 pages of indices, 12 pages of glossary of linguistic terms, and 68 pages for 5 appendices. Among the appendices, one that is particularly useful is the "Comparative table of Cantonese romanization systems;" and another is the "Cantonese syllabary," which contains the full set of syllables in the dialect. In addition, the book contains 65 figures. Most of the figures show displays of pitch contours obtained from sound files that are digitized from prerecorded tapes, or recorded directly, using WinCECIL. The book contains an incredible compilation of over 200 tables that cover a wide spectrum of topics under the rubric of Cantonese phonology. All figures and tables are painstakingly listed at the beginning of the book. Those compilations, alone, make the book a valuable and convenient source of information. For completeness, one should also note that the volume contains two linguistic maps in the introductory pages before the main chapters begin. One map shows China's major dialect groups distributed by province, and the other the Yue dialect areas in the southeastern part of China. The seven-major dialect classification is based in Yuan (1983), and subdivided into Northern, Central, and Southern based on Norman (1988), although the source of the latter, larger grouping is not made explicit. The classification of Chinese dialect grouping follows earlier convention and is not based on the finer-grained classification outlined in Li (1989), and reflected in Wurm et al.'s (1987) Language Atlas of China. The second map, on the classification of the Yue dialects, follows the dialect subgrouping proposed by [Yue-] Hashimoto (1972). Two main groups are proposed, based on phonological and lexical criteria, yielding a larger Pearl River Delta group and a smaller Siyi-Liangyang group, with each group further subdivided. While the authors acknowledge that Map 2 is "not exactly to scale and locations of districts are only approximate" (p. xlvii), the enlargement of the Pearl River Delta region that include Zhongshan and the Siyi districts could have been more accurately drawn with respect to their relative geographical location. The greater geographical isolation of the Siyi districts, for example, is not apparent from the map: Enping, Kaiping, and Taishan are depicted as more centrally located than they actually are, and the fourth district, Xinhui, is somehow overlooked. Clearly, with a book this voluminous, it is impossible to do justice to it in a short review. The book not only provides a detailed description of the phonology of Cantonese, but it also serves as a fantastic reference source. Given the wealth of data described in prose, combined with summaries and comprehensive listings of data, this is a must-have book for anyone seriously interested in Chinese dialectology, and especially those researching in Cantonese phonology, synchronic or diachronic. The amount of time that the first author had pored into producing this book is mind-boggling. It is unlikely that one will see such scholarly dedication to detail for a very long time. And for that alone, this volume will serve as a truly invaluable reference source for many, many years to come.

REFERENCES
BALL, J. Dyer (1924). Cantonese Made Easy. Fourth Edition. Hong Kong: Kelly and Walsh, Ltd. (First edition: 1883)

CHAN, Marjorie K.M. (1980). Zhong-shan Phonology: A Synchronic and Diachronic Analysis of a Yue (Cantonese) Dialect. M.A. thesis, University of British Columbia, Vancouver, Canada. CHAN, Marjorie K.M. (1982). "A response to Boltz' notes on Cantonese dentilabialization." Journal of the American Oriental Society l02.1:107-109. CHAN, Marjorie K.M. (1984). "Initial consonant clusters in Old Chinese: evidence from sesquisyllabic words in the Yue dialects." Fangyan (1984) 4:300-313. CHAN, Marjorie K.M. (1996). "Gender-marked speech in Cantonese: the case of sentence-final particles je and jek." Studies in the Linguistic Sciences 26.1/2 (Spring/Fall 1996):1-38. GUO, Kailun Zhang and Lu Jingguang [= Helen Kwok and Kang Kwong Luke] (1986). "Yueyu yudiao chutan" (Preliminary study of Cantonese intonation). Yuwen Zazhi 13:32-40. HASHIMOTO, [Anne] Oi-Kan Yue (1972). Phonology of Cantonese. Cambridge: Cambridge University Press. HUANG, Borong (1986). "The sound system of the Yangjiang dialect." Computational Analysis of Asian and African Languages 26:53-74. JOHNSON, Keith (1987). "Intonation in Cantonese." OSU Working Papers in Linguistics 36:1-15. [Ohio State University, Columbus, Ohio] KAO, Diana L. (1971). Structure of the Syllable in Cantonese. The Hague: Mouton. KARLGREN, Bernhard (1934). "Word families in Chinese." Bulletin of the Museum of Far Eastern Antiquities 5.9-120. KWOK, Helen (1984). Sentence Particles in Cantonese. Hong Kong: Centre of Asian Studies. LI, Rong (1989). "Hanyu fangyande fenqu" [Classification of the Chinese dialects] Fangyan (1989).4:241259. LI, Xingde (1985). "Guangzhouhua yuanyinde yinzhi ji changduan duili" [Opposition of vowel quality and length in Cantonese] Fangyan (1985) 1:28-38. MATTHEWS, Stephen and Virginia Yip (1994). Cantonese: A Comprehensive Grammar. London and New York: Routledge. NORMAN, Jerry (1988). Chinese. Cambridge: Cambridge University Press. OUYANG, Jueya (1993). Putonghua Guangzhouhuade Bijiaoyu Xuexi. (Comparison and study of Putonghua and Cantonese) Beijing: Zhongguo Shehui Kexue Chubanshe. PULLEYBLANK, Edwin G. (1973). "Some new hypotheses concerning word families in Chinese." Journal of Chinese Linguistics 1.1:111-125. VANCE, Timothy (1976). "An experimental investigation of tone and intonation in Cantonese." Phonetica 33:368-392. VANCE, Timothy and Carol A. Walker (1976). "Tone and intonation in Cantonese." Papers from the Regional Meeting of Chicago Linguistic Society (CLS) 12:640-655.

WURM, Stephen Adolphe et al. (general editors) (1987). Language Atlas of China. Part 1. Hong Kong: Longman (Far East) Ltd. (Part 2 was published in 1991.) YUAN, Jiahua (1983). Hanyu Fangyan Gaiyao. [Survey of Chinese dialects] Second edition. (First edition: 1960) Beijing: Wenzi Gaige Chubanshe. ZHAN, Bohui and Yat-Shing Cheung (1987). A Survey of Dialects in the Pearl River Delta. Volume 1: Comparative Morpheme-Syllabary. Hong Kong: New Century Publishing House.

Marjorie K.M. CHAN Department of East Asian Languages and Literatures The Ohio State University Columbus, Ohio 43210 U.S.A. Homepage: http://people.cohums.ohio-state.edu/chan9