ZHONG-SHAN PHONOLOGY: A Synchronic and Diachronic Analysis of a Yue (Cantonese) Dialect by MARJORIE KIT MAN CHAN B.A., Univ e r s i t y of B r i t i s h Columbia, 1974 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in THE FACULTY OF GRADUATE STUDIES Department of L i n g u i s t i c s We accept t h i s thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA September, 1980 © Marjorie K i t Man Chan, 1980 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of Brit ish Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. The University of Brit ish Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 Depa rtment ABSTRACT Zhong-shan i s a county i n Kwangtung Province i n southern China. What i s normally referred to as the "Zhong-shan d i a l e c t " i s the speech of S h i - q i , the administrative centre for the county. For the present thesis, data were c o l l e c t e d from native Zhong-shan speakers from Shi-qi and neighbouring v i l l a g e s where the speech can be equated with the S h i - q i , Zhong-shan d i a l e c t . The data e l i c i t e d consist of two main types: (1) c o l l o -q u i a l vocabulary, for which graphic representation (an the form of standard Chinese characters) do not e x i s t , and (2) a lexicon based on the reading of a standard word l i s t for Chinese d i a l e c t surveys (namely, the Fang-yan Diao-cha Zi-biao), which contains some 3,700 Chinese characters. The synchronic study, which used both sets of data, i s based on an amalgamation of Western struc-t u r a l i s t and Chinese ( t r a d i t i o n a l and modern) approaches. For the diachronic study, the d i a l e c t survey l i s t , arranged according to h i s t o r i c a l phonological categories, was indispensible. The diachronic study e s s e n t i a l l y mapped the pat-tern of correspondences of the d i a l e c t against the h i s t o r i c a l categories to which each word belonged. Against such a backdrop, i t i s possible to observe the development of a given d i a l e c t with respect not only to e a r l i e r s trata of the Chinese language, but also to other modern Chinese d i a l e c t s . Thus, i n Zhong-shan, some features may reveal certain mergers with reference to a p a r t i c u l a r stratum of the language, whereas other features may show survivals of yet older d i s t i n c t i o n s . References to previous studies on the Zhong-shan d i a l e c t are also made when differences between (.or among) data seem s i g n i f i c a n t . C r o s s - d i a l e c t a l l y , since the Can-tonese d i a l e c t i s the standard for the Yue d i a l e c t group to which Zhong-shan belongs, a comparison between Zhong-shan and Cantonese i s made throughout the study. Other southern Chinese d i a l e c t groups, such as Min and Hakka, are also c i t e d where relevant. The thesis i t s e l f i s divided into two main parts: the f i r s t part i s the synchronic study, and the second part the di a -chronic analysis. In order that the thesis may better serve future research endeavours, both the c o l l o q u i a l lexicon and the lexicon of character readings are included: the c o l l o q u i a l data appear . at the end of Part I, while the d i a l e c t word l i s t occurs at the end of Part I I . The reading of the characters i s recorded d i r e c t l y onto the format of the Fang-yan Diao-cha Zi-biao that the Chinese L i n g u i s t i c s Project at Princeton had prepared express-ly for fieldwork purposes. Immediately following the d i a l e c t survey material i s an index to the d i a l e c t material. The index i s likewise prepared by the Chinese L i n g u i s t i c s Project, with, the words arranged according to Pin-yin romanization. i v TABLE OF CONTENTS T i t l e Page i Abstract i i Table of Contents i v L i s t of Tables v i i L i s t of Charts v i i i L i s t of Maps i x Acknowledgement x Part I. Synchronic Study CHAPTER 0. INTRODUCTION 1 0.1. Terminology, T r a n s l i t e r a t i o n and Other Conventions 1 0.2. H i s t o r i c a l , Geographical and L i n g u i s t i c Setting 3 0.3. Data Base 6 0.3.1. The Informants—Biographical Sketch 6 0.3.2. Data-Collection Techniques 8 0.3.3. Additional Sources for Zhong-shan Data 11 Notes to Chapter 0 14 CHAPTER 1. PHONETIC DESCRIPTION AND PHONEMIC ANALYSIS. 17 1.1. I n i t i a l s 21 1.2. Finals 32 1.2.1. Nuclear Vowels 40 1.2.2. Medials 50 1.2.3. Endings 62 1.2.4. S y l l a b i c Nasals 63 1.3. Tones 63 1.3.1. Tonal System 64 1.3.2. Tone Sandhi 6 7 1.3.3. Tone Change 68 1.4. Combination of I n i t i a l s and Finals 77 1.4.1. Labia l D i s s i m i l a t i o n 77 V 1 . 4 . 2 . Syncope 78 1 . 4 . 3 . S e s q u i s y l l a b i c S t r u c t u r e s 79 1 . 4 . 4 . C o l l o q u i a l Versus L i t e r a r y Forms 95 Notes to Chapter 1 101 CHAPTER 2. SYLLABARY AND LEXICON 107 2 . 1 . S y l l a b a r y Arranged Accord ing to Modern Zhong-shan F i n a l s 110 2 . 2 . L e x i c o n of C o l l o q u i a l Terms Arranged Accord ing to Modern Zhong-shan F i n a l s 126 P a r t I I . D i a c h r o n i c Study CHAPTER 3 . ANALYSIS OF MODERN REFLEXES OF HISTORICAL CATEGORIES 161 3 . 1 . I n i t i a l s 164 3 . 1 . 1 . Reconst ructed Values of M idd le Chinese I n i t i a l s 164 3 . 1 . 2 . Zhong-shan Correspondences to Midd le Chinese I n i t i a l s 175 3 . 1 . 2 . 1 . B i l a b i a l s (LMC) 175 3 . 1 . 2 . 2 . L a b i o d e n t a l s CLMC) 180 3 . 1 . 2 . 3 . Denta ls (LMC) 183 3 . 1 . 2 . 4 . Denta l S i b i l a n t s (LMC) 191 3 . 1 . 2 . 5 . R e t r o f l e x e s (LMC) 197 3 . 1 . 2 . 6 . R e t r o f l e x S i b i l a n t s (EMC) and P a l a t a l s (EMC) 201 3 . 1 . 2 . 7 . V e l a r s (LMC) 216 3 . 1 . 2 . 8 . G u t t u r a l s (LMC) 232 3 . 2 . F i n a l s 241 3 . 2 . 2 . Modern Zhong-shan Correspondences to LMC F i n a l s 260 3 . 2 . 2 . 1 . Guo-she 261 3 . 2 . 2 . 2 . J i a - s h e 265 3 . 2 . 2 . 3 . Yu-she 266 3 . 2 . 2 . 4 . X i e - s h e 271 3 . 2 . 2 . 5 . Z h i - s h e 277 3 . 2 . 2 . 6 . X i a o - s h e 282 3 . 2 . 2 . 7 . L i u - s h e 283 v i 3.2.2.8. Xian-she 285 3.2.2.9. Shen-she 289 3.2.2.10. Shan-she 290 3.2.2.11. Zhen-she 297 3.2.2.12. Dang-she 304 3.2.2.13. Jiang-she 307 3.2.2.14. Zeng-she 309 3.2.2.15. Geng-she 311 3.2.2.16. Tong-she 317 3.3. : Tones • 319 3.4. Concluding Remarks 333 Notes to Chapter 3 334 CHAPTER 4. LEXICON (AS ARRANGED IN THE FANG-YAN DIAO-CHA ZI-BIAO) 337 Rhyme Groups: 1. Guo-she 1 339 2. Jia-she 7 345 3. Yu-she 14 352 4. Xie-she 28 366 5. Zhi-she 47 385 6. Xiao-she 67 405 7. Liu-she 81 419 8. Xian-she 92 430 9. Shen-she 110 448 10. Shan-she 115 453 11. Zhen-she 148 486 12. Dang-she 166 504 13. Jiang-she 180 518 14. Zeng-she 184 522 15. Geng-she 191 529 16. Tong-she 212 212 Index 565 Bibliography 655 v i i LIST OF TABLES Table 1. The I n i t i a l s i n Zhong-shan 21 Table 2. (a) Analysis of the Fi n a l s — C h a n 35 (b) Analysis of the Finals—Chao 36 (c) Analysis of the F i n a l s — E g e r o d 37 (d) Transcription of Finals i n (Macao) Zhong-shan—Ball 38 (e) Phonetic Transcription of the Fi n a l s i n Cantonese 39 Table 3. (a) Tonal System of Zhong-shan 6 5 (b) A Comparison of Cantonese and Zhong-shan Tones 66 Table 4. Coll o q u i a l Versus L i t e r a r y Forms: (a) Tone /13/ Aspirated I n i t i a l and /22/ Unaspirated I n i t i a l 98 (b) /a:ng/ and /ang/ Finals 98 (c) /ia:ng/ and /ing/ Finals 99-100 Table 5. Words with Alternate H i s t o r i c a l Tonal Categories: (a) Words L i s t e d as Ping-sheng 327 '(b) Words L i s t e d as Shang-sheng 327 (c) Words L i s t e d as Qu-sheng 328-329 (d) Words L i s t e d as Ru-sheng 329 Notes to Table 5 329 v i l i LIST OF CHARTS Chart 1. Syl l a b l e Structure of Chinese 17 Chart 2. Syl l a b l e Structure of Zhong-shan 19 Chart 3. (a) Analysis of the Vowels—-Chan 40 (b) Analysis of the Vowels—Chao 41 (c) Analysis of the Vowels—Egerod 41 Chart 4. The 36 I n i t i a l s of Late Middle Chinese 165-166 Chart 5. Early Middle Chinese I n i t i a l s 169 Chart 6. Middle Chinese I n i t i a l s i n the Fang-yan Diao-cha Zi-biao 172 Chart 7. Zhong-shan Correspondences to the Middle Chinese I n i t i a l s i n the Fang-yan Diao-cha Zi-biao 173-174 Chart 8. D i s t r i b u t i o n of the LMC Glides and Medials 244 Chart 9. The 16 Rhyme Groups 248 Chart 10. Late Middle Chinese Finals 249-252 Chart 11. Rhymes and Grades Within Each Rhyme Group 256-258 Chart 12. Zhong-shan Correspondences to the LMC Fi n a l s 259 Chart 13. (a) Zhong-shan Correspondences to the H i s t o r i c a l Tones 320 (b) Cantonese Correspondences to the H i s t o r i c a l Tones 320 Chart 14. Ru-sheng Correspondences i n Cantonese and Zhong-shan 323 Chart 15. Tonal Correspondences i n Zhong-shan (and Cantonese) 323 Map 1. Kwangtung Province Map 2. Zhong-shan County X ACKNOWLEDGEMENT The data for the present study of the Zhong-shan d i a l e c t was co l l e c t e d at various i n t e r v a l s , beginning with a short project i n the spring of 1977; the bulk of the data for the thesis, however, was gathered i n the spring of 1978. For these e l i c i t a t i o n s , the p r i n c i p a l informants were my parents, Chen Gui-hong P ^ L - ^ j ^ l ^ anc^ Yang Zhi-fang ^ / c V ^ f . They responded t i r e l e s s l y to my seemingly endless questions. To them, I owe a depth of gratitude that words cannot f u l l y express. Their willingness, patience and support have made t h i s experience a very personally rewarding one. I am grateful to the L i n g u i s t i c s Department at the Univer-s i t y of B r i t i s h Columbia for the academic t r a i n i n g that I have received, and for t h e i r supportive role i n my endeavours. Con-cerning the thesis-writing i n p a r t i c u l a r , I am indebted to my Committee members, Dr. M. Dale Kinkade and Dr. Sarah B e l l , for th e i r reading of an e a r l i e r draft and he l p f u l suggestions. I am also appreciative of Dr. Kinkade's assistance i n the f i n e r phonetic transcriptions noted i n the study. To my advisor, Professor E.G. Pulleyblank, I owe deep gratitude for his many constructive c r i t i c i s m s and suggestions throughout the thesis-writing, and for his c l a r i f i c a t i o n of v a r i -ous points that were unclear to me. His helpfulness, together with his quiet support and immense patience during that entire period, cannot be over-emphasized. I would also l i k e to thank Professor Jerry Norman, at the University of Washington, who kindly read through an e a r l i e r d r a f t and made a number of help f u l comments, p a r t i c u l a r l y with regard x i to Min d i a l e c t material. Also h e l p f u l i n my thesis endeavours has been Karl K. Lo, a native Zhong-shan speaker, who supplied a few of the forms that are recorded here. I am also grateful to my husband, Gary R. White, for his moral support, encouragement, and general a s s i s t -ance that f a c i l i t a t e d the writing and f i n a l type-up of the thesis. I wish to acknowledge with h e a r t f e l t thanks support from the University of B r i t i s h Columbia (U.B.C. Graduate Fellowship 1976-1978). I am also g r a t i f i e d that the University of B r i t i s h Columbia allowed me to use the Summer Research Grant (.1977) to attend the L i n g u i s t i c Society of America's 1977 Summer L i n g u i s t i c I n s t i t u t e . That year, the I n s t i t u t e was held at the University of Hawaii where special emphasis was on Asian and P a c i f i c languages. The study programme was also supported by the L i n g u i s t i c Society of America (LSA Fellowship 1977), to which I wish also to express my deep gratitude. Needless to say, although I owe much to a l l those who have guided me i n the thesis, I am s o l e l y responsible for whatever errors that remain. X I 1 Map 1. Kwangtuna Province. (A modification, of. Yue, 1979:2.) x i i i - 1 -PART I. SYNCHRONIC STUDY CHAPTER 0. INTRODUCTION Studies of. the Yue d i a l e c t s of Chinese have generally concentrated on standard Cantonese and Taishanese, while other di a l e c t s have received peripheral attention. To counterbalance th i s general trend, the Yue d i a l e c t which i s investigated i n the present thesis i s the Zhong-shan"*" <Xf d i a l e c t . The primary goal of t h i s study, however, i s to supplement e x i s t i n g works on the Zhong-shan d i a l e c t with a larger corpus of f i e l d data and a more detailed analysis of i t s phonology on a synchronic and diachronic l e v e l . The hope i s that both the raw data and the analysis w i l l contribute to future l i n g u i s t i c research. 0.1. Terminology, T r a n s l i t e r a t i o n and Other Conventions The term "Cantonese" has often been used to i d e n t i f y both the standard Cantonese d i a l e c t and the d i a l e c t group, thereby creating some confusion for those less f a m i l i a r with the l i n g u i s t i c s i t u a t i o n i n Kwangtung Province i n southern China. To eliminate th i s source of ambiguity, the d i s t i n c t i o n between the terms "Yue" and "Cantonese" drawn by Oi-kan Yue Hashimoto (1972a:1) i s adopted here: "Yue" i s used to refer to the d i a l e c t group, and "Cantonese" to the group norm. Yue i s the d i a l e c t group representative of Kwangtung Province, although found there are also other major Chinese d i a l e c t groups such as Hakka and Min. Cantonese, or Standard Yue, i s the d i a l e c t of the majority of people i n Canton and Hong Kong, located on the Pearl River Delta. "Standard Cantonese" w i l l be a term used only for further c l a r i f i c a t i o n - 2 -or emphasis. The new Pin-yin romanization system, which has been o f f i c i a l l y adopted i n the People's Republic of China to transcribe the national standard, pu--tong-hua ('common d i a l e c t ' , or what i s usually regarded as the Peking d i a l e c t of Mandarin), w i l l be used here for the t r a n s l i t e r a t i o n of Chinese terms, including personal and place names, with a few minor modifications. Hyphens w i l l be inserted between s y l l a b l e s within a word, and an occasional tone mark w i l l be used for disambiguation i n cases where the romanized form of several words, for example, would have been i d e n t i c a l except for tone. The four tones i n Mandarin are marked thus: '"' for l e v e l , " 1 r i s i n g , l v ' dipping ( f a l l i n g - r i s i n g ) , and , v ' f a l l i n g . Pin-yin i s used for Chinese personal and place names except for well-known geographical names, such as Canton and Hong Kong, for which the romanization established i n the China Postal Atlas w i l l be retained. Moreover, rather than attempt to over-standardize personal names to a single romanization system, the romanization that has already been established for the names of well-known Chinese l i n g u i s t s such as Yuen Ren Chao and Fang Kuei L i w i l l likewise be used here. The use of the terms "character" and "word" needs some c l a r i f i c a t i o n . The Chinese language distinguishes between what i s a "character" ( z i ^jl ) and what i s a "word" (c i ^s) ). A cha-racter i s simply the i n d i v i d u a l ideograph, which i s a monosyllable. Often a character constitutes a word; the character da , for instance, i s a word meaning "big". However, i t can also combine with the character xue ^ (which, as a monosyllabic word, means 'to study') to form the d i s y l l a b i c word da-xue ^ ^ / which -- 3 -means ' u n i v e r s i t y ' . In a grammatical a n a l y s i s , one should f u r t h e r in t roduce the l i n g u i s t i c term 'morpheme', which i s 2 l o o s e l y d e f i n e d as a "min imal meaningfu l u n i t " . U s u a l l y there i s a o n e - t o - o n e correspondence between a s y l l a b l e (or charac te r ) and a morpheme; t h a t i s , most morphemes i n Chinese are m o n o s y l l a -b i c , as e x e m p l i f i e d by da and xue above, which can now be analyzed as two m o n o s y l l a b i c morphemes c a r r y i n g the meanings ' b i g ' and ' t o s tudy ' r e s p e c t i v e l y . There a r e , n o n e t h e l e s s , a few r a r e cases of d i s y l l a b i c morphemes whose o r i g i n i s no longer known. The word h u - d i e tg^ JJ ' b u t t e r f l y ' i s a case i n p o i n t ; the f i r s t s y l l a b l e hu c o n t a i n s no meaning i n and of i t s e l f . Other t e r m i n o l o g i e s and convent ions w i l l be i n t r o d u c e d as they are met i n the f o l l o w i n g s e c t i o n s . 0 . 2 . H i s t o r i c a l , Geograph ica l and L i n g u i s t i c S e t t i n g . Dur ing the Han Dynasty (206 B . C . - 2 3 A . D . ) , what i s now the county (x ian ^j^. ) ^of. Zhong-shan was p a r t of Pan-yu county "I) % * I n T a n 9 t imes (618-907) i t became a p a r t of Dong-guan county jj^ . I t was a t the beg inn ing of the Song Dynasty (960-1279) t h a t i t became e s t a b l i s h e d as a separate county c a l l e d X iang -shan county |j J, |tj^ , now over e i g h t hundred years ago. In the f o u r t e e n t h year of the R e p u b l i c of China ( i . e . , 1925) , i n honour of Dr . Y a t - s e n Sun, the f a t h e r of the Chinese r e v o l u t i o n , the name of h i s b i r t h p l a c e was o f f i c i a l l y changed by the p r o v i n c i a l government of Kwangtung from X i a n g - s h a n to Zhong-shan VJ* based on the name which Dr . Sun adopted w h i l e seek ing p o l i t i c a l asylum i n Japan, "Zhong-shan" be ing the Chinese - 4 -pronunciation of "Naka-yama" ^ 4 Geographically , Zhong-shan i s one of the coastal counties of Kwangtung Province. I t i s situated west of the Pearl River delta and immediately north of Macao, thus partway between Canton and Macao (see Map 1). The county i s over 1,800 square kilometres i n area—70 kilometres long running north and south, and 35 k i l o -metres wide east and west. Within t h i s circumference, the county i s divided into nine administrative sections c a l l e d qu (3* . Shi-q i (.otherwise romanized as "Shekki") fa , the administrative centre and the largest town i n the county, i s located i n the f i r s t qu. The Zhong-shan d i a l e c t i s here i d e n t i f i e d and equated with Shi-qi speech, which i s considered the standard for the county. Besides S h i - q i , also entering into the present study i s the neigh-bouring v i l l a g e of Ku-chong jfjjj , situated i n the fourth qu immediately behind the boundary southeast of Shi-qi (see Map 2). A number of the more c o l l o q u i a l expressions c o l l e c t e d i n the present study r e f l e c t Ku-chong speech rather than that of the more educated townspeople of S h i - q i . In terms of i t s l i n g u i s t i c c l a s s i f i c a t i o n , Zhong-shan i s one of the Yue d i a l e c t s , consequently sharing many of the features found i n Cantonese. At the same time, the Shi-qi d i a l e c t i n China i s also influenced by the surrounding d i a l e c t s . Although small i n area, a l l three major d i a l e c t groups i n Kwangtung—-namely Hakka, Min and Yue—are spoken i n the Zhong-shan county. I t i s therefore not surprising to f i n d each d i a l e c t i n turn influenced by the others, i n a process which has been described by Egerod (1956:76) as "balkanization", using the term i n the l i n g u i s t i c sense of "the gradual merging of geographically close, etymologically far-removed - 5 -speech forms". Egerod (p.77) notes, for example, that the simpli-c i t y of Shi-qi tonal pattern, i n contrast to that i n Cantonese, renders the Shi-qi d i a l e c t closer to Min than to Yue with respect to the number of tones. Zhong-shan i s , i n fact, the only exception to the pattern of eight or more tones and the dichotomy of the Yin-ru tone which constitute two of the c h a r a c t e r i s t i c features of the Yue d i a l e c t s (Hashimoto, 1972a:44). Hence, i t would not be unreasonable to suggest balkanization as a possible factor i n the uniqueness of the Zhong-shan tonal system amongst the Yue d i a l e c t s . Although the Shi-qi d i a l e c t i s the standard for the county, i t i s predominant only within i t s own qu. In the remaining eight qu, Cantonese, Hakka, and various forms of Min constitute the 6 major d i a l e c t s . Hence, the Shi-qi d i a l e c t i s actually spoken i n a very li m i t e d area, i t s prominence f e l t only as a r e s u l t of i t s economic and p o l i t i c a l status. In t h i s p osition also, i t has stringent competition from Macao, which i s geographically and h i s -7 t o r i c a l l y , though no longer p o l i t i c a l l y , a part of Zhong-shan county. The l i n g u i s t i c scene i n Macao has undergone d r a s t i c changes since J.D. B a l l ' s fieldwork before the turn of the century when the county was s t i l l c a l l e d "Xiang-shan" (or "Hong-shan", i n g B a l l ' s Cantonese t r a n s c r i p t i o n ) . According to B a l l (1897:550) the d i a l e c t spoken i n Macao was i d e n t i c a l to the Zhong-shan di a -l e c t , with exceptions a r i s i n g primarily from the desire of the educated class i n Shi-qi to emulate the more prestigious Cantonese forms. By mid-twentieth century, Egerod (p.3) observes that Stan-dard Cantonese has become the main d i a l e c t i n Macao. Neverthe-les s , the Cantonese spoken there retains a few traces of the Zhong-- 6 -shan d i a l e c t , such as the fusion of the p l a i n and l a b i a l i z e d velar stops as a r e s u l t of which Cantonese /kwa/ and /ka/, for instance, are both pronounced /ka/ i n Macao, with l a b i a l i z a t i o n l o s t . Meanwhile, because of low y i e l d i n a g r i c u l t u r a l production i n the past i n Zhong-shan"^, i t has been a t r a d i t i o n for the l o c a l people to seek a means of l i v e l i h o o d away from home. Among the places which attracted many immigrants was the Hawaiian Islands. As Chao (1948:49) 1 1 commented, the Chinese population i n Hawaii was predominantly speakers of the Zhong-shan d i a l e c t . Since the publication of Chao 1s a r t i c l e i n the middle of the century, i t i s possible that the i n f l u x of Chinese immigrants from Hong Kong, Taiwan, and other areas may have reduced the proportion of Zhong-shan speakers i n Hawaii. 0.3. Data Base. The study of Zhong-shan phonology i s based on data c o l l e c t -ed by the writer at various i n t e r v a l s from 1977 to 1980, the bulk, of which was gathered during the spring of 1978. Published works on the d i a l e c t have also been consulted, as w i l l be further e l a -borated subsequently. 0..3.1. The Informants--Biographical Sketch. Data were e l i c i t e d from two main informants: Chen Gui-hong fjL $h a n c ^ Y a n 9 Zhi-fang $ ' t n e writer's parents. Chen was born i n 1923 i n the v i l l a g e (cun ^ ) of Ku-chong ^ fj ^ , where he acquired the early part of his education. He com-pleted grammar school i n S h i - q i , which i s west of the v i l l a g e , about half an hour's walk away. His mother, who i s l i v i n g with - 7 '-the family, was born i n the v i l l a g e (xiang ) of Yuan-feng walking distance from the town (see Map 2). Chen attended busi-ness college i n Hong Kong, where he also studied Mandarin, English and Japanese. He t r a v e l l e d between Ku-chong and Hong Kong several times before immigrating to Vancouver, B r i t i s h Columbia i n 1952 to j o i n his parents. Chen's father, who was born i n Ku-chong, spoke Zhong-shan and a few words of English, while Chen's mother only knows the Zhong-shan d i a l e c t . Contact with fellow v i l l a g e r s from Ku-chong i s maintained through friends, r e l a t i v e s and annual gatherings organized by residents of the "Ku-chong Home", a house open to Ku-chong v i l l a g e r s who wish to drop i n from time to time, or who need a place to stay. Yang was born i n 19 27 i n S h i - q i , where she received four or f i v e years of education before i t was disrupted by the invasion of China by Japan. She continued to l i v e i n S h i - q i , working there as a nurse during the war years. Yang l i v e d i n Ku-chong for several years before spending two or three years i n Macao and Hong Kong, a r r i v i n g i n Vancouver, B r i t i s h Columbia i n 1958. She has had exposure to Mandarin and Japanese. Her knowledge of English has been acquired informally, primarily through contact with cus-tomers i n a small, family-operated grocery store. to a much more limited extent since she l e f t Ku-chong at the age of four or f i v e and was then exposed to Cantonese i n Macao and Hong Kong, and i n Canada subsequently. , north of S h i - q i , also approximately half an hour's The writer herself also served as an informant, although - 8 -p.. 3 . 2 . D a t a - C o l l e c t i o n Techniques A p r e l i m i n a r y se t of da ta was c o l l e c t e d i n the s p r i n g of 1977 i n the attempt to e l i c i t c o l l o q u i a l versus l i t e r a r y read ings of the same c h a r a c t e r s based on those found i n Cantonese ( i . e . , i n Hashimoto, 1 9 7 2 a : 1 6 9 - 1 7 0 ) . I t was recogn ized by the w r i t e r t h a t such an approach has i t s l i m i t a t i o n s , the pr imary one b e i n g t h a t a d i s t i n c t i o n between l i t e r a r y and c o l l o q u i a l r e a d i n g of some words found i n Zhong-shan but not p resent i n Cantonese would be missed . Given l i m i t e d t i m e , however, i t was n e v e r t h e l e s s a c o n -v e n i e n t means to o b t a i n a s i z e a b l e l i s t w i thout r e s o r t i n g t o an e x t e n s i v e l e x i c o n . The method used was as f o l l o w s : a character ; , was f i r s t read by Chen, f o l l o w e d by a d i s c u s s i o n of whether or not there e x i s t s a c o l l o q u i a l c o u n t e r p a r t . Some e f f o r t was a l s o made subsequent ly to produce l i t e r a r y versus c o l l o q u i a l forms, a l though not by any s y s t e m a t i c or c o n s i s t e n t approach. The major task of e l i c i t i n g Zhong-shan data was, however, based on the Fang-yan D i a o - c h a Z i - b i a o ^ -^ /l) ^ ^- jfc^ 'A Table of Charac te rs f o r D i a l e c t Surveys ' ( h e r e a f t e r r e f e r r e d to as the " d i a l e c t survey l i s t " f o r s h o r t ) . That source p r o v i d e s a s tandard l i s t of j u s t over 3,700 c h a r a c t e r s arranged a c c o r d i n g to t r a d i t i o n a l Chinese p h o n o l o g i c a l c a t e g o r i e s , and i s used f o r com-p a r i s o n s among the Chinese d i a l e c t s and f o r s t u d i e s of the h i s t o -r i c a l phonology of a p a r t i c u l a r d i a l e c t . A l though d i f f e r e n t e d i t i o n s of t h i s survey l i s t e x i s t , they are e s s e n t i a l l y the same, w i t h but minor v a r i a t i o n s i n the cho ice of c h a r a c t e r s and format . The e d i t i o n used i n the p resent study i s p u b l i s h e d by P r i n c e t o n U n i v e r s i t y (1970) , and i s p a r t of the Chinese L i n g u i s t i c P r o j e c t a t P r i n c e t o n . Th is e d i t i o n , e s p e c i a l l y arranged f o r f i e l d w o r k - 9 -purposes, i s based on the 1955 character l i s t of the same name compiled by the Research Inst i t u t e of L i n g u i s t i c s , Chinese Academy of Sciences i n Peking. The survey l i s t provided data for the study of the phonological system of Zhong-shan, as well as y i e l d i n g the necessary material for a diachronic analysis of the d i a l e c t . In preparing for the d i a l e c t survey on Zhong-shan, each character on the l i s t was assigned a number consisting of two parts separated by a dash with the part preceding the dash in d i c a t i n g reference to the page number i n the d i a l e c t survey l i s t , and the part following the dash that of the character's p o s i t i o n 12 h i n the column on a given page. The word duo ^ 'many', for example, i s assigned the number "1-1" to specify page one, and the f i r s t character on the page. S i m i l a r l y , tuo ftfcj 'to drag along' i s assigned the number "1-2" since i t i s the second charac-ter on page one i n the survey text. (See Chapter 4.) After each character had been given a number, the order of the characters i n the survey l i s t was randomized. What the informant actually saw. i s the character, the number assigned to i t , and, where provided by the compilers, such information as one-word d e f i n i t i o n s , part of speech, environments i n which the charac-ter may occur, and alternate s p e l l i n g , for the purpose of a s s i s t -ing the informant i n r e c a l l i n g the character and/or making the correct choice for a character with multiple pronunciations. A p a r a l l e l case i n English would be to i d e n t i f y "export" as a verb or noun i n order to e l i c i t the form with the stress pattern sought. The d i a l e c t survey l i s t was read by Chen and recorded on a r e e l - t o - r e e l tape recorder. The writer transcribed the data phonemically during the e l i c i t a t i o n and used the tape for double-- 10 -checking afterwards. Questions concerning some of the f i n e r phonetic d i s t i n c t i o n s were brought to the attention of M.D. Kin-kade i n the L i n g u i s t i c s Department at the University of B r i t i s h Columbia. The data c o l l e c t e d was then compared with the Cantonese 13 forms given i n Hashimoto (1972a) . A l i s t of characters was drawn for double-checking pronunciation. Each of these characters was accompanied by additional information to f a c i l i t a t e r e c a l l , or to avoid confusion with other characters which graphically appear quite s i m i l a r . Such information usually consists of d e f i n i t i o n s (in English or Chinese) and the combination of these characters with others to form p o l y s y l l a b i c words. Variant forms co l l e c t e d often r e f l e c t d i a l e c t a l influence from Cantonese, esp e c i a l l y i n those cases where the informant was uncertain of the pronunciation. It i s l i k e l y that some of the more l i t e r a r y words have received Cantonese pronunciation as a r e s u l t of contact with Cantonese speakers i n Hong Kong and Vancouver. This may account for some of the differences found between the present set of data and that obtained by Chao i n Sh i - q i . For the second set of data, Yang served as primary i n f o r -mant, with Chen joi n i n g i n on some of the occasions. The second task was stimulated by the observation of many gaps i n Chao's repertoire of Zhong-shan s y l l a b l e s , including d i s t i n c t i o n of tone. The aim was to f i n d c o l l o q u i a l Zhong-shan words to f i l l i n as many of these empty spaces as possible. The writer soon found that to simply ask whether such-and-such a s y l l a b l e exists i n Zhong-shan produced few responses. The next approach was for the writer her-s e l f to systematically go through each potential s y l l a b l e , - 11 -p a r t i c u l a r l y those l e f t blank i n Qiao's study, i n order to f i n d a word which would l a t e r be confirmed by Yang as a word spoken i n Shi - q i , i n both Shi-qi and Hong Kong, or s t r i c t l y as a Hong KOng colloquialism. The majority of the words suggested by the writer were i d e n t i f i e d by Yang as Shi-qi expressions. Yang was frequently able to elaborate on the meaning of a word suggested by the writer, and sometimes produced additional meanings or usages for the word or s y l l a b l e . These were necessarily informal sessions, conducted whenever the opportunity arose. Chen's mother, who i s now i n her seventies, understands Cantonese, but speaks only Zhong-shan. A few d i s t i n c t i v e expres-sions which she uses are also recorded for t h i s study. 0.3.3. Additional Sources for Zhong-shan Data As far as the writer i s aware, to date only three works have been published which contain f i e l d material on the Zhong-shan d i a l e c t . Of these, "Zhong-shan fang-yan" k % ~$ ('Zhong-shan dialect') (1948) by Y.R. Chao, and portions of The Lungtu Dialect (1956) by S. Egerod contain data on the Zhong-shan d i a l e c t spoken i n S h i - q i . A t h i r d source i s an a r t i c l e by J.D. B a l l i n 1897 e n t i t l e d "The Hong Shan or Macao d i a l e c t " , which describes the Zhong-shan d i a l e c t as spoken i n Macao, allegedly i d e n t i c a l to Shi-q i speech. Shi-qi forms which d i f f e r from those found i n Macao are recorded by B a l l i n his footnotes. A comparison with Chao's a r t i c l e reveals more differences between B a l l ' s Macao data and Chao's Shi-qi data than were reported by B a l l . At le a s t some of these differences may have resulted from sound changes during that i n t e r v a l between B a l l ' s c o l l e c t i o n - 12 -of data on the. Zhong-shan d i a l e c t and Chao's fieldwork i n Sh i - q i , a. period of at lea s t t h i r t y or forty years. B a l l ' s a r t i c l e i s primarily of h i s t o r i c a l s i g n i f i c a n c e ; i t s usefulness for compara-t i v e purposes i s rather l i m i t e d . Tone indicators, for instance, are omitted by B a l l i n his tran s c r i p t i o n s . Furthermore, lacking an international alphabet by which to transcribe with greater precision and conciseness the phonetic sounds of the d i a l e c t at that time, B a l l had to struggle with the inadequacies of the English, language and the limi t a t i o n s of the English alphabet. Consequently, he resorted at times to circumlocutory phrases to describe a p a r t i c u l a r sound. This i s es p e c i a l l y true when he attempted to explain the manner i n which some of the vowel clus t e r s 14 and diphthongs were a r t i c u l a t e d . B a l l ' s choice of format i n his syllabary Ci.e., repertoire of the s y l l a b l e s i n the dialect), also presents a weakness: by using Cantonese s y l l a b l e s to show Zhong-shan counterparts, the structure of Zhong-shan phonology i s obscured. Sounds which, contrasted h i s t o r i c a l l y and were preseryed i n Zhong-shan are not readi l y discerned i n B a l l ' s syllabary i f they merged i n Cantonese. In such, cases, only one of the sounds i n Zhong-shan •iwast presented i n the syllabary and the other merely entered i n the footnote as .exceptions:,-, Egerod's contribution rests primarily cm his f a i r l y detailed phonetic description of the Shi-qi d i a l e c t , as well as the provision of background information on the d i a l e c t and the region. The most valuable and most frequently c i t e d source on the Zhong-shan d i a l e c t i s Chao's a r t i c l e on the phonology of the Shi-q i d i a l e c t . Chao's material i s based mostly on data e l i c i t e d from one informant i n Shi-qi i n 1929.. Supplemental material was - 13 -gathered on another informant i n Hawaii ten years l a t e r . Although te r s e l y written, Chao's a r t i c l e i s a comprehensive work containing a description of the phonological system of Zhong-shan, a s y l l a -bary, a l i s t of some exceptions to the sound changes i n the modern d i a l e c t on the basis of h i s t o r i c a l phonological c l a s s i f i c a t i o n s , and a sample dialogue. Besides the aforementioned works, there i s also a Sino-Portuguese glossary compiled by Guang-ren Yin and Ru-lin Zhang c i r c a mid-eighteenth, century i n Macao which i s of h i s t o r i c a l i n t e r e s t . The glossary, consisting of 395 l e x i c a l items, i s part of the text Ao-meri J j - l u e f*\'$£j (-'Record of Macao') of which, several editions ex i s t today. Studies which have been con-ducted on the glossary include CR. Bawden (1954)., R.W. Thompson (.1959), W. Boltz (.1977), and M. Chan (forthcoming). - 14 -Notes to Chapter 0. .1. Also transcribed as "Chungshan" i n Wade-Giles romanization, and as "Hbng Shan" according to Cantonese pronunciation. 2. For a more detailed discussion of Mandarin morphological structure, see, for example, Kratochvil (.1968:55-88). 3. The h i s t o r i c a l information i s derived from various sources: Hsu, 1970:540; Zhao, 1955:135-136; Xie, 1933:100; and Scenery Publishers' "Zhong-shan qiao-xiang feng-guang" r}? ^ | ^ ('Scenic spots of Chungshan'—English t i t l e provided) (no date—c.1975?). 4. Much of the geographical and l i n g u i s t i c information presented i n t h i s section i s thanks to the description of the county by Egerod (1956:3). (Note that Egerod refers to Zhong-shan as a " d i s t r i c t " . ) 5. Unless noted otherwise, references to Egerod are to his book, The Lungtu Dialect (1956). For Egerod (1956), only page reference w i l l be given hereafter. 6. Hashimoto (1972a:8) notes that although Zhong-shan i s predo-minantly Yue-speaking, a quarter of i t s inhabitants speak a Hakka d i a l e c t while a number speak a Southern Min d i a l e c t . 7. For example, i n Zhao (1955:136) i t i s stated that Macao, while belonging to Xiang-shan county, was leased to Portugal i n the Sino-Portuguese treaty of 1888 i n the thirteenth year of Emperor Guang-xu )\j \M • (See also Bu, 1977:1; Hua-qiao Zhi, 1964:1). 8. A l l subsequent references to B a l l w i l l be to his 1897 a r t i c l e ; therefore, only page number w i l l be s p e c i f i e d hereafter. 9. Karl Lo, a native Zhong-shan speaker, received his elementary - 15 -education i n Macao during the 1940's, at which time the dia-l e c t of i n s t r u c t i o n i n his classes was Zhong-shan, not stan-dard Cantonese (Lo., personal communication) . The emergence of Cantonese as the main d i a l e c t must have been more or less contemporaneous with Egerod's fieldwork. 10. Hsu (1970:540) attributes poor a g r i c u l t u r a l output to i n f e r -t i l e s o i l , whereas the p i c t o r i a l magazine on "Zhong-shan qiao-xiang feng-guang" (n.d.:17) claims that "(t)here are several m i l l i o n mu of f e r t i l e a l l u v i a l sandy land i n Ch.ungsh.an county ... (b)ut i t s annual grain output reached only 200 to 250 k i l o s i n the past because of i t s backward agriculture and shabby water conservancy". (.'Mu1 i s a Chinese land-measure of area.) 11. Unless otherwise indicated, a l l reference to Chao pertains to his publication on Zhong-shan phonology (1948); therefore, only page number w i l l be provided i n subsequent references. 12. After the survey was conducted, a few words were added to the survey l i s t using l e t t e r s or additional numbers. This w i l l be discussed i n greater d e t a i l l a t e r . 13. Unless s p e c i f i e d otherwise, a l l subsequent c i t i n g s of Hashi-moto are of Hashimoto (1972a). For t h i s work on Cantonese which i s c i t e d throughout the present study for comparative purposes, only page number w i l l be given subsequently. 14. An excerpt from the description of the sound / o i / , as i n the word gai / k o i / 'ought', i s c i t e d here as an example: "An approximation of i t may be got by pronouncing the two Canton-:. ese sounds ko ( & ) and y i (_3» ) rapidly together. Having now got th i s sound, then s l i g h t l y open the l i p s further and - 16 -pronounce i t a l i t t l e more open, at the same time taking great care not to get too near to the open koi sound of the Canton-ese c£jj^ ) • In t h i s way and by l i s t e n i n g to a native from Macao pronouncing i t , i t i s possible to a r r i v e at the sound of the word "f$ x , koi i n Macao." (B a l l , p.509). (Underlining of the transcriptions mine.) - 17 -CHAPTER 1. PHONETIC DESCRIPTION AND PHONEMIC ANALYSIS The phonology of the Zhong-shan d i a l e c t w i l l be described according to an amalgamation of American s t r u c t u r a l i s t and Chinese ( t r a d i t i o n a l and modern) approaches. Since i t i s the l a t t e r approach which i s unfamiliar to most Western l i n g u i s t s , some of the terminologies and conventions based on Chinese a n a l y t i c a l categories w i l l be introduced below. F i r s t of a l l , the s y l l a b l e i s analyzed by Chinese l i n g u i s t s into two major parts: the i n i t i a l and the f i n a l (cf. Chart 1). The i n i t i a l i s simply the beginning consonant whose absence con-s t i t u t e s the "zero i n i t i a l " . The f i n a l consists of an optional medial, which i s a glide or v o c a l i c segment and i s ignored i n Chinese rhyming practices. The rhyming part - ( i . e . , the rhyme)•con-s i s t s of an obligatory s y l l a b i c segment, the nucleus, followed by an optional ending, which i s a glide, or a nasal or stop conson-ant. Also considered part of the f i n a l by some Chinese l i n g u i s t s i s the tone, which i s usually described i n terms of p i t c h contour and duration. Other l i n g u i s t s simply regard tone as a supraseg-mental feature and treat i t separately. Chart 1. S y l l a b l e Structure of Chinese. S Y L L A B L E FINAL RHYME INITIAL MEDIAL NUCLEUS ENDING - 18 -This analysis of the Chinese s y l l a b l e i n terms of an i n i t i a l - f i n a l dichotomy i s an inter p r e t a t i o n of a t r a d i t i o n which has i t s source i n a method of " s p e l l i n g " used since late Han, or second century A.D., c a l l e d fan-qie fg^ ^ jj , l i t e r a l l y , 'turning and c u t t i n g 1 . I t i s a means of obtaining the pronunciation of a character by using two other characters, the f i r s t of which bears the same i n i t i a l ( e s sentially the non-rhyming portion), and the second the same rhyme, or f i n a l . The character dong 'east', for example, i s defined phonetically by the gloss de and hong faj- : that i s , d(e) - (h) png = dong. The f an-qie method works well when the f i n a l of a word contains only the rhyme, but breaks down when the f i n a l includes a medial 1, which i s not consistently found i n the second character i n the fan-qie. While fan-qie i s involved i n grouping words which contain the same rhyme, the niu £.3- (.'knot') p r i n c i p l e i s used i n categori-zing words which have the same i n i t i a l ; that i s , words containing the same i n i t i a l which are grouped together belong to the same niu (Pulleyblank, 1977). Beyond the d i v i s i o n of a character into i n i t i a l and f i n a l , the further decomposition of the f i n a l into smaller component parts such as nucleus and ending i s the i n f l u -ence of Western approach, which analyzes a s y l l a b l e i n terms of a li n e a r sequence of sounds arranged according to t h e i r temporal pos i t i o n i n a spoken chain. The canonical shape of the s y l l a b l e i n Zhong-shan (Chart 2) can be analyzed using the terms introduced above. The i n t i a l i n Zhong-shan i s a consonant which, when omitted, constitutes the "zero" i n i t i a l , and i s assigned the phoneme /0/ ( i . e . , the n u l l element, for convenience of reference; the zero i n i t i a l i s omitted - 19 -Chart 2. Syllable Structure of Zhong-shan. T (I) + (M) + N + (E) Examples: £T r 55, L t e a i IJ J / t + i + a: + 55 ng /= 55 /tia:ng / ' n a i l Lkwax J / k + w + a + _ 5 5 /= /kwa 5 5/ 'melon' - U ? r 5 1 i IWAin J / w + — + a: + n 5 1 / = /wa:n 5 1/ 'yet' [ J A i k 2 ] / j + — + a: + k 2 /= / j a : k 2 / 'eat' t [ixp 2J / — + — + i + 2 P /= / i p 2 / ' l e a f 3- In J / — + — + ng + _ 1 3 /= / n g 1 3 / ' f i v e ' where T = Tone I = I n i t i a l (Consonant or "zero" i n i t i a l ) M = Medial . ( / i / or /w/) N = Nucleus (Vowel or s y l l a b i c nasal) E = Ending (Glide, stop, or nasal) i n actual t r a n s c r i p t i o n s ) . With regard to the f i n a l , the f i r s t element i s the medial, which i s an optional segment. Of central importance i s the nucleus, which i s a vowel or a s y l l a b i c nasal i n Zhong-shan. The ending i s optional, consisting of a gl i d e , stop or nasal consonant. The tone, indicated by the use of superscripts, following the s y l l a b l e , i s treated as a suprasegmental feature over the entire s y l l a b l e . With regard to conventions for indi c a t i n g tone, Y.R. Chao has devised two systems, one using tone l e t t e r s and the other .... - 20 numerals. Both systems continue to be used by Chinese l i n g u i s t s . In the f i r s t , tone l e t t e r s graphically represent the p i t c h height and the contour of a tone. The tone l e t t e r s for the various l e v e l tones, for instance, are "1 1 -\ -I J", with a gradual descent i n pitch height represented by the gradual lowering of the horizontal stroke. Tonal length i s also graphically represented. A long . l e v e l tone, for example, i s indicated by a comparatively longer horizontal stroke than a short l e v e l tone (e.g., "1 versus 1 ) . Short l e v e l tones are generally used for s y l l a b l e s ending i n a stop consonant. The second system for tonal t r a n s c r i p t i o n i s a numerical one which uses the scale of "1" to "5" wherein "1" indicates the lowest pitch and "5" the highest. Thus, a high l e v e l tone may be represented by "55", for instance, and a high f a l l i n g tone by "51", and so forth. A single number i s used for denoting short s y l l a -bles, e s p e c i a l l y relevant to the t r a n s c r i p t i o n of s y l l a b l e s with stop consonant endings. For typographical and comparative reasons, i t i s t h i s l a t t e r system of tone notation which w i l l be employed i n the present study. Transcription of consonants i s b a s i c a l l y i n accordance with the International Phonetic Alphabet (IPA), with a few modifi-cations. Aspiration, for example, i s indicated by the d i a c r i t i c ['] (e.g., [p'] instead of IPA [phj) . The segment I t s J represents an a f f r i c a t e rather than a sequence of two segments; I t s ' ] i s the aspirated counterpart. - 2 1 -1 . 1 . I n i t i a l s The consonants i n Table 1 represent the inventory of i n i t i a l consonants i n broad phonetic t r a n s c r i p t i o n . Those con-, sonants accompanied by a dash. "-" occur i n s y l l a b l e - i n i t i a l p o s ition only. Table 1 . The I n i t i a l s i n Zhong-shan. Plosive Unasp*. Asp'd. Nasal Lateral F r i c a t i v e Glide B i l a b i a l p p i - m w Labiodental f -Dental t t'~ n 1-Alveolar t s - t s ' - j Velar k k'- q G l o t t a l (?) h-The broad phonetic t r a n s c r i p t i o n i n Table 1 above i s e s s e n t i a l l y i d e n t i c a l to a phonemic inventory that can be arrived at by applying the p r i n c i p l e s of complementary d i s t r i b u t i o n and phonetic s i m i l a r i t y . Thus, Table 1 also serves as the phonemic system of Zhong-shan i n i t i a l s . For typographical reasons, /ng/ w i l l be used to represent the velar nasal [q] i n the phonemic system and should be considered a unit phoneme. Likewise, / t s / and / t s ' / constitute unit phonemes. In the following paragraphs i s a more detailed discussion of some of the phonemes and t h e i r allophones. The phonemes /p/, / t / , /ts/.and /k/ are unaspirated, and generally quite weak, voiceless plosives, often perceived phonetically as [bj , [dj , {d4z] - 22 -and [g], and sometimes even with s l i g h t voicing. Note that t h i s observation d i f f e r s from that made by Chao (p.51) who asserts that, unlike Mandarin, the di a l e c t s of Cantonese, Zhong-shan and Wu are a l i k e i n having strong, unaspirated stops i n s y l l a b l e -i n i t i a l p o s i t i o n . There i s strong aspiration i n the aspirated s e r i e s , p a r t i c u l a r l y apparent i n the careful enunciation of in d i v i d u a l characters containing them. Some lip-rounding accompanies the i n i t i a l s when they occur before the high rounded vowel [u:J (e.g., ^ [p w, p < W / t s W , k w, h W ] ) . I t was further observed that s l i g h t a f f r i c a t i o n may it 5 22 occur, as i n the word bu tf? 'step', phonetically [p ux. J . There i s some disagreement among l i n g u i s t s concerning the nature of the s i b i l a n t s i n Zhong-shan. Chao (p.51) feels that there i s c l e a r l y no p a l a t a l i z a t i o n i n the Shi-qi d i a l e c t , i n contrast to his observation of strong p a l a t a l i z a t i o n i n Cantonese. Egerod (p.13), on the other hand, posits a p a l a t a l series for both Cantonese and Sh i - q i , with the phonemes /c/, /ch/ and /s/ 3 s' representing the phonetic segments [c J, [c J and [s] respectively. B a l l ' s analysis may constitute a compromise i n that complementary d i s t r i b u t i o n can perhaps be in f e r r e d from his t r a n s c r i p t i o n s : i n general, B a l l ' s / t s / occurs with back vowels while his /ch/ occurs with front ones, which suggests that the p a l a t a l /ch/ i s the 2 res u l t of assimilation to the following front vowel. . ' ..It was observed i n the writer's data that Chen's pronun-c i a t i o n tends to be s l i g h t l y more p a l a t a l than Yang's, but i s nonetheless not as strongly p a l a t a l i z e d as found among some speak-7 ers of Cantonese. In her study of Cantonese, Hashimoto (p.88) 23 -describes the s i b i l a n t s i n her i d i o l e c t as dental i n a r t i c u l a -t i o n , but adds that the place of a r t i c u l a t i o n of t h i s series of sound, ranges from dental to p a l a t a l among Cantonese speakers, with a tendency for some degree of p a l a t a l i z a t i o n preceding the high front vowels. Hashimoto (p.l7,fn.8) further c i t e s an obser-vation made by D.C. Lau at the University of London who found that there i s a greater tendency f o r male speakers than female speakers to p a l a t a l i z e . This would concur with our observations on Chen and Yang's speech with regard to the r e l a t i v e degree of p a l a t a l i z a t i o n of the s i b i l a n t s e r i e s . As noted by both Chao and Egerod, the i n i t i a l nasal con-sonants are often accompanied by homorganic stops: [m3, n^, n~^ ] . However, i n the case of a non-labial nasal followed by the high, back vowel [u:], i t was found that the homorganic stop may be replaced by l a b i a l [b] i n a n t i c i p a t i o n of the following l a b i a l segment, as i n one of the repetitions i n the pronunciation of the word nu Ijp 'anger', which was rendered [n u: ] by Chen. At times, instead of the nasal consonant being accompanied by a homorganic stop, the i n i t i a l segment becomes i n fact a prenasalized stop, as i n Chen's pronunciation of the word men 'door'. I t 51 i s [mum J i n broad phonetic t r a n s c r i p t i o n ; i n narrow phonetic t r a n s c r i p t i o n , however, the word should be recorded as { mbu:n^ ij. The alternation between /n/ and /!/ observed i n the speech of a number of Hong Kong speakers i s not found i n Zhong-shan, which maintains a c l e a r d i s t i n c t i o n between these two i n i t i a l s . None of the words which had the h i s t o r i c a l Ni yf<J 3 (*n-) i n i t i a l are pro-nounced with the l a t e r a l [11 i n Zhong-shan. There are only three cases i n the survey of an In] corresponding to the L a i - 24 -13 i n i t i a l : (.1) the character l i a n g fa which i s pronounced [n0ot) ] 13 with the meaning of 'two (173-12) and [loan ] when i t i s used as a measure word to mean 'a t a e l ' (173-13); (.2) the character l i n g 4f| C201-7) for ' c o l l a r ' and 'to apply for which has the l i t e r a r y 13 13 reading of I l i n ] , while a c o l l o q u i a l reading of [neairj ] only 13 applies to the f i r s t meaning, and that of [leain J only to the second meaning; and (3) the character l i (1.13-13) which i s a c l a s s i f i e r for small, round objects such as seeds, grain, pearls, 5 buttons, etc., and i s pronounced InBp J. Although any explanation of why these three characters have acquired a nasal i n i t i a l can only be speculative, i t i s possible that the /n/ i s a r e s u l t of borrowing i n a l l three cases. Two of the three words, liang and l i n g , can be found i n the very useful c r o s s - d i a l e c t a l reference source, the Han-yu Fang-yan Zi-hui y& 7f "if % (.'Chinese Dialect S y l l a -bary', editted by the B e i j i n g Daxue Zhong-guo Yu-yan Wen-xue-xi Yu-yan-xue Jiao-yan-shi (.1962)—hereafter referred to as "Zi-hui" for short, with page reference included only). There i s only one entry for the word l i a n g . Since the Zi-hui does not provide glosses, one can only suppose that the two meanings of the word, 'two' and ' t a e l ' , are combined i n the single entry i n that s y l l a -bary. While most of the d i a l e c t s have the l a t e r a l 111 as the r e f l e x of the h i s t o r i c a l Lai i n i t i a l , a few of the d i a l e c t s represented therein show an [n] i n i t i a l . In Amoy (a Southern Min dialect), two forms contain an In] i n i t i a l , and one an 11J i n i t i a l , p a r t i a l l y r e f l e c t i n g s t y l i s t i c differences: .[lion] i s . l i t e r a r y 4 whereas [niu]. and Inp] are c o l l o q u i a l readings. However, [n] and [1] are i n complementary d i s t r i b u t i o n i n Amoy, with In], occurring - 25 -5 before nasalized vowels and [q] and [1] elsewhere. Although there may be contact between Zhong-shan and Amoy, or some other si m i l a r Southern Min d i a l e c t , stronger evidence for possible d i a l e c t a l influence comes from Fu-zhou, a Northeastern Min d i a l e c t (data provided by J. Norman). Not a l l Fu-zhou speakers maintain a phonemic d i s t i n c t i o n between /n/ and / l / ; however, among those who do, 'two' i s [naq 2] , while ' t a e l ' i s I - l i o r ) ] . ^ The /n/-~> / l / s p l i t i n Fu-zhou not only serves the same function as that found i n Zhong-shan, but the same assignment of i n i t i a l to meaning occurs i n the two d i a l e c t s . There are no data on the Long-du d i a l e c t ("Lungtu" i n Wade-Giles romanization), which i s spoken i n the second qu i n Zhong-shan, and i s regarded by both. J. Norman and N. Bodman as a Northeastern Min d i a l e c t . Neverthe-les s , one would suspect that Zhong-shan borrowed the / n / ~ / l / s p l i t for the two meanings of the word l i a n g xfc) from Long-du—if such a d i s t i n c t i o n occurs i n that d i a l e c t — o r from Fu-zhou or another Northeastern d i a l e c t where the same d i s t i n c t i o n i s found. It might also be noted that although Hashimoto (p.570) records a l a t e r a l i n i t i a l f o r both meanings of the character l i a n g 1$) , whein i t i s useid to mean ' two1 , the character i s pronounced i n the Yang-shang tone 1 2 4 ] , while i t s use to mean 'tae l ' i s rendered in the Yin-shang tone I 3 5 J . Hashimoto Cp-668) suggests that since lj.ang as a measure word i s a c o l l o q u i a l term, i t s Yin-shang tone may be the product of either changed tone phenomenon ( i . e . , an o r i g i n a l tone changed to a high r i s i n g I 3 5 J ) , or influence from the c o l l o q u i a l layer. The evidence thus far demonstrates, ..firstly,. that the two meanings of the character l i a n g do diverge i n t h e i r suscepti-- 26 -b i l i t y to external influences, with. Zhong-shan and Fu-zhou show-ing an alternation of i n i t i a l s , while Cantonese exhibits a tonal d i f f e r e n t i a t i o n . Secondly, and more importantly with regard to Zhong-shan, given the paucity of exceptions i n the modern r e f l e x of the Lai i n i t i a l i n Zhong-shan and the evidence from Fu-zhou, i t would be l o g i c a l to suggest that the reading of l i a n g which has the /n/ i n i t i a l i n Zhong-shan i s l i k e l y to have been the one borrowed into the d i a l e c t , whereas the one with the / l / i n i t i a l i s the native form. A p a r a l l e l case may be argued for the character l i n g , which has both an / l / and an /n/ i n i t i a l i n the c o l l o q u i a l reading 13 13 i n Zhong-shan: the form [leanj J means 'to apply for' and [neain J ' c o l l a r ' . Again, there are several d i a l e c t s i n the Zi-hui (p.255) which show an [n] i n i t i a l for the character l i n g . Of the southern Chinese d i a l e c t s recorded i n the Zi-hui, only Chao-zhou (Southern 52 Min) shows an [n] i n i t i a l for the character: [nia J (with no other form given). In general, the / n / ^ / l / d i s t i n c t i o n i s not very well maintained i n Southern Min d i a l e c t s . I t i s noted that such a d i s -t i n c t i o n i s likewise not always maintained i n Northeastern Min. It i s therefore conceivable that the nasal i n i t i a l for the c o l l o -q u i a l reading of ' c o l l a r ' i s a borrowing from one of the Min dia-l e c t s . Chao's data on the Zhong-shan d i a l e c t show a l a t e r a l .. i n i t i a l for the c o l l o q u i a l reading of the character l i n g 4%. r which 13 7 i s transcribed by Chao phonemically as / l i a : n g /. I t i s s i g n i f i -cant that Chao leaves the s l o t for the potential s y l l a b l e / n i a : n g x 3 / i n his syllabary empty, which quite strongly suggests that the 13 pronunciation of / l i a : n g / for the character l i n g encompasses - 27 -both, the meanings of ' c o l l a r 1 and 'to apply f o r ' . The present data on Shi - q i , which y i e l d s a nasal form, would lead to the proposal that the borrowing may be a recent one. In contrast, the character l i a n g xfi) i n Chao's data agrees with the present corpus i n haying both a l a t e r a l and a nasal i n i t i a l , which probably r e f l e c t s an e a r l i e r borrowing. The character l i %.JL [n^p^j , which i s a measure word f o r small, round objects, only has the /n/ i n i t i a l i n Zhong-shan. I t i s l i k e l y that t h i s i s also a borrowed form, although the writer has no concrete c r o s s - d i a l e c t a l data on which to support t h i s claim. The general observations made above on the Min di a l e c t s and the r a r i t y of /n/ as a r e f l e x of the Lai i n i t i a l i n Zhong-shan would suggest a sim i l a r borrowing i n t h i s t h i r d and l a s t case of exceptions to the Lai i n i t i a l i n the Zhong-shan data. Moreover, although Hashimoto (p.513) records [lsp ] for the character, i t i s noteworthy that Huang (1970:425)., for example, posits only an /n/ i n i t i a l for the same character, carrying the meaning of 'measure of seeds, grains, etc.' i n his Cantonese Dictionary. The word does not occur elsewhere i n the dictionary, or more precisely, i t does not appear i n the section under / l / . (Huang transcribes the word as nap, i n Yale romanization.) S.L. Wong (.1954)_ likewise records the word L i under the s y l l a b l e [nsp] for Cantonese. However, Wong does note that the character i s also read s i m i l a r to that of the h i s t o r i c a l l y homophonous word, l i j£ . 5 (.113-12) ; that i s , l i has an alternate pronunciation of [ l B p J. One can, however, assume that the reading of the word l i %JL with the /n/ i n i t i a l i n Cantonese i s the primary one i n Wong's Chinese syllabary based on the pronunciation i n Canton. Given Wong's decision to record l i . %JL. only, under .the /n/ i n i t i a l category, and Huang's recording of the word only with an /n/ i n i t i a l , i t appears that not only Zhong-shan, but Cantonese speakers as well, who do not normally "confuse" /n/ and /!/ i n i t i a l s would neverthe-less pronounce the word l i . JJC^L with a nasal i n i t i a l . The i n i t i a l /h/ i s a g l o t t a l f r i c a t i v e . Chao (p. 51). how-ever states that when /h/ precedes the back vowels /u/ and /o/, there i s some velar or uvular f r i c a t i o n , r e s u l t i n g i n a sound which i s almost .[;xj . This allophonic v a r i a t i o n was not observed i n the speech, of Chen and Yang to any s i g n i f i c a n t degree. g The zero i n i t i a l /#/ , according to Chao (p.51), i s a smooth onset before high vowels, but i s usually a g l o t t a l stop [jj before mid and low vowels. This allophonic d i s t r i b u t i o n of the zero i n i t i a l was not consistently observed i n the present data. The zero i n i t i a l occurs before f i n a l s beginning with a nuclear vowel, as well as before s y l l a b i c nasals. In the representation of a s y l l a b l e containing a zero i n i t i a l , the n u l l symbol "ft" i s actually never transcribed. Egerod records a phonemic g l o t t a l stop /?/ rather than a zero i n i t i a l , which he posits before vowels and s y l l a b i c nasals. Egerod makes no note of allophonic v a r i a t i o n s . Before discussing the semi-vowel i n i t i a l s , i t should be made clear that the phonemic analysis of both Egerod and the writer distinguishes between vowels and glides and, moreover, allows the l a t t e r to serve as i n i t i a l s . Chao, on the other hand, does not d i f f e r e n t i a t e glides from vowels i n his analysis. He treats [j] as an allophone of the phoneme/i/, and [wj as that of /u/. His rationale i s founded on the s t r i c t application of the p r i n c i p l e of complementary d i s t r i b u t i o n ; since whether the segment - 29. T / i / , for instance, i s the vowel {ij or the semi-vowel {jj can be predicted on the basis of i t s po s i t i o n i n the s y l l a b l e and/or what rhyme or f i n a l accompanies i t , Chao combines the front vowel and p a l a t a l glide under the phoneme / i / . The corresponding back vowel [u] and the l a b i a l glide {wj are incorporated under the phoneme /u/ on s i m i l a r grounds. Since vowels and glides are i n complementary d i s t r i b u t i o n i n Zhong-shan, contrasting them phonemically would inevitably create redundancies i n the system. Egerod's d i s t i n c t i o n of glides and vowels i s probably prompted by the desire to set up the vowel clus t e r /ua/, which he records phonetically as [ox>] or [ D : J . Set-ti n g up the phonemes /w/ and /u/ allows such minimal pairs as /kwan/ and /kuan/, which are phonetically [kwerij and [kom] respec-t i v e l y (using our phonetic t r a n s c r i p t i o n s ) . Although the present analysis does not have such an objective i n mind, i t i s i n the treatment of medials that the recognition of vowels versus glides becomes a p a r t i c u l a r l y important issue, as we s h a l l see l a t e r . Chao also d i f f e r s from Egerod and the writer i n that he does not treat the glides as i n i t i a l segments. They are not d i s -cussed i n the section on i n i t i a l s . Chao (pp.51,53) refers to them e x p l i c i t l y as medials ( j i e - y i n ^ 'medial sounds'). The > / 22 phoneme /u/ i n the word hua /ua / (in Chao's t r a n s c r i p t i o n , 22 phonetically [WA: ] ) , i s analyzed by Chao as a medial, whereas both Egerod and the writer would analyze that segment as an i n i t i a l . Although on the whole, there i s h i s t o r i c a l basis for analyzing a glide i n such a position i n the modern d i a l e c t as a medial pre-ceded by the zero i n i t i a l , i t i s nonetheless a very costly step i n that i t would substantially increase the number of f i n a l s . - 30 -At the same time, many of these f i n a l s would actually have a very low functional load. F i n a l s such as IvoinT., IwDikJ, [jun] and I q 0 q j , for example, would only occur with the zero i n i t i a l . Chao solves the problem by putting the two glides (his / i / and /u/ phonemically) i n brackets i n his chart of i n i t i a l s and, in his syllabary, he posits these two medials under what i s i n : fac t the category of i n i t i a l s . Thus, although Chao analyses them as medials, he i s actually treating them as i n i t i a l s without actually acknowledging i t . Only i n the case of his f i n a l s / i a / , /ia:ng/ and /ia:k/ does Chao treat / i / as a medial i n his syllabary, the reason being that in these f i n a l s , medial / i / does co-occur with, most of the i n i t i a l s and hence carry a heavy functional load. However, by allowing' / i / to occupy both i n i t i a l and medial po s i t i o n i n his syllabary, Chao also creates c e r t a i n redundancies; a s y l l a -ble such as * / i i a / , for example, would not be possible i n the J d i a l e c t , but i t could be generated from the combination of " i n i t i a l + f i n a l " i n h i s syllabary. In the present analysis, a si m i l a r problem i s avoided by stating the following r e s t r i c t i o n : the two medials, /w/ and / i / , never co-occur with the zero i n i t i a l . (The two semi-vowel i n i t i a l s are /w/ and / j / . ) . According to t h i s co-occurrence r e s t r i c t i o n , the phoneme /w/ i n the s y l l a b l e /kwa/, fo r example, i s analyzed as a medial whereas, i t i s an i n i t i a l i n the s y l l a b l e /wa/. The phoneme / i / i s a medial i n the s y l l a b l e /kia:k/, for instance, while / j / i n the s y l l a b l e such as /ja:k/ i s an i n i t i a l . (The phoneme / i / can only occur s y l l a b l e - i n i t i a l l y when i t i s the nuclear yowel Ii,:].) The reasons f o r choosing an asymmetrical pair of phonemes (/w/ and / i / ) . for the medials w i l l be discussed - 31 -l a t e r . Returning now to the discussion of glide i n i t i a l s , /w/ i s simply the back, rounded semi-vowel [wj, while / j / i s usually the front glide I j ] . Before the front rounded vowel [ 0 ] , however, the glide assimilates to the rounding of the following segment, and thus becomes the front, rounded semi-vowel [qj. The only occur-rence of t h i s regressive assimilation i n the present data i s i n the s y l l a b l e / j 0 j / , phonetically [i[0uj , with the nuclear vowel af f e c t i n g both the preceding and the following segment. With regard to a comparison of the i n i t i a l s i n Zhong-shan and Cantonese, i n general, there i s very l i t t l e difference between the two d i a l e c t s . One observation discussed e a r l i e r was the d i f f e r -ence i n degree of p a l a t a l i z a t i o n present i n t h e ^ s i b i l a n t s i n the two d i a l e c t s . Another, concerns the nasal i n i t i a l s . Unlike Zhong-shan and some of the other Yue d i a l e c t s (e.g., S i - y i d i a l e c t s ) , Canton^ ese does not have homorganic stops accompanying the a r t i c u l a t i o n of nasals. Another difference that i s found i n Cantonese i s the presence of homorganic glides preceding the high vowels. Thus, the Zhong-shan s y l l a b l e [u:n] would be [wu:n] i n Cantonese. Like-wise, Zhong-shan [ix] corresponds to Cantonese [ j i x ] , and Zhong-shan [yx] i s Cantonese [qyx]. Moreover, Cantonese i s often treated as containing a phonological contrast between a p l a i n versus a l a b i a l i z e d velar i n i t i a l ( i . e . , /k/ versus /kw/, and /k 1/ versus /k'w/). The l a b i a l i z e d counterparts are treated i n Zhong-shan as a sequence of a velar i n i t i a l plus a l a b i a l medial to correspond to the non-labial medial / i / . The absence of a medial / i / i n Cantonese, combined with a heavier functional load of the l a b i a l -ized i n i t i a l s i n Cantonese make the option of treating /kw/ and - 32 -/k'w/ as i n i t i a l segments a p r a c t i c a l one for Cantonese. 1.2. Finals. A broad phonetic t r a n s c r i p t i o n of the f i n a l s i n Zhong-shan, together with a phonemic analysis, i s presented i n Table 2 (a) on page 35. In the table, the phonemic renditions of the medials and s y l l a b l e nuclei are placed i n the left-most column, while the endings are recorded across the very top. J u s t i f i c a t i o n for the pa r t i c u l a r phonemic solution proposed here w i l l be detailed l a t e r . In the meantime, for the purpose of comparison, Chao and Egerod's treatment of the f i n a l s are re-interpreted i n Tables 2 (b). and (c) i n order to f i t into the format of Table 2 C a ) . The f i n a l s i n B a l l ' s study are given i n Table 2 (d). For cross-dia-l e c t a l comparison with Cantonese, the f i n a l s i n Cantonese, based on Hashimoto (p.90), are placed i n Table 2 (e). To conform with, the format established here for Zhong-shan, the l a b i a l feature of the l a b i a l i z e d velar i n i t i a l s i s treated as a medial i n the table. Such a treatment w i l l also prove useful i n the l a t e r comparison of h i s t o r i c a l reflexes of various rhyme groups. Egerod's data cause some d i f f i c u l t y since he did not provide a syllabary for Shi-qi and his corpus of data on the d i a l e c t i s extremely limited. I t w i l l be assumed here that the f i n a l s found i n Chao's study are also present i n Egerod's. More-over, some segments which Egerod has analyzed as "vowel c l u s t e r s " have been redi s t r i b u t e d i n Table 2 (_c). : his /aa/ c l u s t e r i s regarded as a simple "vowel nucleus", while the clusters / i a / , /0a/ and /ua/ as well as the sequences /wa/ and /waa/ (_with /w/ o r i g i n a l l y part of the velar consonant c l u s t e r ) , are considered a combination of - 33 -"media l + vowel n u c l e u s " , based on h i s t o r i c a l and c r o s s - d i a l e c t a l d a t a . I t i s more d i f f i c u l t to know the p r e c i s e phonet i c va lue J . D . B a l l in tended i n h i s t r a n s c r i p t i o n of Macao Zhong-shan. In some i n s t a n c e s , there i s an attempt to t r a n s c r i b e the words p h o n e t i c a l l y : f o r example, h i s vowels / i / and / ! / correspond to [ i ] and [ i f ] r e s p e c t i v e l y i n modern ( S h i - q i ) Zhong-shan. However, /'£/ i s a l s o used f o r a sound which i s now the f r o n t g l i d e [ J J . In o ther s i t u a t i o n s , i t ^ a p p e a r s to be s imply an o v e r - d i f f e r e n t i a -t i o n of forms. The i n i t i a l s / t s / and / c h / , mentioned e a r l i e r , are t r a n s c r i b e d as though they may be c o n t r a s t i v e i n the d i a l e c t . I t i s l i k e l y t h a t such a c o n t r a s t between a d e n t a l and an a l v e o l a r (or p a l a t a l ) s e r i e s — a l s o found i n a number of e a r l i e r works on Cantonese ( i n c l u d i n g those by B a l l ) — i s i n f a c t an a r t i f i c i a l c r e a t i o n which became q u i t e w ide ly accepted s i n c e i t had f u n c t i o n a l v a l u e . The o v e r - d i f f e r e n t i a t i o n i s based on a d i s t i n c t i o n found i n the n a t i o n a l s tandard wherein the d e n t a l s i b i l a n t s i n Cantonese g e n e r a l l y correspond to the d e n t a l / p a l a t a l s e r i e s i n Mandar in , and the Cantonese a l v e o l a r s (or p a l a t a l s , as the case may be) t h a t of the r e t r o f l e x s e r i e s i n the Mandarin d i a l e c t . I t i s p r i m a r i l y f o r the purpose of a i d i n g Cantonese speakers i n l e a r n i n g the n a t i o n a l d i a l e c t t h a t the two s e r i e s had o f t e n been kept apar t i n Cantonese, even though such a d i s t i n c t i o n i s founded on pedagog ica l r a t h e r than l i n g u i s t i c grounds (see, f o r example, Chao (.1947:18-19); Yuan (.1960:183), e t c . ) . The f a c t t h a t a s e r i e s of d e n t a l versus p a l a t a l s i b i l a n t s i s phonemic i n Shun-de, which i s one of the Yue d i a l e c t s but i s by no means r e p r e s e n t a t i v e of the d i a l e c t - 34 -family as a whole, may have contributed to the o r i g i n a l d i f f e r e n -t i a t i o n recorded i n e a r l i e r writings on Cantonese. Some of B a l l ' s f i n a l s also appear to be overly d i f f e r e n t i -ated; for example, B a l l transcribes the f i n a l s of the three words ming ^ (.200.-2) 'name', j i n g fcfa (200-3) 'quick-witted', and l i n g ^ (201-8) ' h i l l ' i n three d i f f e r e n t ways, as shown i n (.1) below. The f i n a l s of a l l three words are pronounced £ea:y] i n the c o l l o q u i a l reading i n Zhong-shan. H i s t o r i c a l l y , these f i n a l s not only belong to the same rhyme group ( G e n g ^ ^ ) ^ , but also to the same rhyme (Qing ^ ) and the same grade ( I I I ) 1 ( \ There i s no reason to expect that these f i n a l s recorded by B a l l should have re f l e c t e d a contrast i n t h e i r pronunciation s u f f i c i e n t to warrant three d i f f e r e n t means of t r a n s c r i p t i o n . (1) ..Ball Chan .Gloss ^ /meeng/ /mia:ng 5 1/ ([mea:n 5 x]) 'name' 55 ,,rj_ 55-L 5 /tsieng/ /tsiasng /([tseaxq ]) 'quick-witted' ^ /leng/ / l i a : n g 1 3 / ( [ l e a i n 1 3 ] ) ' h i l l ' In the recording of B a l l ' s f i n a l s i n Table 2 (d), some of the d i a c r i t i c a l marks that he used i n his a r t i c l e are not very d i s t i n c t i n the microfiche of his a r t i c l e from which a photocopy was made for t h i s study. As a r e s u l t , i t i s possible that the writer may have recorded a circumflex {*) where i t should have been an umlaut ("), a grave accent (>) instead of a circumflex, etc. An attempt i s nevertheless made to be consistent. F i n a l s /ing/ and / i k / are found i n Shi-qi speech only. Macao and the surrounding areas use /ang/ and /ak/ instead (phonetically {BIJJ and Ink] i n modern speech). - 35 -T a b l e 2 ( a ) . A n a l y s i s o f t h e F i n a l s — C h a n . \ e n d i n g voweTS. nucleuses - j -w -m -n -ng "P - t - k i i i i i w i im i i n if) i i p i i t i k y y * y i n y i t u u : ux j u i n un U I t uk 03 I- 0U_ 0n 0 t o 3 X 31 j ow o :n o i n D i n D i p D l t D i k a BJ e n BQ Bp B t Bk a: A : A I j A I W Aim A i n A i n A I p A l t A i k 1 medial • vowel nucleus i o 0 31} 0Dk i a : ea.i- e a: w eaim eai.y ea,i;p ea i t ea i k w i w i k wa WBJ wBn WBQ WBt WBk wa: WA : WA I j WA i n WA i g WA I t WA i k s y l l a b i c nasal m ? - 36 -Table 2 (b). Analysis of the Finals—Chao. N e n d i n g v o w e r v n u c l e u s ^ - i -u -m -n -P ~ t - k i i i i i w i i m i i n e q i i p i i t e k y y x y i n y i t u u x u : j u i n U l t CE 031 n 031 q 03 i k O e y o w eri o n et o k o: D i =>* j D : m D i n D i n D i p D i t D i k a A W •em Bn Bn Bp B t B k a: A I A X j A : w A i m A i n A i n A I p A I t A i k m e d i a l * v o w e l n u c l e u s i a : i A : i a i i ) i a i k wa WB j w e n WBt wa: WA I WA I j W A i n WAin WA I t WA i k s y l l a b i c n a s a l m J ] * - 37 -Table 2 (c) . Analysis of the FihaTs--KEgerod. sending vowePv nucleus^ - j -w -m -n -q -P.. - t - k i i : i i w iim i i n eq ixp ixt ek y V- y xn y i t u u X ux j u xn OJ} u I t ok 0 oey oen (Bt a a j aw am an aq ap at ak aa a i axj axw aim a in a xq a xp a i t a i k medial+ vowel nuc] Leus i a ea eaq e ak 0a (BO oeoq oeok ua oo oon Doq DOt DOk wa wa j wan waq wat waa wa i wax j wain wax q wa i t wa i k syllabic nasal m i r) - 38 -Table 2 Cd) . Transcription of Finals i n (Macao) Zhong-shan-- B a l l """-^ ^ ending vower\^ nucleus ~ ~2- -u - i i -m -n -ng -p - t -k * [ i ] ( i ) , (ing) . (ik) [ i i i • . 1 : . - s X / s / 1 U l f f l xn xp x^ t [u] u ung uk [yx] ii ii iin ut [ U l ] u • u un lit [ U l j uu / / U U l [ 0,ux] uo uou [oex ,ceo] 6 6 oii ong ok loi] 0 0 on ong ot ok I D : ] 6" A f O l [owj V 0 b [ B ] a • a i au am an ang ap at ak [ A X ] • a • // a ax au am an ang ap at ak medial+ vowel nucleus I e a x ] e e eng ek lea x J i e xeng xek [ e a x J ye yeu yem yep [ eax] A ee eeng leaxj ya (ya) yam syllabic nasal m Modern Zhong-shan equivalents are recorded on the left-most column. Zhong-shan [q] = Ball's /ung/. > - 39 -T a b l e 2 (e) . Phohe t i c T r a n s c r i p t i o n o f t h e F i n a l s i n C a n t o n e s e . Nending vowelv n u c l e u s \ y \J v - i - y - u . -m - n - 0 - p - t - k i i n i k i : i x i : u i x m i xn i x-p i x t y* V- y xn y i t u UI) uk UI U I Y U X 1 uxn UX-t e e i e xn e xk e V e i 0 0y 0n 0 t cs ce: cexrj ce x k 0 ou D 1 D X V 3X1 o xn O I t ) o x t oxk B B l BU Bm Bn Bl) Bp B t Bk A I A : A l l AXU A i m A xn AXt) AXp A X t AXk medial+ vowel nuc! .eus W I w i n w i k wi x wi x W D A wo x. wo x q WD xk WB WB1 WBn WBq WBt WBk WA I WA X WA X 1 WA xn WA X 0 W A I t WAlk s y l l a b i c nasal m 1 0 • - 40 -1.2.1. Nuclear Vowels Of the f i n a l s , the nuclear vowels w i l l be discussed f i r s t . In the present analysis of Zhong-shan, they form a t h r e e - t i e r , six-vowel system with a phonemic contrast of length i n the low vowels only, as diagrammed i n Chart 3 (a). It may be observed that, generally, tense and lax vowels—or long and short vowels— sharing the same vowel height (high, mid, low) serve as allophones of the same phoneme. The chart also includes an analysis of the combination of "medial + vowel". Comparative charts of Chao and Egerod's vowel system for Zhong-shan are shown i n Charts 3 (b) and (c) respectively. As i n Table 2, part of t h e i r phonemic system i s re-interpreted i n terms of "medial + vowel" for which Chart 3 Ca) serves as the model. Chart 3 (a). Analysis of the Vowels—Chan. Phonetic Representation: I i i u: u 0 ce: rzi Phonemic System: i y u 0 o a/a: A : Medial + Vowel: tea:! = / i a / [03] = / i o / IW B3 = /wa/ [WAX.]' = /wa:/ - 41 -Chart 3 ( b ) . Analysis of the Vowels—Chao. Phonetic Representation: y* u: ce i Medial + Vowel; [ia:] = / i a / . [we] = /ua/ [ W A:] = /ua:/ Phonemic System: i y u CE o/o: a/a: Chart 3 (c). Analysis of the Vowels—-Egerod. Phonetic Representation: 0 l : e u: o Medial + Vowel: [ea] = / i a / [ceo] = /0a/ [ob]/[oi] = /ua/ [wa] = /wa/ [wax] = /waa/ Phonemic System: i y u a/aa - 42 -There are three phonemic h i g h vowels i n Zhong-shan: / i / , /y/ and /u/. The phoneme / i / i s pronounced Ii:.] i n s y l l a b l e -f i n a l position, and when i t i s followed by the l a b i a l glide /w/f the l a b i a l o r dental n a s a l ( i . e . , /m/ o r /n/) , o r the correspond-i n g s t o p s C/p/ and / t / ) . Chao describes the vowel as being almost the cardinal I i ] . Before the stop endings /p/ and / t / , there i s some tendency towards pronouncing the vowel with a schwa o f f - g l i d e thus: I i 9 p ] and I i 9 t ] , f i r s t noted by Chao (p.53). Sometimes i t i s a case o f the laxing o f I i : J , r e s u l t i n g i n the pronunciation o f the word j i e 'knot', f o r instance, as I k i ' t 2 ] (phonemically 2 / k i t /).. At other times, the o f f - g l i d e i s further lowered, p r o -ducing the segment [ i e ] , as observed i n the pronunciation o f the word j i e 'outstanding' as I k i e t 2 J . The phonetic description above i s based on Chen's pronunciation. In general, however, the vowel i s simply a l o n g I i : ] . Egerod (p.14) records the l o n g vowel Ii:. J without elaboration. C r o s s - d i a l e c t a l l y , the si t u a t i o n i s quite d i f f e r e n t : what i s merely a l o w - l e v e l schwa o f f - g l i d e found i n some / i p / and / i t / f i n a l s i n Zhong-shan i s a f u l l vowel i n other d i a l e c t s such as Mandarin, Southern Min and Hakka. The: character j i e 'knot' , 35 32 f o r example, i s pronounced [tcsie ] i n Mandarin, I k i a t ] i n Amoy 21 11 (Southern Min), and I k i e t ] i n M e i - x i a n (Hakka) ( Z i - h u i , p.34). This f i n a l i s reconstructed by Karlgren as Ancient Chinese ( o r "Anc." f o r short) * i a t i n D i v i s i o n I I I , and * i e t i n D i v i s i o n IV; and by Pulleyblank as Late Middle Chinese (.LMC). * i a t f o r both grades, o r d i v i s i o n s . Cross-dialectal data f o r what are the / i m / and / i n / f i n a l s i n Zhong-shan p a r a l l e l ' the above observations, r e f l e c t e d i n the h i s t o r i c a l reconstructions (e.g., LMC *iam and - 43 -*ian). Similar c r o s s - d i a l e c t a l and h i s t o r i c a l observations as above can be made of f i n a l s which correspond to Zhong-shan / y t / and /ut/ f i n a l s i n which a schwa o f f - g l i d e i s only heard sporadic-a l l y . Usually, the two f i n a l s / y t / and /ut/ are pronounced l y i t j and [u:t], without a perceptible g l i d i n g e f f e c t . Again, i t was Chao (p.53) who f i r s t made the observation of a schwa o f f - g l i d e , while Egerod simply recorded a long l y i ] and a long [ u i j , with no discussion of allophonic v a r i a t i o n s . The phoneme / i / i s pronounced [i] before velars. (The preferred IPA symbol for [i] i s [i].) Chao also analyzes t h i s vowel as / i / , which he describes as being quite open before /o/ and /k/, almost becoming an [ e ] , which i s the broad phonetic form he uses for the vowel. Egerod also transcribes / i / as [e] i n the same environment. This vowel i s i d e n t i c a l to the one found i n Cantonese, usually transcribed as either Ie] or [ 1 ] . A further point must also be added concerning the special complementary d i s t r i b u t i o n found i n the s y l l a b l e /ing/. In a l l the tones except the r i s i n g one, the s y l l a b l e has a smooth onset, and begins simply with the vowel [ 1 ] , as i n the word ying jfe 55 'distinguished 1, which i s phonemically /ing / and i s pronounced 55 . ]. In contrast, the s y l l a b l e i n the r i s i n g tone begins with the front on-glide [ j ] , as i n the words ying ^ 'shadow' and ying 'to r e f l e c t ' , both of which are /ing / phonemically, but are 13 i n fact pronounced [jig ], with an i n i t i a l p a l a t a l on-glide. The reason for t h i s phenomenon i s not clear to the writer. Pulley-blank (personal communication) speculates that i t may be the r e s u l t of the laxness of {1] combined with some c h a r a c t e r i s t i c of the r i s i n g tone, since a l l other f i n a l s with / i / i n the r i s i n g - 44 -tone have the tense, long [ i x ] , and are produced with a smooth onset before the nuclear vowel; for example, y i 'chair' i s / i 1 3 / , phonetically [ i x 1 ^ ] , and yan 'to conceal* i s /im 1^/, 13 phonetically lixm J . The phoneme /y/ i s pronounced somewhat more open than cardinal ly] i n Chen's speech, which agrees with Chao's observa-ti o n of the vowel. In Zhong-shan th i s vowel i s always long, occurring as a f i n a l either alone or followed by a dental. How-ever, r e c a l l that the f i n a l / y t / i s sometimes pronounced [y t ] , with an o f f - g l i d e following the vowel. The phoneme /u/ i s pronounced [uxj i n s y l l a b l e - f i n a l p o s ition and when i t occurs before / j / and the dentals, /n/ and / t / . The vowel i s actually s l i g h t l y more open than cardinal [uj, a point also made by Chao (p. 53). . As mentioned e a r l i e r , the f i n a l /ut/ i s sometimes pronounced [u t j . In the environment before the velar endings, /ng/ and /k/, the phoneme /u/ i s somewhat more open than lu] (= luj or the preferred IPA symbol ]>J), which i s i d e n t i c a l to i t s Cantonese counterpart, variously transcribed as I.oJ . IuJ and lu] i n Cantonese. In both Egerod's and the present study of Zhong-shan, what i s transcribed here as the high, lax vowel [u] i s analyzed as the allophone of /u/, i n complementary d i s t r i b u t i o n with the corresponding tense, long vowel [ux]. Chao, however, treats our -[u] as /o/, moreover regarding the analysis of th i s vowel as /u/ in Cantonese being due to the influence of English and German, wherein the l e t t e r "u" i n the English word "put", for example, i s , according to Chao, extremely open and quite close to cardinal [o] (p.53, fn.9). Whether or not Chao's surmise of influence from - 45 -English and German i s true, the analysis of the high, lax, back rounded vowel [u] as /u/ i n the f i n a l s provides congruity to the pattern established e a r l i e r i n the analysis of the high, lax, front unrounded vowel [ i ] . as / i / i n the f i n a l s [ i n ] and [ i k ] (see Chart 3 (a)). Furthermore, the f i n a l s [ug ] and [ukj correspond cross-d i a l e c t a l l y and h i s t o r i c a l l y to [ i n ] and [ i k ] , and should be s i m i l a r l y analyzed. The symmetry between tre a t i n g lax [ i ] as an allophone of / i / and lax [u] as an allophone of /u/ can be b r i e f l y elaborated as follows. H i s t o r i c a l l y , the Late Middle Chinese f i n a l s *-im, * - i p , * - i n , * - i t y * - i j ( < * - i a j ) and *-iw gave the lax f i n a l s -am, -ap/ - s i n ; -at/ -aj # and -aw, now [em], [ep ] and so fort h i n modern Zhong-shan and Cantonese. Such a development did not occur before *-q (and the corresponding Ru-sheng ending, * - k ) . LMC f i n a l * - i n resulted i n -en (now [ i n ] (or [eg] i n alternate transcriptions) i n the two Yue d i a l e c t s ) . As a comparison, LMC f i n a l * - i g gave Mandarin - i n . In the case of the Late Middle Chinese f i n a l *-og (*-ok i n Ru-sheng), i t remains -og i n Zhong-shan and Cantonese, now phonetically [ug] (or [og] i n alternate t r a n s c r i p t i o n s ) . In Mandarin, LMC f i n a l *-og y i e l d s -ug. Thus, the symmetry of -eg and -og (and -ek and -ok) j u s t i f i e s p a r a l l e l treatment of [ i ] and [T J ] on h i s t o r i c a l grounds. The same -eg~ - o g p a r a l l e l l i s m i s observed i n l i t e r a r y Min and appears to be a general southern development. In the north, the p a r a l l e l l i s m between the LMC f i n a l s * - i q and *-og i s observed i n the Mandarin reflexes of - i q and -ug corresponding to the two h i s t o r i c a l f i n a l s . Thus, the tense and lax d i s t i n c t i o n , i n conjunction with h i s t o r i c a l and cross-- 46 -d i a l e c t a l arguments supports a phonemic analysis which recognizes the p a r a l l e l l i s m or symmetry that exists between the vowels [ i l . 12 and l u j , and treats them accordingly. Turning now to the mid vowels, phonemically. there are two mid vowels: /</>/ and /o/. Regarding the f i r s t phoneme, /</>/ i s pronounced [ce:] as a s y l l a b l e - f i n a l segment. There are very few words i n Zhong-shan and Cantonese with the f i n a l Ice:] , and these may be vestiges of an e a r l i e r layer of the language as suggested by Pulleyblank (personal communication). Chao (p.53) chooses phonemic symbol /oe / for Ice:] when i t occurs i n s y l l a b l e -f i n a l position. Before the endings /ng/ and /k/, Chao notes that /ce / tends to break into a [ 0 0 ] c l u s t e r , which i s treated phone-mically as / i o / i n the present analysis. (We w i l l return to t h i s point later.) Egerod posits the vowel c l u s t e r /0a/, which, he transcribes phonetically as [ceo] , occurring with an i n i t i a l con-sonant only, or before /q/ (= /ng/) and /k/. In the present phonemic analysis, /0/ i s recorded as [cei] s y l l a b l e - f i n a l l y , and elsewhere ( i . e . , before the endings /n/, / t / and /j/) as a rounded vowel which i s s l i g h t l y more open and centralized than [ 0 ] . The vowel [ 0 ] i s recorded as the broad phonetic form i n Table 2 (.a). Chao (p. 52) describes the same vowel as a rounded, mid central [el, which i s the same as i t s Cantonese counterpart. Chao analyzes his Ie] as /o/ phonemically, adding an umlaut above the phoneme ( i . e . , /6/) i n order to f a c i l i -tate recognition and r e c a l l of i t s actual phonetic value. More-over, Chao (p.52,fn.8) makes the comment that t h i s vowel i s trans-cribed by Daniel Jones as [ce] , which i s the same phonetic form suggested by Kinkade for the vowel i n question. Egerod (p.14) also - 47 -hears the vowel Ice] i n his Zhong-shan data, analyzing i t phonemic-a l l y as On a preceding page however, Egerod (p.12), i n describing the phonological system of standard Cantonese, records the phoneme /0/ which he states i s I0] before /n/, / t / and / j / . Recent studies on Cantonese (.e.g., Kao, 1971; Hashimoto, 19.72a; Cheung, 19.72) have v a c i l l a t e d between selecting the form Ice] or [0] for transcribing the vowel. The phoneme /of i s pronounced [D:J except i n the environ-ment before the l a b i a l glide /w/ where i t i s pronounced somewhat lower than cardinal l o ] . Whereas i n the present analysis length i s non-contrastive except i n the low vowels, Chao treats short .'[.o.J as /o/ phonemically, and long ID:.] as /o:/, thereby i m p l i c i t l y considering length phonemic i n Zhong-shan. He i s not consistent, however, since he also subsumes his mid central vowel Ie] under the phoneme /o/, and provides no corresponding long phoneme to his / C B/, which only has the long Ice:] as i t s basic phonetic shape. It i s observed that the vowel ID:] sometimes s p l i t s into loa] (in which the i n i t i a l segment of the b i v o c a l i c nucleus i s s l i g h t l y lower than cardinal ID] and the second somewhat more centralized than cardinal [a.]}. Chao (p.53) describes the break-ing of the vowel ID:] (his phoneme /o:/) as [OD] , p a r t i c u l a r l y notable aft e r velars and laryngeals. Egerod (p.14). sets up a 13 vowel c l u s t e r /ua/, with. I DO] as the p r i n c i p a l phonetic form, presumably occurring aft e r velars, laryngeals and / f / , since i t s allophone I D:], which i s enclosed i n round brackets, occurs most commonly after dentals, s i b i l a n t s and l a b i a l s with the exception of / f / . Concerning the allophone [oil, Egerod states that i f i t i s considered a phoneme, i t should be written /o/. This would - 48 -p a r a l l e l his treatment of the vowel I ox,] as /o/ on the preceding page for Standard Cantonese (p.12). It should be observed that i n Cantonese there i s a contrast between IkJ and [kw] (and between the aspirated series, I k ' J and Ik'wjX before the vowel I D : ] , namely a f f e c t i n g the f i n a l s [ o i l , I D :..rjJ and [31k]. Such a contrast i s not found i n Zhong-shan. Consider, for example, the Cantonese pronunciation of the word ge \fy ' ( . c l a s s i f i e r ) ' , which i s {k 3: 4 4.]. 1 4,. and that of guo j$l 4 4 'to cross', which i s Ikwsx *].. The pronunciation of these two words i s merged i n Zhong-shan such that both ge and guo are pro-22 nounced Iks': J . The contrast found i n Cantonese between the two sample words i s r e f l e c t e d h i s t o r i c a l l y i n the d i s t i n c t i o n between Kai-kou V2 (.'open mouth', associated with the absence of l i p -rounding), and He-kou ^ & (.'closed mouth', associated with the presence of lip-rounding), reconstructed by Pulleyblank as LMC *ka and *kwct respectively. As noted above, Chao (p. 52) describes the breaking of the vowel [01] i n Zhong-shan into the sequence [osj a f t e r velars and laryngeals. While one can say that Zhong-shan does, not have a Kai-He d i s t i n c t i o n , Chao suggests that one might equally take the vowel i n Zhong-shan as o r i g i n a l l y between Kai and He. Egerod's treatment of [ 3 : ] , and even more so that of his c l u s t e r [3t>] as /ua/ rather than /o/, as i n our analysis, or /o:/, as i n Chao's, i s obviously an attempt, however i m p l i c i t , to r e f l e c t the He-kou category. There are two low vowels i n the data, contrasting i n length and tenseness. Short /a/ i s phonetically a very lax vowel, des-cribed by Chao (p.52) as one raised towards [e], except before /w/, where i t i s a low, central vowel for which he adopts the symbol - 49 -[ A ] . In our broad phonetic t r a n s c r i p t i o n i n Table 2 (a), the vowel [BJ i s used throughout the set of /a/ f i n a l s . Kinkade per-ceives short /a/ i n the data as simply the low, back vowel [a], although he observes that the vowel occasionally s h i f t s to midway between Is] and lae] . Egerod simply records a. .low central .[&.•] with-out any comment concerning allophonic v a r i a t i o n s . This vowel i s i d e n t i c a l to the corresponding short, low vowel i n Cantonese. Long /a:/ i s a low, central vowel, transcribed here using the symbol [A:]. Chao's phoneme /a:/ i s also [A:] i n Zhong-shan. The same vowel i s also found i n Cantonese. Hashimoto, for instance, transcribes the long, low vowel i n Cantonese as [A : ] . Egerod . records his long, low central vowel as [a:], which he treats phonemically as the vowel c l u s t e r /aa/. Chao (p.52) notes that before stop endings, the difference i n vowel length, between the two low vowels /a/ and /a:/ i s not sharp; for the most part, i t i s the vowel quality which d i f f e r e n t i a t e s them. Egerod (p.14) makes the observation that i n his s y l l a b l e /jaa/, the phonetic form [ja:.J and Ijaei] occur i n free v a r i a t i o n . This i s not noticed i n the speech of the informants used i n the present study wherein only the low, central [AI] i s found. Note i n Table 2 (a) that a l l the vowels i n s y l l a b l e - f i n a l position are long i n Zhong-shan. While there i s a phonemic con-t r a s t between /a/ and /a:/ i n non-final position, such a contrast does not e x i s t i n s y l l a b l e - f i n a l p o s i t ion. Thus, to mark length i n the l a t t e r case would be redundant, and i s therefore omitted in such an environment (e.g., ma ^ 'mother' i s /ma55/, phonetic-55 a l l y [JIA i ] . In the discussion of nuclear vowels i n th i s section, no - 50 -account i s taken of the influence which the medials may have on them. The following section on medials w i l l include allophonic variations of the nuclear vowels r e s u l t i n g from assimilation of certai n features of the medials, and conversely, of the e f f e c t of the nuclear vowels on the medials. 1.2.2. Medials Two medials are proposed f o r the Zhong-shan data, namely / i / and /w/. Although i t might have been more symmetrical to posit either a pair of medial vowels or a pair of medial gl i d e s , the choice of the vowel / i / and the semi-vowel /w/ i s based on synchronic data, supported by c r o s s - d i a l e c t a l and h i s t o r i c a l e v i -dence. Synchronically, for example, vocalic medial / i / i s a vowel i n Zhong-shan, phonetically [e] before /a/, and [0J before /o/. The glide /w/ i s i n fact the semi-vowel [w] i n Zhong-shan. Thus, phonetically, the two medials i n Zhong-shan are not symme-t r i c a l with respect to vowel qua l i t y . Diachronically, there are also good reasons for di s t i n g u i s h -ing the v o c a l i c nature of the two Zhong-shan medials. Medial / i / i n Zhong-shan i s derived from the LMC medial * i , while medial /w/ has i t s source i n medial *w. Thus, h i s t o r i c a l l y , the two medials also d i f f e r i n v o c a l i c q u a l i t y . Pulleyblank (1970-71), for example, analyzes the four grades i n h i s t o r i c a l Chinese phonology i n terms of differences found i n the medials, since Chinese rhyming practices do not take either the Kai-He d i s t i n c t i o n s or the various grades into consi-deration. Pulleyblank reconstructs both medial vowels and medial gl i d e s . Evidence for such a d i s t i n c t i o n i s found i n Cantonese and Zhong-shan. The word guan '|p (134-1) ' governraent o f f i c i a l ' , 55 for example, i s LMC *kuan, Zhong-shan [ku:n ] and Cantonese 5 3 [kuin ], wherein the medial i n Grade I of the Shan LU rhyme group (He-kou series) i s reconstructed as the vowel *u. In modern Zhong-shan and Cantonese, the nuclear vowel *a was l o s t , r e s u l t -ing i n medial *u becoming the new nuclear vowel which was length-ened to compensate for the l o s t segment. In contrast, the word guan (138-7) 'to shut', which i s found i n Grade II of the same rhyme group and reconstructed as 55 53 LMC *kwam, i s [kwAin ] i n Zhong-shan and [kwAin ] i n Cantonese. This time the reconstructed medial i s a g l i d e , *w, which i s pre-served as a glide medial i n both present day Zhong-shan and Can-tonese. In the modern reflexes of many di a l e c t s of Chinese, the two reconstructed forms, *kuan and *kwa:n, have merged (e.g., both, words 'government o f f i c i a l ' and 'to shut' are pronounced guan i n Mandarin). In the Yue d i a l e c t s such as Zhong-shan and Cantonese, however, one can see that the d i s t i n c t i o n between the two forms has been maintained. On the basis of Zhong-shan and Cantonese, i t i s evident that the reconstruction of a medial vowel versus a medial gl i d e i s important i n h i s t o r i c a l phonology. Karlgren also distinguishes a "vocalic u" as opposed to a "consonantal w" i n the two words discussed above, reconstructing guan '^f as Anc: *kuan and guan j^£J as Anc. *kwan. (Note that Karlgren also recon-structs a q u a l i t a t i v e l y d i f f e r e n t nuclear vowel.) In the Zhong-shan data, i t i s therefore reasonable to posit a medial /w/ which 15 arises from an h i s t o r i c a l medial *w. Medial / i / i n Zhong-shan, i n contrast, has i t s source i n an h i s t o r i c a l * i . On the basis of Pulleyblank'. s reconstructions, - 52 -while medial * j i s l o s t i n Zhong-shan, medial * i i s s t i l l present i n the d i a l e c t . It i s maintained as the medial vowel / i / i n the c o l l o q u i a l layer of Grades III and IV words i n the Geng rhyme group. In the corresponding l i t e r a r y layer, medial * i had become the nuclear vowel / i / i n Zhong-shan. Consider the word j i n g t (199-6) 'mirror', which occurs i n Grade II I , Geng )$L rhyme group. Jing /4JfLi i s reconstructed by Pulleyblank as LMC *kiajn (..= Anc. 22 *kiang). I t i s /kia:ng / i n the c o l l o q u i a l layer of Zhong-shan and / k i n g 2 2 / i n the l i t e r a r y layer. Now, contrast j i n g /jjftj with geng ^ (192-8) 'watches of 55 the night', which i s /ka:ng / i n the c o l l o q u i a l reading, and 55 /kang / i n the l i t e r a r y layer. No medial i s present i n the Zhong-shan forms for geng , which i s a Grade II word. Pulley= blank reconstructs i t with a medial * j : LMC * k j a i j n (.= Anc. *kong) . Thus, i n terms of Pulleyblank.'s reconstructions, Late Middle Chinese medial * j . i s l o s t i n Zhong-shan, as demonstrated by the word geng . In contrast, medial * i i s preserved i n the d i a l e c t either as a medial vowel or nuclear vowel, as shown i n j i n g . The preservation of the LMC glide medial *w as a medial /w/ i n Zhong-shan, and the p a r a l l e l preservation of a v o c a l i c medial i n Zhong-shan corresponding to the LMC medial * i lends h i s t o r i c a l support for not pos i t i n g the same vo c a l i c q u a l i t y to the pa i r of medials i n Zhong-shan. Cr o s s - d i a l e c t a l l y , medial * i i s also preserved as a medial segment i n some d i a l e c t s . Regarding the word qing ^ £ (.201-1) 44 ' l i g h t (e.g., i n weight)', for example, i t i s pronounced [tcsliai) .] i n Wen-zhou (a Wu d i a l e c t ) . In Nan-chang (a Gan d i a l e c t ) , the 42 same word has a c o l l o q u i a l reading [te'ian ] and a l i t e r a r y read-- 53 -ing [to.'in ] . In Mei-xian (a Hakka dialect) , qing has a 44 44 c o l l o q u i a l form [ k i a q ] and a l i t e r a r y form I k i n J . (Zi-hui, p.257,) The medial / i / i n the f i n a l s /iong/ and /iok/ i n Zhong-shan i s also derived from medial * i i n Late Middle Chinese. The two f i n a l s are reconstructed by Pulleyblank as *ian and * i a k respectively, (= Anc. *i,ang and * i a k ).. The medial i s also pre-served i n most Chinese d i a l e c t s . The word lia n g ftj (174-18) 'bright' (cf. Zi-hui, p.232), for example, has the s y l l a b l e I l i a y J i n (Peking) Mandarin and a number of other Mandarin d i a -l e c t s , Su-zhou (Wu), Chao-zhou (Southern Min), and the l i t e r a r y layer of Amoy (Southern Min). The s y l l a b l e [ l i a n ] i s found i n Nan-chang (Gan), Mei-xian, and the c o l l o q u i a l layer of Amoy. >• 22 Zhong-shan pronounces the word lia n g faj as [loon ] , which i s 22 phonemicized as / l i o n g /. Pulleyblank (1977) proposes that i n Cantonese c e r t a i n LMC f i n a l s (viz., Grade I f i n a l s i n the Dang % rhyme group), f i n a l s *-ag and *-tik rounded to -og and -ok, just as *-a rounded to -o (in the Guo ^ rhyme group). He fur-ther.postulates that i n Grades III and IV (of the Dang rhyme group), * - i a g and *-iak also rounded to - i o g and-iok as an i n t e r -mediary stage, aft e r which the p a l a t a l medial became fused with the following vowel by a process of umlaut, r e s u l t i n g i n the f i n a l s -cag and -ask. Zhong-shan data would suggest the following development: the vowel *a i n the LMC f i n a l s *-Q, *-aq and *-ak rounded to :D, The same process affected the *-d i n the f i n a l s * - l a g and *'-.iak. In the l a t t e r case with medial * i , the medial underwent a lower-ing and rounding to -0 as a r e s u l t of the following back, rounded - 54 -mid vowel -o. The result is the present Zhong-shan finals, [0oql and l0o.k] . Cantonese, on the other hand, may be a case of the lower-ing and rounding of medial *i to -ce. The nuclear vowel is subse-quently lost, causing compensatory lengthening of -ce to - c e : , yielding present day Cantonese finals IceiijJ. and {cerkj . Liang rftj , for example, is pronounced I ice i n J in Cantonese. This lowering of medial *i in the Dang ^ rhyme group in Cantonese parallels the lowering of medial *i to -e in the Geng rhyme group. What is proposed here is that in Cantonese, medial *i lowered to -e in the finals *-iajn and. *r-iajk. In this case, rounding of the medial does not occur since there is no condition for such regressive assimilation. However, as in the previous case, the nuclear vowel becomes lost, and -e becomes the nuclear vowel. It is lengthened in compensation for the loss of the original nuclear vowel *-a. The finals that emerge in Cantonese are -e:g and -eik- Thus, the colloquial reading of jing^Jb is 44 . • [keiq ] in Cantonese. The nuclear vowel *-a is not lost in Zhong-shan; jing is pronounced [keaxn J (phonemically , . 22 /kiarng / , as noted earlier). Turning now to a synchronic analysis of the interaction between the medials and other segments in the syllable, medial /w/ will be discussed first. In the present data, the only two initials that occur with medial /w/ are /k/ and /k'/. Egerod and Chao, however, also include the velar nasal as another initial which can precede the labial medial. The cluster /ngw/ is merely mentioned by Egerod. In Chao's article it occurs only in the word 51 wan .jf.fi . (137-U4) 'obstinate', phonemicized by as /ngua:n / . - 56 -l a b i a l segment i n Shi-qi depended on the i n d i v i d u a l speaker, although Macao shows a.complete loss of /w/ after velars. Addi-t i o n a l s y l l a b l e s containing a l a b i a l velar cluster recorded by B a l l for Shi-qi i s l i s t e d i n (2) below. B a l l states that these words are spoken by some people i n S h i - q i , and attributes that to the imitation of Cantonese speakers. The asterisk (.*) marks the column ...... which B a l l regards to be the pronunciation of some Shi-q i speakers. Regarding the c h a r a c t e r ^ " , i t i s used i n Cantonese to mean 'to be t i r e d ' , and i s a word which i s not used c o l l o q u i a l l y 22 i n Mandarin. Zhong-shan has i t s own c o l l o q u i a l word /na:j /, with, b a s i c a l l y the same meaning of 'to be t i r e d ' . (In the l i s t below, "Cant." i s "Cantonese", and "Zh." i s "Zhong-shan".) (2) B i l l , ' Cant.* s Macao . Sample Word Modern Cant. Modern Zh. Gloss 1. kwe ke '%, ? ? ? 2. kwing kang fa 35 [ k w 11) J •p 'to b o l t ' 3. kwit k i t *&'\ ? ? ? 4. kwo ko * r 5 3 i [wo:. ' j [koi J 'spear' 5. kwok kok i i Ikwo z.k^ ] Ik a i k 2 J 'nation' 6. kwong kong to IkwDiq J Ikoxn J 'bright' 7. k'wong k' ong [k ' o i n 2 1 ] Ik'oxn 5 1] 'crazy' 8. kwGii koii [ k u xj 3 3] I k u i j 2 2 ] 'to be t i r e d ' 9. k' woii k' ou I k ' u i j 3 5 ] Ik'uxj ] 'middleman' 10. kwu ku [kux.53] [kux 5 5] 'father's s i s t e r ' 11. k' wu k'u ** Ikux 5 3] ? 'wheel' 12. kwun kun f Ikui.n 5 3] I k u i n 5 5 ] 'government o f f i c i a l ' - 55 -Wan fify^ occurs i n the He-kou series of the Shan ih rhyme group. Apparently, i t i s the only common word from that set where the s y l l a b l e /ngwa:n/, or the potential s y l l a b l e /ngwa:k/ can occur. In consulting the p r i n c i p a l informants as well as another Zhong-shan speaker concerning the word wan 7 ^ , i t was found that none of them preserve a l a b i a l segment i n the word: they simply gave /nga:n 5 1/. For the present data at least, i t w i l l be concluded that the only i n i t i a l s which can occur with medial /w/ are /k/ and /k'/. The sequence /ngw/ i s assumed to have merged with the plain / v e l a r nasal. In other words, the Kai-He d i s t i n c t i o n has been l o s t a f t e r the velar nasal i n the Shan rhyme group. The word wan ffy i s i n fact the only word present i n the d i a l e c t survey l i s t which can be reconstructed as LMC *nwa.in. In contrast, there are a few common words with the s y l l a b l e /nga:n/ from LMC *D.ja:n and a small number of indigenous (characterless) words i n Zhong-shan also containing the s y l l a b l e /nga:n/. I t i s therefore not surprising to f i n d that the word wan has l o s t i t s l a b i a l segment, and has merged with, the more common s y l l a b l e without l a b i a l i z a t i o n . In Cantonese (as i n Mandarin), the word wan has l o s t i t s velar nasal i n i t i a l , but has maintained the l a b i a l segment, and i s pronounced r 2 1 i 1WA in J-Concerning the clusters /kw/ and /k'w/, these combine only with the nuclear vowels / i / (found solely i n the rhyme / i k / ) , /a/ and /a:/. Gaps i n the f i n a l such as the non-occurrence of /*wa:m, *wam, *wap/ and so forth are the r e s u l t of l a b i a l d i s s i m i -l a t i o n , which i s a phenomenon found i n many di a l e c t s of Chinese. According to B a l l ' s data, the loss or preservation of the - 57 -The l i s t i n (2) prompts a low^level phonetic observation that might otherwise have been l e f t unstated. In modern (Shi-qi). Zhong-shan and Cantonese, some lip-rounding occurs i n the pro-17 nunciation of i n i t i a l s before the rounded vowel _ L u z J . I t i s possible that t h i s feature i s absent i n Macao, as r e f l e c t e d i n B a l l ' s recording of numbers (.10) to (12) i n the l i s t above. A si m i l a r lack of lip-rounding on the i n i t i a l s may be interpreted from B a l l ' s t r a n s c r i p t i o n of the Macao forms i n (8) and (.9). Numbers (4). to (.7) show a presence of l a b i a l i z a t i o n which i s not eyident i n the Zhong-shan speech recorded by Chao, Egerod and the writer, although i t i s s t i l l preserved i n Cantonese. That i s , there i s a phonemic contrast i n Cantonese between /k/ and /kw/ and between /k.'/ and /k'w/ before the vowel I D Z J , as we have men-tioned e a r l i e r , exemplified by the words g_e and guohjffi^ . Egerod (.p. 3) makes the observation that although the d i a l e c t spoken i n Macao has now s h i f t e d to the standard Yue dia-l e c t , Cantonese, there are s t i l l a few vestiges of the l o c a l d i a -l e c t , among which i s the fusion of /k/ and /kw/. In studying the Sino-Portuguese glossary of c i r a 1750, nevertheless, i t can be strongly substantiated that although the medial /w/ was l o s t a f t e r /k/ when followed by the mid back vowel which i s now ID:.] , the 'li.-. d i s t i n c t i o n between the presence or absence of a l a b i a l a f t e r the velar stop was s t i l l maintained before low vowels at the time that the three hundred and ninety-five Portuguese entries were trans-cribed into Chinese. While the characters ge , ge gffi^ and ge (Cantonese [ k o i ] ) , were used most frequently to represent Portuguese /ko/, the characters guo ^ and guo jyffi^. (.Cantonese [kwDz.]) were nonetheless used for the same purpose, thereby - 58 -suggesting a loss of d i s t i n c t i o n between /k/ and /kw/, at l e a s t i n that p a r t i c u l a r environment. As i n the (Shi-qi). Zhong-shan d i a l e c t today, however, the Sino-Portuguese data show that t h i s d i s t i n c t i o n was preserved before low vowels at that time, r e f l e c t -ing either the speech of the transcriber only, or the eighteenth century Macao d i a l e c t i n general. In the glossary, the characters j i a , j i a and j i a were consistently used to transcribe Portuguese words which do not contain l a b i a l segments. A l l three characters belong to the Kai-kou series of the J i a 4 £ s L rhyme group and are pronounced IkAiJ i n both present day Cantonese and (Shi-qi). Zhong-shan. I t i s therefore s i g n i f i c a n t that gua CtkwAiJ i n both Cantonese and Shi-qi Zhong-shan), which i s the only He-kou word from the J i a rhyme group, should be used to transcribe the word quatro 'four' (entry number 328 i n the glossary). In turn, quatro i s the only Portuguese word reconstructed i n the glossary that contains a velar i n i t i a l followed by a l a b i a l segment. Assuming that the transcriber's speech r e f l e c t s the n: " at the time, one can quite safely conclude from the observation above that, with, respect to low vowels, a Kai-He d i s t i n c t i o n was s t i l l preserved i n the eighteenth century Macao data. Turning now to a''.discussion of medial / i / , the medial com-bines with one of two possible nuclear vowels: /a:/ or /o/. When i t combines with /a:/, medial / i / may follow l a b i a l s (except / f / and /w/), dentals, s i b i l a n t s , velar stops and laryngeal /h/. Re-interpreting Chao's data from th i s perspective, his medial / i / i s phonetically somewhat more open than the vowel {ij (cf. fn.16 of the present chapter). Egerod and the writer transcribe t h i s sound as l e j . At the same time, medial / i / i s sometimes heard as - 59 -a p a l a t a l [j] . Thus, the word ji n g 'neck', for example, may 13 13 Be pronounced Tkeaig J or [kjain J i n free v a r i a t i o n . I t i s only i n the environment a f t e r laryngeal /h/ that p a l a t a l i z a t i o n of the medial does not occur. As suggested by Pulleyblank upon observing t h i s phenomenon i n the data, i t may be a case of the d i a l e c t attempting to produce a p a l a t a l medial to correspond to the l a b i a l one, thereby creater greater symmetry i n the system. The f a c t that medial / i / i s never pronounced as a p a l a t a l g l i d e when i t follows /h/ i s probably due to the e f f o r t s of the speakers not to p a l a t a l i z e /h/, since the r e s u l t of p a l a t a l i z i n g /h/ would be perceived by the native ear as being quite aberrant from the o r i g i n a l laryngeal i n i t i a l . In general, Yang tends to produce more p a l a t a l [j] medials than Chen. At the same time, i t i s also Yang who pronounces the medial / i / vowel higher than Chen. The more open pronunciation of the vowel by Chen may be an influence from Cantonese i n which the Zhong-shan sequence I ear], o r i g i n a t i n g from the sequence ,vi,v V. "medial / i / + nuclear vowel /a:/", corresponds to the Cantonese long (nuc1earI vowel I e : ] . In the environment before /o/, medial / i / i s lowered and rounded as a r e s u l t of regressive assimilation. In t h i s environ-ment, medial / i / i s pronounced [0]. The sequence / i o / only occurs i n the f i n a l s /iong/ and /iok/, phonetically [00^] and [00k] respectively. These two f i n a l s are treated by Chao as /aeng/.and /aek/. Note that Egerod also treats the f i n a l [OBI.] as the vowel clus t e r /0a/. The writer analyzes t h i s f i n a l as phonemic /0/ rather than the c l u s t e r / i o / . Chao treats i t phonemically as /CE/. D i s t r i b u t i o n a l l y , the f i n a l s /iong/ and /iok/ occur with - 60 -dentals, s i b i l a n t s and velar stops. I n i t i a l /h/ combines only with, the /iong/ f i n a l . Present i n both. Chao and Egerod' s data are the s y l l a b l e s lj0oql and I j 0 D k J , which correspond to [jaxgj and IjDikJ i n the present data. B a l l (p.531) records the s y l l a b l e /yong/, examplified by the word yang ^ (193-5) 'ocean1, for both Cantonese and Macao, but records /yong/ for Shi-qi (fn.120). Presumably, the corresponding stop ending r e f l e c t s the same pattern, although i t was not made e x p l i c i t i n B a l l . The present Zhong-shan data and modern Cantonese concur with B a l l ' s observa-t i o n : Zhong-shan has the s y l l a b l e s Ijonj] and [joik] while Canton-ese has [uosir)] and [qcexk] . Based on B a l l ' s observations and the writer's data, one could suggest that s y l l a b l e s [jong] and [joik] are the indigenous forms i n the Shi-qi speech, and that the s y l l a b l e s [j0orjj and [J03kJ recorded by Chao and Egerod are due to Cantonese influence. The writer, however, would prefer to suggest that the difference observed between the present data and those of Chao and Egerod i s s t r i c t l y a case of a sub-dialectal v a r i a t i o n . F i r s t of a l l , Chao (p.51) e x p l i c i t l y describes the i n i t i a l segment of his s y l l a b l e /icEng/ phonetically as the unrounded front g l i d e [ j ] . Thus Chao's s y l l a b l e appears to be mid-way between the s y l l a b l e found i n the present Zhong-shan data and that i n Cantonese. Phonetically, Chao's s y l l a b l e i s [J0oq], whereas the Cantonese s y l l a b l e i s [xifflin ] . Secondly, except for t h i s s y l l a b l e Ij0onj, and the s y l l a -ble with the corresponding stop ending, the phonological system recorded i n Chao and Egerod i s i d e n t i c a l to the present Zhong-shan one. There i s no reason to expect such a unique borrowing. - 61 -Third and l a s t l y , there are a few c o l l o q u i a l terms i n Chao's a r t i c l e that show s l i g h t phonological variations between his informant's speech and the speech of the writer's informants. In the d e i c t i c word ' t h i s ' , for example, Chao records the variant 22 55 55 forms of /ko / and /ko / whereas the present data show /ku /. (On one occasion, nonetheless, the writer heard Chen's mother use 55 /ko / for 'this'.) In the reading of is o l a t e d words, Chao's informant also pronounces some words with an i n i t i a l or f i n a l that d i f f e r s from the pronunciation of the writer's informants. The word ku ^ (.16-24) ' b i t t e r ' , for instance, i s recorded by Chao 13 13 as /k'u / and by the writer as /hu /. In terms of the o v e r a l l pattern of correspondences to the h i s t o r i c a l phonological catego-r i e s , however, the data from Chao, Egerod and the writer are i n agreement. In summary, there i s evidence of low-level, sub-dialectal variations that, i n general, do not a f f e c t either the Zhong-shan phonological system on the whole, or the d i a l e c t ' s h i s t o r i c a l correspondences. One can see no reason to suppose that there has been an is o l a t e d case of substituting a Cantonese s y l l a b l e for a Zhong-shan one. The conclusion, therefore, i s that among some Zhong-shan speakers, the s y l l a b l e [joon] i s used, whereas others use [ j o i n ] . One would suspect, given Chao's observation of an ... i n i t i a l IjJ that the following segment i s probably s l i g h t l y less rounded than cardinal J>]. Cr o s s - d i a l e c t a l l y , the s y l l a b l e [jog] i s also found i n other d i a l e c t s . The word yang p| (173-8) 'bright' (Zi-hui, p. 236) has the s y l l a b l e [jon] (or [iaq]) i n Nan-chang (Gan), Mei-xian (Hakka), and Amoy (S. Min). The Mandarin d i a l e c t s generally - 62 -pronounce the word as Jax.n'l (transcribed i n the Zi-hui as {iag]). Of the d i a l e c t s represented i n the Zi-hui, only Cantonese has the s y l l a b l e [ u 3 i >j] (.[jcenj i n the Zi-hui) . Fu-zhou (_N. Min) has the s y l l a b l e fyon] for yang ^ . Both yang 1 s ^ and f-|j are recon-structed as LMC * j i a g . 1.2.3. Endings The inventory of endings present i n Zhong-shan i s i d e n t i -c a l to that found i n Cantonese. There are three nasals: /m/, /n/ and /ng/; three corresponding stops: /p/, / t / , /k/; and two g l i d e s : /w/ and / j / . The stops are unreleased, and often p r e g l o t t a l i z e d i n Zhong-shan. The ending /k/ i s moreover often simply reduced to a g l o t t a l stop. Syllables with stop endings usually occur i n l e v e l tones only. Exceptions r e s u l t from changed tones, to be elaborated upon l a t e r . Chao notes that a f t e r long nuclear vowels ( i . e . , i n the endings IA I W ] and [i:w]), /w/ i s more open.than a f t e r short nuclear vowels (as i n the endings [ew] and [ow]). The glide / j / i s phonetically [j] with one exception: following the front rounded vowel / 0 / , / j / i s phonetically the rounded semi-vowel [q] . Thus, the word r u i ,gjp^ (61-15) 'stamen' 13 13 / J 0 j /r for example, i s phonetically [^0^ J, i n which both the preceding and the following semi-vowel segments are rounded to {uj as a r e s u l t of assimilation to the rounding of the nuclear vowel. - 63 -1.2.4. S y l l a b i c Nasals There are two s y l l a b i c nasals i n Zhong-shan which c o n s t i -tute the only segment i n the s y l l a b l e , and a f f e c t a very r e s t r i c t e d lexicon. These two segments are /m/ and /ng/, pronounced ImJ and [gj respectively. The s y l l a b l e /m/ only occurs c o l l o q u i a l l y , as * 51 ' the negative marker B "gr /m / 'not' (also found i n Cantonese). Words with the s y l l a b l e /ng/ include: wu /ng"*"3/ ' f i v e ' , wu ^ / n g 5 1 / 'Wu (a surname; a Chinese d i a l e c t group to which the Shang-i < 22 hai d i a l e c t belongs) ', and wu •fQ' /ng / 'to r e a l i z e ' . The s y l l a -ble /ng/ occurs i n a l l but the high l e v e l tone. Words possessing s y l l a b i c /ng/ h i s t o r i c a l l y bore a velar nasal i n i t i a l , and are reconstructed as LMC *nua. Some of the d i a l e c t s s t i l l preserve the velar nasal i n i t i a l , as exemplified by the pronunciation of 331 the word wu f^- i n such d i a l e c t s as Su-zhou (Wu): [gsu J , 23 42 33 Shuang-feng (Xiang): [qu ], Mei-xian: [gu ], Xia-men: [rjo J , 35 242 Chao-zhou: Igo ], and Fu-zhou: [qua ] (Zi-hui, p.94). 1.3. Tones Pitch v a r i a t i o n which i s used to d i f f e r e n t i a t e the meanings of words i s c a l l e d tone. In i t s function as a tonal phenomenon, pitch, c a r r i e s r e l a t i v e value only, with the height and range of the pitch dependent upon such factors as the age and sex of the speaker. Chao's numerical system, which i s used here for trans-c r i b i n g the tones i n Zhong-shan, i s based on the subdivision of the normal pi t c h range of a speaker into f i v e p i t c h l e v e l s , with "5" designating the highest pitch and "1" the lowest. The tones i n Chinese may be s p e c i f i e d by a single pitch l e v e l or the move-ment of the pitch from one l e v e l to another. (In the case of the - 64 -t h i r d tone i n Mandarin, for instance, which has a f a l l i n g - r i s i n g contour, a m u l t i - d i r e c t i o n a l movement i s involved.) Besides d i f f e r e n t i a t i n g l e x i c a l meaning, tone also d i f f e r -entiates grammatical (morphological) meaning. It i s important to bear i n mind t h i s second function, to which we w i l l l a t e r return. For the present, however, the focus w i l l be on the f i r s t function, i n continuation of the phonological analysis of Zhong-shan. 1.3.1. Tonal System In t r a d i t i o n a l , h i s t o r i c a l analysis, there are four tones, or sheng i n Chinese: Ping-sheng ^ Fg^ 'even tone', Shang-sheng _H_ jfe 'ascending tone', Qu-sheng f-^_ 'departing tone', and Ru-sheng >v^jF 'entering tone'. While s y l l a b l e s containing the f i r s t three tones end i n a long vowel, a g l i d e , or a nasal, Ru-sheng sy l l a b l e s end i n a stop consonant ( i . e . , /p/, / t / or /k/) and i s sometimes referred to as "checked" s y l l a b l e s . As a r e s u l t of the abrupt closure at s y l l a b l e - f i n a l p osition, Ru-sheng s y l l a -bles normally are l e v e l i n pitch and shorter i n duration than t h e i r non-checked counterparts. In terms of western, s t r u c t u r a l l i n -g u i s t i c theory and analysis, the Ru-sheng i s i n complementary d i s t r i b u t i o n with l e v e l p i t c h tones and should not be i s o l a t e d and treated as phonemically d i s t i n c t . For diachronic studies and d i a l e c t a l comparisons, i t i s nonetheless convenient to distinguish s y l l a b l e s ending i n /p,t,k/ from those containing other endings. Or more generally, the adoption of the Chinese terms i s useful for analyzing the modern reflexes of these h i s t o r i c a l Chinese tones both within a given d i a l e c t and across d i a l e c t s . T r a d i t i o n a l l y , the tones i n Chinese are further divided - 65 -into "upper" (yin , or shang J, ) and "lower" (yang J% , or xia "f ) r e g i s t e r s , which r e f l e c t the nature of the h i s t o r i c a l i n i t i a l s , to be discussed l a t e r . Suffice i t to say for now that the Yin-Yang s p l i t was taken into consideration by Chao i n his tonal analysis of Zhong-shan. The numerical values o r i g i n a l l y assigned by Chao for Zhong-shan, and subsequently by Egerod as well, are adopted here with the minor modification of taking into account the shorter duration of the Ru-sheng by assigning to i t single numbers, as opposed to double numbers for the other tones. In the enumeration of Zhong-shan tones according to the Chinese approach, there are a t o t a l of six tones i n the d i a l e c t ; but according to western phonemic analysis, there are only four of them. The four phonemic tones i n Zhong-shan are given i n Table 3 (a) below. The present enume-rat i o n of the tones as Tone 1 through Tone 4 rather than the t r a d i t i o n a l terms as used by Chao, recorded i n the table for com-parative purposes, i s primarily to avoid confusion with the h i s t o r i c a l tones. Table 3 (b) i s a comparison of Zhong-shan tones with the Cantonese ones transcribed by Hashimoto. T r a d i t i o n a l names for the Cantonese tones are used i n the table. Table 3 (a)• Tonal Sys Chan Tone 1: 55 (or 5*) Tone 2: 51 Tone 3: 35 Tone 4: 22 (or 2*) C * Used for checked ;em of Zhong-shan. Chao Yin-ping f£ -f- /Yin-ru f^T 7 v : 55 Yang-ping f-§ f : 51 Shang : 35 Qu £ /Yang-ru \% ^ : 22 syllables only.) - 66 -Table 3 (b). A Comparison of Cantonese and Zhong-shan Tones. Cantonese Zhong-shan Yin-ping HA 55/53 55 Tone 1 Yang-ping l l f 21/22 51 Tone 2 Yin-shang 35 \ r 1 3 Tone 3 Yang-shang ?&>. 24 J Yin-qu ft'* 44 ] V 22 Tone 4 Yang-qu n i 33 J Shang Yin- r u Xft^ 5 5 Tone 1 Zhong Yin- r u f ft 7^ 4 \ r 2 Tone 4 Yang-ru 3 J Concerning Zhong-shan tones, Tone 1 i s high and l e v e l i n non-checked f i n a l s . Both B a l l (.p. 510) and Egerod (p. 14) observe that the Zhong-shan upper even tone i s s l i g h t l y lower i n pi t c h than i t s Cantonese counterpart. Chao (p.54) notes that i n Zhong-shan, the s t a r t i n g p i t c h of the Yin-ping (Tone 1) i s s l i g h t l y lower than the Yang-ping (Tone 2). I t i s therefore possible to record Tone 1 as /44/. However, for the sake of greater phonemic contrast, /55/ i s used by Chao, Egerod and the writer to represent the phonemic tonal value of the Yin-ping tone. /5/ i s used for Tone 1 i n those Ru-sheng, or checked, f i n a l s which Chao terms "Yin-ru". Because Ru-sheng s y l l a b l e s end i n a stop consonant, these s y l l a b l e s are r e l a t i v e l y short i n com-parison to those with open s y l l a b l e s or those ending i n a nasal or g l i d e . /5/ p a r a l l e l s the high l e v e l /55/ tone i n that i t s pit c h height i s s l i g h t l y lower than the Shang Yin-ru, or the - 67 -high upper entering tone of Cantonese (which Hashimoto and others transcribe as /5/). Tone 2 i s a high f a l l i n g tone, transcribed by Chao and the writer as /51/. Tone 3 i s recorded phonemically as /13/, although Chao observes that i t i s i n fact closer to .1121 . Again, for the pur-pose of maximizing phonemic d i s t i n c t i o n , /13/ was chosen instead. It was noticed i n the present data that Tone 3 was produced with, a minute dip to the tone; that i s , there i s a s l i g h t f a l l before the r i s e . B a l l (.p. 510) also describes the tone i n Macao as a tone which "descend(s) a short space—beginning at the same p i t c h of voice as the Cantonese ~j» ^ , ha hou* (or Yin-qu—Hashimoto' s /44/), lower r e t i r i n g voice, i t f a l l s a l i t t l e lower at i t s end than the Cantonese T ^" , ha p'ing Cor Yang-ping—Hashimoto's /21/)„, lower even tone". Tone 4 i s a mid-low tone, recorded by Chao as /22/. I t i s here recorded as /22/ i n non-checked s y l l a b l e s . The shorter /2/ i s used for stop endings. 1.3.2. Tone Sandhi It i s noted by Egerod (p.14) that one tone sandhi phenome^ non i n Zhong-shan operates i n the following manner: when two or more low l e v e l tones ( i . e . , /22/) occur i n a sequence, the f i r s t tone i s pronounced s l i g h t l y higher than the succeeding, phonemic-a l l y i d e n t i c a l tone(s), and may be transcribed phonetically as ^ / s i kon / 'work, a f f a i r ' would be 33 22 phonetically [ s i : ko:n ]. The same sandhi phenomenon i s observed i n the present data. - 68 -Egerod a l s o notes t h a t i n n o n - f i n a l p o s i t i o n ( i . e . , i n environments not p reced ing open junc tu re or pause ) , the r i s i n g tone (Tone 3) on ly e x h i b i t s a s l i g h t r i s i n g contour , o r even a low l e v e l tone , p h o n e t i c a l l y [12] o r [11] . The p resent Zhong-shan data agree w i t h Egerod 's o b s e r v a t i o n s , w i t h the a d d i t i o n a l remark t h a t the r i s i n g tone i n such environments tends to be s h o r t e r i n d u r a t i o n than when i t occurs before a pause, such as a t the end of a c l a u s e o r i n s e n t e n c e - f i n a l p o s i t i o n . 1 . 3 . 3 . Tone Change "Tone change", o r b i a n - y i n ftfc^ ~fl 'changed t o n e ' , r e f e r s to the m o r p h o l o g i c a l and s y n t a c t i c use of tone d i s t i n c t from i t s l e x i c a l f u n c t i o n . U n l i k e Cantonese, which has a r i c h d i s t r i b u -t i o n of s y l l a b l e s e x h i b i t i n g the tone change phenomenon s e r v i n g v a r i o u s purposes , the grammatical use of tone i n Zhong-shan i s very l i m i t e d . There are two changed tones i n Zhong-shan, as there are i n Cantonese: a h igh l e v e l /55/ and a lengthened , h i g h r i s i n g / 3 5 / , which we w i l l here term " M o d i f i e d Tone 1" and " M o d i -f i e d Tone 3" r e s p e c t i v e l y . Only M o d i f i e d Tone 3 i s d e s c r i b e d by Chao and Egerod. I t i s a h i g h r i s i n g tone which Egerod t r a n s -c r i b e s as / 3 5 / . The examples t h a t Chao and Egerod g i v e i n v o l v e s y l l a b l e s which o r i g i n a l l y had Tone 3 or Tone 4. A l though exper imenta l s t u d i e s need to be conducted on the b a s i c and changed tones i n Zhong-shan befo re one can d e s c r i b e i n more d e t a i l and w i t h g r e a t e r p r e c i s i o n the tone change s i t u a t i o n , p e r c e p t u a l l y , a t l e a s t , the m o d i f i e d tones are s l i g h t l y longer i n d u r a t i o n than t h e i r b a s i c forms, and are more prominant i n the speech of the female in formants than the male i n f o r m a n t . More -- 69 -oyer, i t has been observed i n the present data that the other tones i n the d i a l e c t also can undergo tone change. In represent-ing bian-yin, the o r i g i n a l tone i s given f i r s t , with the changed tone following, separated from the o r i g i n a l by a comma. The charac-ter undergoing the tone change i s also marked by an asterisk (e.g., chi / j a : k 2 / 'to eat', / j a : k 2 , 3 5 / 'to have eaten' (eat + PERFECTIVE)). Modified Tone 1 does not d i f f e r s i g n i f i c a n t l y from the regular Tone 1 i n terms of pi t c h height. It i s the increase i n s y l l a b l e length which i s most prominent. B a l l (p. 511). describes the Modified Tone 1 i n the Macao (Zhong-shan) d i a l e c t as only . . j .'<: s l i g h t l y higher i n p i t c h l e v e l than the regular Tone 1. The Modi-f i e d Tone 1 i s , i n fact, of the same p i t c h height as the ordinary Yin-rping tone i n Cantonese ( r e c a l l i n g that the Zhong-shan Yin-ping has been noted as being lower i n pitch than the Cantonese one). Tone 1 words which underwent tone change w i l l be indicated using 55 55* an asterisk (e.g., l a / l a : j / 'to p u l l ' , 25L * / l a : j / 'to have pulled' (pull + PERFECTIVE)). One of the functions of bian-yin uses Modified Tone 1 to indicate f a m i l i a r i t y , or c o l l o q u i a l speech. Such changes are highly unpredictable. Zuo fl'jte 'yesterday' / t s o k 2 / and r i |3 'day' 2 / j a t /, for example, combine to form the word 'yesterday', which 2 2 5 i s c o l l o q u i a l l y pronounced /tsiok j a t ' /, with a v o c a l i c change i n the f i n a l of zuo 0^ and a tone change i n r i # • These changes r e f l e c t c o l l o q u i a l , d a i l y usage, whereas the pronunciation 2 2 of zuo-ri Qty Q /tsok j a t / r e f l e c t s a more l i t e r a r y form. 51 Likewise, ming $Q ' l i g h t ' /ming / combined with r i 0 2 'day' / j a t / i s highly l i t e r a r y for 'tomorrow'. The Zhong-shan - 70 -c a s u a l word f o r ' tomorrow' i s based on the combinat ion of ming p lus zao J^- ' e a r l y ' / t s o w 1 3 / to produce / m i n 5 1 ' 5 5 t s o w 1 3 / . In t h i s c a s e , the tone change on ming $fl i s accompanied by a change 19 xn the a r t x c u l a t x o n of the n a s a l endxng as w e l l . The word ya j^L ' t o p r e s s ' has the formal p r o n u n c i a t i o n of / a : t 2 / , which would be used i n terms such as y a - l i JjjjjL f) 2 2 ^ ' p r e s s u r e ' / a : t l i k / . C o l l o q u i a l l y , the word i s / a : t ' / t o 2 mean ' t o p r e s s ' , and / a : t / f o r the p a s s i v e meaning of ' t o be p r e s s e d ' , as i n y a - z h u J^L ( ' .press + c e a s e ' ) . In the word y i ^jj^ ' s i s t e r of one 's w i f e or m o t h e r ' , / i 5 1 / , when i t occurs i n d i r e c t address , i t combines w i t h a §U (or i t s 2 2 2 2 5 1 5 5 graph ic v a r i a n t /a /) to form /a i ' / . (A i s an empty p r e f i x a l form which serves to prevent m o n o s y l l a b i c i t y i n a p p e l l a t i o n s . ) In Zhong-shan, a - y i i s 'mother ' s younger s i s t e r ' . The tone change to M o d i f i e d Tone 1 i s o b l i g a t o r y . ' M o t h e r ' s o l d e r s i s t e r ' i n Zhong-shan i s y i - m a / i ma / , w i t h no tone change i n y i . In words such as H^JL - f i n a l l y ' /saw mi / , tone 5 5 13 5 5 change i s o p t i o n a l , wi th , /saw mi ' / e q u a l l y p e r m i s s i b l e . A & /t, 2 2 . 1 3 , . i n , 2 2 , 5 5 . synonym, \ / h a w mx / xs u s u a l l y pronounced /haw 13 5 5 mi ' / i n the speech of the Zhong-shan in fo rmants i n the s tudy , 33 a l though Hashimoto (p. 97) i n d i c a t e s the o p t x o n a l i t y of Ih-ew 2 4 33 2 4 5 5 mej J versus [hnw mej ' J . Note a l s o t h a t i n Zhong-shan, both s y l l a b l e s undergo tone change. We t u r n now to the more s y s t e m a t i c grammatical f u n c t i o n of tone change, a l l of which i n v o l v e M o d i f i e d Tone 3 . One f u n c t i o n of M o d i f i e d Tone 3 , f o r i n s t a n c e , i s to i n t e n s i f y the a d j e c t i v e i n a r e d u p l i c a t i n g , m o n o s y l l a b i c a d j e c t i v e i n which the tone change - 71 -occurs on the f i r s t member of the reduplicating p a i r . Thus, 13 'good', for example, i s hao Jj$- /how /, and 'very good 1 i s 1 ^ ^5 13 22 /how ' how /. Correspondingly, 'big' i s da ^ / t a : j /, j 22 35 22 while 'very big' i s 7 ^ * 7 ^ / t a : j ' t a : j /. As already stated, the derived tone i s longer, containing a high r i s i n g con-tour. Besides i n t e n s i f i c a t i o n , a reduplicated adjective also has the meaning of 'however + ADJECTIVE 1; thus, 7 ^ * 7^- / t a : j ' 22 ta: j / also has the meaning of 'however big', as i n the follow-ing clause C'NEG" = negative): ( 3 ) K* K & h $ (-> ' .22,35 ^ .22 , 55 13 . 22 ta:j ta: j tu mow j ung CBIG BIG STILL NEG-HAVE USE) 'However big ( i t i s ) , ( i t ' s ) s t i l l no use.' In the case of an adjective modified by the character hao /how13/, meaning 'very', emphasis can be placed on the word 'very' by means of tone change to i n t e n s i f y the adjective follow-ing i t . Thus, the phrase 'very f a r ' ^ 3 " l i b - /how 1 3 y n 1 3 / can be \ 13 35 13 stressed by rendering i t as Jfo*/how ' yn /, denoting even greater distance. The character j_i / k i 1 3 / , on the other hand, has the meaning of ' f a i r l y , rather', and only i n the modified tone does i t acquire the meaning of 'very', serving to in t e n s i f y the adjective. 'Quite good 1, for example, i s 3|;4V/ki 1 3 how13/, while / k i ' how /, with the changed tone, means 'very good, stupen-dous'. A rough t r a n s l a t i o n a l equivalent of sentence (4), for example, would be 'It sure tasted good!' or 'It sure was good!' ("S" = sentence, "PRT" = pa r t i c l e . ) - 72 -(4) ° . Yl3,35 u 13 . ,2 . 22 k i how ja:k l a (VERY GOOD EAT S-PRT) 'It was very good (to eat).' Another regular function of the bian-yin i s to compensate for a deleted s y l l a b l e containing a high tone /5/ or /55/; that i s , the pitch of the tone i s absorbed by the immediately preced-ing s y l l a b l e when the s y l l a b l e which o r i g i n a l l y c a r r i e d i t was deleted. One case of such a usage i s when there i s a change i n the tone of a verb. The tone change serves to mark p e r f e c t i v i t y 55 i n l i e u of the post-verbal perfective marker /p'ow / i n Zhong-55 55 55 shan. Variant forms of /p'ow / are /how / and /ow /. The perfective marker may sometimes be the repeat ..of the f i n a l on the verb i n which the repeated f i n a l c a r r i e s the high pitch, of Tone 1. j. 22 In the case of qu 'to go' /hy /, for example, the perfective 22 55 22 55 of the verb may be /hy p'ow / (or i t s var i a n t s ) , or /hy y /. 22 35 P e r f e c t i v i z a t i o n through tone change would y i e l d /hy ' /. In Zhong-shan, the marker of p e r f e c t i v i t y *|J«L (or *fi. ) /tso / i s also used, but i t i s considered a more formal or p o l i t e form bor-rowed from Cantonese. In analyzing the p e r f e c t i v i z a t i o n of verbs by means of tone change, one theory that has been advanced for Cantonese i s likewise proposed here, namely, that the tone change i s a r e s u l t of copying the high pitch of the perfective marker onto the end of the preceding s y l l a b l e , which i s the verb. When the perfective marker i s deleted, a trace remains i n the compensatory lengthening of the preceding s y l l a b l e , with a concomitant r i s e i n the contour of that s y l l a b l e at i t s end point. There are a few cases of tone change connected with the - 73 -5 deletion of jjL 'one' / j a t /. Three cases of y i - d e l e t i o n and concomitant tone change w i l l be discussed. As i n p e r f e c t i v i z a -t i o n by means of bian-yin above, the tone change here also serves a compensatory r o l e . The f i r s t case of y i - d e l e t i o n involves t r i -s y l l a b i c phrases i n which, the f i r s t s y l l a b l e i s a monosyllabic verb, the second yjL ~-~ , and the t h i r d the reduplication of the monosyllabic verb. The deletion of y i results i n the f i r s t verb form acquiring Modified Tone 3. Thus, 'to take a look' kan-yi-kan i / 22 5 22 ^jj -~ ^ jj , for instance, i s /hon j a t hon / (.'look one look'). 5 The deletion of 'one' / j a t / produces the long, high r i s i n g tone . i 2235 22 on the f i r s t s y l l a b l e y i e l d i n g & * j|j /hon ' hon /. As observed by Kwok (.1971:50). for Cantonese, the resultant r e d u p l i -cated verb form denotes a b r i e f duration of action denoted by the verb. In his analysis of Mandarin, Chao (1968:204) refers to this reduplication as the "tentative aspect of verbs". In the second case of Modified Tone 3 a r i s i n g from the 5 deletion of 'one' / j a t /, the singular occurrence of an action 5 13 can be s p e c i f i e d using the expression y i - x i a ^- e"f* / j a t ha / 'one time/occasion'. CXia nrf (or ~f ) i s a c l a s s i f i e r for the number of occurrences of an action.) When an action occurs once, the number 'one' i s usually omitted. Thus, kan y i - x i a ^ *"f 22 5 13 'to look once' /hon j a t ha / (.'look one time'), becomes 22 35 13 /hon r ha /, with a tone change on the verb. The meaning of 5 13 / j a t ha / i s sometimes extended so that i t does not necessarily always haye the l i t e r a l meaning of 'one time/occasion', as shown i n (.5).. (A c i r c l e " O " i s used to designate a c o l l o q u i a l word with, no written character associated with i t . "CL" = c l a s s i f i e r . ) - 74 -<5> a ' ^ V a S ' f 13 ^ 5 5 n 22 hon ha ..sin syn (LOOK+ONE CL BEFORE PLAN) 'We'll see.'/'Wait and see f i r s t . ' D' Zl 13 13,35 . 13 how yn ' ha (GOOD FAR+ONE CL) 'It's kind of f a r . ' / ' I t ' s quite far (contrary to expecta-tion) . ' c* 13,35 , 13 . 22 51 . .22 . . 5 ja:ng ha tsaw m k i tak (SHADOW+ONE CL THEN NEG REMEMBER ABLE) 'One moment l a t e r (I) (already) forgot!' ( i . e . , i n an instant, very quickly) , , 51 . .55* . 1 3 * 22 22 k'y ha:j ha tsaw ha:m (HE/SHE TOUCH CL THEN CRY) 'As for him, just brush against (him) l i g h t l y and (he) c r i e s . ' In the t h i r d case of compensatory tone change a f t e r y i -deletion, the change of Modified Tone 3 occurs when the second y i —~ i s omitted i n expression containing " y i + CL + y i + CL", such as yi-kuai-yi-kuai — "~ 'piece by piece' ( i . e . , one piece after another) / j a t 5 f a : j 2 2 j a t 5 f a : j 2 2 / ('one piece one p i e c e 1 ) . Deletion of the second y i yi e l d s — J$k* £5&>/jat f a : j ' fasj /, This tone change i s a general one af f e c t i n g any reduplicated c l a s s i f i e r i n the above environment. Note that i f a c l a s s i f i e r i s simply reduplicated, the expression has the meaning of "every + CL". In Zhong-shan the f i r s t member of such a reduplicated p a i r J ^ 2 2 2 2 does not undergo tone change. Thus, / f a : j f a : j / 33 22 (phonetically I f A i j f A i j 1 due to tone sandhi noted e a r l i e r ) means 'every piece'. (Contrast t h i s with Cantonese 44 35 44 I f A i j f A i j ] 'every piece'.) There are also some cases of compensatory tone change due - 75 -to s y l l a b l e deletion that i s highly i d i o s y n c r a t i c and hence non-productive. The expression ^ / k i 1 3 t o ^ / 'how "much, how many', for example, i s often reduced to simply ^ * / k i 1 3 ' " ^ / . Thus, the question 'How much did i t cost?' may be uttered as (6a). or (6b). ("Q" = question.) (6) a. * * 1 3 J ? 5 5 5 1 B. 1 3 2 2 k i . . t o t s i n ma:] a (HOW-MUCH MANY MONEY BUY Q-PRT) 'How much did (it) cost?' 13 22 k i ' t s i n " " " ma:j a (HOW-MUCH MONEY BUY Q-PRT) 'How much did (it) cost?' Tone change can also serve to replace the word dou ^ 55 'also, likewise' /tu / as a marker of inc l u s i o n (INCL). (Note that dou ^jjj* has the same meaning i n Cantonese that i t has i n Zhong-shan, but i n Mandarin dou means ' a l l , altogether'.) The sentence 'I'm going, too', for example, may be uttered as 7 (a) or (b) (7) a. & 1 3 lf55 * 2 2 ngo tu hy (I ALSO GO) 'I'm going, too.' nTo13'35 hy2 (I+INCL GO) 'I'm going, too.' The r e s u l t a t i v e or s e r i a l verb 'to return, give back (some-* ^/ 1 3 5 5 1335 thing)' O ife-- /pia:ng fa:n / i s often reduced to /pia:ng ' /, as i n (8) . ngo pia:ng fa:n k'y l u (I GIVE RETURN HIM/HER S-PRT) 'I have returned (it) to him.' - 76 -b - & 1 3 . ° * 1 3 , 3 5 . 5 1 n ° 2 2 ngo pia:ng ' k'y l u (I GIVE+BETURN HIM/HER S-PRT) 11 have returned (it) to him. 1 There are also at least a couple of tone changes to Modi-f i e d Tone 3 which are highly i r r e g u l a r . The verb you 'by 5 1 (someone—in passive constructions)' /jaw / i n the changed tone means 'to allow, to be up to (someone to decide)', as i n sentence (.9) • (9) ** >L o i y ; . 5 1 , 3 5 . , 5 1 , 2 2 jaw k'y l a (ALLOW HIM/HER/IT S-PRT) 'Let i t be./Don't bother with it./Leave i t alone. 1 \- 51 Another example i s the word mang "^ j 'blind' /ma:ng / 5 1 3 5 which, i n the Modified Tone 3 /ma:ng ' /,..is used to describe some action done b l i n d l y , fervently or p e r s i s t e n t l y , as i n ( 1 0 ) . < 1 0 > * 5 1 r 5 1 , 3 5 ° / 2 2 ^ 2 k'y ma:ng ' kam ja:k (HE BLINDLY THUS EAT) 'He kept on eating ( f u r i o u s l y ) . ' The l a s t example involves interrogatives containing the c o l l o q u i a l word dian 'how' / t i m 1 3 / . In Cantonese, a c o l l o -1 **** q u i a l expression such as 'how, i n what way' i s 3 5 3 3 1 3 [ti:m qceirj ]. In Zhong-shan, /tim / undergoes tone change to the high, r i s i n g Modified Tone 3 for the same expression: ,u - jo o r 2 2 1 3 /tim ' Jong /. Likewise, other combinations with /tim / for interrogatives r e s u l t i n a tone change i n the word; for example, 'why' i s / t i m 1 3 ' 3 5 k a : j 1 3 / ; O / t i m 1 3 ' 3 5 t s i 1 3 / means 'how, by what means'. - 77 -1.4. Combination of I n i t i a l s and Finals Phonemically, including the zero i n i t i a l , there are eighteen i n i t i a l s i n Zhong-shan. Of the f i n a l s there are a t o t a l of seventy, i n which tonal d i s t i n c t i o n s have not yet been taken into account. When these are also considered, the combination of the i n i t i a l s and f i n a l s , together with tonal d i f f e r e n t i a t i o n , y i e l d s a sum of at lea s t 1,600 d i f f e r e n t s y l l a b l e s i n the Zhong-shan d i a l e c t . This figure i s , of course, less than the absolute potential number, which, may be the re s u l t of natural gaps here and there, occurring at random i n a language, a r i s i n g from a "defective" d i s t r i b u t i o n of some segments; that i s , there are pot e n t i a l l y permissible combinations that are not actualized i n the d i a l e c t . Some of the gaps are systematic, and may be the consequence of mergers i n the d i a l e c t which are phonologically conditioned, a point which w i l l become clearer l a t e r . Other gaps are the re s u l t of co-occurrence r e s t r i c t i o n s . The most wide-spread case of co-occurrence r e s t r i c t i o n among the Chinese d i a l e c t s i s l a b i a l d i s s i m i l a t i o n , which i s also c h a r a c t e r i s t i c of Zhong-shan. I t i s a topic that w i l l be discussed next. Besides the phenomenon of l a b i a l d i s s i m i l a t i o n , we w i l l also examine that of syncope, the deletion or loss of a phonolo-g i c a l segment r e s u l t i n g i n the contraction of a word. The p a r t i a l reduction of a word creating a sesquisyllabic structure w i l l also be studied, with a discussion of the d i s t i n c t i o n between c o l l o q u i a l versus l i t e r a r y forms i n the d i a l e c t concluding t h i s chapter. 1.4.1. Labial D i s s i m i l a t i o n There are several environments i n which the l a b i a l - 78 -d i s s i m i l a t i o n process occurs i n Zhong-shan. They are e s s e n t i a l l y the same environments as those found i n Cantonese, In both d i a -l e c t s , l a b i a l i n i t i a l s (including velars followed by medial /w/, a l t e r n a t i v e l y analyzed i n Cantonese as l a b i a l i z e d v e l a r i n i t i a l s ) cannot occur together with l a b i a l consonantal endings. Nor can the l a b i a l glide i n i t i a l co-occur with the l a b i a l glide ending. Thus, the following are examples of impermissible s y l l a b l e s i n both Zhong-shan and Cantonese: [*mAi.p, *p'i:p, *moim, *fo:m, *kwi:.m, *k'wAip, *wo:p, *wew] . There i s also a pr o h i b i t i o n of front rounded vowels either preceded by a l a b i a l i n i t i a l , or followed by a l a b i a l ending. The s y l l a b l e s [*0w, *0p, *yip, *p0, *p'y:], for instance, do not occur i n either Zhong-shan or Cantonese. 1.4.2. Syncope A few examples w i l l be c i t e d . The c o l l o q u i a l expression for 'what', for instance, i s % ^ /mat5 j a 1 3 / which, when short-51 ened by syncope, becomes /mia /. The negative imperative 'don't' /m51 hou 1 3/ (.'not good') i s sometimes reduced to /mow13/. k 24 (The alternative term of [mej ] i n Cantonese i s not used i n 22 2 Zhong-shan.) 'Twenty' i s er-shi -\ /ngi sap / ('..two ten') , which can be syncopated to if /ngap 2/. This syncopated form i s more frequently used for figures between twenty-one and twenty-nine i n c l u s i v e than for the number 'twenty' by i t s e l f ; for example, u- -* 2 55 'twenty-three' i s rendered - f /ngap sa:m / f r o m er-shi-san i - 22 2 55 S^— ~T /ngi sap sa:m / ('two ten t h r e e 1 ) . (In Zhong-shan, -* -J, 55 2 ' t h i r t y ' san-shi Ju. T /sa:m sap / (.'three ten') often becomes simply . / s a 5 1 / , since the preceding process of syncope, i f - 79 -2 applied to the number ' t h i r t y 1 , would have yielded /sap /, lead^ ing to confusion with the same s y l l a b l e already being used to mean 'ten' .) There are some words for which only the syncopated form seems to have survived. The d e i c t i c words 'here' and 'there' 5 5 appear to be a fusion of O /ku / 'th i s ' and the diminutive-p a r t i c l e er / n g i 5 1 / to y i e l d 0 / k u j 5 5 / 'here', and O 55 51 55 /nu / 'that' plus /ngi / combine to form O /nuj / 'there'. This i s p a r a l l e l to Mandarin zhei 'jX, for 'this' and na f)^ for 'that', with the corresponding locative terms zher yg\^j(.zhei Vl\ + er 'tfij ) for 'here' and ner (na J)J + er ) for 'there' . In 55 55 Zhong-shan, /kuj / and /nuj / also mean 'this thing/matter' and 'that thing/matter' respectively. The same p a r t i c l e er /ngi"* 1/ mentioned above was 5 5 probably also fused i n the interrogative word O /naj / 'where', 5 5 as was postulated by our informant, Chen, i n which O /naj / i s the syncope of na ^ /no 5 1/ 'where' (which only occurs i n l i t e r a r y • • in 51 usage) and the diminutive p a r t i c l e er /ngi /. (Compare thi s with Mandarin nar (na ^ + er ) 'where'.) 1.4.3. Sesquisyllabic Structures While the examples i n the preceding section show the reduc-t i o n of two s y l l a b l e s to a single one by syncope, there are other s y l l a b l e s i n Zhong-shan where th i s merger i s only p a r t i a l l y accom-plished, r e s u l t i n g i n a "sesquisyllabic structure", a term used by J. Matisoff (.1973:86) to refer to morphemes i n Proto-Austro-A s i a t i c (a l i n g u i s t i c superstock which includes, among i t s members, Mon-Kmer and Viet-Muong) which were "a s y l l a b l e and a h a l f " i n - 80 -20 length : "(t) hat i s , the prevoealic consonant was often preceded by-a ' p r e - i n i t i a l ' consonant, as i n the modern Cambodian words psaa 'market', tkiam 'jaw', ckat 'dog', kr}aok 'peacock'". In Zhong-shan, there are a few cases of words which can be said to contain sesquisyllables. Such words i n the d i a l e c t have consonant clusters which are separated by an epenthetic schwa. From a synchronic l e v e l , these sesquisyllabic structures can be analyzed as reductions of f u l l form which are p o l y s y l l a b i c i n o r i g i n . This process of schwa-reduction i s moreover r e s t r i c t e d to words i n which the second s y l l a b l e undergoing reduction contains the i n i t i a l / l / . The schwa-reduction process actually affects a very l i m i t e d vocabulary. With the exception of a small handful of c o l l o q u i a l expressions found i n the data thus f a r , other mani-festations of t h i s phenomenon involve onomatopoeic types of words. Both the f u l l and reduced forms are found i n Zhong-shan, with the f u l l form e s s e n t i a l l y c i t a t i o n forms. I t i s the reduced forms that are normally used i n d a i l y , conversational speech. Given the s y l l a b l e structure that normally does not admit to a sesquisyllabic structure, the synchronic analysis very naturally seeks to derive the reduced form from a . f u l l , p o l y s y l l a b i c one. Diachronically, however, there are arguments for proposing o r i g i n a l i n i t i a l consonant clusters for at least some of the sesquisyllabic forms i n Zhong-shan; that i s , some of the sesquisyllabic forms can be associated with words which have been reconstructed i n Old Cor Archaic). Chinese as containing i n i t i a l consonant clusters. Consider, f i r s t , the c o l l o q u i a l word for 'corner-(e.g., of a room)_1 . The characters usually given for i t i n Cantonese and Tai-shan (another Yue dialect) are ji.fl . They are pronounced 4 5 35 1 5 35 Ikoxk l o i k t 'Bw J i n Cantonese and [kok Iok hau J i n Tai-shan. The same term i s also found i n Zhong-shan. The c i t a t i o n form for 2 5 51 i t i s Ikax.k l o i k t'-ew J. C o l l o q u i a l l y , the word i s usually reduced to I k 3 l o : k 5 t ' B w 5 1 ] (,/klok5 t'aw 5 1/). Hashimoto (.1972b: 34) expresses doubt concerning the connection between the second • ti s y l l a b l e and the character luo (.169-21) , which i s normally pronounced i n a low tone i n Zhong-shan, as i t i s i n Cantonese. Hashimoto suggests that both, the f i r s t and the second s y l l a b l e i n fact stand for the character j i a o jj) (.183-20) , and further pro-poses that j i a o jl) i s derived from a h i s t o r i c a l * k l - c l u s t e r i n proto-Yue. (Tou £1^ i s simply a word-formative s u f f i x (Kratoch-v i l , 1968:68), and i s not relevant to the present discussion,) Hashimoto's evidence for claiming that the f i r s t two s y l l a b l e s both represent the character j i a o i s based on fan-qie: j i a o $J has two d i f f e r e n t fan-qie's which demonstrate that i t has both a *k- and an *\i i n i t i a l . Accordingly, Hashimoto hypothesizes that when i n i t i a l c lusters f e l l out of usage, the only means whereby these clusters could survive was by having a s y l l a b l e inserted between the two consonants i n the c l u s t e r . (Yang (.1971) , for example, referred to such an i n s e r t i o n — t h a t of a v o c a l i c element between adjacent consonants—as a "process of anaptyxis".) To account for the tonal difference between the f i r s t and second s y l -4 5 21 35 lable i n the word Ikaxk laxk t'BW ' J 'corner', Hashimoto suggests that the high tone of the second s y l l a b l e i s a r e s u l t of tone change, since such a change i s not uncommon i n reduplicated s y l l a b l e s — o r , i n t h i s case, pseudo-reduplicated s y l l a b l e s — i n the Yue d i a l e c t s . Various Chinese phonologists have likewise reconstructed an i n i t i a l consonant cluster for j i a o |2) , which i s a Grade II word. Jiao jt) has been reconstructed by Fa-kao Chou as Archaic Chinese *krewk, for instance, and by Pulleyblank as Old Chinese *krak . The Zhong-shan sesquisyllabic form for 'corner' would therefore lend support for some kind of * k l - or *kr- i n i t i a l consonant c l u s t e r for j i a o i n Old Chinese, a cl u s t e r which had survived i n protp-Yue, and apparently i n e a r l i e r forms of other 55 " 35 In '*£ d i a l e c t s as well, as witness Ika l s r J |*J }%• i n (Peking) Man-darin for 'a hidden corner' (Hashimoto, 1972b:33-34). The t r i -s y l l a b i c " f u l l " form for the word 'corner' i n Zhong-shan i s b a s i c a l l y a c i t a t i o n form which expands the sesquisyllable to f i t the more common CVC s y l l a b l e structure i n the d i a l e c t . The pro-21 cess of "dimidiation" , which p a r a l l e l s the h i s t o r i c a l process proposed by Hashimoto, i s used i n such situations as teaching the word to a c h i l d , or repeating the word slowly to an i n q u i s i t i v e and persistent l i n g u i s t . It should be noted that despite the f u l l form given by Hashimoto for the word 'corner' i n Cantonese, McCoy (1966:185,fn. 27) i n fact argues that t h i s word i s one of a very rare number of words i n Standard Cantonese which possess atonic s y l l a b l e s . McCoy indicates such s y l l a b l e s by using the tone d i a c r i t i c [°]. The nuclear vowel i n these s y l l a b l e s i s also reduced to a schwa and linked to the following s y l l a b l e by a hyphen to show close junc-ture. Of the words i n his data, McCoy found only two which con-4 35 22 t a i n an atonic s y l l a b l e : [ks° - loxk t'B W 1 'corner' i s one 33 33 of them, and [hem b a ° - I A X - O ] ' a l l , completely' i s the other. (.We w i l l return to the second word later.) In a l l p r o b a b i l i t y , McCoy's atonic s y l l a b l e , which has - 83 -close juncture with the following f u l l s y l l a b l e , i s what we are treating here as a sesquisyllable. However, McCoy seems to deny the existence of the f u l l , c i t a t i o n form; he notes that the forms, containing neutral tone s y l l a b l e s c i t e d i n d i c t i o n a r i e s are record-ed "as i f " they were pronounced with, f u l l tones. The fact that both f u l l and reduced forms can be e l i c i t e d from the present informants indicates at least that both variants do e x i s t and are not mutually exclusive. As suggested e a r l i e r , the c i t a t i o n form i s l i k e l y an attempt to expand the sesquisyllabic form to conform with the more orthodox s y l l a b l e structure of the d i a l e c t . Another c o l l o q u i a l expression i n Zhong-shan which ihas a i sesquisyllabic structure i s the word 'knot'. The c i t a t i o n form 5 5 ^ 5 i s [k'ixt l i : t ], and the reduced form [k' lix.t ]. Hashimoto (1972b: 37, fn..'19; 38,fn.26) notes that K. Whitaker (1952:47-48), i n her d i s s e r t a t i o n on the "Characterization of the Cantonese d i a -l e c t with, special reference to i t s modified tones", proposes an i n i t i a l * k l - c l u s t e r for the word 'knot' by making a connection between the l i t e r a r y word for 'knot' j i e (133-1) [kixt^J 2 (./kit / i n Zhong-shan) and the c o l l o q u i a l word for 'knot' i n 4 4 Cantonese, which has the variant forms of [ k i i t ] and [ l i x t ]• Hashimoto (1972b:33). , however, rejects Whitaker's proposal based on the alternative pronunciation of the c o l l o q u i a l word for 'knot 1 on the premise that the l a t t e r has no fan-qie evidence to support her claim. Hashimoto proposes, instead, a connection between the word x i e k*$. [sixt^] 'to t i e ; a b r i d l e ' and the c o l l o q u i a l word [li x t J. She bases her claim on the observation that the character *f!!i|_, for example, was supposed to have had an * i _ i n i t i a l accord-ing to fan-qie s p e l l i n g . (However, the fan-qie for xie js--t£ i n - 84 -both Chou (1973) and the C i - h a i d i c t i o n a r y , f o r i n s t a n c e , shows an * s - i n i t i a l o n l y . ) Nonethe less , on the b a s i s of her r a t h e r tenuous f a n - q i e ev idence , Hashimoto p o s i t s an * s l - c l u s t e r f o r the word x i e Zhong-shan ev idence , on the other hand, would tend to support W h i t a k e r ' s c l a i m of a * k l - c l u s t e r . Whi le Cantonese a l t e r n a t e s between a /k/ and an / ! / i n i t i a l f o r the c o l l o q u i a l word ' k n o t ' , i n Zhong-shan the same word has a d i s y l l a b i c s t r u c -5 5 5 tu re / k ' i t l i t / , or the s e s q u i s y l l a b i c form of / k l i t / . I f one can assume t h a t the words ' c o r n e r ' d i s c u s s e d above, and the present word ' k n o t ' underwent a s i m i l a r process i n the breakdown of i n i t i a l consonant c l u s t e r s , there i s s u f f i c i e n t b a s i s f o r s u g -g e s t i n g t h a t the word j i e / k i t 2 / ' k n o t ' o r i g i n a l l y bore some k i n d of * k l - c l u s t e r . H i s t o r i c a l l y , the r e c o n s t r u c t i o n of a * k l - c l u s t e r f o r j j e l£ , a Grade IV word, has been suggested by P u l l e y b l a n k (.196 2) , then as a * k i - c l u s t e r . P u l l e y b l a n k d i s t i n g u i s h e s between c l u s t e r s i n Old Chinese which c o n t a i n * - l - and those which c o n t a i n * - r - : * - r - occurs w i t h Grade I I words ( e . g . , j i a o ) , and * - i - w i t h Grade IV words ( e . g . , j i e j& )• Very b r i e f l y , P u l l e y b l a n k was mot ivated to p o s t u l a t e a * k l - c l u s t e r f o r such Grade IV words as j i (154-16) ' l u c k y ' / k a t 5 / (Middle Chinese * k j i t ) and j i e ~tj (154-17) ' t o i n v e s t i g a t e ' / k ' i t 2 / (Middle Chinese * k ' j i t ) . . i n o . order to e x p l a i n why they d i d not have p a l a t a l i n i t i a l s i n M idd le Ch inese , whereas Grade I I I words, such as z h i JL (48-12) ' b ranch ' 55 / t s i / (Ear l y M idd le Chinese *t<?ia) d i d develop p a l a t a l i n i t i a l s . I t was the presence of * - l - i n the Old Chinese forms of Grade IV words such as jjL a and j i e ' f S which b locked p a l a t a l i z a t i o n a t - 85 -the Middle Chinese stage. 5 In the modern Zhong-shan form of / k ' l i t / for 'knot', one s t i l l needs to account for the presence of an aspirated i n i t i a l instead of the unaspirated one i f an association i s to be made between the Zhong-shan c o l l o q u i a l term for 'knot' and the l i t e r a r y 2 one of j i e i/£ / k i t /. There i s also the question of the difference i n tone between the two forms. Neither poses a major d i f f i c u l t y . A number of c o l l o q u i a l (and l i t e r a r y ) words which are pronounced with an aspirated stop i n i t i a l i n Zhong-shan have been reconstructed with p l a i n stops. Words containing the Jian |L» (*k-) i n i t i a l , \ 22 for example, include gua jf%\* (45-1) 'to hang' /k'wa / and guang J^i, (180-1) 'to s t r o l l ' /k'wa.ng22/. Cantonese has [kw A: 4 4J and 33 IkwAiy J respectively for the two words. Mandarin, i t may be noted, likewise showsunaspirated i n i t i a l s for gua and guang . It i s therefore not implausible to consider the reconstruct t i o n of a p l a i n * k l - c luster despite the c o l l o q u i a l word 'knot' i n Zhong-shan showing aspiration on the i n i t i a l segment. Concerning the problem of a tonal difference i n the two forms for 'knot', one might f i r s t look at the c o l l o q u i a l form for 2 5 51 'corner'. In /kok Iok t'aw /, the f i r s t s y l l a b l e preserves the b 2 tonal value of the word ji a o ji| /kok /. Presumably, the second 5 s y l l a b l e /Iok / i s the r e s u l t of a tone change such that i t o r i -2 5 5 g i n a l l y bore tone /2/ ( i . e . , /Iok ' / ) . In the word 'knot' / k ' i t l i t 5 / , neither s y l l a b l e has preserved the Zhong-shan r e f l e x of the tone for j i e / k i t 2 / , namely tone /2/. I t i s possible that the word 'knot' may have undergone an intermediary stage during which only the second s y l l a b l e was at f i r s t affected by tone change (as i n the case of the word 'corner'), and i t i s only - 86 -subsequently that the f i r s t s y l l a b l e was also thus affected. As noted e a r l i e r , besides the word 'corner 1, McCoy (1966: 185,fn.27> also mentions the Cantonese word ' a l l / completely', 33 33 which i s phonetically Ihem ba° - IAI-O ] . The same word i s 33 33 33 transcribed as Ihem P A : I A I I J ] . by Hashimoto (p. 333). The phoneticization of Huang's (1970:394) t r a n s c r i p t i o n of 'altogether, 33 33 33 a l l t o l d ' i s Ihem p A i n I A I J J ] . In Zhong-shan, t h i s word ' a l l ' 51 51 51 51 is pronounced /ham pa la:ng / i n c i t a t i o n form and /ham 51 pla:ng / i n d a i l y speech. In narrow phonetic t r a n s c r i p t i o n , /p/ assimilates the voicing of the preceding segment and i s pronounced 5 1 a 51 l b ] : [h-em b I A I H J . Except for tonal differences, McCoy's form for Standard Cantonese and the corresponding Zhong-shan form are i d e n t i c a l . It i s i n s t r u c t i v e to add that two other Yue dia-l e c t s recorded by Hashimoto (p.19) likewise appear to contain sesquisyllables for the word ' a l l ' : Hashimoto records Ihorn2 plan 1] for ' a l l ' i n Teng-xian, and [Sham °plag] i n Tai-shan. Regarding the portion of the c o l l o q u i a l word ' a l l 1 which < : 51 i s the sesquisyllable (e.g., /pla:ng / i n Zhong-shan, or the 33 33 corresponding Cantonese sequence [ p A i 1A:O J)/ Pulleyblank (personal communication) suggests l i n k i n g i t to the l i t e r a r y word fan fa (109-1) ' a l l ' , derived from LMC *ff\aim<EMC *buam. (Fan 51 21 i s pronounced [ f A i n ] i n modern Zhong-shan and If A M I J i n Cantonese.) Pulleyblank points out that fan f\j i s the phonetic in a character such as fan (109-7) (/fa:n 5 1// EMC *b uam)/ which had been used to transcribe the word 'Brahma'. Fan ) \ i s also the phonetic i n lan , which i s LMC *lam, and modern Cantonese It i s not inconceivable that the word fan ]3u once bore a [ l A i m ] - 87 -consonant c l u s t e r * b l - or *br-. In Zhong-shan, the presence of 51 a velar nasal ending i n the modern r e f l e x /pla:ng / can be ea s i l y explained i n terms of l a b i a l d i s s i m i l a t i o n , changing *-m to - n i n the presence of the l a b i a l i n i t i a l *b-. In the s y l l a b l e s / p a 5 1 l a : n g 5 1 / and the sesquisyllable / p l a : n g 5 1 / , the tone i s /51/ thus corresponding to the same tonal r e f l e x that i s i n the l i t e r a r y 51 word fan )\> /fa:n /. Of the Yue di a l e c t s mentioned above, only i n Zhong-shan has the regular tonal r e f l e x of fan ^ been preseryed. 33 33 The tones i n Cantonese [p A x 1A:I) ], for example, dxd not pre-21 serve the Yang-ping [21] tone of fan [ f A i n ]• What i s s t i l l not e n t i r e l y clear i s the role of the s y l l a -51 51 51 ble /ham / i n Zhong-shan /ham plarng /. I t i s possible that the nasal segment /m/ serves to preserve the voicing of the h i s -t o r i c a l voiced *b- i n i t i a l , when voicing was no longer phonologic-51 23 a l l y d i s t i n c t i n the i n t i a l segment /p/ of the s y l l a b l e /pla:ng / The observation by McCoy and l a t e r by the writer that /p/ i s phone-t i c a l l y [bj i n the modern c o l l o q u i a l word ' a l l ' may not be fortu-51 33 itous a f t e r a l l . The s y l l a b l e /ham / C [ h - e m ] i n Cantonese) may be regarded as an i n t r i n s i c part of the h i s t o r i c a l reconstruction of the word fan : i t bears the voicing feature which would otherwise have been l o s t when *b- no longer contrasted with the corresponding voiceless l a b i a l stops i n the phonological system of the d i a l e c t . Another c o l l o q u i a l word i n Zhong-shan which can be associ-ated with a word which once contained an i n i t i a l consonant c l u s t e r 22 22 51 22 51 i s /ka l a wa:ng /, or /k l a wa:ng / 'crosswise; at ri g h t 33 angles to'. Long-du has a si m i l a r form for 'crosswise': /kaa 33 33 laa waaq / (tone numerals are used instead of Egerod's tone - 88 -d i a c r i t i c s , and /q/ - [IJ]) O O \% (Egerod, p.91). . Jv Norman (personal communication) does not recognize t h i s word as a regular Min colloquialism, and can only presume that i t i s borrowed from 24 Zhong-shan. Ignoring tonal differences between the two d i a l e c t a l forms for the moment, one can propose that the p o l y s y l l a b i c form of the word 'crosswise' i n both d i a l e c t s contain traces of an o r i g i n a l i n i t i a l consonant cl u s t e r . More precisely, the complete 22 22 51 33 33 forms of Zhong-shan /ka l a wa:ng / and Long-du /kaa laa waaq 3 3/ can be connected to the word heng , which i s a Xia i n i t i a l , Grade II word reconstructed by F.K. Chou, .for example, 22 as Archaic Chinese *grwang. The sesquisyllable / k l a / then pre-51 serves the i n i t i a l . *gr- cluster, and /wa:ng / i s simply the s y l l a b l e - f i n a l of the word. Regarding i n i t i a l *g- i n the cl u s t e r , phonologists have observed that at lea s t some of the words containing the Xia i n i t i a l i n Middle Chinese had common origins with velar stops, as r e f l e c t e d i n the reconstruction of i n i t i a l *g- i n Archaic Chinese by Chou for heng jfll , for example, and an aspirated *g---by Karlgren. In terms of h i s t o r i c a l phonology, the Zhong-shan po l y s y l l a b i c form for 'crosswise' thus lends support for the reconstruction of an e a r l i e r i n i t i a l c l u s t e r which contains a yelar element. More generally, the Zhong-shan p o l y s y l l a b i c form provides evidence for an i n t i a l consonant c l u s t e r for the word Turning now to the question of the tonal values recorded for the word 'crosswise' i n Zhong-shan as well as i n Long-du, one observation should f i r s t be made concerning heng (Zhong-shan 51 33 /wa:ng /, Long-du /waaq / ) . Both d i a l e c t a l forms are regular reflexes of the h i s t o r i c a l Yang-ping tone with respect to t h e i r respective pattern of correspondences to the h i s t o r i c a l phonolo-g i c a l system. Thus, the tones i n a l l three s y l l a b l e s i n the Long-33 33 33 du form /kaa laa waaq / are completely regular. In Zhong-shan, while / 5 l / i s the regular r e f l e x of Yang-ping, /22/ i n the 22 22 22 s y l l a b l e s /ka l a / and the sesquisyllable / k l a / i s not. One possible explanation for the exceptional tonal r e f l e x i s that 51 51 /ka / and / l a /, which would be the expected reflexes for 51 Zhong-shan, are extremely rare. In the present data, /ka / only occurs i n a c o l l o q u i a l , characterless word which refers to a leg-51 astride, standing po s i t i o n , while / l a / occurs i n the combination 51 55 it: / l a a:j / Q meaning ' d i r t y 1 , and as a sentence-final p a r t i -c l e . In contrast, tone /22/ counterparts of these s y l l a b l e s are much more numerous, and can be found i n regular l e x i c a l items. 51 51 Perhaps the r a r i t y of s y l l a b l e s /ka / and / l a / motivated the tone change to /22/ i n Zhong-shan. No firm conclusions w i l l be attempted at t h i s time. In the next word, Zhong-shan has two c i t a t i o n forms and one corresponding sesquisyllabic form. To describe an instrument or object that i s very straight, or a road or route that i s very 22 5 5 straight or d i r e c t , Zhong-shan uses the phrase /tim pat l a t / 22 5 5 %, O O , or /tim pat pat /. The sesquisyllabic form i s 22 5 22 /tim pl a t /. For the f i r s t s y l l a b l e /tim /, Cantonese likewise 33 has the word [ti:m J meaning 'straight', and coined .the character to represent that c o l l o q u i a l word. 5 What interests us here i s the sesquisyllable / p l a t / which one would l i k e to associate with the word b i 3p (153-12) 'a writing brush; straight, d i r e c t ' , which has been reconstructed - 90 -w i t h an i n i t i a l * p l - c l u s t e r . K a r l g r e n , f o r example, r e c o n s t r u c t s b i -jt as A r c h a i c Chinese * p l i s t ; Chou r e c o n s t r u c t s i t as * p l i w 9 t . A * p l - c l u s t e r h a s : a l s o been p o s t u l a t e d by P u l l e y b l a n k (1962:111) f o r b i , which i s a Grade I I I word be long ing to the s o - c a l l e d chong -n iu J[ £3- (..'double k n o t ' ) or Grade I I I/XV doub le t f i n a l s . X i e - s h e n g evidence i n Midd le Chinese shows contac t of phonet i c "^p w i t h both i n i t i a l *p- as i n b i / p a t 5 / and * 1 - i n l u ^jp 2 (.162-10) ' a law' /10t / . P u l l e y b l a n k a l s o brought the w r i t e r ' s a t t e n t i o n to h i s t o r i c a l ev idence of b u - l u ^ ^ (Zhong-shan 5 2 VK /pat IfSt /) be ing used to render the p r o n u n c i a t i o n of b i jE" . I t would appear t h a t i n Zhong-shan, the i n c l u s i o n of j$i / t i m 2 2 / meaning ' s t r a i g h t ' was used to r e i n f o r c e the meaning of / p l a t 5 / as the a s s o c i a t i o n of b i Jjjf to the * p l - i n i t i a l c l u s t e r became opaque a f t e r the genera l l o s s of i n i t i a l c l u s t e r s i n Ch inese . E v i d e n t l y , s y l l a b l e s c o n t a i n i n g / p i / then became viewed as some s o r t of onomatopoeic s y l l a b l e . One should suspect 22 5 5 22 5 t h a t of the two c i t a t i o n forms / t i m pat l a t / and / t i m pat 5 pat / , the l a t t e r i s a more recent i n n o v a t i o n , where the o r i g i n a l 5 5 5 meaning of /pat l a t / or / p l a t / i s complete ly obscured . Somewhat more t e n t a t i v e i s the l i n k a g e betwen the Grade I I I , He-kou word juan (142-13) / k y n 1 3 / or / k u n 1 3 / ' a r o l l ' and the c o l l o q u i a l word f o r ' s o m e r s a u l t ' s u p p l i e d by K a r l L o , 55 55 13 *L another Zhong-shan speaker : /kwa:n l a : n taw / O O >y i n 55 13 c i t a t i o n form and / k l a : n taw / i n reduced form. A l though no i n i t i a l * k l - c l u s t e r i s u s u a l l y r e c o n s t r u c t e d f o r the word juan ( e . g . , K a r l g r e n : A r c h a i c Chinese *kiwan) due to the absence of x i e - s h e n g e v i d e n c e , i t might be observed t h a t an i n i t i a l * k l -was hypothes i zed by P u l l e y b l a n k (1962:126) f o r the word. In t h i s - 9.1 -case, Pulleyblank treats *-.!- as a derivational i n f i x based on Wulff's theory of a morphological i n f i x * 1 . 2 5 Thus, juan 'turn around' i s Middle Chinese *kiwan, whereas juan ^ ' r o l l ' i s Middle Chinese *kiwen < Old Chinese *kwl5n. 55 55 55 If Zhong-shan /kwa:n la:n / and /kla:n / can be linked to the word juan , i t i s int e r e s t i n g that the c i t a t i o n form would be the one that i n fact preserves the l a b i a l medial. In this case, one would expect that the c i t a t i o n form and the sesqui-s y l l a b i c form both evolved simultaneously from the dimidiation of the i n i t i a l c l u s t e r , with the c i t a t i o n form preserving the He-kou feature. Conceivably, at an e a r l i e r stage, the sesqu i s y l l a b i c form had i n i t i a l *kw 3l-. The f i n a l i n the Zhong-shan c o l l o q u i a l expression s t i l l presents a problem since the regular Zhong-shan re f l e x for juan i s /yn/. The writer w i l l leave t h i s problem for future investigation. F i n a l l y , there i s also the l i t e r a r y word g_e f[%T (170-10) 2 'armpit' /kok /, which Karlgren has reconstructed as Archaic Chinese *klak. A * k l - c l u s t e r has also been postulated by other phonologists, sometimes with the h i s t o r i c a l l y homophonous word ge ^- (170-7) 'each' /kok / representing ge and others i n th i s set of Grade I words. Thus, L i (1974:251) and Pulleyblank (.1962:119), for example, both posit an o r i g i n a l * k l - i n i t i a l c l u s t e r for g_e ^ : Archaic Chinese *klak by L i , and Old Chinese *kiak (since revised as *klak) by Pulleyblank. C o l l o q u i a l data from a number of Chinese d i a l e c t s (cf. esp e c i a l l y Yang, 1971) strongly supports the postulation of an i n i t i a l consonant clu s t e r for ge . Consider f i r s t the Cantonese example. Included among ^he various c o l l o q u i a l forms for 'armpit' i n Cantonese i s that of [ k A i k * * lAik - 1 t - B j J J ] (Hashimoto, pp.242, 330), for which Hashimoto uses the characters ge-le-di Wh/fc* • P a r a l l e l to the case of the word 'corner', one can postulate that 4 5 the f i r s t two s y l l a b l e s , i.IkAik J and IlAi.k ], both represent the word ge , which o r i g i n a l l y bore a * k l - c l u s t e r i n i t i a l , as reconstructed by Karlgren and others. (The s u f f i x d i ^ , which means 'underside, base', does not enter into the reconstruction of the word g_e j ^ ^ o " . ) Also p a r a l l e l to the case of the word 'cor-ner' i s the tone change i n the second s y l l a b l e : the basic tone of the character l e faf) i s Zhong Yin-ru /4/ ( i . e . , { l A i k 4 ] ) . As argued i n the word 'corner', the character used to represent the second s y l l a b l e i s e s s e n t i a l l y a dummy element with both the f i r s t and the second s y l l a b l e having been derived from the f i r s t character h i s t o r i c a l l y . Dong-guan, another Yue d i a l e c t , has a c o l l o q u i a l form for 'armpit' which i s very close to the Cantonese form; Ikak lak haj M/^ "f (from Yang, 1971—no tone marks provided) . (The s u f f i x xi a "p means 'below, under', and does not pa r t i c i p a t e i n the reconstruction.) The same arguments used i n the Cantonese form for postulating a * k l - c l u s t e r can be applied here. Zhong-shan also has a c o l l o q u i a l form for 'armpit' which reveals an i n i t i a l * k l - c l u s t e r . The Zhong-shan form, however, had undergone a t t r i t i o n of the f i n a l /k/ ending and p i t c h - r a i s i n g 55 55 22 /-\ ~r" to a high tone y i e l d i n g /ka l a ha / O O r as the c i t a t i o n 55 22 form, and / k l a ha / as the sesquisyllabic form. I t i s in s t r u c -t i v e to know that the Fu-zhou form (supplied by Norman). , l i k e the Zhong-shan one, shows loss of the stop ending, and i s phonetically [ k o 2 2 r o u ^ a 2 4 2 ] }%T~^ • Since i t i s at y p i c a l for Zhong-shan - 93 -to lose i t s stop endings, i t i s possible that the a t t r i t i o n of f i n a l /k/ i n the Zhong-shan form may be due to influence from Fu-zhou or another Northeastern Min d i a l e c t . On the basis of comparisons with, the Cantonese, Dong-guan and Fu-zhou forms, one can be quite confident that the Zhong-shan form likewise repre-sents an i n i t i a l consonant c l u s t e r for the word g_e jj^j- . The writer i s actually not the f i r s t to use p o l y s y l l a b i c , c o l l o q u i a l forms to support the reconstruction of a * k l - c l u s t e r for g_e . Among the pioneers to do so i s Yang (1971) who has not only c o l l e c t e d such forms from a number of Chinese d i a l e c t s , but has also included cognates i n other East Asian languages i n order to support arguments for an e a r l i e r existence of a * k l -cl u s t e r i n the word g_e . 'armpit'. Thus, cognates which Yang found for the word 'larmpit' include kliak i n Khmer (.Cambodian) and ke1ek i n Malay. Pulleyblank (personal communication) also proposes that 55 the c o l l o q u i a l , p o l y s y l l a b i c word i n Mandarin for 'armpit' [ka 5 5 t§\ WUD ] may be i d e n t i f i e d with the word ge jffir (_*klak) or a closely related form. For instance, i n what Pulleyblank c a l l s his Type B s y l l a b l e s (for our present purposes, they are Grade III and IV s y l l a b l e s which he reconstructs as ..containing medial * - i - i n the Kai-kou series i n LMC), the development of Old Chinese *klak i s *klak > *k-^.iajk > k-tcsiajk > k-tcsi (reconstruction h i s ) . A velar for the i n i t i a l consonant and an a f f r i c a t e (palatal or retroflex) for the second s y l l a b l e seems to be quite prevalent among the Mandarin d i a l e c t s . The same pattern is, also found i n Su-zhou (Wu) . (See, e.g., Yang, 1971; Han-yu Fang-yan Ci-hui (.'A c o l l e c t i o n of Chinese d i a l e c t a l words') by the same editors as the - 94 -Zi-hui (hence, hereafter ' Ci-hui' for short) (1964:195);.) Concluding the subject of ge , i t should be noted that although the Zhong-shan c i t a t i o n form / k a 5 5 l a 5 5 h a 2 2 / merely adds to Yang's already long l i s t of p o l y s y l l a b i c forms i n various 55 Chinese d i a l e c t s for 'armpit', the sesquisyllabic form / k l a / i s s i g n i f i c a n t i n that i t may attest to an e a r l i e r stage of the evo-l u t i o n of the consonant cl u s t e r . In the present as well as pre-viously discussed reduced forms, one may regard the schwa inserted between the i n i t i a l consonants i n the sesquisyllable as the f i r s t stage i n the anaptyxis process; that i s , the ses q u i s y l l a b i c struc-ture may represent the stage at which adjacent consonants i n s y l l a -b l e - i n i t i a l position were separated by means of schwa-epenthesis, thus creating a sesquisyllable. Subsequent changes led to tone alterations i n some cases, and a t t r i t i o n of segments i n others. The breaking up of the i n i t i a l consonant c l u s t e r , nonetheless, i s captured and preserved i n the sesquisyllabic form. As observed i n McCoy's recording of the c o l l o q u i a l word ' a l l ' and Hashimoto's recording of the same word i n Teng-xian and Tai-shan, i t appears that the sesquisyllable exists i n other Yue d i a l e c t s besides Zhong-shan, and may i n fact be present i n non-Yue d i a l e c t s also. The existence of the c i t a t i o n form, which i s the form normally recorded i n fieldwork, i s then a further progres-sion away from the i n i t i a l c l u s t e r s . As suggested e a r l i e r , the c i t a t i o n form e s s e n t i a l l y represents an o r i g i n a l sesquisyllable which had been made to conform to the more pre v a i l i n g CVC structure. - 95 -1.4.4. Co l l o q u i a l versus L i t e r a r y Forms Among many Chinese d i a l e c t s , there i s a d i s t i n c t i o n between what are recognized as c o l l o q u i a l , or bai \;D , forms and what are regarded as l i t e r a r y , or wen , forms of words which are phono-l o g i c a l l y and semantically related. Generally speaking, the ... :c l i t e r a r y forms r e f l e c t influence from the standard language. In the case of a d i a l e c t which i s influenced by both a regional and a national standard, or one which incorporates forms from one or more dia l e c t s at d i f f e r e n t stages i n i t s history, several forms for a word may e x i s t side by side, r e f l e c t i n g the various sources or layers of borrowing. One example i s the Min d i a l e c t i n southern China i n which one often encounters as many as three layers of pronunciation for a given word. As i n Cantonese, Zhong-shan has two major categories of c o l l o q u i a l morphemes. The f i r s t category consists of purely c o l l o -q u i a l terms which generally lack character representation, with the exception of those special characters coined for Cantonese which can also be used i n Zhong-shan to represent c e r t a i n character-less words. A number of these c o l l o q u i a l Zhong-shan terms are included i n the lexicon i n Chapter 3li The second major category of c o l l o q u i a l morphemes consists of characters which have both a c o l l o q u i a l and a l i t e r a r y reading. The two forms can normally share the same syntactic environment. The c o l l o q u i a l form i s normally used i n d a i l y or informal speech, while the l i t e r a r y form only occurs i n formal speech or i n the reading of written texts. In general, what i s analyzed as the c o l l o q u i a l reading r e f l e c t s an e a r l i e r layer of the Chinese lan-guage, and the l i t e r a r y reading a l a t e r layer. - 96 -The d i s t r i b u t i o n of the c o l l o q u i a l versus l i t e r a r y forms i s , as Hashimoto (p.118) emphasizes, not necessarily exclusive. This i s p a r t i c u l a r l y the s i t u a t i o n with regard to p o l y s y l l a b i c words which may u t i l i z e a c o l l o q u i a l form i n one combination and a l i t e r a r y form i n another. Thus., c o l l o q u i a l l y , a 'provincial however, i s cheng-shi ^ j 1 /sing"""1" si" 1"""/; here, the l i t e r a r y pronunciation of cheng i s used instead. There i s no known general rule governing the choice of a c o l l o q u i a l rather than a l i t e r a r y form (or vice versa) i n a p a r t i c u l a r combination. There are also morphemes which do not have s t y l i s t i c alternants, with one form »£. 55 serving a l l functions. The morpheme qing (2-0-7) / t s ' i n g / 'clear, pure', for instance, only has the l i t e r a r y form regardless of usage or combination, whereas the morpheme jin g (202-1) 1 3 /kia:ng / 'neck' only has a c o l l o q u i a l form. c o l l o q u i a l and l i t e r a r y forms, i t i s obvious that the term " c o l l o -q u i a l " does not apply s p e c i f i c a l l y and solely to "spoken" forms, any more than does the term " l i t e r a r y " apply only to "written" forms. Hashimoto (p.119) contends that although the d i v i s i o n of items i n terms of c o l l o q u i a l versus l i t e r a r y i s based primarily on an actual s t y l i s t i c d i s t i n c t i o n between the two sets of forms, t h i s d i v i s i o n should be considered a phonological one since items not only con-form to s t y l i s t i c d i s t i n c t i o n s , but they also follow c e r t a i n patterns of phonological behaviour. I t i s the phonological c r i -t e r i o n which i s p a r t i c u l a r l y important both i n proposing general statements about the h i s t o r i c a l development of Cantonese and Zhong-Given the non-exclusiveness i n the d i s t r i b u t i o n of the - 97 -shan, and i n formulating reconstructed forms for proto-Yue. The h i s t o r i c a l significance of the c o l l o q u i a l versus l i t e r a r y readings w i l l be c l a r i f i e d i n the second part of t h i s study. For the present, three main types of c o l l o q u i a l verus l i t e r a r y forms w i l l be shown i n t h i s section. In the f i r s t set, there i s an alternation of i n i t i a l s and tones. The morpheme containing an aspirated i n i t i a l together with tone /13/ i s the c o l l o q u i a l form, while the morpheme with the unaspirated form i n tone /22/ serves as the l i t e r a r y counterpart. Examples are presented i n Table 4 (a). The second set of alternants r e f l e c t i n g c o l l o q u i a l and l i t e r a r y readings i s /a:ng/ versus /ang/ f i n a l s . (This includes the corresponding stop f i n a l s . ) The c o l l o q u i a l forms contain f i n a l /a:ng/ and the l i t e r a r y forms f i n a l /ang/. A l i s t of such alternations are given i n Table 4 (b). The t h i r d and f i n a l set shows an alternation between /ia:ng/ and /ing/ f i n a l s (and between the corresponding stop f i n a l s ) . The /ia:ng/ f i n a l i s used i n c o l l o q u i a l speech and the / i n g / f i n a l i s used for the reading pronunciation. A l i s t of such s t y l i s t i c alternations i s presented i n Table 4 (_c) . Syllables with the corresponding stop consonant endings are also included i n the three tables. - 98 -Table 4 (a)• Co l l o q u i a l versus L i t e r a r y Forms: Tone /13/ Aspirated I n i t i a l and /22/ Unaspirated I n i t i a l , Number Word Co l l o q u i a l L i t e r a r y Gloss 5-10, / t s ' o 1 3 / / t s o 2 2 / 'to s i t ' 6-9 29-12 96-9 / t s ' o j 1 3 / / t s o j 2 2 / 'to be located, at' 95-13, J* / f a : m 1 3 / /ta:m 2 2/ ' l i g h t ' 134-18 igft / t ' y n 1 3 / / t y n 2 2 / 'broken' 155- 10, j/t /k'an 1 3/ /kan 2 2/ 'near' 156- 2 Table 4 (b). Col l o q u i a l versus L i t e r a r y Forms: /a;ng/ and /ang/ Fi n a l s . Number Word Col l o q u i a l L i t e r a r y Gloss 184-21 t i /ha:ng 5 5/ a /hang 5 1/ ' constant' 192-4 4. 55 /sa:ng / , 55 -/sang / ' raw' 192-5 55 /sa:ng / / 55 , /sang / 'animal' 192-8 55 /ka:ng / /kang5 5/ 'watches of the night 192-13 -? •5 55 ,h /ha:ng / /hang 5 5/ 'to pervade' 192-14 ft /ha:ng / /hang5 1/ 'to walk' 192-16 /ma:ng13/ /mang 1 3/ 'f i e r c e ' 192-18 /la:ng / / l a n g 1 3 / 'cold' 195-5 f 55 /tsa:ng / /4. 55 , /tsang / 'to struggle' 195-7 22 /ts'a:ng / /*. 55, /tsang / 'to open the eyes' 195-8 to 55 /ka:ng / /kang5 5/ 'to farm' a In the name of a v i l l a g e i n the f i r s t qu i n Zhong-shan county /ha: 55 .13, ng mi /. k In the name of a neighbouring v i l l a g e •Z P 55 .13 , : ^ /ha:ng mi /. - 9 9 -Table 4 (c) . C o l l o q u i a l versus /ing/ F i n a l s . Number Word Col l o q u i a l 1 9 7 - 1 4 5 1 /p'iarng / 1 9 8 - 3 /kia:ng / 1 9 8 -1 9 9 -1 5 1 / • 1 3 , /ja.ng / /pia.ng / 1 9 9 - 2 / • 2 2 . /pxasng / 1 9 9 - 3 2 2 /mia:ng / 1 9 9 - 4 . . . 2 2 , /kia:ng / 1 9 9 - 6 / i • 2 2 , /kia:ng / 2 0 0 - 2 / • 5 1 . /mia:ng / 2 0 0 - 3 5 5 /tsia:ng / 2 0 0 - •9 , 4 - . • 5 1 , /ts 'xa:ng / 2 0 0 - •19 / • 5 5 , /sia:ng / 2 0 0 - •20 ^ » , . 5 1 / /sia:ng / 2 0 0 - •21 / • 5 1 / /sxa:ng / 2 0 1 - •1 / u - 5 5 , /hia:ng / 2 0 1 - •5 * / • 5 1 . /Da:ng / 2 0 1 - •7 /nia:ng / / - . • 1 3 / /lxa:ng / 2 0 1 - •8 / T I 3 / /lxarng / 2 0 1 - •9 / 4 . • 1 3 / /tsxa:ng / 2 0 1 - - 1 0 *A 2 2 /tsViarng / 2 0 2 - - 5 2 2 / t s i a r n g / 2 0 2 - - 7 / • 2 2 , /sxa:ng / 2 0 2 - - 9 2 2 /tsia:ng / 2 0 3 - - 1 0 • • f t / s i a : k 2 / L i t e r a r y Forms: /ia:ng/ and Lit e r a r y Gloss /p'xng / • f l a t ' . 5 5 / /kxng / 'to fear, a f r a i d ' / • 1 3 / /xng / 1 shadow1 / p i n g 2 2 / •handle' 2 2 /ping / 1 s i c k 1 /ming 2 2/ • l i f e ' ... 22, /kxng / 'to respect, be c a r e f u l ' / k i n g 2 2 / 'mirror' , . 5 1 / /mxng / 'name' 5 5 / t s i n g / 'sharp, quick-witted' . • 5 1 . /ts 1xng / 'clear sky' / • 5 5 / /sxng / 'sound, voice' / • 5 1 , /sxng / 'accompli shed, ent i r e ' . . 5 1 . /sxng / •c i t y ' / k ' i n g 5 5 / ' l i g h t (e.g., i n weight)' . . 5 1 / /xng / 'to win' / l i n g 1 3 / ' c o l l a r ' / l i n g 1 3 / 'to apply for' / l i n g 1 3 / • h i l l ' /a. • 1 3 / /tsxng / •a well' - 2 2 / /ts Vxng / 'to i n v i t e ' / * . • 2 2 / /tsxng / 'clean' / • 2 2 / /sxng / 'surname' ix. • 2 2 / /tsxng / 'upright' / s i k 5 / 'to be fond of' - 1 0 0 -2 0 3 - 1 6 1 2 / t s i a . k / / t s i k 2 / 'to burn 1 2 0 3 - 1 7 * / t s ' i a . k 2 / / t s ' i k 2 / 'red' 2 0 4 - 8 / p ' i a : n g 5 5 / / . • 5 5 , /P ing / 'to r i s k (e.g. one's l i f e ' 2 0 4 - 9 / i • 5 1 -a /p'ia:ng / /p'xng / 'bottle' 2 0 4 - 1 6 /tia:ng / / t i n g 5 5 / • n a i l ' 2 0 4 - 2 0 / t 1 i a : n g 5 5 / / t ' i n g 5 5 / 'to l i s t e n , hear, obey' 2 0 4 - 2 3 % / * . ! - 5 1 , b /t'ia:ng / / f i n g 5 1 / 'pavilion' 2 0 5 -2 0 5 -2 3 % / l i a : n g / /lxa:ng / / i - 5 1 , / l i n g / / i - 5 1 , / l i n g / 'efficacious' 'remainder, plus, over' 2 0 5 - 8 \ / t s ' i a : n g 5 5 / / t s ' i n g 5 5 / 'blue, black' 2 0 . 5 - 1 1 KL / • 5 5 , /sia:ng / , . 5 5 , /sing / 'smelly (of f i s h , blood)' 2 0 6 - 1 /tia:ng / / t i n g 1 3 / ' top;.(classifier for hat)' 2 0 6 - •7 $i /ts'ia:ng / / • 1 3 / /sing / 'to awake' 2 0 6 - •8 fa / * - • 5 5 , /tia:ng / / t i n g 5 5 / 'to n a i l ' 2 0 6 - 9 /tia:ng / / t i n g 2 2 / 'to subscribe' 2 0 6 - •10 5 5 /t'ia:ng / / f i n g 5 5 / •to l e t ' 2 0 6 - •11 . . . 2 2 . /tia:ng / / t i n g 2 2 / 'to f i x , to order' 2 0 7 - •25 2 /sia:k / / s i k 5 / • t i n ' Additional morphemes not i n the d i a l e c t survey l i s t : 1 . / l i a : n g 5 1 / / l i n g 5 1 / '(a kind of f i s h ) ' 2 . ^ / t s i a : k 2 / / t s i k 2 / 'straw mat' 3 . / l i a : k 2 / C / l i k 5 / 'to drip, to t r i c k l e ' Only i n the combination /jaw p'ia:ng / 'child/brought to a second marriage'. In such combinations as £dLi •§* / t i tVia:ng / ' the ground i n front of the house i n the v i l l a g e s used for drying grains, etc.' . c In the names of two v i l l a g e s , }$L / t a : j 2 2 l i a : k 2 / and / s a J l i a : k / (from Chao). - 101 -Notes to Chapter 1 1. The segment c a l l e d the "medial" was probably f i r s t extracted from the f i n a l by Bernhard Karlgren, whose reconstruction of Chinese remains the groundwork fo r most studies on h i s t o r i c a l Chinese phonology i n t h i s century. 2. B a l l ' s system of tr a n s c r i p t i o n i s not phonetic, nor i s i t t r u l y phonemic since there are a number of redundancies.X. However, rather than introduce another set of bracketing notation that would be purely ad hoc, B a l l ' s system w i l l be treated as i f i t i s phonemic. 3. The character n i yfcj has conventionally been used by Chinese phonologists to represent a class of words (of which i t i s a member) containing an i n i t i a l which has been reconstructed as *n. The names of other i n i t i a l s are derived i n a sim i l a r manner—a member containing a certa i n i n i t i a l i s used to represent the group possessing the same i n i t i a l . 4. A discussion of " c o l l o q u i a l " and " l i t e r a r y " readings of characters i s presented i n section 1.5. 5. This complementary d i s t r i b u t i o n of the dental nasal and the l a t e r a l i n i t i a l s i n Amoy was brought to the writer's atten-t i o n by Pulleyblank. 6. The semi-circle, which may or may not be underlined, i s used by some t r a d i t i o n a l Chinese phonologists to mark tone. The presence of underlining indicates the Yang r e g i s t e r , and i t s absence the Yin r e g i s t e r . The semi-circle at the bottom-left corner of the tr a n s c r i p t i o n designates the Ping tone, top-l e f t corner the Shang tone, top-right corner the Qu tone, and bottom-right corner the Ru tone. The names of the tones - 102 -refer to h i s t o r i c a l phonological c l a s s i f i c a t i o n s , while the Yin-Yang dichotomy r e f l e c t s reflexes of h i s t o r i c a l i n i t i a l s . These terms w i l l be further elucidated i n t h i s study at the appropriate place. To conform with the present study, Chao's numerical system for tones rather than his tone l e t t e r s i s used for t r a n s c r i b -ing h i s Zhong-shan data. The velar nasal i s transcribed as /ng/ for typographical reasons. The zero i n i t i a l /#/ should not be confused with the vowel /0/ , which w i l l be introduced l a t e r , since the zero i n i t i a l never occurs i n the transcriptions. Moreover, they do not occur i n the same environment. As i n the case of the names of the i n i t i a l s (see footnote 3 above), the character used to represent a given rhyme group also belongs to that group. Thus, Geng i s both the name used for the rhyme group and i s i t s e l f a member of that rhyme group. Furthermore, a rhyme group i s composed of a number of "rhymes", which also possess names established by convention. As i n the above cases, the character bearing the name of the rhyme i s also a member of that rhyme. The f i n a l s h i s t o r i c a l l y exhibit a four-way contrast of deng ^ 'grade, d i v i s i o n ' (Karlgren translates the term as ' d i v i -sion', and Pulleyblank translates i t as 'grade'). P r i o r to Pulleyblank's (1970-71) theory of the system of the four grades, i n which the contrast among the grades i s i n terms of phonological d i s t i n c t i o n s i n the medials, reconstruction of the grades had been accounted for p a r t i a l l y i n terms of d i s t i n c t i o n s i n the medial and p a r t i a l l y i n terms of d i s t i n c -- 103 -tions i n the rest of the f i n a l . The system of the grades w i l l be explained i n greater d e t a i l i n chapter 2. For the present, l e t i t s u f f i c e that the grade system i n Pulleyblank.'s (1970-71) analysis concerns a four-way d i s t i n c t i o n of the medials involving the presence or absence of p a l a t a l i z a t i o n . 11. The Zi-hui i s not s t r i c t l y phonetic. Length, for example, i s not marked, nor are glides consistently kept d i s t i n c t from vowels. Moreover, i n i t i a l glides which can be predicted from features i n the following vowel may be omitted, as i n the various Mandarin d i a l e c t s . 12. Thanks i s due to Pulleyblank for explaining the symmetry . i between I i ] and [ u ], and hence providing additional support for the phonemic analysis proposed here i n which they are treated as corresponding high vowels, / i / and /u/ respectively, 13. The sequence [ob] i s presented as the p r i n c i p a l form for the c l u s t e r /ua/ i n Egerod's vowel chart, with l o i j included i n round brackets. However, only the sequence [UD] occurs i n the discussion of the various phonemes i n the d i a l e c t . This l a t t e r form i s probably a typographical error, and should i n fact be [ob] , which only appears on the vowel chart. 14. Unless stated otherwise, Cantonese pronunciations follow the phonetic t r a n s c r i p t i o n used by Hashimoto. A few minor changes, are made i n order that her t r a n s c r i p t i o n system conform with the present one; e.g., Hashimoto's glides [1], lyj and [u] are transcribed as [ j ] , IH] and Iw] respectively, and her [k] and [k'] are transcribed as {kw] and Ik'w]. w Regarding [kw] and Ik'w], whether one wishes to consider them as consisting of a sequence of sounds or a single l a b i a l i z e d - 104 -segment i s b a s i c a l l y a phonological decision. In terms of h i s t o r i c a l phonology, i t i s advantageous to have the l a b i a l element separate from the i n i t i a l and treated as a medial. Kai-He d i s t i n c t i o n s do not a f f e c t the h i s t o r i c a l i n i t i a l s but they do the medials, such that the l a b i a l element occurs in He-kou series, for example, whereas the p a l a t a l element (e.g., Zhong-shan medial / i / ) occurs i n the Kai-kou s e r i e s . Synchronically, i n terms of minimizing the t o t a l number of i n i t i a l s and f i n a l s , i t i s i n fact more economical for both Cantonese and Zhong-shan to treat the l a b i a l segment as part of the i n i t i a l by creating an extra pair of i n i t i a l s , namely l a b i a l i z e d velars [kw] and [k'w]. Zhong-shan also has to contend with a p a l a t a l segment which has a very d i f f e r e n t d i s t r i b u t i o n from /w/. Medial / i / co-occurs with less than half of the rhymes ( i . e . , f i n a l minus the medial), but with almost a l l the consonantal i n i t i a l s . In t h i s l a t t e r case, i t i s more economical to treat the p a l a t a l medial as part of the f i n a l than to create a p a l a t a l i z e d series of i n i t i a l s , which would double the present inventory of i n i t i a l 3 * Thus, an alternate solution to what has been proposed here that would be more sympathetic to a synchronic analysis would be to create a l a b i a l i z e d velar series for Zhong-shan, and recognize one medial only—namely, the v o c a l i c medial / i / — which would be part of the f i n a l . A comparable solution i s i n fact adopted i n setting up the Zhong-shan syllabary and the lexicon of c o l l o q u i a l Zhong-shan terms i n chapter 3. The segments [kw] and Ik'w] are treated there are part of the i n i t i a l . - 105 -15. Obviously, among h i s t o r i c a l Chinese phonologists, there i s not always consensus i n interpretations and reconstructions. This i s a case i n point. J. Norman (personal communication), for instance, does not believe, as do Karlgren, Chao and Pulleyblank, that there was a d i s t i n c t i o n of medials between guan j|J and guan ^ . Norman proposes that guan ^* (*kuan). passed through a stage *kuon and then became [ku:n] i n Can-, tonese and Zhong-shan. Guan j|ifj (*kuan) , on the other hand, Norman feels did not evolve i n the same way because the nuclear vowel was front and hence much more re s i s t a n t to rounding. Thus, Norman posits the difference between the two words i n terms of a difference i n the vowel rather than a difference i n the medial. 16. Chao's medial / i / i n his s y l l a b l e s / i a / , / i a i t j / and /ia:k/ i s the f u l l vowel [ i ] . These s y l l a b l e s are transcribed i n the present study as / j a / , /ja:ng/ and /ja:k/ respectively, with / j / analyzed as a f u l l glide i n i t i a l . 17. Norman pointed out that lip-rounding of the i n i t i a l before [u:] also occurs i n Peking Mandarin and probably i n most other Chinese d i a l e c t s . 18. Sheng also has the dictionary d e f i n i t i o n of 'sound, voice'. 19. Chao (p.68), on the other hand, records the phrase 'tomorrow morning' as ^ ^ [men.51 t s i :w55 tsow 1 3] , with ming 0$ 51 pronounced [men ] i n t h i s context. 20. The writer i s indebted to David Strecker for bringing to her attention the concept of the "sesquisyllable" discussed i n Matisoff's a r t i c l e . Another term for the phenomenon i s that of the "minor s y l l a b l e " , which i s used i n England (Pulley-- 106 -blank, personal communication), 21. Pulleyblank informs the writer that the term "dimidiation" was used by Broodberg to re f e r to his theory that such rhym-ing compounds as kun-lun ^ , for the Kunlun Mountains i n Tibet, were derived from monosyllables; thus, kun-lun f v ^ , for example, would be derived from *klun, etc. 22. Except for r e t a i n i n g the following i n McCoy,'s t r a n s c r i p t i o n system: schwa [ a ] , symbol for the neutral tone [°], recording of a voiced [b],and use of a hyphen to indicate close junc-ture, his data has been phoneticized based on Hashimoto 4s system. A couple of variant forms for 'corner' recorded by McCoy (1966:185,fn.27) are: [ k a ° - l D i k 3 5 ] and [ka°- l o x k 4 ] . 51 23. The i n i t i a l suggestion for the s y l l a b l e /ham / i n Zhong-shan was made by Pulleyblank. 24. Because Long-du i s spoken i n the Zhong-shan county and i s moreover i n the qu, just west of Shi-qi (Qu II on Map 2) , not only does the phonological system of (Shi-qi) Zhong-shan have the unique status of forming the Long-du l i t e r a r y pronuncia-tions , but many Shi-qi forms have also been incorporated into the c o l l o q u i a l layer of the Long-du d i a l e c t . 25. Chinesisch und T a i , by K. Wulff, Copenhagen: Levin and Munksgaard, 1934. - 107 -CHAPTER 2 . SYLLABARY AND LEXICON A syllabary representing the repertoire of Zhong-shan sy l l a b l e s i s presented i n the following section. Wherever possi-ble, a character i s used to represent a given s y l l a b l e . Single underlining of a character indicates c o l l o q u i a l reading, and double underlining s p e c i f i e s l i t e r a r y reading. C o l l o q u i a l words with no character counterpart (or only coined characters to repre-sent them) are marked using a c i r c l e " O " , except for onomatopoeic sy l l a b l e s and foreign loans. The l a t t e r are distinguished by the use of an "X" i n the syllabary since what i s normally a rare or impermissible sequence can occur i n an onomatopoeic expression or borrowed term. However, i n the lexicon of c o l l o q u i a l Zhong-shan words that follow the syllabary, a c i r c l e i s used to represent any characterless word, including onomatopoeic s y l l a b l e s and loan-words . In both the syllabary and the lexicon, medial /w/ i s treated as part of the i n i t i a l i n order to reduce the combinatory p o s s i b i l i t i e s of i n i t i a l s and f i n a l s . Thus, besides /k/ and /k'/ there are also the clusters /kw/ and /k'w/. Medial / i / continues to be treated as part of the f i n a l . Besides the words e l i c i t e d by the writer, the syllabary also includes some words recorded by Chao. Co l l o q u i a l terms entered into the lexicon from Chao's a r t i c l e are also given e x p l i -c i t acknowledgement. In the d i a l e c t survey l i s t , also, the source of data i s likewise s p e c i f i e d i f no confirmation i s made by the writer that a given word i s also used by her informants or i s pro-nounced i n the same way by them. For convenience of comparison, Chao's data are treated as - 108 -i f they were transcribed using the present phonemic system. Hence, the f i n a l [uk], for example, i s transcribed /uk/ for both the present data and Chao's. In Chao's o r i g i n a l phonemic system, [uk] would have been transcribed as /ok/. (Note that i n the present system /ok/, i n turn, i s phonetically [axk].) Besides the ease of comparison, using a single phonemic system has another advantage: i t f a c i l i t a t e s collapsing of two sets of data with no r e a l loss i n information. Where the two sets of data d i f f e r , as i n the case of the s y l l a b l e [jsin] versus [jcexrj]., the present form of [joxg], phonemically /jong/, i s used i n the syllabary. S i m i l a r l y , since the present data do not show the i n i t i a l c l u s t e r /ngw/, no space i s a l l o t t e d for i t i n the syllabary. As regards the lexicon, i t s primary purpose i s to assign meaning to the many c i r c l e s and X's scattered throughout the syllabary. Other items included merely serve as a sampling of Zhong-shan c o l l o q u i a l usage. No attempt i s made here to pro-duce an extensive lexicon. I t should also be made cl e a r that a number of the expressions, p a r t i c u l a r l y the more vulgar or abusive terms, are not actually used by the writer's informants, but are expressions that can be found i n the d i a l e c t and are used by the less educated. In terms of format, the lexicon i s arranged according to the Zhong7Shan f i n a l s presented i n Table 2 (a). The lexicon begins with the set of f i n a l s containing the high, front nuclear vowel / i / ( i . e . , / i / , /iw/, /im/, et c . ) , followed by the set of f i n a l s containing /y/, and so forth. The following symbols and abbreviations appear i n the lexicon: - 109 -r>^ = r e p e t i t i o n of the s y l l a b l e that occurs at the left-most column S = sentence Q = question CL = c l a s s i f i e r PRT = p a r t i c l e s i . = slang l i t . = l i t e r a l l y esp. = especially s.o. = somone s.t. = something Moreover, i f no characters are assigned to a p o l y s y l l a b i c word, .'<• i t w i l l be assumed that the word contains characterless words, with or without the exception of the repeated s y l l a b l e from the left-most column. Thus, there w i l l be no case of "0~", " ^ O O " , etc., actually recorded i n the lexicon; they w i l l simply be implied. - 1 1 0 -2 . 1 . S y l l a b a r y A r r a n g e d A c c o r d i n g t o M o d e r n Z h o n g - s h a n F i n a l s i y u 5 5 5 1 1 3 2 2 5 5 5 1 1 3 2 2 5 5 5 1 1 3 2 2 p p ' m f ** *& | if f # $ o # •* £ t t ' n 1 o x </o ft it hi f if «t A IS fti & t s t s ' s *« 0 tin- fa 0L & i*t» 'X jjb $ fc & A & A $ fg tit §& k k " n g h 0 A* * % -£6 o A 4 o * 4 -# i & * f /il Jk* a ^ 4" li» » ti t k w k ' w w j - I l l -m i 0 o a 5 5 5 1 1 3 2 2 5 5 5 1 1 3 2 2 5 5 5 1 1 3 2 2 p p ' m f ft 1k & i& ik h % t t ' n 1 0 o *$ I i rt #f *f 4 ft tt. o a o ft O o O O O O t s t s ' s o X ft it i l •fc 41 £L i x -ft IL & ^ » + % o k k ' n g h 0 o o % 0 V\ o o % 0 ML WK * 0 % *fH f f -f k w k ' w w ' $ & it * i& ** 1 * j 1 i |j f^" -e> /IL - 1 1 2 -i a u j 5 5 5 1 1 3 2 2 5 5 5 1 1 3 2 2 5 5 5 1 1 3 2 2 p p * m f X o o O O O o * ft i i & t t ' n 1 % o o *fe ii i t i l Rt i l o # O t s t s ' s i t < f i t l f k k ' n g h 0 o o hi o •1 1 g$ 0^ *t o % 1— / k w k ' w w j - 113 -Vs. a j a;,j 55 51 13 22 55 51 13 22 55 51 13 22 p p.' m f X ft & & i & o i i * . i i i i f o < # Ik * *£ t i ft t t' n 1 ft & £> ft t i t » * i f t « f tt 4rt 0 & 4# 4# 48 o o II ts t s ' s t f 4 o A A & fa 1 i t o $ f £ t m f t 4 i o H BI k k' ng h 0 W a # i *^ It) t & & % o K'h A o | © S to M ^ ft ;S o £ ° ft &i 4& kw k'w w i £ * ^ IS) £ ft j - 114 -i w o w a w 5 5 5 1 1 3 2 2 5 5 5 1 1 3 2 2 5 5 5 1 1 3 2 2 p p ' m f if fja. ik % 4 Is o 4t o fa #8> o ±> o ti f o o t t ' n 1 flfc \'L % IP o % M, o ft 3 ft AS i i f 1 o f % t s t s ' s i '£) B i l M 4ft * i is * 1 i t * '£ A f ? j i ft A i f *\ £ & # & {i £ k k ' n g h 0 & # IL it *. f t 4 ^ ^ t ft 0 ft 0 ^ Sit -6 ;& k w k ' w w j 4k # 1 ^ - 1 1 5 -a; w i a ; w ira 5 5 5 1 1 3 2 2 5 5 5 1 1 3 2 2 5 5 5 1 1 3 2 2 P P ' m f & fa to 46J -Jtfe t t ' n 1 I f O V % k xk \k ft % / f k & t s t s ' s •*» ^ | $ o 41 % & » -ft £ ik f l k k ' ng h 0 X o & (IL tk rt i o | o o o % %r *f o ii 4 t k w k ' w w j - 1 1 6 -am am a :m 5 5 5 1 1 3 2 2 5 5 5 1 1 3 2 2 5 5 5 1 1 3 2 2 p p ' m f X t t ' n 1 O X O O 0 O O O O o fa ;f &t iii $ * t il i i i l o If :^ t s t s ' s ^ -3T >M1.- f ft # t #1 i f *) k k ' ng h 0 A % # O vfy O 0 O it o ° & k w k ' w w j 1 4 & & o - 117 -i a ; j n i n y n 5 5 5 1 1 3 2 2 5 5 5 1 1 3 2 2 5 5 5 1 1 3 2 2 p p ' m f i t 4 $ 01 ° f t f n 1 II * * O 4f te. ft. III tfty t s t s ' s i i . $ ft % ® A J & 3 t & k k ' n g h 0 f t a* « & i t t ft m 1 £ ik ri k w k ' w w j - 1 1 8 -u n jzSn o n 5 5 5 1 1 3 2 2 5 5 5 1 1 3 2 2 5 5 5 1 1 3 2 2 p p " m f * * t % £ # in n yk n t t' n 1 / k W 0 ts ts' s i £ >| o k fa h o ill 4 «ft /'II k k ' n g h 0 f f % 1 "ft 1 If It 1 * % fh k w k ' w w j - 1 1 9 -a n a : n i n g 5 5 5 1 1 3 2 2 5 5 5 1 1 3 2 2 5 5 5 1 1 3 2 2 p p ' m f 1 1 * * * * M M f 4 w tit ° afc »H if) ^ t JCJ ^ I O t t ' n 1 % o £ ^ £" o o •hi? %V Ar/~ 5? ^ jfc « ft *a ® « « T JS i l ft a 4 t s t s ' s -I ft i m ?t / f - % 4 / i 'At J « % tk ik i_ si n k k' n g h 0 ^ f it ft & *t 4 t *)t « ft w & | l C /fi W T Ft kw k'w w ^ £® $ * 9 *£. & ftS 'it f & & ^ . . . j & A. & if - 120 -u n g o n g a n g <>N? v\ 55 51 13 2 2 55 51 13 2 2 55 51 13 2 2 p p' O % * i t if) /m -J-X M Ai m t \\ 0 k f 0 t t ' t i t >t * * n O i 1 I o ti o 1 i t ft t s * f If t s " s *} fa 0 1 0 i HP 8 o k 'A i - f i r i t k' $ 0 o n g h & i f an I o 0 1 ik H x2 0 0 B o kw 0 k'w w * * j ft * 1 9 4 - 121 -a ; n g l o n g i a : n g 55 51 13 22 55 51 13 22 55 51 13 22 p p' m f o \ h%. o -f X I % t f n 1 o O L & $fu o i i t s t s ' k s i » if i o M 4 /» *t ^* *l 4 « $ ft *8 l i i £ k k' n g h 0 JL o o I ^ 1 51 31, 1 t- ft I $ 1 41 0 kw k'w w O X 0 Bi o y[c j 4 11 - 1 2 2 -i p o p a p a : p i a : p 5 2 5 2 5 2 5 2 5 2 P P ' m f t %& 0 o O t ' ft 0 n U o 1 t s & # 1 f i t t s ' . 'S fab i f . + to k X k ' t J - l o # 0 h A-a 4*. 0 t 0 k w k ' w w j o - 1 2 3 -i t y t u t o t a t 5 2 5 2 5 2 5 2 5 2 5 2 p p' m f O t t' n 1 o IK o H o If X g?u a t 0 o o 0 t s t s * s ~ KK ° q A o & \& 1 il 1 k k ' n g h 0 o H o % & ¥*• S i A X 44 ;4 X k w k ' w w h ii j - B - 1 2 4 -a;t l a : t ik uk ok ak 5 2 5 2 5 2 5 2 5 2 5 2 p p' m f /V 0 X X h « o « 0 ^ o If i l t t' n 1 & X ft & it as 0 0 * « $ o o o ^ ts t s ' s it X i t f W ;f 31') I* k k" ng h 0 O X * 1 $ ft 4-o f Jl. i/E. kw k'w w ft o o j - 1 2 5 -a ; k i o k i a : k ( S y l l . ) 5 2 5 2 5 2 5 . 5 5 1 1 3 .. 2 2 p P' m f O f o t X t ' 0 n • 1 o o o i t t s f f t s ' & o 8 4 * k fa k ' ng h ft 0 X k w k ' w o 0 w j *1J - 126 -2.2. Le x i c o n of C o l l o q u i a l Terms Arranged A c c o r d i n g to Modern Zhong- shan . F i n a l s . . 55 P i o i 55 55 / p i p i / = a w h i s t l e . , . 55 p ' l o = worn. . 55 mi IL = t o s l e e p . . 51 mi \L = / f a : t 2 m i 5 1 mung 2 2/ ^ ) = to have a dream. ..55 t i 55 = some (Chao). (More f r e q u e n t l y : /naj /.) ,..51 t i 0 / t i 5 1 t i 5 1 t a r n 5 1 t a m 5 1 / = s a i d o f s.o. who i s qu i c k -f o o t e d o r nimble. . 55 n i = Q-PRT. . 55 l i o / l i 5 5 t s a j 1 3 / (~ Ah ) = a s m a l l basket; a bamboo-type c o l l a n d e r or s i e v e . o / l a : w 5 1 l i 5 5 / ($Jjp ~ ) = a wired s t r a i n e r w i t h . long handle. / k i 5 5 l i 5 5 k w a r j 5 5 l a : j 5 5 / ("^ ~ * ) = d e p i c t s sound of speech t h a t i s u n r e l a t e d and unconnected (Chao). l i 2 2 m = tongue. t s i 5 1 0 = to poke s.o. wit h f i n g e r (from \% / t s i 1 3 / ? ) . o / s a p 5 t s i 5 1 t s i 5 1 / {}$. ) = very moist. o / t s i 5 1 t s i 5 1 t s a 5 1 t s a 5 1 / = t a l k a t i v e , g o s s i p y . o / t s i 5 1 t s i 5 1 t s a m 5 1 t s a m 5 1 / = m u f f l e d sound of v o i c e s . , • 55 t s ' 1 $ = t o s t i c k ; / t s ' i 5 5 saw 1 3 t s ' i 5 5 k i o k 2 / ( ~ % ^ jjtfV) = very s t i c k y ( l i t . , s t i c k s to hands, and f e e t . 4. i-51 t s 1 1 = s m a l l pond.,., (Note: / t ' o n g 5 1 / ^ = l a r g e r pond; / t s ' i 5 5 t'ong / (;&*,# ) = pond ( n e u t r a l ) . 9$ / s i 5 1 man 5 1/ (<~$ ) = t a l k ; / s i 5 1 man 5 1 t o 5 5 k o 2 2 f a i n / ( ~ j yjh ) = m o r e t a l k t h a n r i c e " very t a l k a t i v e . #t / t ' i n 5 5 s i 5 1 / ( ^ ) = the weather. - 127 -s i " A / s i 1 3 f a t 5 / C~ % ) = rectum. i 1 3 JL / s i 1 3 s i 2 2 £ / s i 2 2 kwa:n 5 5/ (/~ #j ) = because. k i 5 5 A ' u 1 3 k i 5 5 / (.jjt ^ ) = hungry. 51 51 51 51 51 k i O / k i l i ka:w la:w / = noisy t a l k i n g — a s of a group. k i 2 2 - f £ j / k i 2 2 sam 5 5/ ( ~ / C / ) = memory (Also: / k i 2 2 s i n g 2 2 / = to turn around (e.g., head) (Chao). .55 _ ,, 55 .55 .13, \ •, j. , • ngx O /tsa:ng ngi nga:j / (f ^ O ) = almost ( i . e , lack just a l i t t l e b i t ) . h i 1 3 = to l i f t up. i 5 1 j?p / i 5 1 k a 5 5 / ( ~/>° ) = es p e c i a l l y , even more. iw . 13 miw *ky /miw 1 3 t s t f j 1 3 miw 1 3 s i t 2 / (~ "% ~ k ) = to sneer. 4 . . • 51 t' I W hk = CL for long, slender objects (e.g., necklace, thread, street, e t c . ) ; also used for trees. . 55 niw fa* 55 55 55 = slender; /niw pa:ng pa:ng / = very slender. o 5 R 55 1 3 5 5 5 /niw„ nana 7 = precariously; / k ' i tak niw nang / ( <AE \% ^ O ) = to be standing precarious-ly (e.g., on a high ledge). , . 55 liw o = to poke finger i n corners, small holes, etc. . . 22 tsiw = to chew. . , . 55 ts ' I W o = to pry open; to dig (e.g., a d i t c h ) . . , . 55 k' iw o = to cross the legs; to scoop out (e.g. with spoon). . , . 13 k' iw o = to cross the legs. o = coincidental, extraordinary, unexpected. /k'iw 1 3 t'aw 5 1/ ( <~ Ik ) = of the g a r l i c family, used for p i c k l i n g . 13 hiw o = to understand. , . 22 hiw o = to st i c k out (e.g., l i d of an opened t i n can). (Also: /hia:w 2 2/). - 128 -im %fk = straight; / t i m 2 2 pa£ 5 l a | 5 / ° 22 = v e § y straight (Also: /tim pat pat /, /tim p l a t / ) . -JfU^ = to mind, to be fussy, to d i s l i k e . 2 22 /pa:k im / = peevish (referring to children only) i n O = diaoer. (Also: /niw 2 2 p ' i n ] 3 / (>fic ~ ) , / s i 1 3 p ' i n ± 3 / ). 55 55 55 55 = where. (Also: /min naj /,,/naj /, /min /ts'y / ^ ).. (Phonetic variant: /pin / ) • ri 55 13 fljj /min tsow / ( ^ ) = tomorrow. /how 1 3 t ' i n 5 5 / ( <k-y~ ) = clear day. O /ngaw 5 1 n i n 5 5 / ($ "*".)-•= cow's milk (Chao). f | / t s i n 5 5 h o n 5 1 t s i n 5 5 l a : n g 1 3 / («~ £ ~ ' t" ) = shivering cold, to have c h i l l s . *t /t'aw 5 1 s i n 5 5 / (Sit ^ ) or / t s i a : n g 2 2 s i n 5 5 / (.IE. r~- ) = just a while ago, just a moment ago. lh\ / s i n 2 2 k a j 5 5 / ( ) = castrated chicken. ^ (Buti one,who castrates chicken i s c a l l e d : /im k a j 5 5 low 1 3/ JfctttiL)-0 = to l i f t up (e.g., mat), to peel o f f . /min 5 1 t s ' o n g 5 1 / ( ~ # w ) = bed. ing O = to toss c a r e l e s s l y . O / t i w 2 2 t i w 2 2 f i n g 2 2 / ( i f $ ^ ) = to be hanging loosely or precariously. 0 / t i n g 1 3 t a j 2 2 / = fussy (e.g., to describe the way someone eats). = to hold; to bring. - 129 -k i n g 5 1 O / k i n g 5 1 l i n g 5 1 kam 5 1 lam 5 1/ or / k i n g 5 1 l i n g 5 1 kung . lung / = of noise from stamping feet, mov-ing^ furniture, ^ ^ t c . (7\lso: / k l i n g klam / or / k l i n g klung /) . iP s i p 2 ^ / t ' i n 5 5 s i p 2 / ).= l i g h t n i n g . ^ = to wedge, to s l i d e under a narrow opening. k i p 5 jj£ / p ' i 5 1 k i p 5 / (JL^ ) = suitcase. (Hashimoto £p. 349) acknowledges McCoy's analysis that [ki:p ] for 'bag, suitcase 1 i s a loan word from English 'grip'.) 5 5 13 ngip O = to blink; /ngip nga:n / (~ eK__) = to blink; i n a wink of the eye ( i . e . , very q u i c k l y ) . 5 13 0 /ngip s i / = stingy. O = to bite.Ce.g., of small i n s e c t s ) . i p 2 = to preserve i n s a l t or vinegar; by extension, means: to treat c r u e l l y a daughter-in-law or children of the husband's e a r l i e r marriage. i t 5 fy^ = to pinch; to tear up; to peel. mit ^•,5 /,. .5. ...5 .13 . 22 .55 . .13, . ^ r> t i t O / ( 3 a t ) t : L t S 1 kam na] tsaj / ((.-) ^ ^ v$ 0 ) = a very tiny l i t t l e b i t , a very small amount. O / t i t 2 t i t 5 tow 1 3/ (j0L~/& ,.|) = to f a l l down. 5 5 5 l i t 0 = to be i n a knot (as i n wrestling); / k ' i t l i t / = a knot (Also: / k ' l i t / ) . t s i t O = to t i c k l e ; to squeeze out. 2 22 t s ' i t O = tp^dote on, to l i k e (e.g., a c h i l d ) , /fa:n jan t s ' i t / K ^ ) = lovable (of a c h i l d ^ (But., older-generation use the phrase: /fa:n jan tsang / ( VL> >^ % ), l i t . , ' d e t e s t a b l e ' ) . k i t 0 = to jab, to pierce. k ' i t 5 O (see: / l i t 5 / ) . (ki) . - 130 -|& / n g i t 2 t'aw 5 1/ C~*ft) = sunshine. ik 5 5 O / t i k sik / = small, cute and elegant. O = to make a check mark ' </ ' . kjj* = the 'hand' r a d i c a l . 5 5 5 O /t'iw t ' i k / = fussy, c r i t i c a l . = to take s.t. along. 5 5 5 5 5 5 O /pik l i k pa:k la:k / or / p l i k pla:k / = crackling sound—of f i r e , or s t r i n g of fir e c r a c k e r s . O = to yank; to straighten up (e.g., clothes). /m51 s i k 5 s i n g 2 2 / (»&'•*'fl.) = naive; to be misbe-having. O = to clog, to jam. O / k ' i k 5 l i k 5 k'ap 5 l a p 5 / or / k ' l i k 5 k?lap 5/ = odds and ends; c l a t t e r i n g sound—as of dishes. y ;jt = he, she, i t ; / k ' y 5 1 t i 2 2 / (~ •*€*) = they. O = to keep warm by putting over low heat. O / h y 5 1 tow 1 3/ M£'J ) = to worry (Chao) . / t o 5 5 y 1 3 sgij13/ = a l o t of r a i n . yn = crooked. = to r o l l i n s.t. (e.g. f l o u r , d i r t ) ; to coat. = to handle (esp. a l o t or too much). = to crawl through. 1 3 = c l a s s i f i e r for^kooks (also pronounced /kun / ) . (Also use: /pun / — a more-modern term). - 131 -yt t y t 5 o = to protrude; to pout. t ' y t 5 o /wa:t 2 t ' y t 5 t ' y t 5 / (>f ~ ~ ) = very slippery. i y t 2 o / l y t 2 h i 1 3 j a t 5 kaw 5 1 l a w 5 1 / form a clump. (~&- *% O ) = to t s y t 5 O = to give a k i s s . t s y t o '= a d u l l , gnawing pain. k y t 2 = a^piece that i s one-half or less; l i m .22 22 /ta: j pun 3 0 = thick (of l i q u i d ) . (Cantonese: [ k i : t ].) 2 2 2 55 K5*) ngyt 0 /ngyt / or /ngyt kung / (^ J ) = the moon. u p u 1 3 £ / p u 1 3 / . ^ / p u 1 3 t s ' 0 j 5 1 / (~&iL) = an axe (But: 1 /ts'j2$j 3 J-/ ) = a hammer). P ' u 5 5 /P 'u 5 5 k o j 2 2 / (~S.) = luggage. p ' u 5 1 = to f l o a t ; /la:w 5 1 p ' u 5 1 / (#$~) = to f i s h or scoop out; to l i f t up (e.g., table) by two or more people. 55 22 55 -c mu 0 /a mu / (It ~ ) = woman servant (Chao) . mu 1 3 kfy /a22 mu 1 3/ (ft ~ ) = mother (Also: ^ a 2 2 ma13/.) (Chen's mother c a l l s her mother /a t s i a /ft . - W i ) . / i o 1 mu 1 3/ (^ &-~) = mother's ojLder s i s t e r Colder term; a more modern one i s : / i ma /Jtfc. k$g ) -mu 2 2 ^ = fog; /mu22 s 0 j 1 3 / (~*)<-) = dew. t u 5 5 %f = s t i l l , yet; also. 55 55 55 nu O = that; /nu naj / = those; there. l u 5 5 " f * A u 5 5 k u 5 5 l u 5 5 l u 5 5 / = to talk i n an i n d i s t i n c t manner (Chao). 2 2 0 = inchoative, S - f i n a l marker. (Also: / l u /.) t s u 2 2 O /Verb + t s u 2 2 s i n 5 5 / (Verb + ~/b») = V - f i r s t . = to save, to c o l l e c t . O = t h i s ; / k u 5 5 t s a n 2 2 / O f£ ) = and then. #D / t s i 5 5 ma 5 1 k ' u 5 1 / (/£/& — ) = sesame paste. O /ts'jzfok'u 5 1 §gaj 1 3 / 5 ^ ^ r ) = young c h i l d ' s penis (Also: /tsaj tsaj /.) woman; )L / h u 5 5 n i o n g 5 1 / (~i£j = a married / h u 5 5 n i o n g 5 1 t s a j 1 3 / (~4M3-) = a g i r l . 0 = to throw o f f (.e.g., blanket), ftf = t o Paste; / t s i o n g 5 5 u 5 1 / (.{JJL—) = paste, = d i r t y ; /uDO tsow 3 D l a : t z t ' a : t z / | f 0 O ) = d i r t y . A 55 51 51 vS? /u ham ham / = very dark and murky; / u 5 5 t ' i n 5 5 ha:k 5 t i 2 2 / {~jL&i&) = very dark-as of pending storm. ^ /haw 2 2 p u j 2 2 t a j 1 3 / (^ jL^/fo ) = at the back, behind ^ /muj 5 5 t s a j 1 3 / = a purchased maid. O = there, at that place. 0 = a kind of clam (Chao). >f / l u j 5 1 kung 5 5/ ( r 3 ' A ) |,thunder; /ha:ng 5 1 l u j 5 1 / iff <~ ) or /hiong l u j / («fo ~ ) = to thunder. O = .here, at t h i s place. O = H e y l ung O = sound of h i t t i n g or hammering; sound of f a l l i n g object h i t t i n g a surface. O = a surge of fragrance or odour. O /p'ung 5 5 p'aw 2 2/ = f l u f f y . - 133 -O = t o s t i c k s . t . f i r m l y onto t h e ground. 51 51 "jL O / t ' u n g p'o / "4r ) = shaman o r medium ( f e m a l e ) ; / t ' u n g 5 1 k u n g 5 5 t s a j 1 3 / 43-) = shaman o r medium (male). (Most shamans a r e women. They a r e n o t found i n e v e r y v i l l a g e , a l t h o u g h b o t h Ku-chong and Yuan-f e n g had one.) O = b u r n t . = s t i l l , y e t . (More f r e q u e n t l y use: /wa:n /J^.) • 22 22 O / t s ' u n g ha / (<-^ "F ) = j u s t below, t h e n e x t one down. $fc^ = f o o d t o be e a t e n w i t h r i c e . 22 55 -n; V* /a kung / (sk ~ ) = p a t e r n a l g r a n d f a t h e r ; 55 /Surname + kung / = m a t e r n a l g r a n d f a t h e r . = t o push. O = t o f e r t i l i z e ( C h a o — h e a l s g , . t h i n k s t h a t i t may be a sound change from ,/ung / yjfc. ) . u t 5 5 55 *b O = sound o f a h o r n ; / p u t p u t t s 1 i a / ('v"v'Jf ) = c a r ( o l d term used by v i l l a g e c h i l d r e n ) . <c . , . 2 ^ .22 , , .51 . 22. ^ j - j l / ? * x ^ , / p ' u t t a : ] k'waj s m / {^K'K^ ) = t o be a matchmaker ( l i t . , t o f a n w i t h a l a r g e p a l m - l e a f fan) . 5 5 55 '-0/ / k u t k u t s i a : n g / rfr ) = sound o f s w a l l o w i n g . uk O = d u s t y ; /puk t s ' a n / ('-'J^ ) = d u s t . Hit = t o p u t one' s head down on one ' s arms. "^K = t o t o p p l e ; t o be t u r n e d u p s i d e down. 5 22 Jkt 0 = t o p r e d i c t , t o guess; /muk t s u n g / (.<*- j ) t o guess c o r r e c t l y . 0 = t o p r i c k . - 134 -c 55 5 nuk O = to move, unsteady; / n i nuk / - unsteady, wobbly (e.g., of legs of a table). 5 suk l u k 5 j$!o = to r o l l ; a r o l l , a loa f . 0 / f a 5 5 l i 5 5 l u k 5 l u k 5 / (.#> 0 ~ ~ ) = very bright and c o l o u r f u l . l u k 2 yjjfa = to scald. ts'uk 5 O /ts'uk 5 h i 1 3 t s ' uk 5 h i 1 3 / (~,££~i^ ) = to be bunched up. O = rancid (e.g., of s p o i l t meat, r i c e , e t c . ) . 22 5 /a suk / (i. ) = father's younger brother; a term of address for s.o. younger than one's father; father (used by some, e.g., Chen), when a male i s "adopted" by his father's deceased older brother who i s c h i l d l e s s — i n order that the l a t t e r may have p o s t e r i t y ) . j u k 5 = to move; /juk 5 j u k 5 kung 2 2/ = to wriggle about • ( i . e . , can't s i t s t i l l ) . tjzS55 O = to scold. l e i 5 5 = to remove food from inside mouth with tongue. l e i 5 1 / t ' i n 5 1 1 0 5 1 / (fJ7~) = fresh-water s n a i l s . 13 tsjz* O = to eat up. s 0 5 1 O / s 0 5 1 s 0 5 1 s i a : n g 5 5 / ) = sound of running water. ' 51 51 51 51 O / s i s i sjzJ sfi / = depicts (rude) manner of guzzling down food. k 0 5 5 O = to r o l l up. 13 k0 O = to r o l l up (e.g., cigarette, s c r o l l , mat, etc.) k 0 2 2 JfrfL = to saw (s.t. small). k'izT 1 fn = eggplant; /fa:n 3 k'szTV (& ~ ) = 0 5 1 0 = S - f i n a l p a r t i c l e of assertion. tomato. - 135 -O /tam 2 2 t 0 j 5 5 / ( J l ~ ) = shabby. ")<)L. = to steam-cook; of smoke going up the chimney. P&_ / t s ' 0 j 5 5 haw 1 3 sa:w 2 2/ V ) = to sing (But,in modern usage, i t mean to whistle, esp. at g i r l s ) . 22 22 O /10n ts0n / = clumsy, troublesome, nuisance. ^ /tsjrfn 5 5 s i 5 5 / ( 'v if ' i , ) = v e r m i c e l l i . O (See: /10n 2 2/) O / t i n 5 5 t i n 5 5 ts'jrfn 2 2 ts'tfn22/ ) = insane, crazy. = to s l i d e down. » sound of snoring. o 3P5 = C L f ° r trees. )JL /P'o 5 5 t s ' o j 2 2 / (-I-) = spinach. „ A 22 51 « „, m /a p'o / ) = paternal grandmother; 51 /Surname + p'o / = maternal grandmother; / k a 5 5 p ' o 5 1 / C & . ~ ) = mother-in-law. _ 55 55 51 O /mo s i t'o / = slow, to be a slow-poke. 1* ^ ^ . , 55 u 51 51 13 .13 ., 2, ^ = to (touch; /mo ha:m ngy su ma] t'a:p / (^fyfij* ,$kJLO) = touch salted f i s h , count r i c e bin--constantly counting what one has, how much money one i s making, etc. Jji / t o 5 5 f a : n 5 1 / ) or / t o 5 5 f a : n 5 1 s a : j 2 2 / ("'W'&J = thank you (normally said by rec i p i e n t of a favour). i & J / t ' o 5 5 h a : j 5 1 / = sl i p p e r s . 51 22 0 /t'o l a j / = to involve, to implicate. $ = to rub with the fingers. •jS / l o 5 5 s o 5 5 / (~*jL) = wordy. 55 51 0 / l o jaw / = derriere. O / t a 1 3 ha:m 2 2 l o 2 2 / tfl O ^ ) = to yawn. O / l o haw / (~ = denotes f e e l i n g i n throat afte r eating o i l y , f r i e d food. to axe. = an axe. / 22 "{© = t h i s (Chao, variant form: /ko /) . •jlS = possessive s u f f i x ; CL (general) ; t h i s (Chao and B a l l ) . 0 = to rinse. O = to eliminate (vulgar term). = Q-PRT (phonetic variant: / a 5 5 / 0 ^ ). °$ = Oh! Ah! = nodding p a r t i c l e . oj O = an exclamation of disgust or exasperation. j ) £ / f a w 5 1 f o j 5 5 / ( # L ~ ) or /t'aw 5 1 t s y 5 5 t ' o j 5 5 / ) = Drop dead! rfri = a long time. = to come. (Cantonese: [ I B J 2 1 ] ) / t s ' o j 5 5 k o 2 2 n i 1 3 / (~ O'ftf. ) = Fie on you! 0 = to leave i t up to s.o. 22 22 O /ngoj soj / = i r r i t a t i n g (e.g., as of a canker sore). 53 O = to carry an infant on the back. (Cant.: [me: , ] ) . - 137 -22 51 Je = to love; t o want; to need; /oj s i / C~ 0f ) = sometimes. ow = to cook (in water); a cooking pot. 55 55 O = perfective s u f f i x (Variant forms: /how /, /ow /) = suds. O / f a : t mow / (^~) (Chen, suggests i t came from English 'mold') !to become moldy 1. ^ = not have. = a case, envelope; a s u i t (of clothing). = to mix. tyjp /low 5 1 l o w 5 1 l u k 5 l u k 5 / = hard-working (of one who r t o i l s ) . O = to pour i n and out; to pour through a funnel; a funnel. 22 5 • ^ i /tsow f a t / = a cooking stove—has an opening i n the front for putting i n f u e l , and openings on top of the stove for a wok, k e t t l e , etc. The ones used i n Ku-chong and some houses i n Shi-qi had a large hole towards the front of the stove f o r the wok, and two smaller holes i n the back for k e t t l e s , pans, etc. /fat / i s Jjf ?) . = to shake; /ngow 5 1 ngow 5 1 nap 2/ = wobbly, shaky. )fr /m51 how 1 3/ (A ~ ) or /mow13/ (fr o m : 4 ^ ) = don't 2 (negative imperative). (Zhong-shan also use /mok / 0 (See /p'ow55/) om O = CL for leafy vegetables (e.g., spinach), clumps of grass, bushes, etc. - 138 -on ^ = straw (Also used for f u e l i n the cooking stove.) 22 , 13 /sow kon / (^ ^ ) = (straw) broom. , 55 51 = to put, to place; /on jan / = (^ A-) = mother-in law. ong 51 ^, 51 a. / m a J X f o n g ' V t # K ~ ) = c a n d y , = c r a z y . O = s t a l e , •^ t = t o s h a k e o u t . o p ^ = c o m p a t i b l e . o t O = t o s c o l d , t o . r e p r i m a n d . ok j* = to h i t hard on the head. O = to chop (e.g., wood). A i 5 55 55 jwl = to peel (e.g., s h e l l ) ; /mok fa sang / (~ MJ £.) = to s h e l l peanuts; to be the t h i r d party (si.) = don't (negative imperative); /mok2 f i 5 5 / (^ ^ ) = could i t be possible that . . . (Apgroxima£c|ly simi l a r to the Cantonese use of [m f u g ]»%3aL) • / j o n g 2 2 mok2/ ^) or / s i o n g 2 2 mok2/ (J£, ^ ) = approximately. = to carry with head or shoulder. - 139 -t'ok^ O = to explore 5 Iok O = to extract teeth ... }%r /kok 2 Iok 5 t'aw 5 1/ or /klok 5 t'aw 5 1/ ( $ ^ ) = corner. 2 = inchoative S - f i n a l marker (variant: /ok /). . O = to support s.o. ph y s i c a l l y . O = to quiz o r a l l y . O = to knock on the head. 5| = to raise the head, or t i l t i t up. a O /ham 5 1 p a 5 1 l a : n g 5 1 / or /ham 5 1 p l a : n g 5 1 / = a l l (.&).. p ' a 3 i O = to paddle (e.g., canoe). 55 . . ma ty- = twin. O = /ma 5 5 k u 5 5 i n 5 5 / (/-O^!) = cigar; /ma 5 5 k u 5 5 i n 5 5 t a j 2 2 / (~0 ' A £ ^ ) = cigarette butt; 55 55 (/ma ku / i s borrowed from the l a s t two s y l l a -bles of Portuguese 'tobaco'—term used i n Shi-qi.. Observe that the Portuguese word was recorded i n the Sino-Poirtuguese glossary using the Chinese w o r d s ^ " 5 5 / Pronounced i n Zhong-shan today as /t a : j ma ku / ) . 2 2 ma 0 = a pause-PRT at the end of a clause, ta jft = dozen (from English 'dozen'). t a 5 1 O = s t i l l , yet CAlso: /wa:n51/)J^,) . t a 1 3 jft / t a 1 3 l i 1 3 / = to take care of a place .(e.g. , a room), to keep a place clean. n a 5 5 = a scar. O = to s t i c k ; / t s ' i 5 5 n a 5 5 / (J>$o ~ ) = sticky; /m51 n a 5 5 ka:ng 5 5/ ("%^0) = don't connect (as of unrelated t o p i c s ) . Iok 2 ts'ok 2 k'ok 5 ngok 2 - 140 -22 na = an i n t e r j e c t i o n — H e r e ! (Chao).. l a 5 5 O / k a 5 5 l a 5 5 h a 2 2 / (~ O T ) or / k l a 5 5 h a 2 2 / = armpit ( f l£-) . = S-PRT—exclamatory p a r t i c l e of affirmation (from / l o k 2 a 5 5 / *>£-*f=i ) . RI Ri RR *• l a 3 X O / l a a:j / (~ = d i r t y . vjfc. = S - i n i t i a l p a r t i c l e — T h e r e ! 13 l a O = pungent, alkaline taste. l a 2 2 = a crack (e.g., i n the wall) . t s a 5 1 O = / t s i 5 1 t s i 5 1 t s a 5 1 t s a 5 1 / = yakkety yak, talka-t i v e , gossipy. 22 22 22 tsa O = yet (Informants claim i t i s from /mi a / 0 > . 55 ts'a O = bad. 22 22 £l< sa O = to loosen (from /san / ?) . k a 5 5 O / k a 5 5 ma 1 3 s i 2 2 / = an "X". / k a 5 5 / or / k a 5 5 h a 1 3 / (~ "F ) (Chao has / k a 5 5 ha 2 2/) = now. (There i s a saying that goes:4* *^ »w-*1)L"lr H| "tjj. < fiP : Fo^^han uses , Canton uses ^ ( i . e . , [ j i x k A : ] ) . 22 55 > ^ = elder s i s t e r ; e^g., / t a : j ka / ) = eldest s i s t e r ; /ngi ka / ~ ). = second eldest s i s t e r , etc. (Also used: / t s i / k% i n thc^same-^nviron-ment for 'eldest s i s t e r ' , but /ngi t s i / ) i s only used i n salutations i n l e t t e r s ) . 51 ka O = standing position with-legs .astride. k ' a 5 5 O /k'a:w55 k ' a 5 5 / = to make an "X". O / k ' a 5 5 k ' a 5 5 s i w 2 2 / ( IL) = to laugh. k ' a 5 1 £-(; / k ' a 5 1 t s a : t 2 / H f ) = cockroach. n g a 5 5 ^ /nga 5 5 t s a j 1 3 / if & ) = a baby; 22 55 J . / t a : j nga / (?^ -'~' ) = big baby—but used for the eldest of one's younger brothers; / • 2 2 55/ /ngi nga / brothers, etc. n g i 2 2 5 5 (^ ~ ) = next eldest of one's younger - 141 -2 2 2 2 O /nga tsa / = awkward,, bulky. O /nga 2 2 t s a 2 2 / (~ fllf) = said of s.o. who hogs space food, the road, etc. 55 A' *f*J = Q-PRT (phonetic variant: /o /»4\ ) (Chao) . w[3 = Q-PRT to e l i c i t r e p e t i t i o n of information (.Chao) » ^ = S - f i n a l PRT—affirmative exclamation. (Chao) . aj *JL /maj 5 5 m a j 5 5 s i w 2 2 / 0 ~ ~ ? L ) = to gri n . = to cram; /maj 5 5 s y 5 5 / (~ ) = bookish; to cram. O = then (e.g., i n conditional.Jsentences) . O = precisely, exactly. O = rotted Xe.g., wood); useless (e.g., p e o p l e ) — from / f a j / jfc ?) (Chao). O = to hand s.t. up to s.o. Jj? /a t'aj / (^L^) = term of address used by the older generation to a boy (more formal). 5 3 O = some (Cantonese: [ t i : ] ) • 55 13 O /naj tsaj / (~ /(V) = small amount; 5 55 / j a t naj / (— ~ ) = a small amount (post-verbal) 5 55* / j a t naj / (w~<*) = a l l (pre-verbal) . O = where O / n a j 2 2 ka:k 2/ {~j[%r) = adhering s t r i c t l y to rules and regulations. |\j = fat t y or greasy (of food) . = to place (Zhong-shan usually use: /on / ty , n / f o n g 2 2 / ^ , or / t a t 5 / O ) . = son, c h i l d ; diminuative s u f f i x . " = said of food that i s hard on the digestion. 1 55 5 O /saj t s i k / = a side glance. - 142 -it- / m 5 1 s a J 1 3 / (*«'~) = don't need to, don't have to. O / t o 5 5 k a j 1 3 / C$t~ ) = cunning (from / t o 5 5 k a j 2 2 / 'many schemes'?).. (Chao). \% / k ' i n g 5 5 k a j 1 3 / (>fl~) = to chat. O / l o n g 5 1 k ' a j 5 5 / (4jL~ ) = daring. O = to take a very small b i t e . O = to b i t e using the top and bottom front teeth (e.g., to eat melon seeds). J$z = vagina (Chao) . O /nga:n 1 3 h a j 1 3 h a j 1 3 / ( 0 & ~ ~ ) = expectantly, eagerly (Chao). 5 22 = to be; / j a t haj / (.~- ^ ) = or, or else. 22 22 £ O /aj na:w / (.<"<,' If! ) = noisy and bustling. 22 22 O /aj f a j / = worried, unsettled f e e l i n g . 22 22 O /aj tsaj / = st u f f y (e.g., of a room). ^ = to be able. O = bad, naughty. ft eJQj = to badger. a: j O = mile (from English 'mile'). iJ£ = to be close or near; with, along with, close to. 2 51 jifL = aspect marker of completion (e.g. /ja:k ma:j / n <i >~= to eat up the rest) . y = remaining l i q u i d and substance at the bottom of a bowl or cup. O = slow, pokey. 22 55 Jrfl /siw na:j / (^ ~ ) = wife of a young master. )0 / n a : j 5 1 n a : j 1 3 / ( ~ ) = mothjr-in^law (used i n r i c h families; otherwise /on jan /Q A- )• O = t i r e d ; /nga:n 1 3 n a : j 2 2 / (SjL^ ) = sleepy. - 143 -22 na:j O = to be connected to. 51 51 55 l a : j O / l a : j t a : j / = sloppy. l a : j 2 2 = to pour l i q u i d (e.g., sauce) over s.t, 55 L * t s ' a : j = to knead; to press on top of s.o. 13 ts ' a : j = to step on, to tread on. sa:j = to waste. 22 ka:j = to saw (something large). 1 3 J k'a:j ffy = to take to s.o. = a section (e.g., of an orange). n g a : j 5 5 ^ / n g a : j 5 5 t s i 1 3 / (~ 3~) = very poor people who l i v e near Shi-qi; e.g., such women are c a l l e d : / n g a : j 5 5 t s i 1 3 p ' o 5 1 / ( ~ - V $ ) . n g a : j 1 3 O / t s a : n g 5 5 n g i 5 5 n g a : j 1 3 / (#" O ~ ) = almost (.i.e., lack just a l i t t l e bit). . 5 5 ha:j O = to brush against l i g h t l y , to touch l i g h t l y . 51 ha:j O = itchy. 51 51 51 k'wa:j <*jjf|'\ = bad, michievous, vill a n o u s ; /k'wa:j jan / ( ~ / / 0 = v i l l a i n s . k'wa:j 2 2 $L /k'wa:j 2 2 t s i 1 3 / (.*>4*) = chopsticks. aw paw 5 5 = to squat. 51 paw tf> = to s t i c k out i n a lump; a swelling. 22 22 22 paw 0 /law paw / = clumsy. p'aw 2 2 /p'aw 2 2/ or /p'ung 5 5 p'aw 2 2/ = f l u f f y . maw55 / t i 2 2 maw55/ (^0'°) = ru f f i a n s , b u l l i e s . taw 5 1 0 /low 1 3 taw 5 1/ (£ ~) = father (si.) taw 2 2 O = to be near to; /taw 2 2 h i 1 3 k'an 1 3/ l~ or /taw 2 2 t s y 2 2 / (~&) = to be near. - 1 4 4 -= to rest. O = to unwrap. = to be angry. O = to c l u s t e r around, to hover over. O = plant from which the Annamese use the leaves to r o l l betel-nuts to chew (Chao). ^ = overcoat. = to cover over. •frxh = to p u l l up (e.g., pants); to l i f t up. 2 2 = CL for keys ( i . e . , a bunch) (Also: /nang /• O = to do s.t. taking advantage of circumstances, <Q = CL for stones, etc. ( i e . , a lump). /kaw 2 2 n i n 5 1 / ( ^ ) = l a s t year. O = to watch clo s e l y . O /aw 5 5 l a w 5 1 / = dirty , a:w /na:w k'a:w / {~&K) = to argue, to quarrel i.% = wrinkled. 0 /sa:w 5 5 k o n 5 5 / (.~#Ll ) = to air-dry ( i . e . , as opposed to drying i n the sun). 0 = to grab. 51 51 O /la:w ka:w / = hectic, m a hurry; / k i 5 1 l i 5 1 ka:w 5 1 la:w 5 1/ or / k l i 5 1 kla:w 5 1/ = noisy t a l k i n g — a s of a group of people. 0 /k'a:w55 k ' a 5 5 / = to make an "X". ft? = to scratch; /nga:w ha:j / = to scratch an i t c h . = to rake (e.g., leaves, grassjl °$^= to b i t e . 2 2 O = to enjoy (e.g., talking) (Also: /how /•£$-). 51 51 :i O /ha:w p'o / (~ dp ) = a loose woman; 51 51 51 /ha:w t'an t'an / = said of a loose woman i n reference to her mannerism, behaviour, etc. 13 22 2 22 O = j o i n t ; /saw a:w / ^ ) = wrist; /kiok a:w / = ankle. am = pump. 55 13 /f^ = to pump; /tarn s0j / {r^j^~) = to pump up water. 51 51 51 51 O / t i t i t a i r v t a r n / = said of s.o. who i s quick-footed or nimble. 13 35 ]$i^_= to throw, to throw away; /tarn ' / = to excel. 22 55 = to hang down; /tarn tgij / = shabby. ft? = to trample; /tarn 2 2 saw 1 3 tarn 2 2 k i o k 2 / (~ ^ jJ^ P ) ^ = to stamp the feet up and down (as i n a tantrum). = a pool of water. «jg /t'am 5 1 t'am 5 1 t s y n 2 2 / (<~~ ) = to c i r c l e about. = to lure. O / t a j 1 3 t a k 5 nam 5 5/ (iMf-') = to be able to endure or withstand; 55 13 /nam tsaj / (~"14-) = s.o. who i s w i l l i n g to do favours, etc. for people. O = s o f t . to think. 0 = soaked. 55 51 ^,...51 * 55 51 51 0 = to shade over; /lam p'ung t'aw / tXL. = term for the r a d i c a l mi an " • > 7 " 'a r o o f . O = to collapse j ^ v = to p i l e up. O / t s i 5 1 t s i 5 1 tsam 5 1 tsam 5 1/ = muffled sound of voices. - 146 -O = a callous. O = to sprinkle (e.g., seasoning, f l o u r ) . ^,^/kam 5 5 t s i 1 3 / (.^4-) or /kam 5 5 j o n g 2 2 / C~ <flo = thus, i n t h i s way. 51 51 51 51 51 51 0 / k i l i kam lam / or / k l i klam / = denotes noisy state of a f f a i r s , . O / k i n g 5 1 l i n g 5 1 kam 5 1 lam 5 1/ or / k l i n g 5 1 klam 5 1/ =» noise from stamping feet, moving furniture, dropped objects, etc. /kam 2 2 t o 5 5 / l~ ) = t h i s much, that much. A'am 5 1 l o 5 1 / ) = spider. O = to lean with forearms resting on a surface (e.g., a r a i l i n g ) . 51 51 55 O A'am k'am ts'ia:ng / = denotes doing s.t. i n a big hurry.or f r a n t i c a l l y . {2 = a l i d , a cover; to put a cover or l i d over s.t. O = to slap i n the face. O = toothless. = to whisper; to grumble; /ngam 5 1 ngam 5 1 ts'am 5 1 ts'am / = grumbling sound. O /ham 5 1 p a 5 1 l a : n g 5 1 / or /ham 5 1 p l a : n g 5 1 / = a l l (fKi) . O = to bump against. O = to press down on. O = to f i s h s.t. out of one's pockets. a :m = tender beef; / n g 1 3 f a 5 5 na:m 1 3/ {Z.tw) = side pork. = to hold close to the body (e.g., as of s.t. very dear or precious); to hold s.t. large and bulky (e.g., a p i l e of clothes). = to step over. | /mun51 ts'a:m 1 3/ i f ] " ) = door-step, threshold. - 1 4 7 -$L / l a : p 2 1 ^ a : p 2 ts'a:m 1 3/ or / f a n 2 2 k i 5 5 2 2 y 1^ >'a:m / ( ^ |£ ~> ) = dustpan. (See also: /fan / O = bran for feeding pigs. ~ J u s t ' correct; /nga:m55 nga:m 5 5 s i n 2 2 / (~~ $lL) = just r i g h t , exactly. O /p'uk 5 nga:m 1 3 t s ' y 5 5 / or /p'uk 5 k a : j 5 5 t s ' y 5 5 / = to be i n a rush. ^Note: ^ not an abusive term as i n Cantonese [ p ' u k k A i j ] 'Drop dead!'). . 22 22 (s /ha:m s i n / = telephone. O = to move up and down ( i . e . , standing and squatting i n succession). an I / t a : j 2 2 p a n 2 2 t s i o n g 2 2 / = elephant. 2 2 2 0 /pan ts'ot / = stubborn; sucker ( s i . ) . gill) = to sleep (Chao)—usually / m i 5 5 / (. ;jjL) i n Zhong-shan. j£ = ash u s e d to bury human waste, which i s l a t e r ^ used as f e r t i l i z e r ; / f a n 2 2 k i 5 5 / (~ ^ ) = a shallow bamboo pan; / f a n 2 2 k i 5 5 ts'a:m 1 3/ (-$ %t ) = dustpan. O = to put s.t. down very roughly; to s i t down hard and abruptly. O = to pace; / f a n 5 1 s i o n g 3 1 f a n 5 1 l o k 2 / (~Jt~/S*) = to pace up and down. 0 / f a n 5 1 f a n 5 1 t s a n 2 2 / ( — % ^ ) - shivering, shak-ing; / l a : n g 1 3 f a n 5 1 f a n 5 1 / (;^ ) = shivering cold. 0 / f a n 2 2 haw 2 2/ C ^ f c j = to move backwards. = to exasperate. ^ / l a n 5 5 s i o n g 2 2 l a n 5 5 l o k 2 / ( = to chatter incessantly (Chao). / t s ' a n 5 1 / , / i n 5 5 t s ' a n 5 1 / (>M^) or /puk 2 t s ' a n 5 1 / = dust. - 148 -O = to shuffle the feet Cas of old people). 55 55 55 ^ /ngan / or /ngan tsan / - small, stunted i n ^~ growth. ^ = to j i g g l e the feet while s i t t i n g down; /ngan 2 2 k i o k 2 f a 5 5 / l~ftif&). = the j i g g l i n g of the feet up and down while seated (considered impolite). a:n O = to brush or wave away; to dust. O = to p u l l down. O = to climb, to hold to, to hang on (Cha^ wonders i f i t i s an i n i t i a l change from /p'a:n /, with the same meaning) . 4" / j a 1 3 ma:n 5 1/ {f$~ ) = peevish (referring to c h i l d -^ ren only); 55 51 /tiw ma:n / (.3 ) = unreasonable, s p o i l t , peevish. ^ / t a r n 5 5 t a r n 5 5 / ) = only. jffijj^ = a stand, a s t a l l (e.g., for f r u i t s , f i s h , etc.) (Chao). a& 55 55 2 sMi = to crawl, to creep; /la:n s i kat / = Get out! 22 55 22 55 0 /la:n l i la:n tsaw / = raggedy. 22 51 13 0 /la:n t'aw t'iarng / = s.o. who t r i e s to pre-vent trouble—extended to mean s.o. who wants to be the f i r s t to do s.t. O = quaint (of people). 0 /k'a:n 5 5 k'a:n 5 5 pow 5 5 pow 5 5/ ( — = pots and pans. 0 / k i 1 3 sap 2 nga:n 5 5/ (^xi^~) = said of someone who i s very, very old. = to crush by r o l l i n g . O /kwa:n 5 5 l a : n 5 5 taw 1 3/ (~ O-^ ) or / k l a : n 5 5 taw 1 3/ = a somersault (from Lo; Yang gave: / f a : n 5 5 kam 5 5 taw 1 3/ 0^ ). - 149 -ang pang 22 mang" 55 mang 13 fang tang 51 22 nang 55 22 nang ts'ang 22 sang 55 k 1 ang 55 k 1 ang 22 51 ngang 55 hang" hang 13 ang 51 i . 5 5 5 5 = to lean against; /aj aj pang""" pang"""/ C$Li^ / v "' % ' ). = describess .o. who always leans against s.t. 22 22 = to p u l l , to yank, 55 55 22 O /mang k'ang siong / (~ Oi% ) = a sour, puckered expression. |$L = bold, f i e r c e . 0 = to punch. O = to take care of (e.g., children); to take s.o. some place. O = to knock s.o. on the head. Q = a knob. 55 55 43 5 5 5 55 O /niw _ nang / = precarious; / k 1 I tak niw nang / (4l 1 f i'l"") = to be standing precariously (e.g., on a high ledge) . 51 55 iWi O = a knot, a j o i n t ; / t s * a : j nang / {%. ~ ) = a knot i n a piece of wood. O = said of thread hanging out. 22 55 O /nang kow / (~ & ) = to be on tip- t o e s . O = to moan and groan continuously; to complain. ||| = to blow the nose; /sang 2 2 p i 2 2 k o 5 5 / (~|p^[ ) = I to sneer. O = to l i g h t l y snap against s.t. (e.g., chinaware or glassware to f i n d out i f i t i s of good or poor q u a l i t y ) . 0 = to knock (e.g., door). 0 = powerful. 51 51 55 O /ngang ngang sia:ng / (•—/%. ) = moaning sound. O = to grumble or complain continuously. O = taut. O = to press against; /ang 1 3 k i o k 2 / (r? dtf) = to press against the foot Ce.g., as with a! piece of - 150 -13 55 pebble i n the shoe); /ang sam / (~"/o>) = (s.t.). pressing the heart-—worried. O = to slam (e.g., door). O = to carry a heavy load of things. ^ = to wait; to l e t (Also: / t a n g 1 3 / ^ ). . a:ng O /niw 5 5 pa:ng 5 5 pa:ng 5 5/ (Jf'h^1^) = very slender. O = Bang!—sound of gunfire, f i r e c r a c k e r , slamming of the door, etc.; to slam (the door). 55 55 flfl O /p'a:ng hoj / [~ rfl ) = to divide up. O = to stretch s.t. across or over s.t. else. 55 2 O /ma:ng kiok / (~A*|') = leggings (Chao). $L ~ powerful; /ma:ng13 kwaj 1 3/ (~* %>) = a l i v e l y ghost. O = to kick o f f (e.g., blanket). Q = wool, yarn (from French 'laine'?). 55 55 A jt O /la:ng tsung / ) = alarm clock. O / l a : n g 5 1 l a : n g 5 1 s i a : n g 5 5 / (rJ~ ) = denotes doing s.t. very quickly and e f f i c i e n t l y . O = to s t u f f . • 5 3 - to give b i r t h (Cantonese: [sow ] ) ; /haw 2 2 sa:ng 5 5/ (4^.^) = the young people. = to scour. 51 13 51 2 O = to get i n the way; /ka:ng saw ka:ng kiok / (~% ~ k^f ) = to get underfoot, to get i n the way. 13 22 O / t a ka:ng / ) = a hold-up, a highway robbery. O = h i l l s i d e , mountainside (Chao; he thinks th_at perhaps i t i s a sound change from |§0 /kong /. 'mountain ridge'). = a cooking pot. - 1 5 1 -5 5 kwa:ng O = to lock up, to imprison. i 5 1 kwa:ng o = to slam (e.g., door). 1 3 kwa:ng 0 = stem Ce.g., of a f r u i t ) (Also: / t a j2 2 / ^ ) . , , 5 5 k'wa:ng ! a frame, a framework. 0 = a hoop. k'wa:ng 0. = to e n c i r c l e ; a c i r c l e (as to walk i n a c i r c l e ) . 2 2 k1wa:ng ** = to bump against; to tear one's o u t f i t by catch-ing i t on s.t. o = sound of metal. 5 1 wa: ng 0 51 51 /wa:ng man / = peevish (of a c h i l d ) ; cranky. , 5 1 , . 22 . 2 2 5 1 , , 2 2 , 2 2 /watng.^ / ,. /ka l a wa:ng / or /k l a l a wa:ng / = crosswise. 2 2 wa:ng A /ka:w 5 5 i 1 3 warng 2 2/ = the horizontal bars connecting the legs of a chair. ap t a p 5 o = a small earthen j a r or jug (Chao). u / t a p 5 t a j 5 5 t'aw 5 1/ l-sifoUi) = to lower the head. . 2 tap ik = to s t r i k e or h i t l i g h t l y (as of raindrops, or l i g h t hammering). f a p 5 0 = to cave i n , to collapse; lowered, collapsed. 5 ' " nap GO = dented; /nap5 nap 5 kung 1 3 kung 1 3/ (-~~^ j§ ) rough or bumpy surface, warped. nap 2 yfe = sticky, moist (e.g., of fingers, body); / n i 2 2 n i 2 2 nap 2 nap 2/ (jIL S«Si — ' ) = sticky and moist (of body or body parts). l a p 5 O = to ,!^step into (e.g., water). t s a p 5 i t , = a small amount that can be picked up with the * fingers {e.g., of sand, r i c e , e t c . ) , a pinch of. sap 5 c$ /sap 5 s u j 2 2 / (rs&fy) = miscellaneous; /sap 5 t s a j 1 3 / (^4^-) = a grocery store. kap 5 O /kap 5 t s a j 1 3 / fa ) = a frog. - 152 -k'ap O = to slow-boil or stearn-cook—from sound of slow b o i l i n g water r a i s i n g and lowering the l i d of the pot. 2 2 22 0 /k'ap k'ap tsan / = to shiver. 5 t a. u j-j. • 4 . 1 / 5 55 5 .22, ngap *A^= to chatter i n d i s c r e e t l y ; /ngap sa:m ngap s i / (^js_~* CD ) = to make thoughtless comments. ngap 2 to nod the head. hap 5 <s" = to close the eyes; to doze, ft 5 ap = to cover (e.g., with a medical or herb patch). / * a: p 2 ta:p O = to request s.o. to do one a favour (Also: /t'okV-tb ) • t'a:p 2 O = a container; /maj 1 3 t'a:p 2/ ( i L ^ ) = r i c e bin. 2 na:p O = to catch on s.t. l a : p 2 i& = to c o l l e c t ; / l a : p 2 n a : j 1 3 / (~#) ) = a hoarder. O = to push up (e.g., sleeves). tsa:p ^.^/tsa:p / t s i k / (.MA'] ) = to lean sideways. 2 sa:p 'AlSL = t o c o o k by b o i l i n g . 0 = to sprain (e.g., ankle). 2 nga:p "i^ L = t o f o l d up or r o l l up (e.g., sleeves). 2 2 51 ha:p 0 /ha:p p'a / (~ ^ ) = chin. a:p 2 = to pressure or force s.o. to do s.t. j a : p 2 ^ = to wave the hand. at 5 5 5 13 mat <Li /mat / or /mat ja / (r'J-i ) = what (Chao suggests perhaps sound change f r o m ^ /mat /).; /mat5 k a : j 1 3 / ) = why. f a t 5 0 = to sweep up (.e.g., onto the dustpan). O = a p i l e (e.g., clothes). O = to place, to put. O = to glare at s.o. O = to s l i p down. O = to dip into sauce. O = anxious, uneasy. JtL = to come o f f . O /wa:t 2 l a t 5 l a t 5 / (Jj| ) = very smooth. O = to erase or blot out Ce.g., wrong characters). O /m51 juk 5 t s a t 5 / ("4r^f' v) = not move at a l l (to denote something that i s d i f f i c u l t to move). j|l = a cork; to cork. 13 2 2 O /haw tsat tsat / (<X ^~ ) = to stammer. O = to hesitate; nervous; a j o l t . *<] = to pierce. O = to put the hair up; to get up (vulgar term); to lean over with derriere up. ^/U = to s l i d e over sideways i n a seated position; / s i 1 3 f a t 5 ngat 5 ngat 5/ ) = derridre always moving—can't s i t s t i l l . 5 5 5 0 /ngat ngat tsa t / = not open-minded. £Jfc = blunt, dull-edged. O A'wat 2 t'aw 5 1 l u 2 2 / C^SSlS) = a dead-end street. = kernel, p i t (.of a f r u i t ) ; /ngazn 13 wat 2/ (8fL~) = eye, eyeball; /nga:n 1 3 wat 2 t i n g 2 2 t i n g 2 2 / ( J / L ~%~^ 2\ = f $ a r i n § in t e n t l y , dec|g i n thought (Also: / t i n g h i j a t p'un siong / / 3 L i&-~ O ^ 0 ) • /nga:n 1 3 wat 2 t a t 2 t a t 2 / ( i i ^ O O ) = to glare at s.o. l3 = the sun. - 154 -a: t o p'a:t O = a small amount of l i q u i d . 5] 51 2 2 51 2 O / P ' i l i p'a:t l a : t z / or / p ' l i p ' l a : t V = sound of water being splashed (as by young children) 2 ta : t O = to f a l l (from a great height). = to toss c a r e l e s s l y . ^ = CL for locations. f a : t 2 0 / f a : t 2 h a 1 3 t ' a : t 2 s i a : n g 5 5 / ) = sound of s l i p p e r s . na:t 2 = to have a l i g h t burn (e.g., on the hand from cooking, or by a l i t c i g a r e t t e ) . l a : t 2 O / l a : t 2 t ' a : t 2 / = d i r t y . O = to v i s i t s.o. 2 sa:t O = to keep badgering s.o. for s.t. (as of a c h i l d ) . nga:t = stench of urine. O to s l i d e back and forth i n the seat. 5 5 51 £ ha:t O /ha:t t s 1 i / (r^ *jf J ~ a sneeze. a : t 5 = to press. 2 2 2 a:t = to be pressed; /a:t l i k / fj ) = pressure. ak mak5 /mak5 t'aw 5 1/ (~S-t.) = trade-mark (from English 'mark'); good-looking (Chao). 2 tak 0 = to n a i l . 5 55 55 5 5 nak O / n i n i nak nak / = describes walking on t i p -toes, as on high heels. O = great, great grandchild. 2 -nak 0 = to tread on, to step on. 22 2 0 / n i nak / = sti c k y . 5 5 5 lak %b /lak k'ak / = jerky (of speech); rough (of surface) - 155 -5 rJr" 5 5 sak fjg- /sak ak / = a hiccup. = great grandchild. ngak 5 ofo = to cheat. 2 ak = S - f i n a l PRT (Chao). 2 22 2 2 k'wak O /nga:ng k'wak k'wak / = denotes s.t. very hard and s t i f f (e.g., of a piece of stal e bread). 2 wak O = to swing one's arms (while walking). a :k 2 ma:k O = to open (e.g., mouth, eyes). 2 fa:k O = to whisk, to beat (e.g., eggs); to swing one's arms back and forth. 5 5 5 5 5 ta:k 0 / t i k t i k ta:k ta:k / = c l i c k i n g sound of high heel shoes. t'a:k 5 O /ma:n13 t'a^k 5 h a ^ 5 / (0jfe> ~ ) = night-time (Also: / j a ma:n / ^ Qfij ) . l a : k 5 0 / p i k 5 l i k 5 pa:k 5 l a : k 5 / or / p l i k 5 p l a : k 5 / = crack-l i n g sound—of f i r e , or s t r i n g of fir e c r a c k e r s . l a : k 2 $/) /mok2 ts'a:k 2 l a : k 2 / (.&) 0 ) = bard (to waist). 51 51 2 2 O / l i l i la:k la:k / = low r u s t l i n g sound of paper or si m i l a r sheets of material. 2 sa:k O = to chop (something large). Q = a piece (e.g., f i e l d ) . 2 ka:k Q = to sieve. $r = abnormal or strange (of person). a:k 2 = a bracelet. 2 4 k'wa:k el* = to make a loop (e.g., i n sewing); to wrap around and t i e with s t r i n g or cloth; to la t c h . j a : k 2 o'U = to eat; / j a : k 2 i n 5 5 / ) = to smoke (cigarettes, etc.) - 156 -long = to peck; to h i t against a surface with a small, sharp object (e.g., a small chip of glass).. iok = to chop—a singular action, as to chop s.t. i n h a l f . = to chop repeatedly to small pieces, as to make minced meat. 2 2 5 / t s i o k j a t ' / (/^ 8 ) = yesterday. i a 55 13 = beer (from English word). (Also: /pia tsaw / .~ 1% ) • / s a : n 5 5 p ' i a 5 1 / (ti* — ) = h i l l s i d e . = to carry s.t. that hanejg from the arm (e.g., a purse) (Cantonese: Iwaxn ] ) . = what (from /mat5 j a 1 3 / - t i f f ). = slanted; / w a : j 5 5 w a : j 5 5 m i a 1 3 m i a 1 3 / """^ ) = to denote s.t. that i s askew or slanted. 13 13 / l i a f i a / = sloppy, d i r t y . 13 13 22 / t i a / or / t i a tiw / = to be s p o i l t or pampered. (See: / f i a 1 3 / ) = subordjnative s u f f i x (Chao) (Zhong-shan usually uses /ko /.) = a part of a tree, e.g., a branch. = stool (vulgar term). = to stagger. i a : w = to toss away. 55 55 13 O /lia:w sung low; / O' O v&) = term for north-ern Chinese p e o p l e — a c t u a l l y from lao xiong {JLf^-i) 'elder brother", a p o l i t e term of address. O = to grab. 51 51 55 O /kia:w kia:w sia:ng / (~ f^-e, ) = denotes noisy rowdiness. O = to s t i c k gut (e.g., l i d of an opened t i n can). (Also: /hiw / ) . ia:m O = a s l i g h t cough, or the sound of a l i g h t cough.. ,ia:ng 13 * O /pia:ng fa:n/ (r* 1&>) = to give back. O = to throw s.t. at s.o. 0 = the back (of a c h a i r ) . jfJi /jaw 5 1 p ' i a : n g 5 1 t s a j 1 3 / (jlfc^te") = a c h i l d brought to a second or subsequent marriage by a woman; / t ' o 5 5 jaw 5 1 p ' i a : n g 5 1 / (#b;»£~) = to bring such a c h i l d to a subsequent marriage. (Uncomplimen-tary terms). 0 = a piece or a sheet (e.g., paper, land). 0 = sound (e.g., of a drum). O /tia:ng / or /tia:ng fong / (.-»7> ) = a place. 0 /Verb + t i a : n g 2 2 t i 2 2 / (Verb + ~ = to have just V'd. ^» = an open space, an area of l e v e l ground; 51 51 /wo t'ia:ng / (.^^ ) = area for threshing grain (Chao). 22 51 / t i t'ia:ng / (>€J^) = area i n front of the house for drying grains, etc. i n the v i l l a g e . O = to hide. $#J = pretty. - 158 -kiarng 3 j^^ /kiazng / or /kia:ng sia:k / (~ <ta ) = to be very careful with, to handle with great care and gentleness. k ' i a z n g 5 1 O / k ' i a : n g 5 1 / or / l i a : n g 5 1 k ' i a : n g 5 1 / (yf ~ ) = auspicious. 13 22 51 13 t'ia:ng 0 /la : n t'aw t'iarng / = s.o. who does s.t. f o r another person which he finds embarrassing or disagreeable to do. xa:p ti a : p O = to taste a l i t t l e , ( a s to try i t ) . k i a : p 5 O /ngo 2 2 tow 2 2 k i a : p 5 k i a : p 5 s i a : n g 5 5 / £'] ~ ~ ) = hungry to the extent that the stomach i s pro-testing, i.;e., very hungry. i a : t p i a : t 2 O / l a : n 2 2 p i a : t 2 p i a : t 2 / ('/{j^ ~ ~* ) = soft and mushy. p ' i a : t O / p ' i a : t p ' i a : t / = derriere (speaking to a young c h i l d ) . 2 p ' i a : t O = a mass of soft, mushy substance; / l a : n 2 2 p ' i a : t 2 p ' i a : t 2 / 1%$""* ) = soft and mushy; / p ' i a : t 2 s i 1 3 / {^M-) = said of s.o. who s i t s there and does not f e e l l i k e moving. (Very uncom-plimentary term). t i a : t 2 O / l a : n 2 2 t i a : t 2 t i a : t 2 / #|0 ~ ~ ) = soft and mushy. i a : k p'ia:k O = to throw around car e l e s s l y (e.g., one's clothes) 2 tia:k O = t o chase s.o. away. l i a : k 2 ;i& / t a : j 2 2 l i a : k 2 / (A~ ) and / s a j 2 2 l i a : k 2 / (i.©~) /'-^ ~ = names of v i l l a g e s (Chao) . 5 13 5 -lia:k O = smart, clever; /la:n l i a : k / = smart alecky. t s V i a : k 2 = pa i n f u l ; /t'aw 5 1 t s ' i a : k 2 / ( S J l ^ ) = to have a - 159 -" headache. 13 2 = upset, hurt; to be mad at; /nga:n ts ' i a : k / r J) = to be envious. = to love (e.g., a child) (Can also be used to abuse s.o. by intending the opposite; i . e . , by the use of sarcasm.) (Syllabic). 51 55 >* P£ = not; /m koj / lr/> ) = please; thank you (after s.o. does a p o l i t e deed, such as handing one a cup of tea, a bowl of r i c e , etc.) ffi / n g 1 3 tsok 2 low 1 3/ = undertaker. (Also: /kun 5 5 t s ' o j 5 1 low 1 3/ kf & > • [ e ] 0 S- or c l a u s e - f i n a l PRT. (Chao). [ e j ] fX^= exclamation of affirmation (Chao). [el O = S - f i n a l PRT. (Chao). 0 = S - f i n a l PRT. (Chao). [a] = Q-PRT for yes-no Q's and negative Q's. (Chao). a 0 /mung51 k'wa 5 1 k'wa 5 1/ ($$^~~') = very blurry. - 160 -/pa:k 2 j a 5 5 / (& ~ ) = father (term of address y = to poke with a long object (e.g., a closed umbrella) - 161 -PART I I . DIACHRONIC STUDY CHAPTER 3. ANALYSIS OF MODERN REFLEXES TO HISTORICAL CATEGORIES. Bernhard Karlgren, a pioneer i n applying Western l i n g u i s t i c methods to the study of the h i s t o r i c a l Chinese sound system, established two stages i n the language: "Ancient Chinese" and "Archaic Chinese". As outlined i n his Compendium (.1954:212), Karlgren designates "Ancient Chinese" to be the language of c i r c a 600 A.D., as c o d i f i e d i n the Qie-yun "gift rhyme dictionary, which he i d e n t i f i e s as the d i a l e c t spoken i n Chang-an, Shen-si. "Archaic Chinese", on the other hand, refers to the language spoken i n the He-nan region during the f i r s t Zhou centuries (from 1028 B.C.), based partly on the rhymes i n the S h i - j i n g ('Book of Odes') and other early manuscripts, and p a r t l y on xie-sheng characters (compounds containing a " r a d i c a l " (or " s i g n i f i c " ) and a "phonetic"). Following Karlgren, most Chinese phonologists continue to reconstruct Ancient Chinese as a single stage i n the language based on the Qie-yun. Pulleyblank, who prefers the'terms "Middle" and "Old" Chinese to Karlgren's "Ancient" and "Archaic" Chinese, further subdivides Middle Chinese into "Early Middle Chinese" (EMC) , and "Late Middle Chinese" (LMC). Pulleyblank (.1977:12) emphasizes that the s h i f t from EMC to LMC i s not simply a case of h i s t o r i c a l evolution of the language, but represents a major s h i f t i n d i a l e c t base. Pulleyblank (1970:204, 1977:4) considers "Early Middle Chinese" to be the language of the Qie-yun, which he iden-t i f i e s as the standard Mandarin of the Northern and Southern Dynasties i n the s i x t h century i n the courts of Lo-yang\ and - 162 -Nan-jing, and "Late Middle Chinese" the language of the rhyme tables, representing the speech i n Chang-an during the middle and l a t t e r part of the Tang dynasty (618-907 A.D.). Although h i s t o r i c a l l y the two languages do not constitute a continuum i n the evolution of Chinese, i n formulating deriva-t i o n a l rules, Pulleyblank (1970-71:204, 1977:12) nevertheless treats Late Middle Chinese as though i t has developed from Early Middle Chinese. He j u s t i f i e s t h i s treatment on the following grounds: since the two languages are clo s e l y related and t h e i r phonological categories are on the whole quite congruent, Late Middle Chinese must have descended from a language the phonological system of which was very si m i l a r to that of Early Middle Chinese. Hence, i t makes sense to est a b l i s h rules to derive the former from the l a t t e r even though t h i s does not quite exactly r e f l e c t h i s t o r i -c a l r e a l i t y . According to Pulleyblank, i t i s Late Middle Chinese which formed the basis for both Sino-Japanese (Kan-on) and Sino-Korean. Moreover, with the exception of Min, a l l the modern Chinese dia-l e c t s can be traced back to Late Middle Chinese; that i s , the phonological system of the modern d i a l e c t s — i n c l u d i n g both l i t e r a r y and c o l l o q u i a l l a y e r s — c a n be explained on the basis of the phono-2 l o g i c a l categories of Late Middle Chinese. The primary source for the reconstruction of Late Middle Chinese i s the rhyme tables, which have evolved from e a r l i e r rhyme d i c t i o n a r i e s which, i n turn, are expansion and revisions of the Qie-yun of Fa-yen Lu, completed i n 601 A.D. The rhyme tables have e s s e n t i a l l y extracted the information contained i n the rhyme dic t i o n a r i e s on the pronunciation of the language, and have /•; - 163 -systematically arranged the information i n the form of a g r i d . Thus, tones and f i n a l s , for example, are placed along the v e r t i c a l axis, while i n i t i a l s are arranged along the horizontal one. Although these rhyme tables do not contain phonetic descriptions of the sounds i n the language, they do contain descriptive terms for the place and manner of a r t i c u l a t i o n of the phonological cate-gories, of which some have to be interpreted. Kai-kou versus he-kou, for instance, has been understood to be a d i s t i n c t i o n of the absence of l a b i a l i z a t i o n versus i t s presence i n the f i n a l s . Other terms include ya-yin % % 'back-tooth sound' for velars and cun-yin % ' l i p sound' for l a b i a l s . The term qing \ 'clear' i s used for i n i t i a l s reconstructed as p l a i n , unvoiced segments, and ci-qing )%, }\ 'second clear' for the aspirated ser i e s . Zhuo or 'muddy' i n i t i a l s are usually assumed to be voiced i n i t i a l s , although Pulleyblank (1970-71:210-211) reconstructs such i n i t i a l s as p l a i n consonants accompanied by voiced aspiration i n Late Middle Chinese which have evolved from e a r l i e r voiced i n i t i a l s (e.g., the Ding ^ i n i t i a l i s reconstructed as LMC *tfi-< EMC. *d-). In discussing the reconstructed values of h i s t o r i c a l Chinese phonological categories, Pulleyblank's reconstruction of Early and Late Middle Chinese w i l l be adopted; the following sec-t i o n w i l l therefore be a b r i e f description of Pulleyblank's system. No attempt w i l l be made to discuss t h e o r e t i c a l issues and contro-versies connected with c e r t a i n reconstructions. Following the introductory description of Middle Chinese phonology, we w i l l proceed to analyze Zhong-shan reflexes of these Middle Chinese sounds according to the organization of the phonological categories found i n the d i a l e c t survey l i s t , the Fang-yan Diao-cha Zi-biao. . - 164 -The aim of the chapter i s to demonstrate the general pattern of correspondences of modern Zhong-shan to h i s t o r i c a l categories. As a r e s u l t , rather than try to account for a l l the exceptions present i n the data on modern Zhong-shan, there w i l l be an e f f o r t to discuss only some of these i r r e g u l a r sound changes. Emphasis w i l l be placed on the regular pattern of correspondences. As i n the synchronic analysis, the subsections w i l l focus on the main categories of i n i t i a l s , f i n a l s and tones. S p l i t s and mergers of various phonological categories i n Zhong-shan w i l l be discussed, as well as some of the more intere s t i n g problems and exceptions observed. The compilation of Zhong-shan data for the present diachronic analysis i s placed i n Chapter 4. 3.1. I n i t i a l s 3.1.1. Reconstructed Values of Middle Chinese I n i t i a l s In the rhyme tables representing Late Middle Chinese, there were o r i g i n a l l y t h i r t y i n i t i a l s which were l a t e r increased to t h i r t y - s i x . These i n i t i a l s are l i s t e d i n Chart 4 on the f o l -lowing two pages. The l a t e r s i x i n i t i a l s are marked by asterisks (*). As we have already seen i n Chapter 1, the characters repre-senting the d i f f e r e n t i n i t i a l s serve as the names of these i n i t i a l s . The reconstructed values assigned by Pulleyblank (.1977:64) are also given i n the chart. In the set of t h i r t y - s i x i n i t i a l s , the l i g h t l i p sounds (qing cun-yin), or dental l a b i a l s , arose out of the corresponding l a b i a l s e r i e s , or the heavy l i p sounds (.zhong cun-yin) . The Feng %~ (LMC *ffi-) i n i t i a l had s p l i t from the Bing j L CEMC *pf\-> . i n i t i a l , and the Wei 4^- (LMC *v-) i n i t i a l from the Ming B$ Chart 4 . The 36 I n i t i a l s of Late M i d d l e Ch inese . Qing Zhuo (c lear ) (2nd -c lear ) (muddy)(not -c lear -not -muddy)(Clear) (muddy) Qing C i - q i n g Zhuo Bu-qing-bu- -zhuo CUN-YIN sound) LABIALS: a) Zhong (heavy) Bang Pang B ing / Ming m LABIODENTALS: b) Qing %t ( l i g h t ) F e i * Fu* Feng1* ^ f « p ) | ^ ; f « p « ) ^ f « b) Wei* SHE-YIN (tongue sound) DENTALS: a) She-tou£&Jt Duan (tongue head) t Tou Ding N i RETROFLEXES: b) She -shang ^ J ; Zh i Che (tongue up) t r j^jfc t^r* Cheng Niang* CHI-YIN Si % ( f r o n t -t o o t h sound) DENTAL SIBILANTS: a) Ch i - tou&SS. J i .ng k t : ( tooth head) RETROFLEX SIBILANTS: b) Zheng-chi i fe} Zhao ( t rue tooth) 8§ ts Qing :> t s ' Chuan I t 5' Cong ^ t s f i Chuang* X i n X i e fCP s ^ sfi Shen Chan I 5 ty'1 (Chart 4. cont'd.) Qing Ci-qing BAN-SHE-YIN DENTAL: (half tongue sound) BAN-CHI-YIN RETROFLEX: % ** (half tooth sound) ?i YA-YIN VELARS; Jian (back tooth 7 7 sound) HOU-YIN GUTTURALS: Ying Xiao (throat *"/ 7 sound) * Not included i n the o r i g i n a l 30 i n i t i a l s . Ri lz) r Qun kfi Xia xn Yu »4TV # - 167 -(EMC *m-) i n i t i a l . In the case of i n i t i a l s F e i $ (LMC *f-) and Fu (LMC *f-) / although they are distinguished i n the series of * f t h i r t y - s i x i n i t i a l s as having arisen from the Bang (EMC *p-) and Pang ffi (EMC *p'-) i n i t i a l s respectively, Pulleyblank regards that d i s t i n c t i o n as almost c e r t a i n l y just a h i s t o r i c a l one based on fan-qie d i s t i n c t i o n s i n the Qie-yun and reconstructs both F e i and Fu as LMC * f - . Pulleyblank (1970-71:217-218) argues that the d i s t i n c t i o n between Fei and Fu had i n fac t been t h e o r e t i c a l from the st a r t ; they were kept d i s t i n c t p r e c i s e l y because they had arisen from d i f f e r e n t Early Middle Chinese i n i t i a l s . The Chuang AL i n i t i a l (LMC *(t ) ? f i - ) , according to Pulley-blank, was not phonemically d i s t i n c t from i n i t i a l Chan-'T^ (*gfi-). I t i s placed i n the muddy a f f r i c a t e column i n the Yun-jing, under the dental s i b i l a n t Cong (LMC *tsfi-) and corresponds i n Grade II to EMC *d^-but i n Grade III to EMC *?•<-. The Chan i n i t i a l , which i s placed i n the muddy f r i c a t i v e column, under Xie (LMC *sfi-) i s mainly confined to Grade III where i t corresponds to EMC -d?. Since *d? and *f had merged i n LMC as the f r i c a t i v e *gfi, the authors of the rhyme tables had only the fan-qie spellings to di s t i n g u i s h them and mistakenly assigned the o r i g i n a l a f f r i c a t e to the f r i c a t i v e column and vice versa. Before high vowels ( i . e . , i n Grade III and Grade II of the Zhi rhyme group), the pronunciation of both i n i t i a l s was mostly as a f r i c a t i v e , while before non-high vowels ( i . e . , i n Grade II i n other rhyme groups), the pronunciation was mostly as an a f f r i c a t e ; but there i s evidence of free v a r i a t i o n i n both cases (1970-71:223). A further small complication i s that there i s also a voiced r e t r o f l e x f r i c a t i v e i n i t i a l *\ i n EMC Occurring only i n two words i n the Zhi rhyme group. This would - 168 -have been pronounced *sn-.in LMC. I t was placed as Grade II of i n i t i a l Chan i n the Yun-jing. The Yu (LMC x#-) , or zero, i n i t i a l Pulleyblank regards as probably a weak velar f r i c a t i v e y ( l i k e the "zero" i n i t i a l i n Mandarin), which would perhaps produce less confusion i f omitted altogether i n transcribing reconstructed forms. Although there i s a phonemic d i s t i n c t i o n between i n i t i a l s Ying <(j and Yu '&\ during Late Middle Chinese, i t had generally disappeared between late Tang and Yuan times. Vestiges of the d i s -t i n c t i o n could nevertheless be discerned i n the phonemic contrast between Yin- and Yang-ping tones. (Pulleyblank, 1970-71:227). Since the d i a l e c t survey includes i n i t i a l s which Pulley-blank has reconstructed as part of the phonological system of Early Middle Chinese, i t i s also necessary to discuss t h i s e a r l i e r layer of the language. Pulleyblank (1977:80) posits thirty-nine i n i t i a l s , which are shown i n Chart 5 overleaf. He uses some of the same characters from the standard l i s t of the t h i r t y - s i x Late Middle Chinese i n i t i a l s where possible, and supplements these with additional characters to cover those d i s t i n c t i o n s which are not present i n the l a t e r stage. Early Middle Chinese i s reconstructed as a diasystem which accounts for both the northern and southern d i a l e c t areas. Where the values reconstructed for the two groups d i f f e r , i t i s the southern d i a l e c t a l form which i s enclosed i n parentheses i n the chart. Whereas Pulleyblank (1977:12) regards Late Middle Chinese as seemingly closer to the northern variety of Early Middle Chinese, i t i s the modern southern Chinese d i a l e c t s which he suggests have closer correspondences to cert a i n d i s t i n c -tions i n the southern branch of Early Middle Chinese. - 169 -Chart 5. Early Middle Chinese I n i t i a l s . LABIALS: DENTALS: Bang Pang Bing Ming b 0fl m Duan Tou Ding Ni -fa n Yun L a i RETROFLEXES: Zhi Che Cheng Niang fa t r 4&tr« j f-dr j ^ n r (t) (t<) (d) (n) DENTAL Jing Qing Cong SIBILANTS: $j| t s ; f f t s ' ^ d z t z ) PALATALS: Zhao Chuan Chan Ri $8 t<? j f tQ< ^ d? ® Qi (?) RETROFLEX Zhuang Chu Chuang SIBILANTS : | i t ? : ^ ts « ^ j d z. ( $ Xin /O's Xie Shen Shen Yang (or Yi) Shan S i h 5 / f ^ * VELARS; Yl Jian Qi Qun Xiao Xia GUTTURALS: Ying * #(ft) (. ) : Southern d i a l e c t a l form. What eventually became the muddy i n i t i a l s i n Late Middle Chinese were f u l l y voiced consonants i n Early Middle Chinese. As one may observe from the reconstruction of a g l o t t a l stop for the Ying ^ i n i t i a l i n both Charts 4 and 5, t h i s p a r t i -cular i n i t i a l had remained stable during the two stages of Middle Chinese. - 170 -The reconstructed value of *x- for the Late Middle Chinese i Xiao i n i t i a l projects back to *x- for the northern d i a l e c t and *h- for the southern one, while that of *xfi- for the Late Middle Chinese Xia j^. i n i t i a l goes back to * - for the northern form and * f i - for the southern. The Early Middle Chinese Yang % (Yi vk i n the d i a l e c t survey) and Yun i n i t i a l s are merged as the Late Middle Chinese Yu 9$\ (*#-, or zero) i n i t i a l . In the rhyme tables of Late Middle Chinese, what were Early Middle Chinese Yang and Yun i n i t i a l s are i n complementary d i s t r i b u t i o n i n Late Middle Chinese: Yun occurs i n Grade III and Yang i n Grade IV. y A Pulleyblank reconstructs Y i % as the true zero i n i t i a l for Early Middle Chinese, which consists of a closed class of two e n c l i t i c p a r t i c l e s , y i %, and yan -^j . They are subsumed under the Yu °H'J i n i t i a l i n Late Middle Chinese. As mentioned above, the Early Middle Chinese p a l a t a l s i b i l a n t s became merged with, t h e i r r e t r o f l e x counterparts, y i e l d -ing the Late Middle Chinese "true front-tooth' i n i t i a l s . The two Early Middle Chinese series are i n complementary d i s t r i b u t i o n i n Late Middle Chinese: the p a l a t a l series occurs i n Grade III rhymes, while the r e t r o f l e x series i s found i n Grade IV. I t i s proposed by Pulleyblank (.1970-71:219) that the complementary d i s t r i b u t i o n came about when medial was l o s t i n words which bore an Early Middle Chinese r e t r o f l e x s i b i l a n t i n i t i a l . It should be noted that the Ni j/£j and Niang i n i t i a l s i n the two stages of Middle Chinese are combined i n the survey l i s t , probably as a r e s u l t of accepting the proposal put forth by Y.R. Chao (.1940:210) that these two i n i t i a l s were merely i n - 171 -complementary d i s t r i b u t i o n , and not actually phonemically d i s t i n c t . Pulleyblank (1970-71:214-216), on the other hand, argues for a phonemic d i s t i n c t i o n between the two i n i t i a l s which, by the Mongol period (Yuan dynasty, 1279-1368) was on the verge of disappearing. The Ri l3 i n i t i a l i s reconstructed by Pulleyblank as a p a l a t a l nasal * i > - i n Early Middle Chinese which became, i n Late Middle Chinese, a lax, voiced r e t r o f l e x * r - , as i n the modern Mandarin pronunciation of that i n i t i a l . The s h i f t of the Ri i n i -t i a l from p a l a t a l to r e t r o f l e x i s part of the more general s h i f t i n the language, as witness the merger of the p a l a t a l s i b i l a n t s with those of the r e t r o f l e x series i n Late Middle Chinese. The velar i n i t i a l s remained stable from Early to Late Middle Chinese. It i s important to r e a l i z e that, on the whole, the d i a l e c t survey l i s t maximizes ce r t a i n d i s t i n c t i o n s found i n the two stages of the language, and not others. The f i n a l r e s u l t i s a t o t a l of forty i n i t i a l s i n the d i a l e c t survey. These i n i t i a l s are shown in Chart 6 on the next page. In the chart, Late Middle Chinese serves as the base. In some cases, however, i t i s necessary to make i t clear that a certain series only occurs i n one stage of the language. For the sake of c l a r i t y , the p a r t i c u l a r stage—LMC or EMC—is therefore s p e c i f i e d at the same time that the c l a s s i f i -cation of the series i s given; f o r example, Labiodentals (LMC), Retroflex S i b i l a n t s (EMC), and P a l a t a l S i b i l a n t s (EMC). The l a t t e r two sets of i n i t i a l s , EMC Retroflex ... and P a l a t a l S i b i l a n t s , subsequently merged as LMC Retroflex S i b i l a n t s . Since the diachronic study of Zhong-shan i s based on the l i s t of characters prepared i n the Fang-yan Diao-cha Zi-biao, the - 172 -Chart 6. Middle Chinese I n i t i a l s i n the Fang-yan Diao-cha Zi-biao. BILABIALS: (LMC) Bang Pang Bing Ming 1 P P' £ Pft Aft * LABIODENTALS: (LMC) DENTALS: (LMC) DENTAL i: SIBILANTS: (LMC) RETROFLEXES (LMC) RETROFLEX SIBILANTS: CEMC) PALATALS : (EMC) Fei * f ( * P ) Duan Zhi & » t r Zhuang & tg (>LMC tg-II) Zhao 8 ^ t 9 0 LMC tg: -III) Fu Tou i t * ' Q|ng :,% t s ' CHE ttl.tr' Chu ^ tg' (>LMC tg'-H) Feng £ f ft Ding * tft (<d) Wei (cm) Ni 3/6 n Cong Xin 4*c.tsfi /O' s (<dz) Cheng /Niang^ ( tdr Chuang Shan /W^ dzL it» g (>LMC (>LMC (t)gft g-II) -ID Chuan y to* (>LMC tg' -III) Shen Chan d? (>LMC gft (>EMC g -III) - III Xxe If sft 1 <5z) L a i Shen Ri ft* B i> (> LMC gft (-> LMC r ) -III) VELARS: J'ian Qi Qun Y i Xiao Xia (LMC) j£,k $ k< jpfkfl ( t o g - X \ GUTTURALS: Ying • (LMC) f i ? *B»I# (III; <:EMC w IV c EMC j»A j ). phonological categories set up i n that survey w i l l be the launching ground for the study of modern Zhong-shan reflexes to the h i s t o r i -c a l system. Thus, Chart 6 shows the Middle Chinese i n i t i a l s found i n the survey l i s t , while Zhong-shan reflexes of these i n i t i a l s are summarized i n Chart 7.on the following two pages. - 173 -.Chart 7. Zhong-shan Correspondences to the Middle Chinese I n i t i a l s i n the Fang-yan Diao-cha Zi-biao. ZHONG-SHAN INITIALS COLLOQUIAL LITERARY M.C. InitiaJ?^,,^ Ping > Shang Qu, Ru Ping Oblique Ui Bang P — *c U H Pang P' s § I . I Bing P' P P' P r~1 m Ming m C/) Fei P/P' (LMC) BI0DEN1 Fu Feng P' f < Wei i n Duan t Ui Tou t' a EH Ding f t f t Q Ni/Niang n Lai 1 Ui EH Jing ts LMC) SIBILS Qing t s ' i LMC) SIBILS Cong ts ' ts t s ' ts EH Xin s W Q Xie ts ' 1 ts t s ' ts W W X Zhi ts — H a cx, Che ts i HH1 o EH - 1 Cheng ts' ts t s ' ts - 174 -(Chart 7. cont'd) ZHONG-SHAN INITIALS ^ ^ ^ ^ ^ M.C. I n i t i a l ^ ^ COLLOQUIAL LITERARY Ping, Shang Qu, Ru Ping Oblique (EMC) RET. SIBILANTS Zhuang ts (EMC) RET. SIBILANTS Chu t s ' (EMC) RET. SIBILANTS Chuang ts' ts t s ' ts (EMC) RET. SIBILANTS Shan (Si) s (EMC) 1 PALATALS | Zhao ts (EMC) 1 PALATALS | Chuan ts ' (EMC) 1 PALATALS | Shen Shen Chan s (EMC) 1 PALATALS | Ri n g ~ j (LMC) 1 VELARS 1 Jian k (LMC) 1 VELARS 1 Qi h k' (LMC) 1 VELARS 1 Qun k' k •v k' . k (LMC) 1 VELARS 1 Y i ng LITERARY AND COLLOQUIAL; Kai-Kou He-Kou M.C. I n i t i a T ^ ^ s ^ I II j III | IV Tong, I | IlijIIlj IV Other I | i i f i l i | : i v (LMC) .VELARS Xiao . Jfc. h ~ f (LMC) .VELARS X ia h ~ w I (LMC) 8 1 GUTTURALS! Ying 0 0 j 0^ w I (LMC) 8 1 GUTTURALS! Yu (III) j I (LMC) 8 1 GUTTURALS! Yu (IV) 0~j j • - 175 -3.1.2. Zhong-shan Correspondences to Middle Chinese I n i t i a l s 3.1.2.1. B i l a b i a l s (LMC) The regular correspondent to the Bang ^ (*p-).' i n i t i a l i s /p/, e.g., 8-19 /pa / 'to take' 15-23 ft / 13 / /PU / 'to mend' 40-1 ft /pu: / 'a cup 1 77-1 / • 13 / /piw / 'to display' 186-1 /pak 5/ 'north' There are nonetheless about a dozen words which are pro-nounced with an aspirated /p 1/ i n i t i a l . Most of these exceptions can be explained i n terms of analogical readings of more common characters; that i s , readings based on characters which, appear graphically s i m i l a r . In the case of the word po (4-18) 'lame', 55 the c o l l o q u i a l reading of /paj / has preserved the regular Bang 13 correspondent while the l i t e r a r y form, /p'o /, has acquired an aspirated i n i t i a l . The word bao (.183-1) 'to s t r i p o f f , also contains the aspirated /p'/ i n i t i a l for the l i t e r a r y reading of the word, and 2 i s pronounced /p'ok /. In the c o l l o q u i a l layer, however, i t i s 5 2 pronounced /mok / or /mok /, with an /m/ i n i t i a l . One other word f*. 2 i n the data containing an /m/ i n i t i a l i s bo 'SJ (196-6) /ma:k / 'to break open with hands'. The alternative form has i n i t i a l 2 /p'/ and i s read /p'ia:k /. Zhong-shan and Cantonese do not d i f f e r much with regard to the modern reflexes of the Bang i n i t i a l . Generally, the same word i n the two d i a l e c t s would exhibit the same correspondent. - 176 -There are few differences. One such difference i s i n the word 22 bao jfo (73-16) 'leopard', which i s phonetically IpAi-w ] i n 44 Zhong-shan, and [ p ' A i w ] i n Cantonese. In thi s case, Zhong-shan has preserved the regular r e f l e x of /p/. The Pang C*p'-)' i n i t i a l i s usually pronounced /p'/ in modern Zhong-shan, as i t i s i n Cantonese. Examples are: (2) 5-22 XiiL / P ' ° 2 2 / 'broken' 17-9 Aft /p'u 2 2/ *a store' 111-16 *9 /p'an 1 3/ 'personality' 131-4 ^ / p ' i n 2 2 / 'a s l i c e ' 194-5 /p'a:k 2/ 'to clap' It should be noted that although many of the Middle Chinese i n i -t i a l s enter the same general correspondence patterns i n Zhong-shan as i n Cantonese, with regard to i n d i v i d u a l words the two d i a l e c t s may show some differences, as already observed with the Bang i n i t i a l . In t h i s case, the word pin £a i n (2) can be c i t e d . Zhong-shan has an aspirated / p ' / for the word, whereas i t i s un-35 aspirated i n Cantonese: [pen ]. In addition to the regular r e f l e x of /p'// there i s also a small handful of words which has /p/ as the correspondent to the Pang i n i t i a l i n Zhong-shan (e.g., bo iJL (3-10) 'glass' /po 5 5/. Hashimoto (p.630) c i t e s the word pou (82-18) 'to s p l i t ' as containing an [f] i n i t i a l i n modern Cantonese f o r which she could give no explanation. In Zhong-shan, the same word simply 13 contains the regular /p'/ i n i t i a l : /p'aw /. Zhong-shan reflexes of the Bing jji (*pft-< EMC *b-) i n i t i a l f a l l into the following pattern according to h i s t o r i c a l - 177 -tonal categories: Ping-sheng words have the aspirated /p'/ i n i t i a l , with no d i f f e r e n t i a t i o n of c o l l o q u i a l or l i t e r a r y pronunciation. Qu- and Ru-sheng words, i n both the c o l l o q u i a l and l i t e r a r y layers, contain the p l a i n /p/ i n i t i a l . Shang-sheng words, however, sub-^ divide according to whether they are c o l l o q u i a l (C.) or l i t e r a r y (L.) forms--the former i s aspirated while the l a t t e r i s not. ;-Examples from the various tones are presented i n (3). (3) a. Ping-sheng: 3-11 /p'o / 'an old woman' b. Qu-sheng: 35-3 | ^ / p a : j 2 2 / 'to destroy' c. Ru-sheng: 194-8 & /pa:k 2/ 'white' d. Shang-sheng / p ' i ] 22 49-5 ^ ' i 1 3 / C . 'a q u i l t ' 205-18 $u /ping /L. 'also' There are several points which may be brought up at th i s time. F i r s t of a l l , i n Zhong-shan a word with a muddy i n i t i a l , such as the Bing i n i t i a l , and containing Middle Chinese Ping-sheng has Yang-ping /51/ as the regular correspondent. In more general terms, the Yang re g i s t e r i s a r e f l e x of the muddy i n i t i a l s , while other i n i t i a l s occur with the Yin r e g i s t e r , tone /55/ i n Ping-sheng words i n Zhong-shan. Secondly, with, regard to the h i s t o r i c a l Shang-isheng reflexes, not only do the i n i t i a l s d i f f e r i n the modern form i n Zhong-shan, but the tones as well. The c o l l o q u i a l form has tone /13/, whereas the tone of the l i t e r a r y form has merged with the - 178 -regular Zhong-shan r e f l e x of the h i s t o r i c a l Qu-sheng, namely tone /22/. Lastly, the general pattern of correspondence to words with i n i t i a l Bing i n Zhong-shan and Cantonese i s e s s e n t i a l l y the same. However, whereas Cantonese has [f] as the only other cor-respondent to the Bing i n i t i a l i n a few i s o l a t e d cases, Zhong-shan has / f / as well as /h/ as i r r e g u l a r correspondences, as exemplified i n (4). (4) 7-13 ^ /faw 2 2/ 'a port' 119-13 ^ / f a : n 2 2 / 'a section 1 212-2 j|L /hung 5 1/L. 'a s a i l " (/p'ung51/C.) For the word bu ij£ , which has an alternate pronunciation of fou in Mandarin, Pulleyblank informs the writer that i t i s a c o l l o q u i a l word not found i n the Guang-yun, a rhyme dictionary which i s an enlarged version of the Qie-yun (published i n 1007 A.D. and i s s t i l l extant). Tang dynasty sources use bu ^ for what i s e v i -dently t h i s word, bu 1^ , which i s described as a d i a l e c t word of Wu ^ or Chu jjj^ A* 2 1 2 1 Peng yt- i n (4) above i s pronounced [fug ] or Ip'ug J i n Cantonese. The form with the /p 1/ i n i t i a l i n both Zhong-shan and Cantonese represents the c o l l o q u i a l reading. Cantonese [fj corresponds to Zhong-shan [h] before high back vowels (.luj and [u:]), which accounts for the laryngeal i n i t i a l i n the Zhong-shan form for peng as opposed to the labiodental i n i t i a l found i n „ ^ 3 Cantonese. The modern Zhong-shan r e f l e x of the Ming 0$ (*m-) i n i t i a l i s /m/. Examples are shown on the next page. - 179 -34-3 1 /ma:3 / 1 to buy 1 53-5 /mi / 'beautiful 129-7 ik /mm / 'to sleep' 196-8 /ma:k2/ 'wheat' 215-13 2 /muk / 'wood' While the Ming i n i t i a l i s usually pronounced with the b i l a b i a l nasal i n Zhong-shan and Cantonese, i n three cases i n the data the Ming i n i t i a l has /n/ as i t s modern r e f l e x i n the two Yue di a l e c t s . These three exceptions are l i s t e d i n (_6) , showing the Zhong-shan pronunciation only. (6) 47-15 / n i 5 1 / 'to f i l l ' 49-8 pj- / n i 2 2 / 'the ends of a bow" 72-10 Jjgf /na,:w51/ 'an anchor' With regard to mi and mi jpf above, Hashimoto (:p.631) proposes that these two words are possible vestiges of the d i s t i n c t i o n of the so-called chong-niu ^ 0- ("double knot') pairs of i n i t i a l s , or Grade III/IV doublet i n i t i a l s , since the two words belong to the Grade IV counterpart of these doublets for which some p a l a t a l feature has been suggested. In the Zhong-shan data, mo ^§ (194-11) 'a f i e l d path' i s 2 pronounced /pa:k /, with a /p/ i n i t i a l . The expected i n i t i a l i s i 3 /m/, as i n Cantonese. [mek ] i s recorded for Cantonese i n the Zhong-hua Xin Zi-dian ('Chinese new dictionary') (1977:606) , which agrees with Chen's observation. Hashimoto (p.590) records the word as [met ]. B a l l (p.524) records /mak/ for mo f § (which would be /ma:k/ as the modern, corresponding form) for Zhong-shan, con-t r a s t i n g with his recording of /mak/ ( i . e . , [mek]) for Cantonese. - 180 3.1.2.2. Labiodentals (LMC) There are two regular correspondents to the Fei ffi (*f-<EMC *p-) i n i t i a l i n Zhong-shan: /h/ before the high back vowel /u/, and / f / elsewhere, as exemplified i n (7). The /h/ r e f l e x before /u/ affects the Yu T&s , L i u %J and Tong l | f ^ rhyme groups. Recall that p a r t i c u l a r l y i n the environment before the high back vowel [u:J , /h/ i s d i s t i n c t l y a l a b i a l i z e d f r i c a t i v e £ h w ] . Names of the rhyme groups are included i n round brackets i n (7a). (7) a. Before /u/: 23-1 55 /hu"/ 'a sage' (Yu) 89-7 /hu 2 2/ 'wealth' (Liu) 218-4 /hung 5 5/ 'wind' (Tong) 220-4 * /huk 5/ 'happiness' (Tong) Elsewhere: 46-8 / f a j 2 2 / 'to abrogate' 65-10 / f i 5 5 / •to f l y ' 109-8 / f a : t 2 / 'laws' 164-11 / f a n 1 3 / 'powder' 165-7 / f a t 2 / 'not' Note, however, that the L i u rhyme group has both /u/ and /aw/ as modern Zhong-shan reflexes: /h/ occurs before the former, 2 2 as i n fu ^ /hu / above, and / f / before the l a t t e r , as i n fou £ (.87-23) 'not' /faw 1 3/. In addition to i n i t i a l s / f / and /h/ as regular correspon-dents of the Fei i n i t i a l , there are three words, l i s t e d i n (8), which are pronounced with a b i l a b i a l stop as a r e f l e x . A l l three words i n (.8) come from the Yu rhyme group. Of these, two have the - 181 -aspirated /p 1/ i n i t i a l , and one the p l a i n /p/ i n i t i a l . 18) 25-4 25-5 'dried meat to begin' 25-6 an axe Recalling that the Fei i n i t i a l had developed out of the EMC Bang (*p-) i n i t i a l , the apparent exception can actually be analyzed as forms which have preserved the e a r l i e r i n i t i a l . Hashimoto (p.36) observes that a l l the Yue d i a l e c t s have labiodental f r i c a t i v e s corresponding to the LMC labiodentals, with one exception: the . Chen-cun variety of Shun-de has [p'] as the regular r e f l e x . In the case of the Min d i a l e c t s , a b i l a b i a l stop i s i n fact a regular r e f l e x of t h i s same set of LMC i n i t i a l s . • Fu ^ , for instance, 51 has the c o l l o q u i a l reading of [pu ] and the l i t e r a r y counterpart 51 52 of [hu ] i n Xia-men. The same word i s recorded as [pou ] i n f 31 has a c o l l o q u i a l form of {p'uo ] 31 and a l i t e r a r y one of [xu J. (Zi-hui, p. 77..) I t i s reconstructed as *p- i n Proto-Min by Norman (.1969:260) . In Cantonese, [f] (and only [f]) i s the regular correspon-dent to the Fei i n i t i a l . The singular exception i n Hashimoto's 35 data i s fu , which i s pronounced [p'ow ] (p.425). The Fu (*f- < EMC *p'-) i n i t i a l i n Zhong-shan follows the same d i s t r i b u t i o n a l pattern recorded for the F e i i n i t i a l : /h/ before /u/, and / f / elsewhere. There are no exceptions i n the data. Cantonese likewise has [f] corresponding to the Fu i n i t i a l , with no exceptions observed i n Hashimoto's data. also does not deviate from the pattern occurring i n the F e i Correspondence to the Feng f- (*ffi-C EMC *b-) i n i t i a l - 182 -i n i t i a l — i n either Zhong-shan or Cantonese. Nevertheless, there are two exceptions present i n the Zhong-shan data which have an alternation between / f / ~ / h / and /p 1/ corresponding to a l i t e r a r y -c o l l o q u i a l contrast. As already mentioned, the b i l a b i a l stop can be analyzed as traces of an e a r l i e r stage i n the evolution of the i n i t i a l . C9) 85-5 /faw 5 1/L., /p'u 5 1/C. 'to f l o a t ' 87-24 ^ /hu 2 2/ 'a wife', but /p'u 1 3/ i n : jjty J$? / s a n 5 5 p ' u 1 3 / 'daughter-in-law' Cantonese also shows a l i t e r a r y - c o l l o q u i a l d i s t i n c t i o n for the words i n (9) above, r e f l e c t e d i n a I f J ^ t p ' ] contrast (Hashimoto, pp.488,632). Another exception to [f] as the regular correspondent to the Feng i n i t i a l i n Cantonese c i t e d by Hashimoto (pp .>577,632) is the l i t e r a r y word fu #| (180-5) 'to bind' I p o i k 4 ] . The word 2 i s /fok / i n Zhong-shan, with the regular labiodental r e f l e x . Note, however, that B a l l (p.520), on the other hand, records /pok/ for Zhong-shan, and /fok/ for Cantonese. Chao's data agrees with the present Zhong-shan form. The correspondent to the Wei (*v-< EMC *m-) i n i t i a l i n Zhong-shan has merged with that of the Ming i n i t i a l . The r e f l e x of both h i s t o r i c a l i n i t i a l s , i s /m/. Again, r e c a l l that the Wei i n i t i a l had i n f a c t developed out of the Ming i n i t i a l . I t i s therefore to be expected that the Min d i a l e c t s , for example, would show [mj as a regular r e f l e x of the Wei i n i t i a l rather than iv] or [wj, as i s the case i n the northern Chinese d i a l e c t s . Examples of the Zhong-shan/correspondent to the Wei i n i t i a l are shown i n (10). 183 -(10) 66-8 y£j 85~7 it 146-8 178-18 -jy 220-14 8 / -13 / /mi / /maw / /ma:t2/ /mong 5 1/ /muk2/ 'a t a i l ' 'to scheme' 'stockings' 'to forget' 1 eyes' In the Zhong-shan data on W e i - i n i t i a l words, the only exception appears to be the word wan (145-3) 'to draw back', 13 which has a /w/ i n i t i a l : /wa:n /. Hashimoto (p.632) c i t e s the same word as the only exception to [mj as the regular r e f l e x of the Wei i n i t i a l i n Cantonese. The word wan ffijj also has the 24 l a b i a l glide i n i t i a l i n Cantonese and i s pronounced {wAin ] . 3.1.2.3. Dentals (LMC) The Zhong-shan (and Cantonese) correspondent to the Duan (*t-) i n i t i a l i s / t / , with only a small handful of exceptions Examples of the regular correspondence i n Zhong-shan are shown i n (11): 1-1 \ / 4 . 55 -/to / 'many' 37-4 to /*. -55/ 'to bow the head' 42-10 ** / 4 - -22, 'opposite to' 81-1 ... 22/ /tiw / 'to f i s h (with hook 186-4 /t a k 5 / 'to obtain' Exceptions containing an aspirated dental stop are l i s t e d i n (.12) (12) 16-3 gi. / t ' u 1 3 / 'animal stomach' 37-5 ±Jv^-y<t'aj 5 1/ 'a dike' I \ 55 67-16 p# /t\'ow / 'garrulous' - 184 -69-5 /t'ow 1 3/ 'to pray' 79-8 $|i / t i w 5 5 / "to engrave' (/t'iw 5 5/C.) As i n Cantonese (Hashimoto, p.632), the word du (.16-3) 'animal stomach1 i n Zhong-shan has merged with the word du (16-7) for 'human stomach', which i s written with the same charac-ter . In the word ti. j f t jr (.37-5) , observe that i t has an aspirated i n i t i a l i n conjunction with Yang-ping tone /51/ i n Zhong-shan. Pulleyblank (personal communication) states that many di a l e c t s have readings for t h i s word which imply the muddy Ding J^ L (*tfi-) i n i t i a l rather than the Duan i n i t i a l . (See, for example, Zi-hui, p. 56.) The Zhong-shan reading of aspirated / t ' / i n dao i s based on tao P/7 (67-19) 'to desire', which has the Tou j j ^ . (*t'-) i n i t i a l . The source of aspiration for dao ^ (69-5) i s not cl e a r . Diao jQp. i n (12) above has a variant form /t'iw / which i s s t r i c t l y c o l l o q u i a l . The same exceptions noted above for the Duan i n i t i a l are also found i n Cantonese (Hashimoto, pp.632-633). The words duo (137-1) 'to gather up' and duo $ f (.137-2 2) 'to weigh a thing i n the hand' are both / t s y t / i n modern Zhong-22 shan, with an alternate reading of /ts0j /. An a f f r i c a t e i n i t i a l i s also found i n modern Cantonese; for example, i t i s recorded i n 4 the Zhong-hua Xin Zi-dian as [ t s y i t ]. Pulleyblank advises that there i s an alternate reading i n the Guang-yun that can be recon-structed as LMC *tryat (< EMC *trwiat). The regular r e f l e x of LMC *tryat, with the Zhi #tt (*tr-) i n i t i a l , i s / t s y t / i n Zhong-shan. 2 \4i Hence, the Zhong-shan form of / t s y t / for duo Jjfc i s derived from an alternate reading and i s not an exception to the Duan i n i t i a l . The /n/ correspondent to the Duan i n i t i a l i n the word - 185 -niao Jlj (80 -6 ) 'a b i r d ' i n Zhong-shan (and Cantonese) i s an excep-ti o n widespread among the Mandarin d i a l e c t s . The anomalous i n i t i a l i n the modern di a l e c t s can be explained i n terms of an avoidance of homonymy with a taboo word. The regular correspondent to the Tou 2 S C*t'-) i n i t i a l i s / t ' / i n Zhong-shan, as i n Cantonese. Zhong-shan examples are given i n (.13) . 1-2 / t ' o 5 5 / 'to drag along 37 -6 • » 55 'a ladder' 6 7 - 1 9 55 /t'ow D D/ "to deserve' 1 0 8 - 3 / t ' i p 2 / 'a card' 2 0 7 - 1 7 / t ' i a : k 2 / 'to kick' There are several exceptions containing the unaspirated dental stop corresponding to the Tou i n i t i a l . These are l i s t e d i n (14) below. (14) 9 4 - 5 gfc / t a : p 2 / 'to tread' 1 3 0 - 1 0 fjfc / t i n 1 3 / 'protruding (belly) ' 204 -22 / t i n g 5 5 / 'a (sand) bank' Hashimoto (p.633) suggests that the l i t e r a r y word ta ( .94-5), 3 [ t A i p ] i n Cantonese, i s due to the analogical reading of the word ta t|" (94-7) 'a stack ( c l a s s i f i e r for paper) ', which belongs to the Ding i n i t i a l . A similar proposal cannot be put forth for Zhong-shan since the i n i t i a l f or ta i s aspirated i n modern 2 Zhong-shan: /t'!a:p /. The Zhong-shan pronunciation of p l a i n / t / i n taffi may be due to Cantonese influence. The lack of aspira-t i o n i n both t i a n and tin g ;"J above i s probably the r e s u l t of - 186 -analogical r e a d i n g s — t i a n to dian ^ (.130-9) 'a statute' ('found i n common words such as: zi-dian 'Jr ' dictionary') , and ting tp ding J (.204-15) 'an i n d i v i d u a l ' or ding (204-16) 'a n a i l ' . The pattern of correspondence of the Ding (*tn- < EMC *d-) i n i t i a l p a r a l l e l s that of the Bing i n i t i a l i n Zhong-shan (.and Cantonese) . Middle Chinese Ping-sheng words have the / t ' / i n i t i a l i n the modern Zhong-shan d i a l e c t , with no l i t e r a r y - c o l l o -quial d i s t i n c t i o n . In contrast, the h i s t o r i c a l Qu- and Ru-sheng words are unaspirated i n the modern form. With regard to the Middle Chinese Shang-sheng words, there i s a d i s t i n c t i o n of l i t e r a r y versus c o l l o q u i a l layer: the i n i t i a l i n the l i t e r a r y '. reading i s unaspirated and the tone i s merged with the Zhong-shan ref l e x of the h i s t o r i c a l Qu-sheng; the i n i t i a l of the c o l l o q u i a l form i s aspirated and the tone i s /13/. An example of each i s given i n (15). Dan (95-18) i n (15d) has a c o l l o q u i a l versus a l i t e r a r y reading i n Zhong-shan. (The same s t y l i s t i c d i s t i n c t i o n for the word i s also found i n Cantonese.) (.15) a. Ping-sheng: 81-17 jJLlJ /t'aw 5 1/ 'head' b. Qu-sheng: 54-12 / t i 2 2 / 'the earth' c. Ru-sheng: 132-10 $L / t i t 2 / 'to f a l l down* d. Shang-sheng: 95-18 /t'a:m /C. ' i n s i p i d ' 95-18 yf^ /ta:m 2 2/L. ' i n s i p i d ' - 187 -Aside from exceptions r e s u l t i n g from the unpredicted pre-sence or absence of aspiration on the i n i t i a l (e.g., the Ru-sheng word ta (94-7) /t'a:p / 'a stack' has i n i t i a l / t ' / instead of / t / ) , there are also a couple of words which show an a f f r i c a t e as a re f l e x of the Ding i n i t i a l : (.16) 215-19 i% /tsuk 2/ 'a c a l f 215-20 /tsuk 2/ 'a ditch' The regular Zhong-shan correspondent to the Ni j/Li (*n-) i n i t i a l i n the d i a l e c t survey i s /n/. Recall that Ni and Niang "jflL. (*nr-) i n i t i a l s are combined under the Ni i n i t i a l i n the survey l i s t with no in d i c a t i o n of which word i n fact belongs to the Ni and which to the Niang i n i t i a l i n the rhyme tables which d i s t i n -guish t h i r t y - s i x i n i t i a l s . The merging of the two Middle Chinese i n i t i a l s does not present a problem i n analyzing the Zhong-shan data since the modern r e f l e x of both h i s t o r i c a l i n i t i a l s i s /n/, as shown i n (17). (.17) a. Ni i n i t i a l : 14-16 ^ /nu 5 1/ 'a slave' 2 94-8 /na:p / 'to give' 184-13 |fo /nang 5 1/ 'can, may' b. Niang i n i t i a l : 20-6 /ny 1 3/ 'a woman' 88-1 /naw 1 3/ 'a knot' 104-10 / n i p 2 / 'to tread' (also / s i p 2 / ) A l l the exceptions to the regular correspondence arise from words with the Niang i n i t i a l . In the case of the word - 188 -. 55 nian jUj (.101-9) 'sticky' /nim /, although i t does have /n/ as the regular r e f l e x of the Niang i n i t i a l , when i t enters into the combination nian-mi ^ jjp- 'glutinous r i c e ' , the term i s / t s i m 5 5 13 maj / i n Zhong-shan, with / t s / as the correspondent to the Niang i n i t i a l . (The same alternation of i n i t i a l s occurs i n Cantonese v i s - a - v i s t h i s word.) Another N i a n g - i n i t i a l word which has / t s / as the r e f l e x i n Zhong-shan i s nian ,&JL. (124-9, 125-8) '(.a stone r o l l e r for husking 13 grain)'. The a f f r i c a t e i n i t i a l for the word, which i s / t s i n / i n Zhong-shan, i s probably due to the analogical reading of the character zhan $^ (124-16) 'to open 1, which occurs i n a number of common p o l y s y l l a b i c words i n the language. The word nie (.104-10) has two variant forms i n Zhong-2 2 IL shan: /nip / and / s i p /, while nie jpi. (104-8) ' (a surname) ' has 2 only the form of / s i p /. In discussing a si m i l a r phenomenon of I s ] corresponding to the Niang i n i t i a l i n Cantonese, Hashimoto (p.637) suggests that the reading of a s i b i l a n t i n i t i a l may be based on the analogical reading of the more common word she 4 2 (.105-3) (phonetically [si:p ] i n Cantonese and I s i i p ] i n Zhong-shan), which contains the EMC Sh^n ^ (.*<?- < LMC * s - ) i n i t i a l . It may be noted that B a l l (p.525) records nie ^ as /nip/ for both Zhong-shan and Cantonese. The word nong >J[ (222-8) 'thick (of l i q u i d s ) , dark (.of 51 t i n t s ) ' i s /nung /, except i n reference to strong teas, soup and other l i q u i d s , i n which case the word i s c o l l o q u i a l l y pronounced 51 • • /jung /. (A sim i l a r alternation of nasal and p a l a t a l i n i t i a l exists i n Cantonese with regard to t h i s word.) The three words i n (.18) on the next page exhibit only a - 189 -p a l a t a l / j / corresponding to the Niang i n i t i a l . The f i r s t and t h i r d word i n 0-8) likewise have a p a l a t a l glide i n Hashimoto's data for Cantonese. >-ll $ " 2 2 (18) 112-11 g /jam / 'to rent' 170-19 jM / j o n g 2 2 / 'troubled' 174-17 ^ / j o n g 2 2 / 'to brew' 1* v ti In the case of ren (or -iin) & • Peking l i n for |fj i s also i r r e g u l a r . The alternate pronunciation of re*n for the word i n Mandarin found i n some sources (e.g., Mathews' Chinese-English Dictionary) would correspond to the p a l a t a l i n i t i a l i n the Zhong-shan and Cantonese forms for ren ^ . The / j / r e f l e x may be based 2 2 on the analogical reading of ren ft (.113-2) 'duty' /jam /. An alternate reading for niang J$j__ i n (.18) above can be found under i n i t i a l Ri 0 (EMC *i>-> LMC *r-) . Since / j / i s one of the regular reflexes of the Ri i n i t i a l i n Zhong-shan, the read-ing of niang with the p a l a t a l i n i t i a l i n Zhong-shan may be based on the R i - i n i t i a l reading. I t i s possible, however, that the p a l a t a l glide i n i s o l a t e d occurrences of N i a n g - i n i t i a l words i s a vestige of the o r i g i n a l d i s t i n c t i o n between the Ni (*n-) and Niang C*nr-) i n i t i a l s , as Pulleyblank suspects. Probably s i g n i f i -cant i s the observation that a l l exceptions i n the Zhong-shan data pertaining to the two i n i t i a l s a r i s e with respect to the Niang i n i t i a l , with the cases of the p a l a t a l glide as a r e f l e x of the Niang i n i t i a l being p a r t i c u l a r l y noteworthy. Pulleyblank (1970:214-216) provides evidence from the Tibetan hP'ags-pa s p e l l i n g to propose that *rir- became *n j - ( i . e . , * i > - ) i n front of - i , and that the d i s t i n c t i o n between * n i - and *i>i-- 190 -survived u n t i l early Yuan (.1279-1368) even i n Mandarin. Given the case of the Ri (EMC * i > - ) i n i t i a l , i n which * i > - yielded / j / as one of i t s correspondents i n Zhong-shan, one would expect / j / to be a regular r e f l e x of the pa l a t a l nasal * i > - a r i s i n g from the Niang i n i t i a l . Thus, i t may be the case that some words with i n i t i a l / j / descending from the Niang i n i t i a l are traces of the o r i g i n a l d i s t i n c t i o n between the Ni and Niang i n i t i a l . The L a i C * l - ) i n i t i a l has / l / as the regular r e f l e x i n Zhong-shan. Examples are given i n (.19) . (19) (-20) 3-17 / i t 5 1 / 1 a mule' 18-25 / l u 5 1 / 'a donkey' 69-15 /low / •old' 129-15 * / l i n 5 1 / 'to p i t y ' 189-6 il / l i k 2 / 'strength' There , are five exceptions i n the data: 3-19 p /wo55/ 'lines on the finger' 39-10 22 /tai / 'to belong to' 113-13 /nap 5/ 'a grain; ( c l a s s i f i e r ) 173-12 / n i o n g 1 3 / 'two' 201-7 4k / • 13 / n /nxa:ng /C. 'co l l a r ' (/ling 1 3/L.) The pronunciation of the word luo jffi with a /w/ i n i t i a l i n Zhong-shan may be based on the analogical reading of the common word wo (4-17) 'a nest', which i s a He-kou word, reconstructed i n LMC as *?wa, with the Ying f/p (*?-)' i n i t i a l . The writer has;no explanation to o f f e r for i n i t i a l / t / i n the word l i _ aside from observing that jJL does occur as a phonetic i n a few words - 191 -h i s t o r i c a l l y containing a dental stop i n i t i a l . Di \(a moun-ta i n tree l i k e the cherry).', for example, i s h i s t o r i c a l l y homopho-nous with d i % (38-6) 'younger brother' / t a j 2 2 / . One could, at th i s point, make a couple of suggestions, including possible ana-l o g i c a l readings. The writer w i l l , nevertheless, leave further speculations for the present. L i ||L i s also recorded by Chao as 22 /ta j /. Cantonese likewise demonstrates an xrregular dental 33 stop i n i t i a l for the word (..[tuj J ) . The l a s t three exceptions i n (20), with i n i t i a l /n/, have already been discussed i n Chapter 1.1 i n which d i a l e c t a l borrow-ings was postulated for the rare occurrence of /n/ as a r e f l e x of the Lai i n i t i a l i n Zhong-shan. 3.1.2.4. Dental S i b i l a n t s (LMC) Zhong-shan, l i k e Cantonese, has / t s / as the correspondent to the Jing ^ (*ts-) i n i t i a l . Examples are: (21) 14-23 / t s u 5 5 / 'to rent' 38-8 ffi / t s a j 5 5 / 'to crowd' 124-11 H / t s i n 1 3 / 'to cut with scissors' 175-2 / t s i o n g 2 2 / 'a general' 207-22 £1 / t s i k 5 / 'to spin' There are four words i n the data, l i s t e d i n (22), which have / t s ' / corresponding to the Jing i n i t i a l . (22) 6-5 /ts'o / 'to push down' 69-19 /ts'ow 2 2/ 'to bathe' * 22 71-3 j^ jL /ts'ow / 'easily provoked' 101-15 jfjfl! / t s ' i m 5 5 / 'to destroy' - 192 -The i r r e g u l a r correspondence of an aspirated a f f r i c a t e to the Jing i n i t i a l s i n the four words i n (.22) Calso aspirated i n Cantonese) i s explained for Cantonese by Hashimoto (p.634) i n terms of ana-l o g i c a l readings—an explanation which may equally be applied to Zhong-shan. Hashimoto suggests that cuo may be based on the analogical reading of the more common word cuo (6-6) 'a f i l e 1 , zao •)£ and zao on cao (71-6) 'to grasp', and j i a n A* on gian ^ (101-16) 'a s l x p of bamboo'. The regular Zhong-shan (.and Cantonese) correspondent to the Qing .(..* ts'-) i n i t i a l i s / t s ' / , as i n (.23) . (.23) 25-18 *~ / t s ' y 1 3 / 'to marry, to take 77-11 / t s ' i w 1 3 / 'quiet' 101-16 % / t s ' i m 5 5 / 'a s l i p of bamboo' 132-13 / t s ' i t : / 'to s l i c e ' 173-17 . • 13 /ts'long / 'to snatch' There are a number of exceptions. In the case of both, qu j£HL (19-1) 'maggots' / t s y 5 5 / (alternate pronunciation given i s 13 ¥ 22 /tsu /) and qu 'jl^.. (162-13) 'to burn' /tsan /, however, the unexpected / t s / i n i t i a l i s based on the reading with the Jing i n i t i a l . Observe nonetheless that B a l l (p. 530) records /ch' i i / for qu jfcjj. . The aspirated i n i t i a l recorded by B a l l would corres-pond to the Qing i n i t i a l . The unexpected / t ' / as a r e f l e x of the Qing i n i t i a l i n the / 55 53 word cong ||) (213-6) 'a chimney' /t'ung / ( [ f u g J i n Cantonese) may r e f l e c t the use of cong t§b for what i s actually a c o l l o q u i a l , characterless term i n Yue. Hashimoto (p.634) merely mentions the exception i n Cantonese and does not attempt to suggest a possible - 193 -source for the anomaly. The Cong 4 ^ (.* tsfi-< EMC *dz-) i n i t i a l exhibits the same pattern of correspondence as the previous muddy i n i t i a l s , Bing and Ding, as exemplified by (24): (24) a. Ping-sheng: 140-5 ^ / t s ' y n 5 1 / 'entire 1 b. Qu-sheng: 58-20 %~ / t s i 2 2 / 'a (written) character' c. Ru-sheng: 170-4 jj$ /tsok 2/ 'yesterday' d. Shang-sheng: 29-12 fa / t s ' o j 1 3 / C . 'to be at' 29-12 £ / t s o j 2 2 / L . 'to be at' The same pattern of correspondence i s found i n Cantonese. Hashimoto (p.635) points out two Ru-sheng words i n her data containing the aspirated I t s ' ] corresponding to the Cong i n i t i a l . instead of [ts] . The two words are j i . %3§- (.113-16) 'to compile' [ t s ' * p 5 ] , and zei | | j (186-11) 'a t h i e f [ t s ' A i k 3 ] . They 5 2 are also aspirated i n Zhong-shan, pronounced /tsVap / and /ts'ak / respectively. Hashimoto suggests that the f i r s t word i s based on the analogical reading of the Q i n g - i n i t i a l word, q i jfcjf. (113-14) 'to f e l l a seam', which has the aspirated i n i t i a l (in both Canton-ese and Zhong-shan). For the second word, zei , Hashimoto suggests that the aspiration of the i n i t i a l may have been borrowed from neighbouring Hakka d i a l e c t s . I t should be noted that an aspirated, a f f r i c a t e i n i t i a l i s also found i n the Min d i a l e c t s of Xia-men and Chao-zhou (Zi-hui, p.117). Fu-zhou likewise has a - 194 -/ t s 1 / i n i t i a l for the word: / t s ' e i k / (Norman, personal communica-tion) . There are two words i n the data with i n i t i a l / s / A One i s t , 2 2 cui "j-^ (64-12) 'sad' /s0j /, which likewise has the {s] i n i t i a l i n Cantonese. Hashimoto (p.635) explains the Is] as having r e s u l t -ed from the reading of cu i ^ based on cui ^ (.64-13) 'pure grain', the l a t t e r of which contains the Xin (*s-) i n i t i a l . The second word has a regular correspondence i n Cantonese but i s ir r e g u l a r i n Zhong-shan: cong -(213-8) 'a clump of trees' /sung 5 1/. (Cong jjj^ often occurs i n such combinations as cong-shu Jj^ % 'a c o l l e c t i o n of reprinted works'.) The only other word 51 i n the Zhong-shan data with the pronunciation of /sung / i s chong (219-1) ' l o f t y ' , which contains the Early Middle Chinese muddy i n i t i a l , Chuang ^rw(*dz.-> LMC * ( t ) s f i - i n Grade II).. In Zhong-shan, as i n Cantonese, the Early Middle Chinese r e t r o f l e x s i b i l a n t s have generally merged with the dental s i b i l a n t s . In both series, the r e f l e x of / s / for the muddy i n i t i a l can be regarded as an i r r e g u l a r sound change. One would suspect that an i n i t i a l / s / from these two series i s probably traceable to the same source. The other exception i n Hashimoto's data for the Cong i n i t i a l i s the l i t e r a r y word cun (also d u n ) , (157-6) 'to squat', 21 . which has a [ f ] i n i t i a l : [t'ym ]. The Zhong-hua Xin Zi-dian, however, records two Cantonese pronunciations for the character: 53 21 {t0n ] and I t s ' y i n ], neither of which contains an aspirated stop. I t i s likewise recorded with i n i t i a l s [ t j and I t s ' ] i n Wong 53 (1954); however, [ t o n J i s regarded as a c o l l o q u i a l reading. As i n i t i a l l y suggested by Pulleyblank, given the i r r e g u l a r r e f l e x - 195 -53 of Mandarin dun and Cantonese .[t0n J , i t i s possible that both 53 dun and I t 0 n ] are words which have become associated with the character c u n | | , but had i n fact d i f f e r e n t o r i g i n s . In Zhong-shan, cun | | i s pronounced / t s ' y n 1 3 / . The d i a l e c t survey l i s t records cun as a Ping-sheng word. An alternate reading occurs i n Shang-sheng, which i s what the Zhong-shan r e f l e x i s apparently based on. The Shang-sheng reading i s added i n the Ji-yun rhyme dictionary as a Q i n g - i n i t i a l word i n the Zhun rhyme. In c o l l o -53 qui a l speech, 'to squat' i s [mew ] i n Cantonese, represented by the coined character, ). The Zhong-shan equivalent i s 55 [pew ]. The regular correspondent of the Xin (*s-) i n i t i a l i s /s/ i n Zhong-shan (and Cantonese). Zhong-shan examples are: (25) 10-17 . . 55, / s i a / 'some' 23-16 / s y 5 5 / 1 necessary' 110-6 /sam / 'heart' 170-6 i /sok 2/ 'a large rope 206-7 A , . 13 / T /sing /L. 'to awaken' (ts'ia:ng 1 3/C.) About ten words i n the present data exhibit either / t s / or / t s ' / as the r e f l e x of the Xin i n i t i a l , most of which Hashimoto (p.635) has attributed to analogical reading of more common words i n Cantonese. In some Zhong-shan exceptions, Hashimoto's analysis of those found i n Cantonese may also be applied to Zhong-shan. The word seng (184-20) 'a Buddhist p r i e s t ' , for example, has an unaspirated a f f r i c a t e i n i t i a l , which Hashimoto may be correct i n a t t r i b u t i n g to the analogical reading of the surname zeng ^ 53 55 (184-15) [tseg ] (/tseng / i n Zhong-shan). - 196 -In the case of the l i t e r a r y word ci_ \z\ (.58-22) 'to wait 33 upon', Hashimoto records [ t s i i ]. In the Zhong-shan data, Chen 22 22 pronounces the word as / t s ' i /, and Yang as / s i /. From the h i s t o r i c a l point of view, Yang's pronunciation represents the regular correspondence to the Xin i n i t i a l . Hashimoto suggests the analogical reading of c_i i n Cantonese to such words as s i (58-25) 'to connect' and s i ^ s ] (58-26) 'to feed', which are pronounced [ t s i i 2 2 ] i n Cantonese. In Zhong-shan, si_ r^J i s 22 h 22 / t s i / and s_i/p) i s / t s ' i /. Chen's pronunciation of c i with an aspirated i n i t i a l may be based on the analogical reading of the word si/^JJ . There are also a few cases i n which Cantonese has the regular r e f l e x and only Zhong-shan shows an i r r e g u l a r form, as i n the word sui ffijg^ (61-13) 'marrow'. The word i s pronounced 13 /ts'0j / i n Zhong-shan, with i n i t i a l / t s ' / . Norman (personal communication, hereafter "p.c." for short) points out that the 3 i n i t i a l i s * t s ' ^ i n Min, with [ts'e ] i n Amoy, for example, and suspects that the Zhong-shan / t s ' / i n i t i a l for the word i s a sur-v i v a l of an older stratum. The Ru-sheng word xie ^ (126-12) 'dirty, ragged' i s 22 / t s i / i n Zhong-shan. The highly i r r e g u l a r pronunciation of t h i s word may have been read i n error except that double-checking s t i l l 4 resulted i n the same pronunciation. (It i s [si i t ] i n Cantonese.) Pulleyblank suggests that the Zhong-shan form could be a Qu-sheng derivative. Norman supplied the Amoy word I i a - t s i s J ' d i r t y ' , \i which appears to be a related form. In Zhong-shan, the word xing fyjL (.206-7) 'to awaken i s 13 13 /sing / i n the l i t e r a r y reading, but i s c o l l o q u i a l l y /ts'ia:ng /, - 197 -An aspirated a f f r i c a t e i n i t i a l i s found i n various Min d i a l e c t s ; ~52 , 3 for example, Chao-zhou has [ts'e ], and Amoy [ t s ' i J (Zi-hui, p.259). Again, i t i s possible that the c o l l o q u i a l form i n Zhong-shan i s a survival of an e a r l i e r form. The regular reflexes of the Xie ^ (*sfi-< EMC *z-) i n i t i a l i n Zhong-shan i s the same as those found i n the Cong i n i t i a l (see (.24)): /ts''/ for Ping-sheng and the c o l l o q u i a l layer of Shang-sheng; and / t s / for Qu- and Ru-sheng, and for the l i t e r a r y layer of Shang-sheng. The same pattern occurs i n Cantonese. About a quarter of the X i e - i n i t i a l words, however, have / s / as the r e f l e x i n the Zhong-shan data (Cantonese has s l i g h t l y fewer exceptions of /s/ than Zhong-shan) . The word su (225-2 16) 'common' i s /suk / c o l l o q u i a l l y i n Zhong-shan. The variant 2 form /tsuk / i s borrowed from Cantonese and considered a l i t e r a r y pronunciation. Xun ;d) (.160-14) 'a period of ten days' i s / s 0 n 5 1 / as opposed to an aspirated a f f r i c a t e i n i t i a l i n Cantonese: 21 I ts'on J. Interestingly, of the southern d i a l e c t s represented i n the Zi-hui (p.84,220), only Cantonese has an a f f r i c a t e d i n i t i a l for the words su /fe and xun *$) , as opposed to Mei-xian, Xia-men, Chao-zhou and Fu-zhou, for a l l of which the i n i t i a l i s I s J . Hui (or s u i , old reading) (.45-9, 6 4-17) 'a comet' has 22 i n i t i a l /w/ i n Zhong-shan C/waj / ) , whxch r e f l e c t s the reading with the Yu **%\ i n i t i a l , h i s t o r i c a l l y and synchronically homophon-ous with wei (.46-6) 'to protect' /waj 2 2/. 3.1.2.5. Retroflexes (LMC) With the exception of the r e t r o f l e x nasal, the Late Middle Chinese retroflexes have a l l merged with t h e i r dental s i b i l a n t - 198 -counterparts discussed i n the immediately preceding section. Thus, / t s / i s the regular correspondent of the Zhi £ t t (*tr-) i n i t i a l , and / t s ' / that of the Che (*tr'-) i n i t i a l . The Cheng (*trfi-< EMC *dr-) i n i t i a l undergoes the same d i s t r i b u t i o n of sound correspondence as Cong , exemplified i n (.24). The r e t r o f l e x Niang >j[fL, (*nr-) i n i t i a l has already been discussed i n conjunction with the Ni $J i n i t i a l . A few exceptions to the h i s t o r i c a l r e t r o f l e x series w i l l be discussed. The correspondence of dental stops i n Zhong-shan to t h i s series i s of i n t e r e s t since they may be vestiges of the southern branch of the series which had dentals instead of retro-flexes (see Chart 5). Dental correspondences found i n the present data are l i s t e d i n (26) below. Normally, one would not expect to f i n d s y l l a b l e s /ty/, /tiong/ and / t i o k / i n Zhong-shan i f a l l cor-respondents of the r e t r o f l e x series were a f f r i c a t e s i n the d i a l e c t . 20-14 0 / t y 1 3 / 'to store up 1 20-16 f / t y 1 3 / '(a hemp-growing plant) 154-8 / t i t 2 / 'orderly' 175-8 / t i o n g 2 2 / ' j o y f u l ' 183-9 / t i o k 2 / 'to cut/polish stone' 183-10 2 /tiok / •to peck' (/tiong5 5/C A / t / i n i t i a l for zhu i s also recorded by Chao for Zhong-shan although he includes an alternate form with an a f f r i c a t e i n i t i a l : 13 /ts'y /, which i s not found i n the present data. H i s t o r i c a l l y , the fan-qie for zhu ffi i s ding-lu 3" 5 / t i n g 5 5 l y 1 3 / , implying a dental stop i n i t i a l . I t i s highly probable that Zhong-shan has preserved the - 199 -dental i n i t i a l despite Chao's suggestion that the / t / i n i t i a l i n zhu jffi may be due to d i a l e c t a l influence from Fu-jian ( i . e . , Min). His proposal i s based on the fact that i t i s i n the Min d i a l e c t s that dentals are the regular correspondents of the LMC retroflexes. Chao records only two other words with a / t / i n i t i a l from t h i s h i s t o r i c a l s e r i e s : zhuo (183-9) (see above), and deng (or cheng) (188-6) i n the combination yan-deng-deng SL tit /nga:n 1 3 t a n g 2 2 t a n g 2 2 / 'to stare f i x e d l y , to glare'. ( A l l the di a l e c t s recorded i n the Zi-hui (p.244) i n fact show a dental stop for deng $ft~) . The paucity of data, combined with an alternate a f f r i c a t e i n i t i a l for zhu Jffy , makes Chao's suggestion of d i a l e c t a l borrowing quite plausible. It i s the writer's contention, however, that the dental i n i t i a l s exhibited i n the Zhong-shan words represent a preserva-t i o n of an h i s t o r i c a l dental i n i t i a l rather than the r e s u l t of d i a l e c t a l borrowing. F i r s t of a l l , one would not expect influence of Min i n Shi-qi speech i n the pronunciation of words such as chang $p , for which Chen gave not only the meaning, but various combina-tions i n which i t occurs. Secondly, a / t / i n i t i a l for zhuo and zhuo occurs i n B a l l ' s data for both (Macao) Zhong-shan and Cantonese. A / t / i n i t i a l for these two words continues to be used in modern Cantonese and Zhong-shan. I t would appear that the preservation of a dental for these two words at le a s t exceeds the boundary of influence that Min may claim to exert on Shi-qi speech. F i n a l l y , the s y l l a b l e / t i o k / also occurs i n the characterless, 5 2 c o l l o q u i a l words /tiok / 'to chop (.e.g., i n h a l f ) ' , and /tio k / 'to mince (by chopping)' i n Zhong-shan. Likewise, the s y l l a b l e 13 /t y / occurs i n at le a s t one c o l l o q u i a l Zhong-shan word: /ty / - 200 -'to poke with a long object (e.g., as, accidentally, with, a closed umbrella)'. Such data suggest; that these s y l l a b l e s are probably indigenous to the d i a l e c t . Further investigation i s l i k e l y to y i e l d more evidence to support t h i s position. For the present, as a general observation, there are a number of Zhong-shan words which show v e s t i g i a l traces of an e a r l i e r layer of the language that do not necessarily point to d i a l e c t a l influence. I n i t i a l / t / i n some l e x i c a l items from the LMC r e t r o f l e x series may be part of that set of genuine survivals of an e a r l i e r stratum. A number of words have the i n i t i a l / s / as a r e f l e x of the r e t r o f l e x s e r i e s . Some of the exceptions have already been iden-t i f i e d by Hashimoto (p.636-637) as being due to analogical read-ings i n Cantonese. The same claim could be made i n Zhong-shan i n those cases where the two d i a l e c t s overlap i n exhibiting the same ir r e g u l a r i n i t i a l . In some cases,the exception of an /s/ i n i t i a l only occurs i n the Zhong-shan data, as i n the words chu ^ (.20-15) '(a surname)' / s y 1 3 / , and chi (.189-14) 'to order, to i n s t r u c t ' / s i k 5 / . Both words are derived from the Che i n i t i a l . (Hashimoto 35 gives no pronunciation for these two words; they are [ts'y: j and 5 [ts'ik J respectively i n the Zhong-hua Xin Zi-dian.) Of the small number of Cheng-initial words with the /s/ i n i t i a l i n Zhong-shan, the l i t e r a r y word chu /jjj^ (19-7) "to c o l l e c t ' / s y 5 1 / has the regu-24 l a r a f f r i c a t e i n i t i a l i n Cantonese: [ts'y: ]. C o l l o q u i a l l y , 'to 13 24 c o l l e c t ' i s /ts'u / i n Zhong-shan and {ts'ow J i n Cantonese. C r o s s - d i a l e c t a l l y , an [sj i n i t i a l for the word chu 4 * $ i s found i n Mei-xian. In that d i a l e c t , i t i s pronounced [ s u 1 1 ] (Zi-hui, p.87). - 201 -3.1.2.6. Retroflex S i b i l a n t s (EMC) and Palatals (EMC) As mentioned e a r l i e r , the s i b i l a n t series of retroflexes and palatals of Early Middle Chinese were no longer phonologically d i s t i n c t by Late Middle Chinese. By then, they had merged to become the Late Middle Chinese r e t r o f l e x s i b i l a n t series (see Charts 4, 5 and 6). The o r i g i n a l two series nevertheless remained i n complementary d i s t r i b u t i o n : EMC r e t r o f l e x s i b i l a n t series occur i n Grade II of the LMC r e t r o f l e x s i b i l a n t series, and the EMC p a l a t a l series i n Grade III of the LMC r e t r o f l e x s i b i l a n t series, as shown i n Chart 6. By Southern Song (1127-1279), the r e t r o f l e x series had also merged with the r e t r o f l e x s i b i l a n t s (Pulleyblank, 1970-71:214). Synchronically, Zhong-shan has merged the LMC r e t r o f l e x s i b i l a n t series with the LMC dental s i b i l a n t series, just as i t has that of the LMC r e t r o f l e x series i n the previous section. Thus, i n Zhong-shan the regular correspondent of the Zhuang $L (EMC *ts-) and Zhao fif. (EMC *t<?-) i n i t i a l s i s / t s / , that of Chu %7] (EMC *ts •'•-)• and Chuan jsf (EMC *t<?'-) i n i t i a l s i s / t s ' /, and that of Shan fa (EMC *g-) and Shen -ff (EMC *?-) i s /s/. The muddy Chuang (EMC *dz.-) i n i t i a l behaves s i m i l a r l y to the Cong i n i t i a l , with / t s / or / t s ' / depending on the h i s t o r i c a l tones and the st y l e of speech, l i t e r a r y or c o l l o q u i a l . The same pattern of correspondence i s found i n Cantonese. There are, of course, the other i n i t i a l s i n these two series, which w i l l be discussed l a t e r . We w i l l f i r s t deal with, some exceptions pertaining to the i n i t i a l s just mentioned. Two exceptions are found i n the data for words descending from the Zhuang i n i t i a l . Although one would expect i n i t i a l - 202 -/ t s / , ce (or -zej/fflij (190-1) 'the side' /ts'ak / has an aspirated a f f r i c a t e i n i t i a l i n Zhong-shan. An aspirated i n t i a l i s also found i n Xia-men, Fu-zhou and a number of northern Chinese d i a l e c t s including Peking Mandarin (Zi-hui, p.14). The other word i s zheng (195-7) 'to open the eyes', which has the regular correspondent of the Zhuang i n i t i a l i n the l i t e r a r y layer 55 22 (tsang / ) , but / t s ' / i n the c o l l o q u i a l layer (/ts'a:ng /).. No aspiration i s recorded for the word i n any of the di a l e c t s repre-sented i n the Zi-hui (p.245). There are six words i n the data which do not exhibit :. i n i t i a l / t s / as the regular correspondent of the Zhao i n i t i a l . They are l i s t e d i n (27) below. A l l six words are pronounced with an aspirated a f f r i c a t e i n i t i a l i n Hashimoto's data, with the excep-ti o n of zhun jtfj for which nothing was given. (27) 76-1 0a / t s ' i w 5 5 / 'bright* 151-14 i f / t s ' a n 1 3 / 'to examine' 151-15 ^ / t s ' a n 1 3 / 'a rash' 160-18 fifa / s0n 5 1 / 'gizzard' 160-19 -fj. /sjrfn 5 1/ 'to r e i t e r a t e ' 188-5 / t s ' i n g 1 3 / 'to save' The pronunciation of zhao 0e» with the aspirated / t s ' / may be based on the analogical reading of the word chao j££, (75-24) 'to leap -over', which often occurs i n a combination such as chao-guo 55 22 'to exceed; to surpass' /ts'iw ko /. Hashimoto (p.639) suggests that the aspiration i n the words zhen -f^ and zhen 0- i n Cantonese may be due to the analogical reading of the common word chen fjj^ (152-12) 'to take advantage o f , which has the [ts'.] i n i t i a l . - 203 -i<4~ 22 (In Zhong-shan, chen likewise has i n i t i a l / t s ' / : /ts'an /.) The i n i t i a l / s / i n the words zh.un ffllj and zhun-f %^ i n Zhong-shan may also be based on analogical readings: the former to shun ^}Li (.160-22) 'pure' and the l a t t e r to shun (160-24) ' r i c h , good (as wine)'. The source of / t s ' / i n Zhong-shan and Cantonese for the word zheng \\hr i s not known to the writer, and probably not to Hashimoto either since she offers no explanation for the i r r e -gular sound change;:;in that word. In Zhong-shan, there i s also the word zhi ,$JL. (.48-13) 'a 55 branch', which i s usually pronounced / t s i /, except i n the com-bination l i - z h i jfyj ' lichee (a f r u i t grown i n southern China) ' 22 55 / l a : j k i /. In the l a t t e r case, Zhong-shan has a /k/ i n i t i a l for zhi $ L f as opposed to an a f f r i c a t e d one for the word i n the various d i a l e c t s represented i n the Ci-hui (p.90). Note, however, that i n C. Douglas' Amoy dictionary (.1899:101), f o r example, the word zhi , as a monosyllabic word, i s transcribed as [ki] i n Xia-men (.or Amoy), with, an unaspirated. velar stop i n i t i a l . Chao-zhou likewise has [ki] as the pronunciation for zhi when i t occurs i n i s o l a t i o n (Zi-hui, p.44). As regards the graphic variant, i 55 zhi (.48-12) 'a branch', although i t i s pronounced / t s i / i n Zhong-shan, i n the Southern Min dia l e c t s of Xia-men and Chao-zhou, j, 55 33 for instance, zhi i s pronounced [ki ] and [ki ] respectively (Ci-hui, p.424). It i s highly probable that the word zhi ^ L^ once bore a *k- i n i t i a l i n Zhong-shan. Today, i t i s only i n the p o l y s y l l a b i c word 'lichee' that *k- has been preserved i n the d i a l e c t . In other contexts, the velar i n i t i a l has been l o s t . It i s also s i g n i f i c a n t that z h i 3L. occurs as a phonetic i n such words as q i jkjL A ' i 5 1 / - 204 -(as i n Shi-qi fci^ ) , j i _ ^ (30-3) ' s k i l l ' / k i 2 2 / , and j i $ L 22 C50-4) 'a singing g i r l , a pr o s t i t u t e ' / k i /, which have maintained the velar pronunciation of the Middle Chinese velar i n i t i a l Qun (*kfi-< EMC *g-) . The h i s t o r i c a l velar series, one might note, has since undergone p a l a t a l i z a t i o n i n Mandarin. To claim that zhi (or ) had probably undergone a sim i l a r process i n Zhong-shan, with thus f a r the singular exception of the word i n the frozen form of l i - z h i ^ , i s not far-fetched, especially i n view of the reconstruction of th i s word by Tong-he Dong, for exam-ple, as Archaic Chinese *k'xieg (Chou, 1973:125,141), and by Fang-kuei L i as Archaic Chinese * k r j i g (class notes, 1977). A velar i n i t i a l for zhi faj^ (and i t s graphic variant ) has also been postulated by Pulleyblank for Old Chinese based on foreign trans-c r i p t i o n s i n early texts (.1962:10 5-106). I t i s reconstructed by Pulleyblank as *t$ i a i n Early Middle Chinese, although he suggests that an e a r l i e r form underlying EMC *ti?ia i s * k j i a . He further proposes the same form for proto-Min. I t would appear that the same could also be postulated at least for proto-Zhong-shan, and perhaps proto-Yue as well. With regard to the Chu i n i t i a l , the unaspirated / t s / i n i t i a l i n the word chou 4 $ (86-18) 'to pluck stringed instrument 22 with fingers' /tsaw /, instead of / t s ' / , which i s the normal Zhong-shan correspondent to the Chu i n i t i a l , i s not an i r r e g u l a r development. Chou j$ has a Qu-sheng derivative which contains 22 the Zhuang i n i t i a l , for which Zhong-shan /tsaw / would be a regular correspondent The Chuan % as i t s regular r e f l e x . A l l three are probably based on the i n i t i a l has only three exceptions to / t s ' / - 205 -analogical reading of more common words: shu (24-4) 'a pi v o t 1 /k'y 5 5/ on qu j j (24-11) 'a d i s t r i c t ' , che p (127-3) 'to hinder' / t s a j 2 2 / on zhi (36-8) 'to regulate 1 and zhi ^ (36-9) 'to make', and chong ^ J Q (220-2) 'a blunderbuss' / t ' u n g 1 3 / on tong 'to govern', which has the Tou i n i t i a l . The Shan tin i n i t i a l has /s/ as i t s regular correspondent i n Zhong-shan. There are only four exceptions i n the data. Xia ( l i t e r a r y reading sha) (9-25) 'a great house' / h a 2 2 / i s not distinguished from x i a (9-6) (as i n Xia-men /|_ f^J 'Amoy'. / h a 2 2 mun 5 1/), which i s a descendent of the Xia t£ (*xfi-< EMC *x-) i n i t i a l . The source of i n i t i a l / t s ' / for the word shan (97-13) 22 'name of various species of f i r and pine' /ts'a:m /, which posses-es the h i s t o r i c a l Shan i n i t i a l , i s not known to the writer. The same i r r e g u l a r i n i t i a l also appears i n Cantonese (Hashimoto, p. 638), for which Hashimoto provides no explanation. I t should be noted, however, that the same anomaly occurs i n Hakka. Mantaro Hashimoto (1973:222) records shan jfy as [ts'ami] . (or [ts'am 5 5], conforming with the present use of tone numerals and the trans c r i p -tion used i n the Zi-hui) for Hakka. It constitutes one of the few exceptions to [s] as the regular correspondent of the Shan i n i t i a l i n that d i a l e c t . Whatever the source of the sound change, i t had obviously affected both the Hakka and Yue d i a l e c t s . (This i s contrasted with another southern Chinese d i a l e c t , Amoy, for example, which has the regular / s / i n i t i a l for the word (Douglas, 1899:190) .) 51 >/-The reading of /ts'yn / for the l i t e r a r y word, shuan 4i-'to t i e up', which i s found under both the Shan i n i t i a l (138-6) - 2 0 6 -and the Qing i n i t i a l ( 1 4 0 - 4 ) , may be based on the analogical read-ily 52_ ing of the common word, guan "jZ ( 1 4 0 - 5 ) 'entire' /ts'yn /, which possesses the muddy Cong i n i t i a l . That would account for shuan J £ occurring with a Yang-ping / 5 1 / tone instead of the Yin-ping / 5 5 / tone. (Shuan $Z ( 1 3 8 - 6 and 1 4 0 - 4 ) i s inuPing-sheng i n both entries. Because i n neither case i s i t a muddy-initial word, i t should have had the Yin-ping tone as the regular tonal correspon-dent. ) The word shua $\ ( 1 4 0 - 1 ) 'to brush' / t s ' a : t 2 / i s pronounced with the a f f r i c a t e / t s ' / i n i t i a l i n both Zhong-shan and Cantonese (for Cantonese, see e.g., Hashimoto, p . 5 3 8 ; Z i - h u i , p . 1 1 ; Huang, 1 9 7 0 : 3 6 2 ) . I t i s possible that the pronunciation of shua i n these two d i a l e c t s has fused with the semantically overlapping word ca ( . 1 1 8 - 1 ) 'to scour, to rub i n ' ; for example, 'to brush the teeth' may be given as shua-ya $>\ % or c a - y a ^ i n Chinese (Mathews, 1 9 7 1 : 8 3 6 , 9 7 3 ) . (Shua-ya i s the more common term i n c o l -4 l o q u i a l Mandarin.) The fact that Cantonese I t s ' M t J i s recorded i n the Zi-hui under the pronunciation of shua ffi>\ , whereas ca ^ i s the character recorded as the Cantonese word corresponding to Mandarin shua i n the Ci-hui (p. 2 9 3 ) may be demonstrative of at least some confusion between these two words i n Cantonese, and the same apparently holds for Zhong-shan. What may have happened i s that, i n many instances, Zhong-shan and Cantonese use ca as the term corresponding to Mandarin shuaffi) . Eventually, the word shua ffi] i t s e l f became i d e n t i f i e d i n Cantonese and Zhong-shan (the l a t t e r probably following i n wake of the former's influence) 4 2 as being pronounced [ t s ' A i t ] and I t s ' A i t ] respectively. In contrast to the Cantonese and Zhong-shan s i t u a t i o n , Xia-men, - 207 -although i t also has ca corresponding to the term shua i n Mandarin, nonetheless has retained the pronunciation of shua as [ s u a t 3 2 ] d i s t i n c t from that of ca ||- , which i s [ t s ' a t 3 2 ] (Cj.-hui,p.293) . Zhong-shan has /s/ as the regular descendent of the Shen ^ i n i t i a l , with only two exceptions. The word shi (.49-16) 13 'to l i c k ' has two forms: a l i t e r a r y form of / l a j / and a c o l l o -13 q u i a l one of / l a : j /. An / l / i n i t i a l i s also found i n an a l t e r -35 nate pronunciation for the word i n Cantonese: [ I A I J ]. The l i t e r a r y versus c o l l o q u i a l form given by Chen for the word remains a puzzle. What i s int e r e s t i n g at t h i s point i s that, quite possi-bly, the word for 'to l i c k ' i n Zhong-shan and Cantonese with the / l / i n i t i a l i s etymologically related to the c o l l o q u i a l terra for 'tongue', for which Cantonese has coined the character j^ f'J , 33 22 pronounced [lej ] i n Cantonese and [ l i : ] i n Zhong-shan. This 33 term for 'tongue' i s also found i n Tai-shan ( [ l e j ]) and Yang-454 jiang (..[lej ] ) , and perhaps other southern Chinese d i a l e c t s as 13 It has been suggested by Pulleyblank (p.c.) that / l a : j / well (Wong, 1970:262; Ci-hui, p.194) It has been suggested by Pul for shi i n Zhong-shan i s i n fact a genuine s u r v i v a l of an ear-l i e r form for the word; for example, i t i s Old Chinese *ldj?> *?ia-2, r e c a l l i n g that the Shen ^ i n i t i a l i s EMC *?. The same can also be said of she (127-4) 'tongue' / s i t /, which l i k e -wise has the Shen i n i t i a l . I t may very well be the case that the /!/ i n i t i a l i n the various Yue d i a l e c t s r e f l e c t s a much e a r l i e r form of she containing *1-. Pulleyblank pointed out to the writer that the two words have correspondences i n Sino-Tibetan and Tibeto-Burman according to Benedict. Pulleyblank also notes that - 208 -Vietnames has a word f o r 'tongue' which has i n i t i a l /!/ and appears to be an old Sino-Vietnameseloan: lufcri.,. Tonally, i t would corres-35 pond to Cantonese I I A I J J . There are some, on the other hand, who claim that the use 33 f of [lej ] i n Cantonese i n place of she •£ i s due to the avoidance of homophony with the word t s i i t ] 'to lose money' (Norman, Pulleyblank, p . c ) . Such avoidances of taboo s y l l a b l e s are common. For example, instead of saying du-shu ~%*%,-@ 'to read a book—to 2 55 55 study' /tuk sy /, one often finds the s y l l a b l e /sy / replaced' 22 55 by sheng Wfr- /sing / 'to excel', since /sy / i n Zhong-shan (and Cantonese) i s homophonous with the inauspicious word, shi ^ 'to lose'. Given the h i s t o r i c a l arguments, i t may be the case that t h i s l a t t e r theory .of the o r i g i n of an /!/ i n i t i a l word for 'tongue' i s simply folk etymology. Another exception to /s/ as a r e f l e x of the Shen i n i t i a l i n / 13 Zhong-shan i s dun ffa (161-9) 'a shield' /t'0n /, which represents the Middle Chinese reading with the Ding i n i t i a l (number 158-3) . 13 An unaspirated / t / i s recorded for the word i n Chao (./t0n /) . Aside from the i n i t i a l / t s / and / t s ' / entering into the regular patterning of the muddy Chuang i n i t i a l , there are eight words in Zhong-shan for which the r e f l e x i s that of /s/: (28) 57-16 •£ / s i 2 2 / 'a scholar' 'to f i l l an o f f i c e ' 'to wait for' 'an a f f a i r , a matter' •sad' 'a mountain peak' . .22, i / 57-17 , .22, / s i / 57-19 / -22/ / s i / 59-5 $ , .22, / s i / 86-19 , 51 , /saw / 110-12 #- /sam / - 209 -118-9 / s a : n 5 1 / 'sound of water' 219-1 & /sung 5 1/ ' l o f t y ' Seven of the words i n (.28) likewise have / s / as the i r r e g u l a r correspondent of the Chuang i n i t i a l i n Cantonese according to Hashimoto (p.638). Only s i yj|^ i n Hashimoto's data has [ts] as the i n i t i a l i n Cantonese. S i i s i n fact one of the rare words the i n i t i a l of which i s descended from the Early Middle Chinese Si i n i t i a l which, by Late Middle Chinese, had merged with- the r e t r o f l e x s i b i l a n t i n i t i a l , Chuang (LMC *(t)sfi-) , and occurs i n Grade II of that i n i t i a l . That Zhong-shan has /s/ for si. / j ^ may represent the retention of i t s d i s t i n c t i o n from the Chuang i n i t i a l . As observed by Hashimoto, shi - i , shi and shi ^ are pronounced with a f r i c a t i v e rather than an a f f r i c a t e i n most of the modern Chinese d i a l e c t s . . (See also Zi- hu i , p.51. for shi and shi ^ .) Although Mandarin data would suggest that the Chuang i n i -t i a l has always been an a f f r i c a t e , Pulleyblank (1970:223) notes that ancient evidences were not consistent i n showing an a f f r i c a t e i n i t i a l . Vietnamese, for instance, has an a f f r i c a t e exceptionally; i t most frequently has /s/, thus implying a f r i c a t i v e rather than an a f f r i c a t e for the Chuang i n i t i a l . For example, shi * i s si,. shi ^ su', chou jjjy sou , cen S O B , chan son, and chong sung. Even though Sino-Korean usually has c^ - or cj_-, thus implying an a f f r i c a t e , among the words with i n i t i a l s_- i s shi ^ s_ Tibetan likewise shows both f r i c a t i v e s and a f f r i c a t e s . Among the f r i c a t i v e s are shi si_ and shi ^ s_e, s i . To account for what appears to be an i r r e g u l a r correspon-dent to the Chuang i n i t i a l , Pulleyblank (.1970:223), for example, - 210 -suggests that the Chuang and Chan (LMC . *sn-).' i n i t i a l s forxaed one i n i t i a l phonemically and were pronounced as a f r i c a t i v e before / i / or / i u / , with a tendency toward a f f r i c a t i o n elsewhere. (Re-c a l l that Pulleyblank regards them as nondistinct phonemically i n LMC.) He proposes that i t was l a t e r , when the r e t r o f l e x stop i n i t i a l , Cheng (.EMC *dr-> LMC * t r f i - ) , became a s s i b i l a t e d becoming * t r f i - , that a new phonemic d i s t i n c t i o n emerged between Chuang (*tsn-) and Chan (*sfi-) before / i / . I t was then that the Chuang i n i t i a l i n Grade II became s t a b i l i z e d as an a f f r i c a t e . The s i b i l a n t i n i t i a l i n a number of Chuang-initial words i n Zhong-shan and other d i a l e c t s , as well as cognates i n Vietnamese, etc., can therefore be explained as being based on a f r i c a t i v e pronunci-ation before Chuang became s t a b i l i z e d as an a f f r i c a t e i n i t i a l . Conversely, as can be expected, although /s/ i s the regular r e f l e x i n Zhong-shan for the Chan i n i t i a l , s l i g h t l y over 10% of the data has / t s / and / t s ' / as correspondents. A s i m i l a r set of correspondences i s found i n Cantonese (Hashimoto, p. 640). . Usually, whatever word has an a f f r i c a t e i n Cantonese i s likewise a f f r i c a t e d i n Zhong-shan corresponding to the Chan i n i t i a l . For example, both d i a l e c t s have a / t s ' / i n i t i a l for chou jSj*^ (86-21 30) 'to pledge with wine' (Cantonese [ts'ew ], Zhong-shan /ts'aw 5 1/; she %\ (127-6) 'broken' has / t s / as i n i t i a l i n both Cantonese ( I t s i i t ^ ] ) . and Zhong-shan ( / t s i t 2 / ) . I t i s possible that she $f i n the two d i a l e c t s i s based on the reading with the Zhao $L i n i t i a l ( i . e . , zhe i/f (127-1) 'to break'). Interestingly, she $J , which also has the meaning of 'to lose money', as i n the bimorphemic word she^-ben ^ ^ ' to lose money i n business' ('.to - 211 -2 3 lose + c a p i t a l money') i s / s i t / i n Zhong-shan (Jsi:-t ] i n Can-tonese) , for which Cantonese has coined the characters (Hashimoto, p.307) and (Huang, 1970:440). I t appears quite l i k e l y that the s i b i l a n t i n i t i a l i n Zhong-shan and Cantonese [ s i i t j i s i n fact a preservation of the o r i g i n a l descendent of the Chan i n i t i a l . Moreover, i n the case of Cantonese, the pronunciation of 3 4 tone [3] i n Isixt ], as opposed to [4] i n [ t s i i t J, shows more c l e a r l y than Zhong-shan the preservation of the o r i g i n a l Yang-ru tone for she descending from the muddy Chan i n i t i a l . Con-. t r a s t t h i s with the descendent of the Zhao i n i t i a l zhe ty , which, has a voiceless i n i t i a l and a tense nuclear vowel i n modern Cantonese, thereby conditioning the re f l e x of the Middle Yin-ru tone I4J . 4 In the case of zhi (.190-16) 'trees' / t s i k 2 / , i t has a reading with the Cheng i n i t i a l upon which the Zhong-shan reading may have been based. Shao (.176-15) / t s ' iok /, as i n shao-yao *h Ik )^ 'peony', on the other hand, i s probably a C h i n g - i n i t i a l derivative. There are two additional words i n Zhong-shan which have an a f f r i c a t e i n i t i a l corresponding to Is] i n Hashimoto's data for the Chan i n i t i a l : (29) 86-29 yfft» /ts'aw 5 1/ 'an enemy' 226-6 |jj /tsuk 5/ 'Szechwan' Chou /\fu i s a f f r i c a t e d i n the majority of the Chinese d i a l e c t s , including Cantonese i n the Zi-hui (p.153) The pronunciation of /tsuk 5/ for shu % i n Zhong-shan i s probably due to the analogical reading of a word such as zhu >ji% (226-1) 'a candle', which occurs - 212 -i n such common terms as la-zhu #jfy 'wax candle' / l a : p 2 tsuk 5/. 5 The fact that /tsuk / has the Yin-ru tone /5/ instead of Yang-ru /2/, which would be the expected tonal r e f l e x of a muddy i n i t i a l 5 word, i s additional evidence that the pronunciation of /tsuk / for zhu i s probably based on analogy. The Shen j^j i n i t i a l has / s / as the regular correspondent i n both Zhong-shan and Cantonese. In general, the same words with an a f f r i c a t e i n i t i a l as a correspondent of the Shen i n i t i a l are found i n the two d i a l e c t s . Thus, the word chi (50-17) 'wings, 22 f i n s ' , for example, i s / t s ' / i n both d i a l e c t s : / t s ' i / i n Zhong-44 shan and [ t s ' i : ] i n Cantonese. In most Chinese d i a l e c t s , the i n i t i a l i s [ t s ' ] or I t s ' ] for t h i s word. In fact, none of the di a l e c t s represented i n the Zi-hui (p.48) has a f r i c a t i v e i n i t i a l for chi ffiL . In contrast, the S h ^ n - i n i t i a l word shi (.5 8-5) 'to begin' i s / t s ' / i n both Zhong-shan and Cantonese; with the exception of Mei-xian, which also has the i n i t i a l [ t s ' ] for the word, a l l the d i a l e c t s i n the Zi-hui (p.50) show a f r i c a t i v e [s] or IgJ . In addition to the exceptions c i t e d by Hashimoto, two other words i n Zhong-shan have an a f f r i c a t e as the r e f l e x of the Sh^n i n i t i a l . Although Hashimoto records nothing beside these two words, l i s t e d i n (30) below, the Zhong-hua Xin Zi-dian agrees with the Zhong-shan data i n recording a f f r i c a t e i n i t i a l s for them. (30) 49-17 ^ / t s ' i 1 3 / 'a pig' 123-11 / t s i n 5 5 / 'odour of sheep or goat' The pronunciation of shan ^ i n Zhong-shan (and Cantonese) i s probably based on the analogical reading of the word zhan - 213 -(123-9) ' f e l t ' . Two words with a highly i r r e g u l a r r e f l e x of /h/ for the i 13 She'll i n i t i a l are shang flvO (174-6). 'noon' /hiong /, and xiang (175-17) 'rations or pay for troops' /hiong /. Cantonese likewise has i n i t i a l [h] for the two words which, as Hashimoto suggests (for Cantonese, but also applicable to Zhong-shan), i s probably due to the analogical reading of more common words such as xiang fa (175-21) 'facing towards' or xiang ( = ^ .174-14) 'noise, sound'. The EMC p a l a t a l nasal i n i t i a l , Ri 0 (*r>- > LMC *r-) has one regular r e f l e x i n Cantonese, the high front glide [ j J , which i s rounded to [uj before rounded nuclear vowels. Zhong-shan has two regular correspondents to the Ri i n i t i a l : / j / and /ng/, which are conditioned by whether the f i n a l has a high vowel or not: / j / occurs before non-high vowels and /ng/ before high ones, or more s p e c i f i c a l l y , the high front vowels [ i : J and I y : J. Note, however, that / j / and /ng/ are not i n complementary d i s t r i b u t i o n i n the phonological system of Zhong-shan as a whole, since /ng/ also occurs with non-high vowels when i t functions as the regular r e f l e x of the Y i (*n-) i n i t i a l . i/j/, however, never occurs before the high vowels [i:] and [ y : ] . ) Examples of R i - i n i t i a l words from d i f f e r e n t rhyme groups with / j / and /ng/ as i n i t i a l s i n Zhong-shan are l i s t e d i n (31). The rhyme group to which each word belongs i s also s p e c i f i e d i n (31) within round brackets. (.31) a. /ng/ before high front vowels: 19-16 Jia /ngy 5 1/ ' l i k e , as' (Yu S$S^ ) 55-6 / n g i 1 3 / 'you, your* (cl a s s i c a l ) (Zhi jL ) - 214 76-5 /ngiw / 1 to forgive' (Xiao ) 10.3-5 / - 1 3 , /ngim / 'to dye' (Xian ). 141-17 /ngyn 1 3/ 'soft' (Shan fa ) b. / j / elsewhere: 11-17 t / • 13 / /Da / 1 to provoke' CJia ML ) 26-3 / j 0 j 1 3 / •milk" (.Yu i ^ . ) 61-15 87-1 L / j j r f j 1 3 / /jaw / 1 stamen1 'soft' (Zhi ) (.Liu % ) 114-10 / j a p 2 / 'to enter 1 (Shen :'L ) 154-15 0 / j a t 2 / ' sun 1 (Zhen ffc) 175-20 -dl /• 2 2 / /jong / 'to y i e l d ' (Dang ) 221-12 / j u k 2 / 'flesh, meat' *K r u i , i (.Tong jfU). Recall that the sy l l a b l e / j 0 j / , i n ru and s U [ 0 i l ] phonetically (see 1. 2.3) . In a couple of cases, the alternation of a nasal or p a l a t a l glide i n i t i a l r e f l e c t s a c o l l o q u i a l versus l i t e r a r y reading, as demonstrated i n (32). . (32). 151-17 /ngan 1 3/C, / j a n 1 3 / L . 'to endure' 172-13 | ^ /nong 5 1/C, /jong 2 2/L. 'pulp' The /n/ r e f l e x i n rang may stem from the reading with the Niang i n i t i a l , which i s recorded i n the Guang-yun. For the word ren -f;2> (153-7) 'to recognize', which belongs 22 to the Zhen rhyme group, the l i t e r a r y r e a d i n g i s / j a n / and the 22 c o l l o q u i a l /nging /, with a d i f f e r e n t f i n a l as well as a di f f e r e n t i n i t i a l . The word ren (153-8) ' e l a s t i c ' only has a c o l l o q u i a l 22 reading of /ngan / i n the present Zhong-shan data, but i t may be 33 observed that Cantonese has a l i t e r a r y reading of [JBII J contrast-33 ing with a c o l l o q u i a l reading of [rjen J . Moreover, i n his s y l l a -bary, Chao records /ngan 2 2/ as the c o l l o q u i a l reading for ren implying that there i s a l i t e r a r y counterpart, presumably with a glide i n i t i a l . Another c o l l o q u i a l with i n i t i a l /ng/ i s nian $ 'twenty', which i s h i s t o r i c a l l y homophonous with ru "N- / l i s t e d i n (31b) above. Again, there i s no / j / i n i t i a l counterpart i n the d i a l e c t . I n i t i a l /n/ as a re f l e x of the Ri i n i t i a l occurs i n only a few words i n Zhong-shan. I t was brought to the writer's atten-t i o n by Norman that the Ri i n i t i a l i s reconstructed as *n- for Min. Recall also that the Ri i n i t i a l i s EMC * i > - . I t i s possible that the dental nasal i n a few Zhong-shan words i s a sur v i v a l of an e a r l i e r stratum. A. word with i n i t i a l /n/ i n Zhong-shan i s er (59-12) 'cakes' / n i 2 2 / , from the Zhi rhyme group. The Can-3 3 tonese pronunciation for the word i s [lej ] according to Hashi-3 3 mo to (p. 4 6 0 ) and [nej ] according to the Zhong-hua Xin Zi-dian ( 1 9 7 6 : 6 3 1 ) . Another word with /n/ corresponding to the Ri i n i t i a l i n 1 3 tfo Zhong-shan i s ru yjc ( 2 1 - 8 ) 'you' / n i /, from the Yu rhyme of the Y u r h y m e group. Although Long-du also has s y l l a b l e / n i 2 4 / for ru (Egeriod, p. 8 5 ) , one would not expect the borrowing into Zhong-shan of a c l a s s i c a l word such as ru ylf from Long-du. Although the Yu rhyme i s treated as a He-kou category i n the di a l e c t survey, i t was i n fact c l a s s i f i e d as Kai-kou i n the Yun-ji n g . The reading of f i n a l / i / i n Zhong-shan for ru j\r would suggest possible preservation of Kai-kou which i n v i r t u a l l y every other instance has been l o s t i n the d i a l e c t . Ru 'jfc i s reconstruct-ed by Tong-he Dong, for instance, as Archaic Chinese *niag, and - 216 -by Pulleyblank t e n t a t i v e l y as Old Chinese * n l a ? . In Cantonese, ru JHJf i s pronounced 'Iqyi-2^] . B a l l ' s data (.p. 531) likewise show a He-kou f i n a l ; the i n i t i a l i n t h i s case i s /ng/: /ng{i/. There are four words i n the Zhong-shan data with the zero i n i t i a l as the r e f l e x of the h i s t o r i c a l Ri i n i t i a l , which may be due to Cantonese influence; that i s , normally /#/ i s the Zhong-shan correspondent to the glide i n i t i a l i n Cantonese i n the envi-ronment before high tense vowels. (.32) 24-7 / y 5 1 / "a scholar' 123-16 ft\ / i n 5 1 / 'certainly' 123-17 j $ / i n 5 1 / 'to burn' 187-20 At) / i n g 5 1 / 'as usual' Not i n the d i a l e c t survey, but included i n Chao's data (p.67), i s ru ^ 'a c h i l d ' / y 5 1 / and / y 2 2 / , which also has a zero i n i t i a l as a r e f l e x of i n i t i a l Ri. 3.1.2.7. Velars (LMC) The regular correspondent to the Jian | L (*k-)' i n i t i a l i n Zhong-shan (and Cantonese) i s /k/, as exemplified i n (33). 1-15 55 / k o " / 'older brother' 37-23 t i / k a j 5 5 / 'a chicken' 90-12 22 /kaw / 'to rescue' 177-12 >* /kong1 3/ 'extensive, wide 194-19 /ka:k 2/ 'to reach' There are, however, about forty words which have /k'/ as a r e f l e x of the Jian i n i t i a l , accounting for roughly one-tenth of the J i a n - i n i t i a l words i n the data. Approximately the same number of aspirated exceptions occurs i n Hashimoto's data on Cantonese, a l b e i t not necessarily involving the same words. Thus, while both Zhong-shan and Cantonese exhibit an aspirated velar stop i n i t i a l i n such words as gui (61-4) 'regulations', gei £/o (.114-13) 'to give', jue (148-1) 'to decide' and kuang (208-5) 'metal ore', for example, there are other J i a n - i n i t i a l words i n which only one of the two d i a l e c t s has an aspirated i n i t i a l while the other the regular unaspirated one. For instance, gou ;||. (82-7) 'a ditch' and j i u (.87-3) 'a pigeon' are aspirated i n Cantonese, a sound change which. Hashimoto (p.641) attributes to the avoidance of homophony with a taboo s y l l a b l e . Both words are transcribed as 53 55 [k'Bw J by Hashimoto for Cantonese, i n contrast to /kaw / i n Zhong-shan. Information on these two words i n Cantonese i s not consistent, however; for example, both words are recorded as un-aspirated by the Zhong-hua Xin Zi-dian (19 76) for Cantonese. In contrast, while Wong (.1954:11-12) likewise records an unaspirated i n i t i a l for j i u A.l? > for the word gou , he has an aspirated i n i t i a l as the c o l l o q u i a l reading and an unaspirated one for the l i t e r a r y counterpart. The unaspirated i n i t i a l i n Zhong-shan for gou i s the c o l l o q u i a l reading since the word for 'ditches', -f. 55 51 which, are found i n the v i l l a g e s , i s gou-qu ygi. /kaw k'y /. In the case of the words yu (19-17) 'to dwell', gua (.45-1). 'to hang up', and jju ^ (221-13). ' chrysanthemum' , they are aspirated i n Zhong-shan but unaspirated i n Cantonese. Other i r r e g u l a r reflexes of the Jian i n i t i a l present i n Zhong-shan are also found i n Cantonese. A number of the Cantonese exceptions have been accounted for by Hashimoto (p.641). In ^ - 218 -general, since i t i s the same words which are affected i n the two d i a l e c t s , Hashimoto's suggestions for the source of the i r r e g u l a r development of certain words can also be used for the Zhong-shan exceptions. I n i t i a l /h/ i n the word j i a o jjjy^ r (.74-11) 'leaven, yeast' may be based on the analogical reading of xiao i ^ - (.74-14) ' f i l i a l ' , and those of j i a o "fyj (.80-1) 'to cleanse with water' and j i a o (80-10) 'to be lucky' based on xiao (.80-11) •dawn'. Both xi|o i and xiao fljb are descendents of the Xiao (*x-) i n i t i a l . The presence of Iq] instead of [kj i n the words gou ^) (82-5, 84-19) 'to hook' and gou /$£) (.82-6) 'a hook' i s attributed by Hashimoto to the avoidance of a taboo s y l l a b l e i n Cantonese. The readings with Ik] are, nonetheless, also found i n Cantonese (see, for example, Wong, 1954:11; Zhong-hua Xin Zi-dian, 1976:57, 585). Zhong-shan also has the velar nasal i n i t i a l for these two words despite the fact that the other word with which they should be homophonous, namely gou ^ , mentioned above, has a /k/ i n i t i a l . In some cases, an unexpected i n i t i a l has i t s source i n another reading with a d i f f e r e n t h i s t o r i c a l i n i t i a l . The word chuan «j* (139-7) 'to s t r i n g together' / t s ' y n 2 2 / , with i n i t i a l / t s ' / i n Zhong-shan, for example, i s probably derived from the reading with the EMC Chuan % i n i t i a l . The / t / i n i t i a l i n j i u j^j (87-5) 'to involve' and j i u ^ 13 35 (91-5) 'to correct", both /taw / i n Zhong-shan (Ituw ] i n Can-tonese) i s highly i r r e g u l a r , and may be due to the analogical reading of dou ij- (.83-1) 'a peck (a dry measure) 1 , which has the Duan i n i t i a l , although such a proposal i s quite u n l i k e l y . The zero i n i t i a l i n the word kuai % (.43-4) 'to calculate 22 /uj / probably has i t s source i n the reading with the Xia i n i t i a l hui ^ (43-8). 'to meet' and hui (.43-9) 'to be able'. Another exceptional r e f l e x to the Jian i n i t i a l i s / l / , occurring i n the word l i a n jjjgC (.103-8) 'face'. Apparently, the l a t e r a l i n i t i a l for the word i s common among the modern Chinese dia l e c t s (cf., for example, Zi-hui, p.181). H i s t o r i c a l l y , more-over, there i s evidence that the word had once contained a l a t e r a l i n i t i a l since i t had a fan-qie of l i - j i a n j) xJfy\ , with l i pos-sessing the L a i (*1-) i n i t i a l i n Middle Chinese, (cf. Hashimoto 1972b:29). The pronunciation of a /w/ i n i t i a l for guo ^ f f i (.4-7) 'a cooking pot' and guo (.12-17) 'a kind of wasp' (both, /wo 5 5/ i n Zhong-shan) i s probably due to the analogical reading of the word wo ffi (.4-17) 'nest'. Norman suggests that i n i t i a l /w/ may perhap also be the r e s u l t of contamination from huo a wok' /wok /. With, regard to the Qi 'Jj^ (*k'-) i n i t i a l , almost half of the Zhong-shan data have /k'/ as a correspondent of the h i s t o r i c a l i n i t i a l , and about the same number have /h/. In a few cases, as shown i n (.34), the data demonstrate a clear d i s t i n c t i o n between l i t e r a r y and c o l l o q u i a l readings v i s - a - v i s /k/ and /h/. (.34) 18-14 y# /k'u 2 2/L., /hu 1 3/C. 'a store-house' 71-12 ^ /k'a:w22/L. /how22/C. 'to depend on' 201-1 fyfi A ' i n g 5 5 / L . /hia:ng 5 5/C. ' l i g h t , young' The i n i t i a l /h/ i n ku ^ i s used c o l l o q u i a l l y for the name of the v i l l a g e , Ku-chong J% %j /hu ts'ung /, for instance, while the A - 22 l i t e r a r y reading of ku m i s /k'u /. The tonal change of ku - 220 -as a r i s i n g /13/ p a r a l l e l s other c o l l o q u i a l - l i t e r a r y pairs i n which the c o l l o q u i a l reading has /13/ and the l i t e r a r y form /22/ (.see, for example, (.15) and 24).) In t h i s case, however, the i n i t i a l i s not 'muddy' h i s t o r i c a l l y , as i s true of the previous examples of such a tonal alternation. Most of the words containing the /h/ i n i t i a l descending from the Qi i n i t i a l represent the c o l l o q u i a l layer of the d i a l e c t , as exemplified by (.3 5) below. 22-12 / h y 2 2 / 'to go away' 58-11 / h i 1 3 / 'to r i s e ' 83-12 XX /u 13 , /haw / 'mouth' 117-8 /hon 2 2/ 'to see, to look at 216-5 /huk 5/ 'to weep, to cry' Those words with i n i t i a l /k'/ tend to be l i t e r a r y words, as shown in (36) . 24-12 55 A ' Y / 'to expel' 31-4 22 A ' o r v 'to h i t against' 37-25 55 A ' a j b b / 'a r i v u l e t ' 147-5 A ' y n 1 3 / 'a dog' 154-17 ii / k ' i t 2 / 'to investigate' It might be of in t e r e s t to point out that kan 'to see, to look at' i n (.35) i s a c o l l o q u i a l word i n Zhong-shan, whereas i t i s a 44 l i t e r a r y one i n Cantonese (pronounced [hoi-n ] ) . The c o l l o q u i a l Cantonese equivalent i s d i 9 $ [ t ' s j 3 5 ] . In Cantonese, where two-thirds of the Q i - i n i t i a l words are now pronounced with [h] (or further changed to I f ] i n He-kou words), there are more instances - 221 -of l i t e r a r y words being pronounced with, the Ih.J i n i t i a l i n that d i a l e c t than i n Zhong-shan. The word quan i n (36) i s a case i n point. In both Zhong-shan and Cantonese, t h i s word i s s t r i c t l y l i t e r a r y . The common word for !dog' i s gou (83-9)., pronounced 35 13 Ik-ew ] i n Cantonese, and /kaw / i n Zhong-shan. In the l a t t e r d i a l e c t , the word quan TN. has the /k'/ i n i t i a l , thereby r e f l e c t i n g i t s status as a l i t e r a r y word. In Cantonese, on the other hand, i> 35 quan i s pronounced [hyxn ]. It i s obviously observations of th i s unpredictable sort that prompted Hashimoto (p.642) to simply note that one-third of the Q i - i n i t i a l s i n Cantonese have i n i t i a l [ k ' ] , most of which are l i t e r a r y words, while the two-thirds which have an [h] (or [f]) i n i t i a l are either c o l l o q u i a l or l i t e r a r y words. Accepting the assumption that sound change i s regular, one would analyze the apparent b i f u r c a t i o n of the Qi i n i t i a l into /h/ and /k'/ not as an incomplete process of deplosivization, as Hashimoto wishes to suggest, but as the complete deplosivization of the Qi i n i t i a l . Presumably, the l i t e r a r y words i n the d i a l e c t l a t e r became subject to influence from a northern standard which had not undergone the deplosivization process. Only four words i n Zhong-shan show a change from /h/ to / f / , a f f e c t i n g some words i n the Guo ; $ L rhyme group and the He-kou rhymes of the Xie ^ rhyme group. Unlike Cantonese, the words with the / f / i n i t i a l i n Zhong-shan only occur with non-high (and non-front) nuclear vowels i n the modern d i a l e c t . Contrast, for instance, ku ^ (18-15) 'trousers 1, which i s Cantonese . [ f u : 4 4 ] and Zhong-shan /hu 2 2/, with ke $ (4-9). 'a c l a s s ' , which. 53 55 has i n i t i a l / f / i n both d i a l e c t s : [fo: ] i n Cantonese, and /to / - 222 -in Zhong-shan. A more detailed analysis of the process that changed /h/ to / f / i n the Yue di a l e c t s w i l l be given i n the discussion of i the Xiao Qfa i n i t i a l where the change of /h/ to / f / affects a greater proportion of the Zhong-shan data. Two l i t e r a r y words have / l / as the correspondent of the Qi i n i t i a l : ke $|L (4-11) ' c l a s s i f i e r for trees' / l o 1 3 / (with / k o 1 3 / as a variant form), and ke ^ (3-16). 'a kernel; c l a s s i f i e r 13 for small, round objects' / l o /. There are a few words i n Grades III and IV rhymes contain-ing the Qi i n i t i a l i n which the s y l l a b l e - i n i t i a l segment i s / j / , implying a loss of the o r i g i n a l i n i t i a l , r e s u l t i n g i n the medial serving as the new i n i t i a l i n the d i a l e c t . One such word i s qiu Jl. (.87-6). 'a h i l l o c k ' /jaw 5 5/. However, the word qiu jf^ 'a place} a surname', which should be homophonous with qiu Jc. accord-55 ing to h i s t o r i c a l sources, i s pronounced /hia:w / i n the surname of a Zhong-shan r e l a t i v e of the informants. Another word with L 2 i n i t i a l / j / i s the common word chi vr& (.208-3) 'to e a t ' — / j a : k / 4 in Zhong-shan. I t i s pronounced [heik ] i n Cantonese, and [hjek] i n Rai-Ping another Yue d i a l e c t (from 1977 f i e l d n o t e s ) . (Ball (p.531), however, records /yak/ for both Cantonese and Shi-qi speech, contrasting with /yiek/ for Macao Zhong-shan.) I n i t i a l /w/ occurring i n Q i - i n i t i a l words should also be viewed as the loss of the o r i g i n a l r e f l e x , with the h i s t o r i c a l medial becoming the new i n i t i a l /w/ i n the d i a l e c t . This can be c l e a r l y seen i n the following example. Although i n the present data /wat 5/ i s the only pronunciation given for qu ^ (.165-13) 'a grievance', i t i s s i g n i f i c a n t that Chao (p.65) records both 5 5 /wat / and /k'wat / for the word. - 223 -The zero i n i t i a l also occurs i n a couple of Q i - i n i t i a l words; for example, the personal name of Mencius, ke ij*!] (.1-16), 55 i s /o / i n Zhong-shan. The pronunciation of giu j$k (.88-20). 'parched wheat or 22 r i c e ' / t s 1 aw /, with i n i t i a l / t s 1 / , may be due to the analogical reading of the common word chou ^ (90-7) 'foul odour', descend-ing from the Chuan i n i t i a l . Given the tendency to avoid homophony with, words having unpleasant or inauspicious connotations, i t i s surprising to f i n d the present analogical reading. The Qun ^ (*kfi-< EMC *g-) i n i t i a l behaves s i m i l a r l y to other muddy i n i t i a l s ; i n t h i s case, Zhong-shan has the reflexes of /k/ or /k'/ depending On the h i s t o r i c a l tones and on whether the s t y l e i s c o l l o q u i a l or l i t e r a r y . Examples are given i n (37). (.37) a. Ping-sheng: 3-5 *fjc /k ' j rf 5 1/ 'eggplant' b. Qu-sheng: 90-14 || /kaw 2 2/ 'old' c. Ru-sheng: 199-15 / k i a : k 2 / 'clogs' d. Shang-sheng: 155-10 jfr, /k'an 1 3/C. 'near to' 155-10 /kan 2 2/L. 'near to' Naturally, there are some exceptions to the pattern of correspondences to the Qun i n i t i a l . Ju (19 9-4) 'drama' /k'ia:k /, for example/,has an aspirated instead of the unaspirated form for a Ru-sheng word. In fact, most of the southern Chinese d i a l e c t s , including Cantonese, Mei-xian, Xia-men, and Fu-zhou, have I k ' ] as the i n i t i a l i n the word ju Jp] (Zi-hui, p. 97). The Y i (*n-) i n i t i a l has /ng/ as the regular Zhong-shan correspondent. In contrast, Cantonese has Iy] i n those words descending from Grades I and II rhymes, and I j J or IuJ i n those Grade III and IV rhymes which have a high or front vowel i n the modern d i a l e c t , and Iq] i n other Grade III and IV rhymes. Zhong-shan examples are given i n (.38). . ( . 3 8 ) 2-5 4 /ngo 1 3/ 'I, me1 19-22 k /ngy 5 1/ 'a f i s h 1 56-20 lit /ngi / 'to doubt' 117-9 /ngon 2 2/ 'shore' 198-7 ML / n g i n g 5 1 / 1 to welcome In the case of Grade III Y i - i n i t i a l 1 % and TongArhyme groups, the modern r e f l e x i n Zhong-shan i s / j / , as exemplified i n (.39) below. (39) 174-12 /Jong 1 3/ "to look up to' 176-20 jb / j o k 2 / 'harsh' 226-13 / j u k 2 / 'jade' The Y l - i n i t i a l words i n the f i r s t grade of the Yu rhyme group (or more precisely, i n the Mo ^ rhyme) have the s y l l a b i c velar nasal [n] as the modern Zhong-shan (and Cantonese) r e f l e x . The s y l l a b i c nasal constituting the entire s y l l a b l e i n these words i s without exception i n the data. A few examples, a couple of which have already been c i t e d i n 1.2.4 i n connection with the discussion of s y l l a b i c nasals i n the d i a l e c t , are presented i n (40). (40) 15-6 * /ng / 'Wu' 16-25 X /ng / 'five• 16-27 > / 13 . /ng / ' noon 1 18-16 ik /ng / 'to be mistaken' Observe, however, that i n B a l l ' s a r t i c l e , while a s y l l a b i c nasal i s recorded for Cantonese, the Zhong-shan form has s y l l a b l e /ung/. Aside from the systematic exceptions discussed above, there are very few exceptions to /ng/ as the r e f l e x of the Yx i n i t i a l i n Zhong-shan. The Grade IV word yan (.131-16) ' ink-22 stone 1 / i n / may have been subject to d i a l e c t a l influence. The 33 35 word xn Cantonese, for instance, i s I j i i n ' ]. As noted e a r l i e r , the zero i n i t i a l i n Zhong-shan often corresponds to the Cantonese pa l a t a l [j] i n i t i a l when i t occurs before high front vowels. . 51 Observe that yan %ft (129-24) /ngin /, which can be used as a graphic variant of yan , has i n i t i a l /ng/ i n Zhong-shan. Given the velar nasal r e f l e x i n yan %^ , i t does not appear to be the case that yan underwent the following sound change: *gjian > *jian> j i n , which would have been the case had the zero i n i t i a l been a genuine Grade IV r e f l e x . One can quite safely conclude that the zero i n i t i a l i n yan i s not a true r e f l e x of the Grade IV Yx i n i t i a l , but i s i n a l l p r o b a b i l i t y the product of d i a l e c t a l contamination. The l a s t two velar i n i t i a l s to be discussed are Xiao *x-) and Xia ^ (*xfi-< EMC **-)', both of which, one might have noted from Chart 4, were t r a d i t i o n a l l y c l a s s i f i e d as hou-yin ('throat sound'), or guttural, i n i t i a l s . There are two regular correspondents of the Xiao i n i t i a l i n Zhong-shan (and Cantonese): /h/ and /if, which are conditioned - 226 -by the nature of the f i n a l . In Zhong-shan, /h/ occurs i n Kai-kou words and i n those He-kou words which have the r e f l e x of a high and/or front vowel (v i z . , / i / , /y/, /u/ orffif), while / f / occurs only i n words descending from He-kou rhymes which have a non-high, non-front vowel i n the modern d i a l e c t . Examples of t h i s pattern of d i s t r i b u t i o n i s given i n (.41). (.41). a. /h/ i n Kai-kou words: 60-1 ^ / h i 5 b / 'rare' 70-4 Jfr /how13/ 'good' 172-23 ^ / h i o n g 5 5 / 'fragrant' /h/ i n He-kou words with a high and/or front vowel: 6-17 fa / h c i 5 5 / 'boots' 17-1 H 13/ /hu / 'a t i g e r ' 148-4 & / h y t 2 / •blood' 209-7 h 55 /hing / 'an elder brother' 223-3 m 55 /hung / 'bosom' b. / f / i n He-kou words with a non-high, non-front vowel: 5-17 I 3 / /fo / • f i r e ' 13-1 / f a / 'a flower' 160-4 / f a t 5 / 'suddenly' 177-2 55 /fong / 'wild, barren Although Cantonese (Hashimoto, p.644) also has Ih] and I f ] as the regular correspondents of the Xiao i n i t i a l , t h e i r d i s -t r i b u t i o n d i f f e r s somewhat from that i n Zhong-shan. I n i t i a l I n ] , for instance, occurs i n Kai-kou words and i n only those He-kou words which have the r e f l e x of a front vowel i n the d i a l e c t , while I f ] occurs i n He-kou words with, a non-front vowel, the height of which, i s not relevant. In Cantonese, moreover, the Tong 1^, rhyme group constitutes an exception to t h i s d i s t r i b u t i o n i n that, although, the r e f l e x of the Tong rhyme group i s a non-front vowel in modern Cantonese, the i n i t i a l i s nonetheless I h ] rather than the expected [f] (e.g., xiong i n (41) i s Ihuq 5 3] i n Cantonese, 53 not [fug ] , the l a t t e r being the pronunciation of words descend-ing from Late Middle Chinese labiodentals. I t has been suggested that the fronting of the Xiao i n i t i a l to a labiodental ( i . e . , the so-called " d e n t i l a b i a l i z a t i o n " process) i n the Yue d i a l e c t s i n cer t a i n environments i s quite a recent sound change. W. Boltz (197 8), for instance, proposes to date the denti-l a b i a l i z a t i o n process at c i r c a mid-eighteenth century on the basis of a Sino-Portuguese glossary compiled around 1750. As has been argued elsewhere (Chan, forthcoming), the data Boltz c i t e d gave evidence not for the period when d e n t i l a b i a l i z a t i o n was i n the process of a f f e c t i n g Cantonese (.actually Zhong-shan, to be more 5 correct ), but rather the terminus ad quern of that sound change, for the data show quite c l e a r l y that the d e n t i l a b i a l i z a t i o n of the Xiao i n i t i a l (and, to a more lim i t e d extent, the Qi i n i t i a l , e s p e c i a l l y i n Zhong-shan) had s t a b i l i z e d by the middle of the eighteenth, century. When that sound change had taken place s t i l l remains to be solved. Among the i r r e g u l a r correspondents to the Xiao i n i t i a l i n Zhong-shan i s the l a b i a l /w/ segment, as i n hui "^tf (67-3) 'taboo' 13 /waj /. The pronunciation may be based on the analogical reading of the more common word wei $ (.66-10) 'admirable', with the EMC Yun is (*w-) i n i t i a l . The source of /w/ as the i n i t i a l segment •+ 13 i n the l i t e r a r y word,hui rf (67-4) 'plants' /waj / i s a b i t more - 228 -d i f f i c u l t to explain. Hui -f| is: c l a s s i f i e d under Qu-sheng words, although i t does have a Shang-sheng reading from which, the Zhong-shan form may have been derived, which, would account for the r e f l e x of /13/ instead of /22/; however, the h i s t o r i c a l i n i t i a l of the Shang-sheng derivative i s also Xiao. In discussing /w/ occurring as the i n i t i a l segment i n the Zhong-shan pronunciation of those words descending from the Xiao i n i t i a l i n the paragraph, above, i t should be added that they are He-kou words, and diac h r o n i c a l l y , one can equally claim that Zhong-shan had l o s t the i n i t i a l segment that should correspond to the Xiao i n i t i a l . The l a b i a l medial /w/ by default then becomes the i n i t i a l segment i n the modern d i a l e c t . This remark i s also r e l e -vant to the observation of /w/ or / j / as the s y l l a b l e - i n i t i a l segment of some words descending from the Qi , Xia and Ying i n i t i a l s . Diachronically, /w/ and / j / correspond not to the h i s t o r i c a l i n i t i a l s , but to the medials. In other words, with the loss of the o r i g i n a l r e f l e x of these h i s t o r i c a l i n i t i a l s , the medials became the new s y l l a b l e - i n i t i a l segment i n the modern Zhong-shan d i a l e c t . . *. 2 The word q i (156-5) 'to reach to' /ngat / has a velar nasal i n i t i a l i n Zhong-shan corresponding to the Xiao i n i t i a l . No explanation can be offered at t h i s point, although i t i s observed that Cantonese also has a velar nasal i n i t i a l for the word. Xiu (87-13) 'to rest' /jaw / has a / j / i n i t i a l i n Zhong-shan (and Cantonese). Again, the writer can give no immediate explanation for the i r r e g u l a r correspondence. (This i s , of course, a case of the loss of the o r i g i n a l i n i t i a l , since xiu <44< i s a Grade III Kai-kou word; the i n i t i a l / j / segment i n the modern d i a l e c t would be - 2 2 9 -the o r i g i n a l medial i n the word.) There are two regular correspondents to the Xia Z£ i n i t i a l i n Zhong-shan, namely /h/ and /w/. Their d i s t r i b u t i o n i s simi l a r to that of that Xiao i n i t i a l : /h/ occurs i n Kai-kou rhymes and i n those He-kou rhymes which have the r e f l e x of a high vowel i n Zhong-shan ( i . e . , /y/ or /u/), and /w/ as the i n i t i a l segment i n the remainder of the He-kou rhymes. Cantonese also has {hj and IwJ as the main correspondents of the Xia i n i t i a l . The d i s t r i b u -t i o n of the two reflexes p a r a l l e l s the d i s t r i b u t i o n found i n Cantonese of i n i t i a l s IhJ and [fj i n X i a o - i n i t i a l words. Zhong-shan examples are provided i n ( . 4 2 ) . ( 4 2 ) a. /h/ i n Kai-kou words: 3 - 4 || /ho / 'to congratulate' 3 3 - 2 1 | i / h a : j 5 1 / 'shoes' 2 0 5 - 1 4 7 f ) / h i n g 5 1 / 'a form' /h/ i n He-kou words with a high vowel: 1 5 - 1 3 i /hu 5 1/ 'a pot' 1 4 7 - 3 /hyn 5 1/ 'to hang up' 2 1 6 - 6 j ^ j - /huk 2/ 'a corn measure' b. /w/ i n He-kou'words with, a'non-high vowel.: . 4 - 1 3 / W O 5 1 / 'peace' 1 3 8 - 2 /wa:t 2/ 'slippery' 1 7 7 - 4 -jjj /wong 5 1/ 'yellow' There are also about ten He-kou words i n the data i n which the i n i t i a l underwent a change from /h/ to /#/ i n Zhong-shan, as exemplified by the following: - 230 -(43) 43-8 'If / u j ^ 2 / 'to meet" 134-8 ^ / y n 5 1 / 'to f i n i s h ' 137-11 ft / u t 2 / ' l i v i n g ' 211-6 / i n g 5 1 / 'a glow-worm' These words may have been generally influenced by words descending from the Ying ^ and Yu i n i t i a l s , which have the r e f l e x of the zero i n i t i a l when the following segment i s a high vowel. In Cantonese, these vowels are preceded by homorganic glides (e.g., A" 33 35 >>> 21 hui (43-8) i s pronounced {wu:j ' J, wan yij I q y m J, ying 'JlX r . 21. t . '<£ [j i g ] , etc.) . A number of words i n the data have a velar stop as the r e f l e x of i n i t i a l Xia. Although some may be based on alternate readings with a Middle Chinese velar stop i n i t i a l or on analogical readings, others may suggest possible preservations of an Old Chinese *g- i n i t i a l which had l a t e r merged with Old Chinese *h-to form the Middle Chinese Xia i n i t i a l (see, for example, Pulley-blank, 1962:86-88). A l i s t of velar-stop i n i t i a l words from the Xia i n i t i a l i s given below. (.4 4) 1-24 55 'small plants' 13-4 *1 /ko / 'to punt a boat' 15-18 A ' u 5 1 / 'pudding, (edible) paste' 34-7,8 w / k a : j 1 3 / 'surname; to understand' 42-22 /a / k ' u j 2 2 / 'a stream overflowing i t s banks 43-10 if 22 A - ' U D / 'to draw, to paint' 47-1 a A'waj / 'to lead by the hand' 47-2 51 A'waj / 'a p l o t of land' 74-17 t-, 22. /ka:w / 'a m i l i t a r y or naval t i t l e ' - 231 -10.1-•3 & /ka:p 2/ 135- •7 /k'un 1 3/ 182- •4 / k 1ong 2 2/ 19 5-•13 t /king / 215- •4 /kung 2 2/ 'a casket' 'bright' 'nape of the neck' 'the stalk of a plant' 'the din of b a t t l e ' Hu also has the form of /hu5"'"/, which i s the l i t e r a r y pronun-51 c i a t i o n , and /u /, which i s used to mean '(non-edible) paste' or 'to paste". Apparently, Min likewise has a number of velar stops as a r e f l e x of the Xia i n i t i a l . Hu jfcffi , for example, has a /k/ i n i t i a l i n v i r t u a l l y a l l the Min d i a l e c t s , and i s i n fact recon-structed by Norman (.1969:255) as. containing i n i t i a l *k i n Proto-Min. Xiang ^ f l has an alternate reading of /hong 2 2/, as i n j i - x i a n g 'young chicken' /kaj hong /, which undoubtedly i s derived from Cantonese, where the only pronunciation for xiang ^ i s There are also a couple of words with a velar nasal i n i t i a l i n Zhong-shan as a r e f l e x of the Xia i n i t i a l . These are: 51 C45) 72-24 /nga:w / 'savoury food' j. 51 72-25 yZ /nga:w / 'mixed, muddy' Yao (old reading, xiao) ^ , for example/ has a ve l a r nasal not only i n Zhong-shan, but also i n Cantonese, Xia-men, Chao-zhou and Fu-zhou; and i n Xi-an (a Mandarin d i a l e c t ) , the word has a p a l a t a l nasal i n i t i a l (Zi-hui, p.148). There are a few cases of / l / and / f / corresponding to the Xia i n i t i a l ; for example, j i a n J^ fa, (.100-7) 'a warship' /la:m 2 2/, and huang ^ (177-15) 'bright' / f o n g 1 3 / . It i s possible that the i r r e g u l a r reflexes i n such cases are due to analogical readings; - 232 -2 2 the pronunciation of /la:ra / for j i a n , for instance, may be based on that of lan jH (.96-10.) 'to overflow', or lan jM, (56-11). 'a cable', which have the h i s t o r i c a l L a i i n i t i a l . An / f / i n i t i a l on huang ^ i n Zhong-shan may be based on the analogical reading 13 of huang \%> (177-14) 'wild, mad' /fong /, with the Xiao i n i t i a l . 3.1.2.8. Gutturals (.LMC) Two guttural i n i t i a l s are discussed i n t h i s section, Ying ^ (.*?-) and Yu (*#-) . In Zhong-shan, both i n i t i a l s have /$•/1 / j / and /w/ as regular correspondents, d i f f e r i n g only some-what i n t h e i r d i s t r i b u t i o n . In discussing the d i s t r i b u t i o n of /0/, / j / and /w/ i n Y i n g - i n i t i a l words, a separation between Kai-kou and He-kou words would simplify the description to some extent. In reference to Kai-kou words only, the regular correspondent i s /0/ i n Grades I and II, and i n those s y l l a b l e s which have the r e f l e x of a high vowel; and / j / i n those Grade III and IV s y l l a b l e s where the modern re f l e x i s a non-high vowel. With regard to the He-kou s e r i e s — excluding the Tong rhyme group—the regular correspondent of the Ying i n i t i a l i s /0/ i n those s y l l a b l e s which have the r e f l e x of a high vowel, and /w/ elsewhere. In the case of the Tong rhyme group, /0/ occurs i n Grades I and I I , and / j / i n Grades III and IV. Examples of the d i s t r i b u t i o n of these correspondents, as outlined above, are given i n (46). (46) a. Kai-kou Series: 70/ i n words with a high vowel: 79-3 jf, /iw / 'important' - 233 -2 154-18 ZJ /yt / 'second of the Ten Stems' 188-2 ^ / i n g 5 5 / 'ought, should' /0/ i n remaining Grade I & II words: 8-16 jjltj / a 5 5 / 'a crow' 94-2 & /om22/ 'dark' 122-5 ^ /a:n 2 2/ 'quiet' / j / i n remaining Grade III & IV words: 2 2 91-7 itjt /jaw / 'young' 154-19 — / j a t 5 / 'one' 173-1 / j o n g 5 5 / 'central' b. He-kou Series (excluding the Tong rhyme group) /0/ i n words with a high vowel: 15-20 fe. / u 5 5 / 'a crow' 146-2 / y n 2 2 / 'to f i n d f a u l t with' /w/ elsewhere: 4-17 j£ /wo 5 5/ 'nest' 67-5 ^ /waj 2 2/ 'to dread' 13 8-4 jfa /wa:t 2/ 'to dig out' c. Tong rhyme group: /$/ i n Grade I & II words: -** 22 215- 5 jjfi /ung / 'an earthen j a r 1 216- 7 f% /uk 5/ 'a room' / j / i n Grade III & IV words: 221-18 / j u k 5 / 'elegant' 224-11 / j u n g 1 3 / 'to crowd' In the case of exceptions to the above d i s t r i b u t i o n , some may have resulted from the merging of f i n a l s , and others from ana-l o g i c a l readings. For example, Grade I of the Kai-kou series of 234 -the Zhen x£ rhyme group has merged with those i n Grades III and IV, which may account for the presence of i n i t i a l / j / instead of /0/ i n the Grade I word, en %y (.148-10) 'grace' / j a n 5 5 / . (How-ever, an alternative proposal w i l l be offered l a t e r . ) The pronun-c i a t i o n of / i n g 5 5 / i n the Grade II Kai-kou words ying (195-15) 55 (which has an alternate reading of /ang / 'a parrot') and ying (19.5-16) 'cherry' i s probably due to the analogical reading of the Kai-kou word ying jjp (.201-2) 'an infant', which i s a Grade IV word. A comment should be added at t h i s point concerning the various grades i n the d i a l e c t survey. The Fang-yan Diao-cha Z i -biao often collapses certain grades which i n most d i a l e c t s no longer show any phonemic d i s t i n c t i o n . More w i l l be said about thi s subsequently (in the analysis of the grades per se), but for now, the primary implication of t h i s observation i s that many Grade IV Kai-kou words have been included under the category of Grade III words. One example i s that of ying S^. above, which, although i t i s l i s t e d as a Grade III word i n the d i a l e c t survey, i s i n fact from the Grade IV category. 13 Recall from Chapter 1.2 that the s y l l a b l e /ing / i n the r i s i n g tone i s accompanied by the Ij;.] on-glide. Two words i n the data are thus affected: ying ^hj (.198-15) 'shadow' and ying 13 (199-9) 'to r e f l e c t ' , both of which, are phonemically /ing /, but 13 phonetically I j nj ] . There are two exceptions to the d i s t r i b u t i o n of the Tong rhyme group; weng (213-20) 'an old man' / j u n g 5 5 / , and wo 5 (.218-3). 'to water' /juk /. According to the pattern of correspon-dences, these should have had the zero i n i t i a l instead of a - 235 -pa l a t a l onset. However, the pronunciation of an i n i t i a l g lide i n the word weng ^ , for instance, seems to be a p e c u l i a r i t y not only to Zhong-shan, but to the Yue d i a l e c t as a whole, and appears more-over to be r e s t r i c t e d to that d i a l e c t group (cf., e.g., Zi-hui, p.271; McCoy, 1966:82). The p a l a t a l glide preceding en (.148-55 10) /jan / mentioned e a r l i e r likewise appears to be exceptional 53 to the Yue di a l e c t s (e.g., Cantonese Ij/en ]; see also: Zi-hui, p.209). As suggested by Pulleyblank, i t may be the case that Cantonese and Zhong-shan underwent a sound change such that *#sn > *#j an (as i n the case of en fOr1 ) , *#orj > *#jorj (e.g., weng $ ), and the corresponding stop *#ok > *#jok (e.g., wo ) . (Apparently, wo jjL. i s also i r r e g u l a r i n i t s development i n Min (Norman, p.c.), and may be part of the same sound change observed i n Yue.) Cantonese, l i k e Zhong-shan, has [ j], [w] and the zero i n i -t i a l corresponding to the Ying i n i t i a l . What differences e x i s t between Zhong-shan and Cantonese with respect to the d i s t r i b u t i o n of these modern reflexes are only very s u p e r f i c i a l . B a s i c a l l y , where Zhong-shan has the zero i n i t i a l preceding high vowels, Can-tonese has a glide which i s homorganic with the following high vowel. Among the miscellaneous exceptions corresponding to the Ying i n i t i a l are: /k'/ (e.g., ye <%} (129-3) 'to v i s i t a superior' / k ' i t 2 / , / t s / (in ya | U (120-10) 'to crush' / t s a : t 2 / , /ng/ (e.g., y_i #f (190-22) 'or' /ngik 5/, and /m/ (in yao (or miao) ^ (80-12) ' 13 'obscure' /miw / ) . - 236 -The Yu i n i t i a l has, to a greater or lesser extent, merged with Grades III and IV of the Ying i n i t i a l . According to Pulleyblank (1970-71:227), the d i s t i n c t i o n s between i n i t i a l s Ying and Yu had by and large disappeared between the l a t t e r part of the Tang dynasty (618-907) and the Yuan dynasty (.1279-1368) . D i s t i n c -t i o n between the Yin- and Yang-ping tones, however, has been pre-served (e.g., i n Peking tones 1 and 3). I t should also be brought 13 rt to mind that the e n c l i t i c s y i ^ (58-16 / i / and yan T*j (.123-23) 51 / i n /, which represent words with the true zero i n i t i a l s i n Early Middle Chinese, are analyzed as Grade III Y u - i n i t i a l words i n Late Middle Chinese. The Late Middle Chinese Yu i n i t i a l i t s e l f arose out of two Early Middle Chinese i n i t i a l s (see Chapter 3.1.1): Yun -JR C*w-) and Y i (or Yang J|- , i n Pulleyblank' s reconstructions) (,*j-). The d i s t i n c t i o n of the two e a r l i e r i n i t i a l s i s maintained i n Late Middle Chinese insofar as they occur i n d i f f e r e n t grades: Yun i n Grade III rhymes, and Yi i n Grade IV rhymes. In the modern Zhong-shan d i a l e c t , there i s no longer a phonological d i s t i n c t i o n between them, save i n the He-kou series of the Xie ^ rhyme group, exem-p l i f i e d by the pair i n (47). (47) 46-6 / w a J 2 2 / 'to guard' (EMC *w-) 46-7 4-h / j 0 j 2 2 / 'a sharp-pointed weapon' (EMC *j-) Both wei 4fl a n d r u i $\J are Qu-sheng, He-kou words occurring i n the J i rhyme of the Xie rhyme group, d i f f e r i n g only i n that wei ffi originated from the EMC Yun ( > Yu III) i n i t i a l , and r u i 10^ from the EMC Y i (. > Yu IV) i n i t i a l . There i s , nonetheless, another difference between the two - 237 -Early Middle Chinese i n i t i a l s , and that i s i n t h e i r d i s t r i b u t i o n with respect to the f i n a l s . Only the Y i i n i t i a l , for example, occurs with the Xiao ^Uf^ , J i a j|jL , Shen and Tong rhyme groups. (There i s no rhyme group which takes only the Yun, and not the Y i , i n i t i a l . ) Within a given rhyme group, the d i s t r i b u -t i o n of the two i n i t i a l s may also d i f f e r * The Yun i n i t i a l , for instance, combines with the He-kou series of the Dang ^ rhyme group, while the Y i i n i t i a l combines with the Kai-kou portion of the- same rhyme group. Phonologically speaking, however, the differences c i t e d above do not a f f e c t the d i s t r i b u t i o n of /0/, / j / and /w/ corres-ponding to the two EMC i n i t i a l s , which became the Late Middle •': Chmese Yu *jj>} i n i t i a l . The d i s t r i b u t i o n of the correspondents of the Yu i n i t i a l i s as follows: i n the Kai-kou series, /0/ occurs with front vowels ( i . e . , [ i x ] , [ i j and [ y : ] ) , and / j / elsewhere; i n the He-kou s e r i e s — w i t h the exception of the Tong Atf) _ rhyme group, which takes i n i t i a l / j / — / 0 / occurs with tense high vowels (viz., Iu:J and [ y i ] ) , and /w/ elsewhere. Examples are given i n (48) . (48) a. Kai-kou Series: /0/ i n words with a high vowel: 76-20 i% / i w 5 1 / 'to shake' 105-6 jj^ / i p 2 / 'a l e a f 201-5 / i n g 5 1 / L . 'to win' / j / elsewhere: 89-2 fa /jaw 1 3/ 'to have' 173-4 % / j o n g 5 1 / 'a sheep' - 238 -201-5 jSJk / j a : n g 5 1 / C . 'to win' b. He-kou Series (excluding trie Tong rhyme group): /0/ i n words with a tense high vowel: 26-6 /y / 'rain' 67-11 I 22 A O / 'a class, a s e r i e s ' 141-6 IS /yn / 'round' /w/ elsewhere: 165-5 / 2 2 / /wan / 'to transport goods' 179-5 / 51 , /wong / 'a prince, king' 211-5 /wik 2/ 'a j a i l o r ' Tong rhyme group - / j / (.without exception) : 221-19 1 / j u k 2 / 'to nourish, to bring 223-8 * /Jung / 'appearance' 225-10 .. 22, /jung / 'to use' In addition to / j / occurring i n Tong-rhyme words, r e c a l l that i t also occurs i n the fourth grade of the Xie rhyme group which, i n the present Zhong-shan data, only involves the word r u i JjL* (in (.47).) . As regards to exceptions, there are three He-kou words i n the data from the Geng rhyme group which have the zero i n i t i a l instead of /w/. One i s from the EMC Yun i n i t i a l (Yu III) and the other two from the EMC Y i i n i t i a l (Yu IV). They are l i s t e d below. (49) 209-8 §5 / i n g 5 1 / 'glory' if i 51 210-7 "(^ /ing / 'an encampment' 210-8 ij} / i n g 5 1 / 'a grave' At present, the writer i s not e n t i r e l y c e r t a i n as to why these words i n (49) have no l a b i a l on-glide. I t may be that the i n t e r -- 239 -action between a cert a i n combination of segments and a given tone serves as a contributing factor. There may be an avoidance of the sequence /wing/ i n the Yang-ping /51/ tone i n Zhong-shan, for 51 instance, such that the s y l l a b l e /wing / does not occur i n the d i a l e c t . There i s not enough evidence to determine whether t h i s i s an h i s t o r i c a l accident or a phonological constraint i n the d i a l e c t . In Cantonese, a phonological d i s t i n c t i o n between the two EMC i n i t i a l s i s preserved i n the Geng rhyme group: [w] i s the regular correspondent of the Yun (Yu III) i n i t i a l , and [jj the regular correspondent of the Y i (Yu IV) i n i t i a l . Thus, rong '^p , containing the h i s t o r i c a l Yun i n i t i a l , has a l a b i a l i n i t i a l i n 21 modern Cantonese, and i s pronounced .[wig J, i n contrast to Zhong-shan / i n g 5 1 / for the word, as demonstrated i n (49) . Ying 't^ and ying , on the other hand, contain the EMC Y i i n i t i a l and, thus, have a p a l a t a l i n i t i a l i n Cantonese. The two words are pronounced 21 51 [j ii) ]. i n that d i a l e c t ; again, Zhong-shan has /ing / for these two words, also, as indicated i n (49). Observe from the Pin-yin that Mandarin, l i k e Cantonese, has preserved a phonological d i s - . t i n c t i o n between the EMC Yun and Y i i n i t i a l s : rong for , but ying for ^ and tyj . Given the phonological conditioning i n Cantonese and Man-darin for the three words i n (.49) , i t might be tempting to suggest the same conditioning to be i n e f f e c t i n Zhong-shan by a t t r i b u t i n g the presence of / j / i n rong to analogical readings. Such a proposal, however, would lead to other complications. If /w/ and / j / were to be considered the regular correspondents of the EMC Yun and Y i i n i t i a l s r e s p e c t i v e l y , i n the Geng rhyme group i n Zhong-- 240 -shan, there would be two more exceptions to be accounted for i n the d i a l e c t : both y i ^^L (211-4) 'pestilence' /wik 2/ and y i 2 v (211-5) 'a j a i l o r ' /wik / are Y i - i n i t i a l words i n the He-kou series of the Geng rhyme group, and yet have /w/ as the i n i t i a l 3 segment. (They are pronounced [j ik ] i n Cantonese, with [j] being the regular r e f l e x of the Y i i n i t i a l . ) The writer w i l l not attempt to draw any conclusions at t h i s time. Perhaps more i n f o r -mation at a l a t e r date may c l a r i f y certain points, and determine more conclusively the pattern of correspondences of the Yu i n i t i a l i n the Geng rhyme group i n the Zhong-shan d i a l e c t . A few exceptions to the regular correspondents to the Yu i n i t i a l w i l l now be discussed. The words xiong Ht C219-11) 'a bear' /hung 5 1/ and xiong (219-12) 'a male b i r d ' /hung 5 1/ both, have /h/ as the correspondent to the EMC Yun i n i t i a l . The i r r e -gular [h] i n i t i a l i n the two words i s also observed by Hashimoto (p. 6 46)i for Cantonese. Hashimoto, moreover, claims that the ex-ceptional correspondent i s suggestive of traces of an e a r l i e r form of the Yun i n i t i a l . She supports her argument with data from the northern d i a l e c t s , which also point to a f r i c a t i v e rather than a glide i n i t i a l . In fact, i f one accepts the Han-yu Fang-yin Zi-hui as a r e l i a b l e source for making generalizations about the pronun-c i a t i o n of words i n the Chinese d i a l e c t s as a whole, then one might simply state that a f r i c a t i v e pronunciation for the two words i.in question occurs i n an overwhelming majority of the Chinese d i a l e c t s today. That would lend even stronger support for arguing i n favour of an e a r l i e r form of the i n i t i a l as some sort of f r i c a t i v e or laryngeal before i t was l o s t . Evidence from rhyme di c t i o n a r i e s and rhyme tables, however, - 241 -suggests that the reverse i s i n fact the case. Our sources so far show that xiong , for example, was a Grade III Y u - i n i t i a l word, a categorization based on the Qie-yun rhyme dictionary of 601 A.D. By the time of the e a r l i e s t rhyme table extant, namely the Yun-jing -f|,^J[j^, xiong fc% had shi f t e d categories and had become a Grade III word with the i n i t i a l Xia Jj. . That would make the /h/ i n i t i a l the correspondent one would have expected i n the Zhong-shan pronunciation of the word. Presumably, the word xiong had also undergone the same category/sound change which i s ref l e c t e d i n the modern d i a l e c t s . A highly i r r e g u l a r r e f l e x of the EMC i n i t i a l i s / s / in. the f 51 (102-12) 'eaves of a house' /sim /. The same i r r e -21 gular r e f l e x i s found i n Cantonese ([siim ] ) , as well as i n Min (Norman, p.c.),e.g., Fu-zhou has [«sig]C, [«siei]]L., and Amoy has [ esiam] and [ t t s ' i ] . . Norman also informs the writer that a s i g -n i f i c a n t number of words i n Min have, i n i t i a l / s / (or other s i b i -lants) corresponding to the Yi i n i t i a l . As a consequence, he reconstructs the i n i t i a l i n proto-Min as *z- for these words. The /s/ i n i t i a l i n Yue for the word yan j£ i s therefore c l e a r l y another survival from the old Min-like substratum. On t h i s basis, i t i s also possible that yang (.173-11) 'to pretend' / t s ' i o n g 5 1 / , with i n i t i a l / t s ' / (also i n Cantonese) has a si m i l a r o r i g i n . 3.2. Finals 3.2.1. Reconstructed Values of Late Middle Chinese Finals In discussing Zhong-shan reflexes of Chinese f i n a l s based on the rhyme tables, there are several t r a d i t i o n a l categories which have already been introduced but w i l l now be further - 242 -elaborated. F i r s t of a l l , the f i n a l consists of a non-rhyming part (which does not a f f e c t rhyming pr a c t i c e s ) , and a rhyming part, Of the former, we w i l l f i r s t mention the two-way d i s t i n c t i o n c a l l e d hu , or 'voices': Kai-kou fif\ & ('open mouth') and He-kou 'a" & ('close mouth'). Chinese phonologists have agreed that the d i s -t i n c t i o n concerns l a b i a l i z a t i o n : Kai-kou pertains to the absence of l a b i a l i z a t i o n , and He-kou to i t s presence. Of greater controversy among the phonologists i s the i n t e r -pretation of the four-way d i s t i n c t i o n c a l l e d deng ^ , or 'grade, d i v i s i o n 1 . Karlgren, for instance, reconstructed the system of the four grades--yi — , er , san 5, / s i y£? , or I, I I , I I I , IV i n Roman numerals r e s p e c t i v e l y — p a r t l y i n terms of d i s t i n c t i o n s i n the medial segment and partly i n terms of differences i n the nuclear vowel i n order to account for his Ancient Chinese, the language of both the Qie-yun and the rhyme tables. The main c r i t i c i s m of his solution i s that no clear phonological c r i t e r i a could be stated from the o v e r a l l system of the grades. I t i s t h i s f a i l u r e to account for a systematic phonological contrast of the grades that motivated Pulleyblank (1970-71) to propose, for the rhyme table language (LMC), the d i s t i n c t i o n of the grades s t r i c t l y i n terms of phonological d i s t i n c t i o n s i n the medials. Although Pulleyblank has since made modifications to his 1970 theory of the grades, his basic premises remain the same. We w i l l f i r s t look at Pulleyblank's o r i g i n a l proposal. E s s e n t i a l l y , Pulleyblank constructs the system of the grades p a r a l l e l to the d i s t i n c t i o n of the Kai- and He-kou hu; that i s , whereas the l a t t e r involves l a b i a l i z a t i o n , the former concerns p a l a t a l i z a t i o n . Unlike the Kai-He d i s t i n c t i o n , the - 243 -contrast of the grades cannot be explained simply i n terms of a binary opposition of presence or absence of p a l a t a l i z a t i o n . Pulleyblank accomplishes a fourfold phonological system by f i r s t borrowing a d i s t i n c t i o n introduced by Karlgren of consonantal versus vocalic medials, namely / j / and /w/ versus / i / and /u/. Pulleyblank then divides the grades into two subsets, A and B, with A containing Grades I and I I , and B Grades III and IV. Sub-set A i s characterized by the absence of medial / i / and subset B by i t s presence. Further d i s t i n c t i o n s of the grades into each subset Pulleyblank relegates to differences i n the Middle Chinese i n i t i a l s , as summarized i n Chart 8 (from Pulleyblank (1970-71:231), which also takes Kai-He d i s t i n c t i o n s into consideration. Note that although the r e t r o f l e x glide - r - i s an element which i s fused to the r e t r o f l e x i n i t i a l s , since i t shares cert a i n c h a r a c t e r i s t i c s 7 with the other glides , - r - i s treated as a medial i n the chart. The following conventions are used to i d e n t i f y the various classes of LMC i n i t i a l s : K = velars and gutturals, T = dentals (excluding *1-) and dental s i b i l a n t s , Tr = retroflexes and r e t r o f l e x s i b i l a n t s (the l a t t e r series from EMC r e t r o f l e x s i b i l a n t s and p a l a t a l s , excluding * r - ) , P = b i l a b i a l s , F = d e n t i l a b i a l s (or labiodentals), and L = l a t e r a l s (*1- or * r - ) . (Chart 8 appears on the following page.) In Chart 8, note f i r s t of a l l that not every class of i n i t i a l s occurs with, every grade, as shown by the blank entries for Grades II and IV under the Kai-kou section of P (b i l a b i a l s ) and L ( l a t e r a l s ) . The chart also demonstrates the complementarity between b i l a b i a l s and labiodentals ( r e c a l l i n g that the l a t t e r had i n fact arisen out of the former): labiodentals occur s t r i c t l y i n - 244 -Chart 8. Di s t r i b u t i o n of the LMC Glides and Medials. •Subset Grade KAI--KOU HE--KOU K T P L K T P L .. A 1-0- I ' R- T- P- L- Ku- Tu- Pu- Lu-II Kj- Tr- P j - L- Kw- Trw-B - i - III K i - T r i - P i - L i Kiu- T r i u - *Fiu- L i u -IV K j i - T i - P j i - L i - K j i u - T i u -* Fiu> Fa- , Fius> Fus-, except that Fiu-j > F j i - j (From: Pulleyblank, 1970-71:231) Grade III He-kou rhymes, and b i l a b i a l s i n the remaining rhymes. What Chart 8 does obscure, however, i s the complementary d i s t r i b u t i o n of the dental s i b i l a n t s , EMC r e t r o f l e x s i b i l a n t s and the EMC palatals with respect to the grades. As mentioned i n the preceding section on i n i t i a l s , EMC r e t r o f l e x s i b i l a n t s occur i n Grade II only of the LMC r e t r o f l e x s i b i l a n t series, while the EMC palatals are found only i n the corresponding t h i r d grade of the same set of LMC i n i t i a l s . On the other hand, the dental s i b i l a n t s , although, a d i s t i n c t set of Middle Chinese i n i t i a l s , occur only i n Grades I and IV. The complementarity of these i n i t i a l s i s c l e a r l y one of the motivations behind combining certain grades i n the Fang-yan Diao-cha Zi-biao. Many of the so-called Grade III rhymes i n the survey l i s t , for example, are actually composed of rhymes i n Grades I I , III and IV, as witness the inclus i o n of the EMC retro-f l e x s i b i l a n t s , EMC palatals and the LMC dental s i b i l a n t s under the category of "Grade I I I " . In Zhong-shan, the collapsing of - 245 -certain rhymes i n these grades often r e f l e c t s t h e i r mergers i n the modern d i a l e c t . What we have thus far neglected to mention are the re t r o -f l e x and dental series. According to Pulleyblank (.1970-71:214), the r e t r o f l e x series had merged with the corresponding s i b i l a n t s by Southern Song (1127-1279), as we had mentioned previously, and i s found i n both Grade II and III rhymes. The dental series, kept d i s t i n c t from the two LMC r e t r o f l e x series i n most modern Chinese d i a l e c t s , occurs with Grade I and IV rhymes only, thus sharing the same d i s t r i b u t i o n with dental s i b i l a n t s , as indicated i n Chart 8. One might add as an aside that, i n the case of the Ri 0 (*r-).' i n i t i a l , i t i s not apparent from Chart 8 that t h i s i n i t i a l actually occurs i n Grade III Kai-kou rhymes only. I t i s the L a i j|L. C*l-) i n i t i a l which enjoys a wide d i s t r i b u t i o n . This point, however, i s only relevant to the study of the int e r a c t i o n between the Ri i n i t i a l and the various f i n a l s , serving to help explain certa i n observations, and i s not of d i r e c t import to the theory of the grade system per se. Turning now to Pulleyblank's r e v i s i o n of his theory of the grade system, the changes are i n fact quite s l i g h t , but the revisio n i s made easier to comprehend with the o r i g i n a l proposal having been presented f i r s t . To begin with, i n the l a t e r model (Pulleyblank, .1977:65) the sequence - i u - has been coalesced, forming the medial -y-, present i n Ky-, Kjy-, Try-, Ty- and Fy-. These were the e a r l i e r Kiu-, k j i u - , T r i u - , Tiu- and F i u - respectively. Since - i - and -y- are i n complementary d i s t r i b u t i o n with respect to Kai-He d i s t i n c t i o n s , the change consists of representing - i u - , - 246 -which i s actually treated l i k e a unit phoneme anyway, simply by a unitary front, rounded vowel, -y-Of greater t h e o r e t i c a l significance i s the withdrawal i n the l a t e r t r e a t i s e of the assumption that medial -u- characterizes Grade I, He-kou rhymes. In his r e v i s i o n , Pulleyblank (.1977:65). reconstructs both medials -u- and -w- i n t h i s category ( i . e . , Grade I, He-kou: Ku-/Kw-, Tu-/Tw-, Pu-/Pw-, Lu-/Lw-). This modi-f i c a t i o n serves to simplify the accounting of l a t e r developments of certain rhyme groups i n d i f f e r e n t d i a l e c t s . At the same time, Pulleyblank also puts forward the assumption that He-kou f i n a l s were actually d i s t r i b u t e d between Grades I and II by analogy with the corresponding Kai-kou f i n a l s , and were not independently moti-vated. Thus, the contrast between a v o c a l i c and a consonantal l a b i a l medial, according to Pulleyblank, was not noted as such i n the rhyme table analysis. (The revised formulation w i l l be shown l a t e r i n conjunction with the reconstruction of the rhyming portion of the f i n a l s i n Chart 10.) With regard to the rhyming part of the f i n a l , the various f i n a l s i n Late Middle Chinese are subcategorized into sixteen major di v i s i o n s c a l l e d she Jjfrlf. , or 'rhyme groups'. As mentioned e a r l i e r , the name of a given rhyme group i s derived from one of i t s members. The rhyme groups were i n fact not e x p l i c i t l y named u n t i l the Si-sheng Deng-zi \£? ^ -J~ , a rhyme table i n which mergers that had taken place i n Late Tang and Northern Song (Pulleyblank, 1970-71:236) were already indicated by the collapsing of certain rhyme groups. The fact that the rhyme table showed such mergers suggests quite strongly that the names of the sixteen she must have by then been established as t r a d i t i o n a l terms. - 247 -The sixteen rhyme groups are further divided into those which belong to the nei-zhuan (^) ^ (.'inner turn") and those to the wai-zhuan ^ (.'outer turn"). The d i v i s i o n of the she into the two zhuan by Pulleyblank (1970-71:232,336) do not cor-respond pr e c i s e l y to t h e i r d i s t r i b u t i o n to the Yun-jing rhyme table. His rationale for the r e d i s t r i b u t i o n of the f i n a l s under Nei- or Wai- i s based on his agreement with Chang-pei Luo, that the o r i g i n a l d i s t i n c t i o n of Nei and Wai concerns the r e l a t i v e closeness or openness of the nuclear vowel. Interpreting Nei to mean close vowels and Wai to mean open ones, Pulleyblank (.1970-71) reconstructs *-a- and *-a- for the nuclear vowel i n the respective zhuan i n his f i r s t set of reconstructions of Late Middle Chinese f i n a l s . As i n the case of the medials, i t i s thi s set of recon-structions that we w i l l present f i r s t , leaving aside for the time being Pulleyblank's revisions. To begin with, the sixteen rhyme groups, divided according to whether they belong to the Nei- or Wai-zhuan, are presented i n Chart 9 on the next page, together with Pulleyblank 1s reconstructed values for them. Observe also that the majority of the rhyme groups form pairs, with one member of the pa i r belonging to the Nei-zhuan and the other to the Wai-zhuan. There i s one major revisi o n Pulleyblank made i n his 1977 reconstruction of the rhyme groups. He recognizes a tense/lax d i s -t i n c t i o n of the low central vowel, a d i s t i n c t i o n which i s not found i n his e a r l i e r work. He posits for the low vowels i n his system: (a) lax a_, and (b) tense ai_ / a / a i / , and o_ /au/. The non-low vowels are: (a) lax a_, and (b) tense / s i / and 6 /au/. Not only does th i s tense/lax d i s t i n c t i o n a f f e c t the various rhyme - 24 8 -Chart 9. The 16 Rhyme Groups. WAI-ZHUAN NEI- ZHUAN I. I I . Guo J i a ^ ai I I I . Yu ji§. oi IV. Xie a J V. Zhi Jfc- oj VI. Xiao aw VII. L i u jftj aw v n i i : Dang % air) /aik IX. Zeng aiq'/aik X. Geng airj/aik • •• -XI. Jiang jjf- auq/auk XII. T o n g j ^ auij/auk XIII. Shan an/at XIV. Zhen $jjjr on/at XV. Xian am/ap XVI. Shen am/ap (From: Pulleyblank, 1970-71:236) groups, but i t also affects the grades within these rhyme groups, as shown i n Chart 10, which appears on the next four pages. In the chart, the f i n a l s are reconstructed taking into consideration rhyme groups, grades, Kai-He and Nei-Wai d i s t i n c t i o n s , and the various classes of LMC i n i t i a l s discussed e a r l i e r Csummarized i n Chart 8).. Labiodentals (F) (occurring only i n Grade III He-kou rhymes), however, are kept d i s t i n c t i n Chart 10. The order of the rhyme groups correspond to the order that they appear i n the di a l e c t survey. Aside from the observation that some grades have been collapsed i n the d i a l e c t , a comparison of Chart 10 with the d i a l e c t survey reveals other differences not yet mentioned. For instance, C h a r t 1 0 . L a t e M i d d l e C h i n e s e F i n a l s . K a i - K o u He-Kou Rhyme Gp./fer, P : F T L K P F T L K 1 . Guo I I I I a a a r i a i a a wa wa wa ya > a ya 2 . J i a I I I I I I V ax r a i ax j a i r i a : j i a i iax j i a i rwa: wax 3 . Yu I I I I I I I V •g r i a i a i a i a j i a ua ua ua ua rua ya> ua rya ya ya ya j y a 4 . X i e I I I I I I I V aj aj aj axj raxj jaxj r i a j i a j i a j j i a j i a j i a j j i a j uaj uaj uaj uaj rwaxj waxj yaj > j i a j ryaj yaj yaj yaj j y a j 5 . Z h i I I I I I I V r i i r i i i j i i C i ) * j i rwsi yj > j i r y j yj yj yj j y j - 250 -o « i X Em PH 3 rt 3 3 3 v. B rt A •p •P v. •P •v v.. C rt c c rt rt rt X 3 X •T—> •P — rt e rt rt X +-> 4-> rt" •p rt rt rt rt 3 rt rH :x 4^ rt • 4 rt §3 x s 3 o « I •H res ft 3 rt 3 •4 3 rt rt • n -H 3 3 3 •« ctf rt rt - H s rt • H rt r l - H 3 CD 3 CD 3 CD 3 •H PH e •4 rt PH rt - H • H -r-> 3 •H ^ * -CD *H 3 U H -H PH S *4 rt rt PH PH >» S CD rH PH s PH PH N s •H rH PH s •H 3 3 CD 3 •H PH PH PH PH V. e •H rt rt rt rt •p •H rt rt •< rt rt - H •p c rt r l rt rt rt 4-> c rt C rt •H rt rt •H c rt •H O / ^ H H H H H > H • OH 0 0 (0 •H <U X a >i • «§ vo H 3 -H H H H H H H H •H X oo H H > H H H 0 X ! CO CTl H H H > H H H X ! CO Kai-Kou .is He-Kou , Rhyme Gp./Gr P F T L K P F T L K 11. Zhen I II III IV in/t jin/t an/t ran/t rin/t in/t in/t an/t in/t jin/t an/t wan/t rwat yn/t,>un/t ryn/t yn/t wan/t yn/t wan/t yn/t jyn/t 12. Dang I II III IV aq/k iaq/k aq/k raq/k r iaq/k iaq/k aq/k iaq/k aq/k iaq/ k j iaq/k waq/k yaq/k > aq/k waq/k waq/k yaq/k 13. Jiang II orj/k k oq/k joq/k 14. Zeng I II III IV a q/k iq/k a q/k rak rig/k in/k aq/k iq/ k a q/k i<]/k jiq/k waq/k yk 15. Geng I II III IV a: jq/k iarj/ k jiajg/k aijq raijq/k riajq/k iajq/k aijq/ik iajq/k jaijq/k iajq/k j iajq/k ryajq/k yaj q/k waijq/k yajq/k jyajq/k Rhyme Gp./Gr Kai-Kou He-Kou . P F T L K P F T L K 16. Tong I oq/k oq/k oq/k oq/k II roq/k III ywg/k> ug/k riog/k ioq/ k ioq/k yog/ k 013/k ryoq/k yoq/k yoq/k IV ioq/k jioij/k yoq/k jyoq/k ^ O c c u r s w i t h s i b i l a n t s . i un to I - 253 -not a l l the categories i n the chart are included i n the d i a l e c t survey, since some of the categories have few words representing them, and these are often very obscure words (e.g., the rare occasion of "F+ a" i n Grade III Guo-she has no representation i n the d i a l e c t survey). Another, more important, observation concerning Chart 10 i s that Kai- and He-kou rhymes are i n complementary d i s t r i b u t i o n v i s - a - v i s LMC b i l a b i a l i n i t i a l s . Evidences from various stages of the language, from modern reflexes of these rhymes i n Chinese d i a l e c t s , and from Chinese loans i n other Asian languages, support-ed by the complementarity observed i n the rhyme tables, suggest that the b i l a b i a l i n i t i a l s must have been accompanied by l i p -rounding, and the po s i t i n g of a rhyme with a b i l a b i a l i n i t i a l as Kai or He was somewhat arbit r a r y i n the rhyme tables. Pulleyblank 1s reconstruction of non-labial f i n a l s i n some He-kou rhymes on the one hand, and l a b i a l f i n a l s i n some Kai-kou rhymes on the other, was therefore an attempt to r e f l e c t the development of the lan-guage rather than a b l i n d adherence to the Kai-He categories which contradicted the evidences found. With regard to the Yuj - J ^ rhyme group, the d i a l e c t survey only shows He-kou rhymes for t h i s she. The Yu rhyme was i n fact o r i g i n a l l y Kai-kou i n the Yun-jing; i t was l a t e r rhyme tables which r e c l a s s i f i e d the Yu rhyme as He-kou. Likewise, the d i a l e c t survey shows the Tong rhyme group as consisting of only He-kou rhymes, which, was not the case i n the Yun-jing. There, the Dong rhyme was categorized as Kai-kou, and the combined Dong and Zhong it rhymes as Kai-He, which Pulleyblank has interpreted as r e f e r r i n g to Kai-kou Dong and - 254 -He-kou Zhong ^J^_ . The Kai-He categories of the Tong 3J2_ rhyme group had merged by Southern Song times, with the r e s u l t that the Kai-He d i s t i n c t i o n has largely disappeared among the Chinese d i a l e c t s . Only i s o l a t e d cases of contrast have survived i n Chinese. In Wen-zhou, for example, the word long ^ (218-9). 'eminent 1, from the Dong ^ rhyme, i s [ l o g 3 1 ] (.< * l i o g ; Zhong-shan / l u n g 5 1 / ) . The word long (222-9) 'a dragon', d i f f e r i n g from long only i n that i t belongs to the Zhong rhyme, has the l i t e r a r y pro-31 nunciation of [log ] and, more sxgnxficantly, a c o l l o q u i a l one 31 51 of [ l i e ] (.< *lyog; Zhong-shan /lung /) (Zi-hui, p.263). In the case concerning the Xiao 5cX^ rhyme group, i t might be observed that although the rhyme group i s recorded here as Kai-kou only, agreeing with both the Yun-j ing and the d i a l e c t survey, Grade I f i n a l (Hao rhyme), with the LMC b i l a b i a l i n i t i a l , has been reconstructed by Pulleyblank as *-uaw. His reconstruction of a l a b i a l medial agrees with Shao Yong's eleventh century tables i n which this part of the f i n a l was categorized as He-kou. The.-Zhong Yuan Yin-yun-"-shows, a d i s t i n c t i o n between Grade I *puaw:;arid Grade II *paw (< paaw) although i t does not e x p l i c i t l y l a b e l Grade I as He-kou (Pulleyblank, p . c ) . Treating t h i s set as He-kou also serves to explain Kan'on glosses. Consequently, the above obser-vation of He-kou i n the environment given should be kept i n mind since, for the sake of s i m p l i c i t y , the present thesis w i l l treat the Xiao-she as containing only Kai-kou rhymes. Si m i l a r l y , the L i u ^ftj-she i s also treated i n the Yun-jing and the d i a l e c t survey as having only Kai-kou rhymes, although Pulleyblank reconstructs the f i n a l s with a labiodental i n i t i a l (which only occurs i n Grade III) as He-kou. Again, for simpliciaty, - 255 -the Liu-she w i l l be analyzed as containing only Kai-kou rhymes. I t i s the revised reconstruction i n Chart 10 that w i l l be used i n the following discussion of Zhong-shan correspondences of LMC f i n a l s . Before we actually begin that section, however, there i s one subdivision within each rhyme group found i n the rhyme tables and i n the survey l i s t which should be elaborated upon, and that i s the various yun -^ jfr , or 'rhymes', within a rhyme group. Each rhyme group i s composed of one or more rhymes. S t r i c t l y speaking, members of the same rhyme are also of the same tone. Nevertheless, for the sake of economy and greater potential for general remarks concerning rhymes d i f f e r i n g only with respect to tone, the rhyme i n the h i s t o r i c a l Ping-sheng i s used here to ; represent the set of rhymes d i f f e r i n g i n tone but otherwise i d e n t i -c a l . Recall that Ru-sheng words end i n a stop consonant, but are nonetheless treated i n the same set as t h e i r counterparts i n the Ping, Shang, and Qu tones. The Fang-yan Diao-cha Zi-biao i n fact gives both the Ping and Ru-sheng rhymes i n each set of rhymes which contains Ru-sheng f i n a l s . In the case of a set of rhymes lacking a Ping-sheng rhyme as one of i t s members, the rhyme containing the next t o n e — i n the conventional order of r e c i t i n g the tones i n Chinese, namely, "Ping, Shang, Qu, Ru"—serves as the name of that set of rhymes i n both the d i a l e c t survey and the present study. Chart 11 records the sixteen rhyme groups and breaks them down further into rhymes. Each rhyme further represents a set i n which the only s i g n i f i c a n t v a r i a t i o n i s that of tone. Whether these rhymes are Kai- or He-kou, or both, i s also noted. The four grades are indicated by the use of Roman numerals only. Those grades which are combined under - 256 -Chart 11. Rhymes and Grades Within Each Rhyme Group. Rhyme Group Kai-Kou 1. Guo Gefj-JL I Ge ^ III 2. J i a Ma j f i c I I ; H I (IV) 3. Yujjgj^ (Yu & (II) III (IV) - treated as He-kou i n the d i a l e c t survey) 4. Xie iM Tai ^ II Hai x>£ I; [III,IV] J i a y j i II J i e $ II Guai ^ II [II]; III (IV) III Qi ^ [ i n ] ; i v 5. Zhi jt. Zhi ^ (II) III (IV) Zhi ^ (II) III (IV) Zhi (II) III (IV) Wei III 6. Xiao 5J(^ Hao ^; I II Xiao III (IV) Xiao j£- IV 7. L i u j * u Hou/(£ I You £ (II) III (IV) You ^ IV He-Kou Ge ^ I, III Ma fa II Mu $ I Yu y| (II) III (IV) Tai J i a ^ j l II J i e $ II Guai II J i jg- [II]; III (.IV) F e i ^ III Qi % IV Zhi (II) III (IV) Zhi ))§ (II) III (IV) Wei 4jS8L 1 1 1 - 257 -Rhyme Group 8. X i a n ^ 9. S h e n ; $ ^ 10. Shan d\ 11. Zhen 12. Dang % Kai-Kou He'-Kou & Tan ^ I — Tan I I — Xian fa I I — Xian I I — Yan j f c , [ I I ] ; I I I CIV) — I I I Fan Tian J^. IV I I I Qin ( I I ) I I I (IV) — Han I — Shan jgfy I I Shan I I Shan JJJ I I Shan J | I I Xian [ I I ] ; I I I (IV) Xian [ I I ] Yuan ^ t i I I I Yuan fii I I I Xian $ J IV Xian jkj IV Hen ^ I Hun % % I Zhen & [ I I ] ; I I I (IV) [Zhen |L I I , (Zhen I I included under — Zhen |L I I I ) [Zhun^|_ I I , I I I , IV] Zhun (ii) : Yan ^ I I I Wen jC^ i n Tang £ I Tang J(r i Yang f| j (ii) I I I Civ) Yang f | i n I I I , IV] 13. Jiang ^ Jiang II 14. Zeng ^ 15. Geng ^ Deng I Zheng jg. (;il) III (jv) Geng $ [I]; II Geng jj^ I I ; III Qing ^ III (IV) Deng ^ I Zheng III Geng ^ II Geng ^ II; III Qing I l l CIV) - 2 5 8 -Rhyme Group Kai-Kou He-Kou ( . 1 5 . Geng) Qing ^ IV Qing ^ IV 1 6 . Tong^jjl^ Dong Dong ^ I; (II) III (IV) Zhong I l l (IV) ( ): merged with Grade III i n the d i a l e c t survey, [ ]: not recorded i n the d i a l e c t survey. another grade i n the d i a l e c t survey, but are kept d i s t i n c t accord-ing to h i s t o r i c a l sources, are enclosed i n round brackets "( )" to indicate that they do not actually appear i n the d i a l e c t survey as a separate category. Grades which have no character represen-tati o n i n the d i a l e c t survey are set o f f with square brackets " I ] Zhong-shan correspondences to Late Middle Chinese f i n a l s with respect to rhyme groups are summarized i n Chart 1 2 overleaf. The correspondences w i l l be discussed i n d e t a i l i n the following section. In Chart 1 2 , reflexes marked with asterisks (*) indicate that very l i t t l e actual data are present i n the given d i v i s i o n . Stop endings corresponding to nasal endings are implied. Segment /w/ i s treated as though i t i s a medial whether i t occurs as an i n i t i a l segment i n the modern Zhong-shan s y l l a b l e or as a medial following a velar stop. Medial / i / i n the chart i s / j / i n s y l l a b i i n i t i a l p o s i t i on. This i s s i g n i f i c a n t since, i n general, i t only occurs i n Grades III and IV Kai-kou rhymes. Segment / j / i s i n -cluded i n those d i v i s i o n s where i t occurs as an i n i t i a l segment - 259 -Chart 12. Zhong-shan Correspondences to the1 LMC F i n a l s . KAI-•KOU HE-KOU I II III IV I II I l l IV G u o o — 0* — (w) o — — J i a — a i a — (w) a • --Yu — o y u~ng u u Xie oj~a:j a: j aj uj (w) a: j 0 j ~ (w) aj Zhi -- i — 0j~(w) aj Xiao ow a: w iw --L i u a w (j) aw --Xian a: r t w o m arm im — a: n — Shen — am (j)am~ an --Shan a:n~on a: n i n un~yn (w) a: n yn-a: n Zhen a n (j)an u n ~ y n ~ (w) an 0n ~- (w) a n Dang ong iong (w)ong — (w)ong — Jiang — ong — --Zeng an g ing wang*, wa:k* — • ' . wa:k* --Geng a* ang-L f a:ng-C ing-L, i a :ng-C — (w)ang-L, (w)-a:ng-C (w)ing Tong un g (j)ung — (j)ung * Very l i t t l e actual data. - 260 -i n the modern Zhong-shan s y l l a b l e . Round brackets are used to indicate that the enclosed segment only occurs with cert a i n h i s -t o r i c a l i n i t i a l s . The symbol " ~ " i s used to indicate alternation of f i n a l s due to phonological conditioning within a given h i s t o r i c a l rhyme and with respect to certain series of Middle Chinese i n i t i a l s . The round brackets i n fact serve as an abbreviated version of alternations of f i n a l s (e.g., "(w)ong" i s a shortened alternative to "wong~ong"). 3.2.2. Modern Zhong-shan Correspondences to LMC F i n a l s . •. ; In analyzing the modern Zhong-shan f i n a l s as reflexes of those i n Late Middle Chinese, the l a b i a l and p a l a t a l glides i n the modern Zhong-shan s y l l a b l e are treated as elements of the f i n a l even though i n many cases they actually occur i n s y l l a b l e - i n i t i a l p o s ition i n the modern d i a l e c t . As we have observed, the grades play a role i n determining the nature of the LMC medial, as do Kai- and He-kou d i s t i n c t i o n s . The Zhong-shan l a b i a l and p a l a t a l segments, as reflexes of h i s t o r i c a l medials, are therefore r e l e -vant to the analysis of modern Zhong-shan correspondents to h i s -t o r i c a l f i n a l s , and should be included i n the discussion. More-over, reference to a p a r t i c u l a r grade i s based on rhyme table phonology/ and not on the mergers found i n the Fang-yan Diao-cha Zi-biao. The study i n t h i s section w i l l also focus on regular correspondences, and i n general w i l l only touch upon those excep-tions which may have h i s t o r i c a l implications. Thus, analogical readings, readings of a word.based on another h i s t o r i c a l phonolo-g i c a l category i n the rhyme tables, etc., w i l l not be included. - 261 -One quite obvious reason, of course, i s that often one would be dealing with the same exceptions mentioned i n the previous section on the i n i t i a l s . 3.2.2.1. Guo ^ -she In the discussion of the present and subsequent rhyme groups, the r e f l e x of the various grades i n the Kai-kou series w i l l be presented f i r s t , then those i n the He-kou serie s . Thus, the regular Zhong-shan correspondent of the Guo-she i s /o/ (.<. *-d) i n Grade I of the Kai-kou rhymes. Examples are shown i n (50) below. (.50) 1-15 / k o 5 5 / 'elder brother' 1-21 y\ /ho / 'a r i v e r ' 2-3 / t s o 1 3 / ' l e f t ' Cantonese likewise has ID:] corresponding to t h i s rhyme. There are only a couple of exceptions i n Zhong-shan. The l i t e r a r y word ta /\0 (1-3) 'he, she' / t ' a 5 5 / has f i n a l /a/ corres-ponding to t h i s rhyme. A couple of c o l l o q u i a l words have f i n a l / a : j / as the correspondent: (51) 1-12 / t s ' a : j 5 5 / 'to knead' 2-7 A a : j 2 2 / 'big' There i s also another reading for cuo /ts'o /, which has the meaning of 'to rub between the fingers', corresponding to the standard, dictionary d e f i n i t i o n . Although f i n a l / a : j / i n da may be based on the reading i n the Tai ^fc rhyme (.number 30-20) , i t i s possible that the / a : j / f i n a l i n the Guo-she i s a genuine r - 262 -survival of an e a r l i e r *-aj f i n a l . I t has been observed by Hashimoto (p.647), for example, that there are a number of words i n the Guo-she which have f i n a l [oi] or {ai] i n other Yue d i a l e c t s , which may suggest that l a i ] i s the c o l l o q u i a l correspondent of th i s f i n a l i n Yue. It may i n fact be the case that the glide f i n a l i s a v e s t i g i a l trace of an older layer of the language, since i t i s not r e s t r i c t e d to Yue, but occurs i n Min as well. Wo 33 22 (2-5) 'I, me', for example, i s Iijoj ] i n Tai-shan, Igoj ] i n Kai-ping, both Yue d i a l e c t s (Cheng, 1973:275; 1977 f i e l d n o t e s ) . 44 It i s {guai ] i n the c o l l o q u i a l layer of Fu-zhou, a Northeastern Min d i a l e c t (Zi-hui, p.31). (Another word i n the Zhong-shan data with, an o f f - g l i d e corresponding to the Guo-she i s the He-kou word, £o jML (4-18) 'lame' /paj 5 5/.) There i s only one Kai-kou word i n Grade III of the Guo-she 51 51 in the survey, namely qie fia (3-5) 'eggplant' /k'jzs" / (Ik 'cei ] ; 21 Lk'e: ] i n Cantonese), which Pulleyblank has reconstructed as *kfiia<EMC *gia. The Zhong-shan form represents the more regular development from the rare * - i a f i n a l as compared to i t s Cantonese counterpart. The reason i s explained by Pulleyblank to the writer as follows: by analogy with the f i n a l i n one should expect *kfiia to become *k'io > k'ce i n Cantonese, as i n jiang (172-15) 53 'boundary1 (Cantonese [kceig ] ) : *kioq > kiDq>koeq, while i n the Geng jjj^-she, *-iaajg gave *iaag > -eq. In Cantonese, the back vowel of * - i a had i n fact merged with the central vowel i n *-iaa i n Grade II I , Ma fjfc. rhyme, with the subsequent umlauting to *-e, ^ 35 as i n qie JL (.11-11) 'moreover' (LMC * t s ' i a a , Cantonese [ts'e: ] ) . Pulleyblank further suggests a si m i l a r development i n Mandarin may have occurred (or conceivably, the d i r e c t laxing of * - i a to - 2 6 3 -8 2 1 -V~ -ia) . Hence, while the Cantonese form of Ik'ex. J i n qie fle 5 1 shows Mandarin influence, the Zhong-shan form of Ik'ce: ] preserves the more d i r e c t derivation from *k'ia < *kfiia. Note that Min has /kio/ i n a number of i t s d i a l e c t s for qie jfc , thus representing the intermediary stage between the LMC and Zhong-shan forms. Not surpr i s i n g l y , Norman ( 1 9 6 9 : 2 4 9 ) reconstructs qie as Proto-Min *kio. In Zhong-shan, the Grade I He-kou of the Guo-she has /o/ (< *-a after LMC b i l a b i a l s , and *-wa elsewhere) as the re f l e x a f t e r a l l the Middle Chinese i n i t i a l s except the gutturals. After the h i s t o r i c a l gutturals, the f i n a l i n Zhong-shan i s /wo/. Examples of these two f i n a l s are given i n ( 5 1 ) . He fofi /wo51/ has the Xia i n i t i a l , and contrasts with the X i a - i n i t i a l word he y\ /ho 5 1/ 'river' i n ( 5 0 ) , which i s i n the Kai-kou series. 3 - 6 / 5 5 . /PO / 'waves' 4 - 1 3 /wo 5 1/ 'peace' 6 - 1 0 •A / i 2 2 , /ko / 'to cross 6 - 1 1 / f o 2 2 / 'a lesson Cantonese has the same set of reflexes as Zhong-shan, with a simi-l a r d i s t r i b u t i o n . In Cantonese, however, the f i n a l [war] also f o l -lows those h i s t o r i c a l velar i n i t i a l s the modern r e f l e x of which i s • ta. 4 4 [kj. Thus, guo ^ i s [kwoi ] i n Cantonese. The l a b i a l segment has been l o s t i n Zhong-shan, a point which had been raised much e a r l i e r (see Chapter 1 . 2 . 1 i n the discussion of nuclear vowels). Hashimoto (p. 6 4 7 ) also mentions the loss of the l a b i a l medial i n syl l a b l e s with If] as a re f l e x of the Qi (*k'-) i n i t i a l , as in ke -<J|L i n ( 5 1 ) above (.[fo: 4 4].in Cantonese), and postulates the -264 -f o l l o w i n g sound change to have o c c u r r e d i n Cantonese: *hw-y f - . In Zhong-shan, one would then expect the d e n t i l a b i a l i z a t i o n t o have taken p l a c e before the l o s s o f *-w- i n words such as guo above. In other words, the change of *hw->f- i s needed to d e r i v e / f / i n k e i l ^ /foZZ/. In Zhong-shan, the d e n t i l a b i a l i z a t i o n pro-cess must precede the l o s s o f l a b i a l *-w- i n words descending from LMC v e l a r s . I t i s the subsequent l o s s o f the l a b i a l medial t h a t 22 i«L y i e l d s Zhong-shan /ko / i n guo Thus, the d e n t i l a b i a l i z a t i o n r u l e bleeds the c o n d i t i o n f o r the a p p l i c a t i o n of the medial d e l e -t i o n i n Zhong-shan. There are s e v e r a l e xceptions t o /o/ and /wo/ as f i n a l s i n the He-kou Grade I of the Guo-she which are of h i s t o r i c a l i n t e r e s t . The word bo (4-18) 'lame' / p a j 5 5 / (/p'o 1 3/L.) has a l r e a d y been mentioned. Apparently, the p r o n u n c i a t i o n of [pi^jj i s widespread (Norman, p . c ) , and i s probably another example of v e s t i g e s of an o l d e r form of the language. The words l u o 4t C3-17) 'a mule' 51 & 51 /10 / and l u o C3-18) 'a conch' /ljz* / may a l s o c o n s t i t u t e t r a c e s of a pre-Middle Chinese * - j , as P u l l e y b l a n k (p.c.) has suggested f o r the e x c e p t i o n a l /jz5/ f i n a l i n these two words. Norman, f o r example, r e c o n s t r u c t s them as * l h o i f o r Proto-Min. The two "21 9 words are pronounced [ I D : ] i n Cantonese. With regard t o the t h i r d grade i n the He-kou s e r i e s o f the Guo rhyme group, the r e g u l a r r e f l e x i n Zhong-shan (and Cantonese) i s /0/ (.< * - y a ) . Very few words occur i n t h i s p h o n o l o g i c a l c a t e -gory. There i s on l y one word i n the Zhong-shan (and Cantonese) data f o r which a p r o n u n c i a t i o n has been obtained. That word i s xue $0 (6-17) 'boots', / h j ^ 5 5 / i n Zhong-shan, and [hoe: 5 3] i n Cantonese, from LMC *xya, which i s i n t u r n descended from EMC *xua. - 265 -3.2.2.2. J i a she The Jia-she contains only the Ma Jjfa. rhyme. In Grade II Kai-kou, the r e f l e x of the Ma rhyme i s /a/ (< * _ a) i n Zhong-shan (and Cantonese) , as exemplified by (.52) . (52) 7-8 lik /ma / 1 hemp' 9-1 ML /ka / •false' 9-13 46 22 /p ' a 2 2 / 'to fear' Grades III and IV Kai-kou of the Ma rhyme have the r e f l e x of / i a / ( < *-iax.) as the regular correspondent i n Zhong-shan. After the h i s t o r i c a l Ri (EMC p a l a t a l **>-) and Yu IV (EMC *j-) i n i t i a l s , the Zhong-shan s y l l a b l e i s / j a / , which^should not be considered an exception. Phonemically, the s y l l a b l e / j a / can equally be represented as / i a / due to the complementarity of f i n a l / i a / and s y l l a b l e / j a / . Consequently, for s i m p l i c i t y ' s sake, only the f i n a l / i a / i s represented i n Chart 12, implying s y l l a b l e / j a / i f no consonantal i n i t i a l precedes / i a / . Cantonese has [ex.] and [jex] corresponding to Zhong-shan / i a / ([ea:]) and / j a / ( [ J A X ] ) respectively. Zhong-shan examples are presented below. 'a chariot' (53) 11-2 ^ / t s ' i a b V 11- 16 4i. / s i a 1 3 / 'god of the s o i l ' 12- 14 /(L / j a 2 2 / 'night' P r a c t i c a l l y a l l the words i n the He-kou series of the Jia-she are words descending from h i s t o r i c a l velar and guttural i n i t i a l s . The regular Zhong-shan r e f l e x i s /wa/ (< *-wax), with two sets of exceptions. After the velar nasal (which descends from the Y i i n i t i a l ) , the correspondent i s /a/, with l a b i a l i -- 266 -zation l o s t i n Zhong-shan. As a general rule, Zhong-shan loses i t s l a b i a l /w/ following velar nasal /ng/, but s e l e c t i v e l y pre-serves i t after velar stops i n f i n a l s containing a low vowel i n the modern d i a l e c t . (.Contrast He-kou Jia-she, for instance, with the corresponding He-kou series i n the Guo-she, where l a b i a l i z a -t ion after a l l velars has been lost.) The only exception to th i s general rule i s words with the LMC Qi and Xiao i n i t i a l s , where the re f l e x of these two i n i t i a l s i s / f / . In such cases, the f i n a l i s simply /a/ i n Zhong-shan, the l a b i a l segment having been taken care of by the process of *hw-> f. Cantonese has the same corres-pondents to this rhyme as Zhong-shan. Examples to the pattern of correspondences described above are given i n (54) for Zhong-shan. (54) a. /a/ a f t e r /ng/ and / f / : 13-13 & /nga 1 3/ ' t i l e ' 22 13-16 /J£J / f a / 'to transform' b. /wa/ elsewhere: 5 5 12- 16 ^ /kwa / 'melon' 13- 2 % /wa 5 1/ 'flowers' 3.2.2.3. Yu laJ- -she In the d i a l e c t survey, i t appears as though Yu-she only occurs i n He-kou. As noted e a r l i e r , the Yu rhyme, i n which Grades II, III and IV have been combined under Grade III i n the d i a l e c t survey, was i n fact o r i g i n a l l y Kai-kou. Zhong-shan cor-respondents to Grades III and IV (*-ia) of the Yu '$L rhyme have merged with the corresponding grades of the Yu rhyme (*-ya), which i s He-kou. Only Grade II of the Yu J ? » rhyme remains - 267 -d i s t i n c t . As a r e s u l t of the merger i n Grades III and IV, these two grades of the Yu &s rhyme w i l l be discussed i n conjunction with the He-kou serie s . In the case of Grade II, Yu &w rhyme, which, contains only words descending from EMC r e t r o f l e x s i b i l a n t s , i t has only /o/ (..< **-)' as the regular r e f l e x i n Zhong-shan, as exemplified by the words i n (55). Cantonese likewise has [DI] as the r e f l e x of t h i s grade of the Yu rhyme. (.55). 19-8 /^j / t s ' o 5 5 / 'the beginning' 19-9 / t s ' o 5 1 / »a hoe'; / t s o 5 1 / 'to hoe' 20-20 / s o 1 3 / 'that which' With regard to He-kou rhymes i n the Yu-she, only the Mu rhyme has Grade I f i n a l s . The regular Zhong-shan correspondent of t h i s rhyme, which moreover occurs i n Grade I only, i s /u/ (< *-ua). . The p r i n c i p a l exception i s Y i - i n i t i a l (*o-) words, which have a s y l l a b i c velar nasal as the entire s y l l a b l e . Examples of the Grade I He-kou rhyme are given below. (.56) a. S y l l a b i c /ng/ i n words descending from Y i - i n i t i a l : 16-25 Ju / n g 1 3 / 'f i v e ' 18-7 \k / n g 2 2 / 'to r e a l i z e ' b. F i n a l /u/ elsewhere; 14- 9 / t u 5 5 / 'metropolis' 15- 23 fjfi /pu 1 3/ 'to mend' 16- 16 ii / k u 1 3 / 'ancient' Cantonese i s s i m i l a r to Zhong-shan with respect to Y x - i n i t i a l words i n t h i s grade. For the remaining f i n a l s , however, there i s a phonological conditioning involved i n Cantonese: [u:J a f t e r - 268 -h i s t o r i c a l gutturals and velars (excepting Y i *q-) , and Low] else-. 35 where. Thus, l i k e Zhong-shan, gu "5 i n (56) i s Iku: ] m Can-i . 35 tonese, with f i n a l [u:]. Bu ffi , on the other hand, i s [pow ] 13 i n Cantonese, i n contrast to Zhong-shan [pui ]. There i s one exception to the Mu rhyme which may be worth mentioning. The word mo 1.14-8) 'to f e e l for with f i n -55 13 51 gers 1 i s c o l l o q u i a l l y /mo / i n Zhong-shan, with /mo / and /mo / 35 l i t e r a r y counterparts. Tonewise, Cantonese has [HID: J for the pronunciation of t h i s word, and may be the source of the r i s i n g tone i n one of the two l i t e r a r y readings i n Zhong-shan. The other l i t e r a r y form with tone /51/ f i t s into the regular correspondence of a Yang-ping tone for the word i n question. While we have just accounted for the various tones of the word, the f i n a l /o/ instead of the regular /u/ has not yet been discussed. It appears possible that the /o/ f i n a l i s a vestige of an e a r l i e r form. Pulleyblank (.1977:101), for example, has reconstructed rhyme Mu 0. as EMC *-o which resulted from the rounding of Old Chinese *-d during the Later Han dynasty (25-220 A.D.). A f i n a l [o] (or [uo]) i s also found i n a number of other Chinese d i a l e c t s , including Peking Mandarin, Mei-xian, Xia-men, Fu-zhou, etc. (Zi-hui, p.23). Another word with f i n a l /o/ i n the Mu rhyme i n Zhong-shan i s cuo (18-4) 'wrong'. As i n the case o f the word mo^^ , cuo $3 likewise has f i n a l [o] or [uo] i n a large portion of the d i a l e c t s represented i n the Zi-hui (p.27), and may represent v e s t i g i a l traces of an older layer of the language. The Yu ^ rhyme of the Yu-she has Grades I I , III and IV even though only Grade III i s indicated per se i n the d i a l e c t survey. Grade II i s r e s t r i c t e d to words descending from EMC - 269 -r e t r o f l e x s i b i l a n t s , with /u/ (< *-ua) as the regular correspon-dent i n t h i s grade i n Zhong-shan, thus contrasting with Grade II Kai-kou Yu J> rhyme, which has r e f l e x /o/ i n Zhong-shan. Four words belonging .to t h i s grade appear ; i n the survey, only two of which exhibit the regular Zhong-shan correspondence. The other two words have /o/ as the f i n a l . A l l four words are l i s t e d i n (.57) . Cantonese has [owj corresponding to Zhong-shan /u/ i n this f i n a l . The i r r e g u l a r i n chu || and chu also occurs i n Cantonese. (57) a. /u/ as the regular f i n a l : 25-23 | ^ / s u 1 3 / 'to count' 27-4 | ^ / s u 2 2 / 'a number' b. /o/ as the i r r e g u l a r f i n a l : 5 5 23-23 U /ts'o / 'to cut grass or hay' 23-24 / t s ' o 5 5 / 'a chick, a f l e d g l i n g ' As noted e a r l i e r , Grades III and IV of the Yu %L rhyme have merged with the He-kou rhyme, Yu/Jl , i n the corresponding grades i n Zhong-shan. The regular Zhong-shan correspondents to the two rhymes i n Grades III and IV are: /u/ (<. *-ua< *-ya) aft e r the h i s t o r i c a l labiodentals, and /y/ (< *-ya) elsewhere, as exem-p l i f i e d by (58). . (The b i l a b i a l stop i n fu ^ marks the r e f l e x as a survival from a pre-LMC layer, as noted i n the section on i n i t i a l s . ) (58) a. /u/ a f t e r LMC Labiodentals: 23-1 /hu 5 5/ 'a sage' 13 25- 6 ^ /pu / 'an axe' 26- 18 <H /mu22/ 'fog' - 270 -b. /y/ elsewhere: 19-4 ft / t s ' y 5 1 / 'di g n i f i e d ' 19-13 / 55 , /sy / ' a l l ' 19-14 •i / 55, /sy / 'a book' 24-11 & A ' y 5 5 / 'a d i s t r i c t 25-22 / t s ' y 1 3 / 'a p i l l a r ' 27-15 /ngy 2 2/ 'to meet' 27-19 / 2 2 / /y / 'a parable' The set of correspondences i n Cantonese i s somewhat more compli-cated: [u:J occurs a f t e r descendants of the h i s t o r i c a l labiodentals, except aft e r the Wei (_*v- <, EMC *m ) i n i t i a l where f i n a l [ui] had & 33 diphthongized to low] (e.g., wu *£. i n (.58) i s [mow ] i n Canton-ese). In a p a r a l l e l case, lyi] occurs i n the remaining f i n a l s except for the diphthongization of [y;] to [0q] a f t e r words with i n i t i a l s descending from LMC dental s i b i l a n t s and velars. Thus, xu ^ (Xie *sh- i n i t i a l ) , xu ^ (Xin *s- i n i t i a l ) , and qu |£ (Qi *k'- i n i t i a l ) i n (58) are [ t s < 0 q 2 1 ] , [ s 0 q 5 3 J and [k'0q 5 3] respectively i n Cantonese. The rest of the words i n (5 8) have f i n a l {yij i n Cantonese, as i n Zhong-shan. The sole exception to the above description of Cantonese f i n a l {0uj aft e r LMC velars concerns the Yx (*rj-) i n i t i a l : Canton-ese s y l l a b l e s descending from the Yx i n i t i a l i n t h i s set of rhymes are pronounced [qyt] , with a homorganic, rounded on-glide; for example, the Cantonese pronunciation of yu Vjlo, i n (58) above i s 33 A-[qyx ] , and i s homophonous with yu TF^.J , the i n x t i a l of whxch tr ... descends from the Yu xnxtxal. It should be noted that s t r i c t l y speaking, there are no LMC d e n t i l a b i a l s i n the Yu J?„ rhyme. Consequently, the only correspondent of Grades III and IV Yu ^ i s simply /y/, as shown in Chart 1 2 . 3 . 2 . 2 . 4 . Xie 1%, -she The Xie-she has a number of rhymes, i n both Kai- and He-kou, and i n a l l grades. Discussing Kai-kou series f i r s t , there are two rhymes which occur i n Grade I: Hai u a and Tai ^ . The b i l a b i a l series i n the Tai ^ rhyme w i l l be treated separately l a t e r . Of the remaining words i n the Hai and Tai rhymes, Zhong-shan has / o j / C < *-aj < EMC *-oj ) as the regular correspondent for both rhymes with the exception of words i n the Tai rhyme with LMC dentals. In the l a t t e r case, the correspondent i s / a : j / . Example are presented below. (.59). a. / a : j / aft e r LMC dentals i n the Tai rhyme; 30-17 $p / t a : j 2 2 / 'a belt* 2 2 30-18 / t ' a : j / 'excessive 1 30-22 jj / l a ; j 2 2 / 'to rely on' b. / o j / elsewhere; 28-7 ^ / l o j 5 1 / 'to come' 29-4 i / t ' o j 1 3 / ' i d l e ' 31-6 % / h o j 2 2 / 'to injure' The same correspondences are found i n Cantonese. Hashi-moto (p.650) seeks to explain the |> J-j ] •—I>:j] alternation i n Cantonese i n terms of an incomplete merger of the two rhymes, and suggests that the Hai and Tai rhymes may o r i g i n a l l y have been [ a i j ] and J>:j] respectively. Pulleyblank (1977:100) , on the other hand, explains the - 2 7 2 -observation i n terras of the merger of d e n t a l - i n i t i a l words i n Grade I with Grade II words, which had f i n a l * - a : j . Pulleyblank points out that the lengthening of * - a to * - a i . a f t e r dental i n i t i a l s was a general feature of Late Middle Chinese i n Northern Song ( . 9 6 0 - 1 1 2 7 A.D.), and suggests that i t may have been the case that the vowel * - a - lengthened to * - a i - after dental i n i t i a l s , before the lowering of * - 9 J to *-aj i n Grade I i n Cantonese, thereby establishing minimal pairs containing a dental i n i t i a l and an *-aj versus * - a : j f i n a l . Two points are implied i n th i s proposal: f i r s t l y , that must have lowered to * - a - after * a had length-ened to.:*a.i a f t e r dentals; and secondly, that the Hai and Tai rhymes must have d i f f e r e d i n t h e i r course of sound change v i s - a -v i s d e n t a l - i n i t i a l words i n order for there to emerge a new opposition such as * t a j ^ t a : j , * l a j ^ l a : j , etc. That i s , i f a l l Grade I f i n a l s i n t h i s Xie-she had become * - a : j a fter dental i n i t i a l s , there would not have been any s y l l a b l e s such as *taj and * l a j , which l a t e r became *toj and respectively, when *-aj became * - o j . What appears to be suggested for Cantonese i s that d e n t a l - i n i t i a l words i n the Tai rhyme merged with Grade II * - a : j f i n a l s , while the corresponding d e n t a l - i n i t i a l words i n the Hai rhyme did not undergo vowel lengthening, but l a t e r underwent a sound s h i f t from *-aj to *-oj. A similar solution would also account for the Zhong-shan data. The Tai rhyme afte r the LMC b i l a b i a l i n i t i a l s has f i n a l / u j / i n Zhong-shan (and Cantonese). Although t h i s set of words—two of which are recorded for Zhong-shan, l i s t e d i n ( . 6 0 ) — i s recorded as Kai-kou i n the d i a l e c t survey, the Yun-jing and other sources, such words have also been reconstructed as He-kou by Karlgren (Anc. Ch. *-wdi< Arc. Ch. *-.wAd) , Pulleyblank (LMC * - u a j ) , and others. (60) 3 0 - 1 4 £l /puj / 'cowries (was used as currency) ' 3 0 - 1 5 jij? /p'uj / 'copious' Fi n a l / u j / (< * - u a j ) i s also the r e f l e x of the Tai rhyme i n what i s ostensibly regarded as Grade I, He-kou Tai rhyme i n the survey. The treatment of t h i s set of b i l a b i a l descendants as He-kou would also be congruent with the Zhong-shan data. It might be added that i n other rhymes as well, the positing of words descending from LMC b i l a b i a l i n i t i a l s as Kai-kou or He-kou sometimes seems to contradict evidences from the modern d i a l e c t s . The a r b i t r a r y assignment of Kai or He stems from the nature of the i n i t i a l s involved. Pulleyblank, for instance, regards the b i l a b i a l i n i t i a l s as having had simultaneous lip-rounding. I t i s this l a b i a l feature which became a separate phoneme i n some cases. It appears p a r t i -c u l a r l y anomalous among Kai-kou rhymes where one does not expect to f i n d a l a b i a l medial or f i n a l , as i n the case of the two words i n ( .60 ) , which were treated as Grade I, Kai-kou i n the Tai rhyme. Of course, the converse may also be true; that i s , one may f i n d , as i n the Guo-she, that the He-kou f i n a l of Grade I words with b i l a b i a l s i s reconstructed as LMC *-a, i n contrast to *-wa for words with other i n i t i a l s i n t h i s rhyme group (see Chart 10 on LMC f i n a l s ) . The various Grade II Kai-kou rhymes i n the Xie-she have a l l merged i n Zhong-shan, with the f i n a l / a : j / ('.< * - a i j ) i n the d i a l e c t for th i s set of f i n a l s , as shown i n ( 6 1 ). Cantonese l i k e -wise has f i n a l I A L J ] , (61) 31-9 ift / P ' a : j 5 1 / 'a row, a l i n e ' 33-17 $ J / k a : j 5 5 / 'a street' 35-6 jf^ / t s a : j / 'a stockade' The Grade III and IV Kai-kou rhymes of the Xie rhyme group have also merged i n the Zhong-shan d i a l e c t : / a j / C< * - i a j ) i s the Zhong-shan correspondent to the J i and Qi rhymes i n Grades III and IV Kai-kou. (The Qi rhyme only occurs i n Grade Grade IV of the J i rhyme has been combined under the Grade III category i n the d i a l e c t survey.) Examples from the two rhymes are given below. (Cantonese likewise has I B J ] as the correspondent.) (62) 37-3 /maj / 'to deceive 38-10 /sa: / 'to wash' 39-1 .22, /tau / 'a r u l e r ' The present data have no pronunciation for y i (36-15) 'to mow', the only word i n the survey representing the Grade III rhyme, F e i jjj^ . However, y i V'J i s recorded i n Chao as / n g a : j 2 2 / , 33 agreeing with Cantonese [OA:J ] for the word. There i s i n s u f f i -cient evidence to determine whether f i n a l :/a:j/ i n the F e i ^ ^ rhyme has preserved a d i s t i n c t i o n from Grade III Qi rhyme, which has / a j / as i t s regular correspondent i n Zhong-shan. In Zhong-shan, the regular correspondent of Grade I He-kou i n the Xie-she i s / u j / (.*• *-uaj), as mentioned i n connection with LMC l a b i a l - i n i t i a l words descending from the Tai rhyme. There i s no d i s t i n c t i o n between the two Grade I rhymes, Tai ^ and Hui fc^ . Examples are given i n (63). (63) 40-16 '§ / l u j 5 1 / 'thunder' - 2 7 5 -4 1 - 1 A 5 5 / l u r T V 'ashes 1 4 1 - 9 /nu: / 'hungry' 4 2 - 8 /muj 2 2/ "a younger s i s t e r 4 2 - 1 7 a / - 2 2 , /su] / 'fragments' In Cantonese, the f i n a l i s [0x(\ a f t e r dentals and dental s i b i l a n t s , and l u i j ] elsewhere. Thus, l e i ^ , nei fflj- and sui i n ( 6 3 ) above have f i n a l I>q] i n Cantonese, whereas the remaining two examples i n ( . 6 3 ) . have f i n a l I u i j J , as i n Zhong-shan. A couple of Zhong-shan exceptions are nei l^) 1 4 2 - 1 4 ) 2 2 1 2 2 'inside' /noj / and wai #r ( . 4 3 - 7 ) 'outside' /ngoj /, which have f i n a l / o j / corresponding to Grade I He-kou series of the Xie-she. No s y l l a b l e /nguj/ exists i n Zhong-shan. I t would appear that Kai-and He-kou words i n Grade I of the Xie-she have coalesced with respect to the velar nasal i n i t i a l . Cantonese likewise has f i n a l [oij] for the two words, nei |^) and wai . Each of the Grade II He-kou rhymes of the Xie-she i s represented by only a handful of words. As a r e s u l t , the picture which i s formed must take that into consideration. In the J i e & rhyme, the majority of the words i n Zhong-shan have /wa:j/ (. ( *wa:j) as the correspondent, as exemplified i n (.6 4 ) . ( 6 4 ) 4 3 - 1 1 ?jc /kwa:j 5 5/ 'good (of child) ' 4 4 - 2 f £ /kwa.j 2 2/ 'strange' 4 4 - 5 /wa:j 2 2/ 'bad, s p o i l t ' Three words have f i n a l /waj/, perhaps r e f l e c t i n g a pre-LMC layer (Pulleyblank, p . c ) : ( 6 5 ) 4 3 - 1 2 1^1 /waj 5 1/ 'bosom' 43-13 /waj 5 1/ 'kind of tree' 43-14 ^ j i /waj 5 1/ 'name of a r i v e r ' The J i a rhyme has /wa:j/ as the Zhong-ishan f i n a l i n Ping-and Shang-sheng, but /wa/ i n Qu-sheng. Thus, wai ^ (44-6) 'aslant', and guai til (.44-8) 'an old man's s t a f f are /wa:j 5 5/ 13 -t\ and /kwa:j / respectively, but gua JJ- (.45-2) 'to divine' i s 22 /kwa /, although both Zhong-shan f i n a l s descend from LMC *-wa:j. The / a : j / ^ / a / alternation noted here i s also found i n Mandarin (.as observable i n the Pin-yin romanization above) and other dia-o. I'ects. Pulleyblank (1977:120) suggests that the sporadic loss of * - j i n the Jia.rhyme must have taken place quite early, as e v i -denced i n some of the rhyming found i n Tang poetry (.618-907) as well as i n other sources - c i t e d by Pulleyblank. The Guai rhyme has three words in the d i a l e c t survey, only two of which are found i n the rhyme tables. These two are: kuai (45-4) 'fast' / f a : j 2 2 / and hua -f4 (45-6) ' to speak' 22 /wa /; again, both are descendants of LMC *-wa:j. Recall that *k- > *h-, and *hw-> f. The absence of an overt l a b i a l segment i n the f i n a l of kuai "fA i s therefore not s i g n i f i c a n t since the l a b i a l medial has been absorbed into the i n i t i a l . The observation made above for the loss of * - j i n some words from the J i a ^ rhyme i s also applicable to hua . The He-kou rhymes of Grades III and IV Xie-she have / a j / (< * - j i a j < *-yaj) a f t e r / f / descending from labiodentals, /waj/ ( 4 *-yaj) af t e r LMC velars and gutturals, and / 0 j / (.<- *-yaj) else-where ( i . e . , after LMC dental s i b i l a n t s and retroflexes) i n Zhong-shan, as shown i n (66). Cantonese has the same pattern of corres-pondence i n t h i s set of rhymes. - 277 -(.66) a. / a j / after / f / from LMC labiodentals; 46-8 jjfo / f a j 2 2 / 'to abrogate 1 46-9 *jC / f a j 2 2 / 'to bark' b. /waj/ afte r LMC velars and gutturals: 46- 12 £ /kwaj 5 5/ 'a jade tablet or baton* 4 7 - 4 % /waj 2 2/ 'favour' c. /<&j/ elsewhere : 45- 8 |fc / s j r f j 2 2 / 'age' i* . 22 46- 1 /tsp'j / 'to repeat' Observe that while the Grade I f i n a l / u j / i s d i s t i n c t from the Grade IV f i n a l /©'j/ afte r LMC dental s i b i l a n t s i n Zhong-shan, as A 2 2 j£ 2 2 exemplified by sui % ^ /suj / i n (63) versus s u i a? /sjzJj / i n 44 C66) above, the two f i n a l s have merged i n Cantonese, with [sou. ] the pronunciation of. both words. 3.2.2.5. Zhi -she The rhymes of the Zhi-she occur i n a l l but the f i r s t grade. In the d i a l e c t survey, Grades II and IV of both Kai-kou and He-kou series of the Zhi JL and Zhi fi% rhymes have been com-bined under t h e i r respective Grade III rhymes. The Zhi iL- rhyme i s Kai-kou only. Grades II and IV of that rhyme are likewise i n -corporated under i t s Grade III rhyme i n the survey l i s t . The Wei fyjbL rhyme only has Grade I I I , but does occur i n both Kai- and He-kou seri e s . The Kai-kou rhymes have a l l merged i n Zhong-shan. Even grade d i s t i n c t i o n s have, disappeared. The Zhong-shan correspondent to Kai-kou Zhi-she i s / i / (. <. * - i a f t e r Grade III s i b i l a n t s , * - i elsewhere) i r r e s p e c t i v e of grade or rhyme. Examples are given - 278 -(67) below, with rhyme and grade included i n brackets. 48-6 / t s i 5 5 / 'to know1 I £ H I ) 49-2 n / i 5 1 / 'to move across' (. 1 IV) 54-12 A i 2 2 / 'ground, earth' C %\ iv) 57-20 / s i 1 3 / 'to employ' (. 2- i i ) . 58-6 / s i 1 3 / 'a market' t. 2- i n ) 60-1 * / h i 5 5 / 'rare' ( III) One i n t e r e s t i n g exception concerns the Grade IV word, b i & 22 <£ (54-8) 'nose'. In Zhong-shan, the word i s pronounced / p i / 2 when i t refers to 'nose' i n general, but /pat / when i t refers s p e c i f i c a l l y to the 'trunk' of an elephant. The form with the f i n a l Ru-sheng ending bearing the specia l i z e d meaning i s also found i n other Chinese d i a l e c t s (.cf. Hashimoto, p. 6 52; Zi-hui, p.52). Pulleyblank informs the writer that the Mandarin reading of b i implies a Ru-sheng reading (LMC *pfijit< EMC * b j i t ) that i s not attested i n the Guang-yun. Pulleyblank (1973:372) stated that 'nose' was EMC *bj i x ( < b j i s ) from which the current pronunciation 22 of various d i a l e c t s (including Zhong-shan / p i /) i s descended; that i s , the Qu-sheng i n EMC *bji developed out of the loss of f i n a l *-s. Zhong-shan tone /22/, for example, i s the regular r e f l e x of the h i s t o r i c a l Qu-sheng. Pulleyblank postulates that i n the northwestern d i a l e c t s , there was a d i a l e c t a l s h i f t of *-s >*-t. I t i s the borrowing of * b j i t (<*bjis) into the standard c o l l o q u i a l of Tang, supplanting the o r i g i n a l Qu-sheng reading, that would account for the r e f l e x of an implied Ru-sheng reading i n modern Mandarin b i . In Zhong-shan and Cantonese, and perhaps other d i a l e c t s , the Ru-sheng reading from * b j i t and the Qu-sheng reading - 279 -from * b j i v have been preserved by having [p-et] , descending from * b j i t , acquire a meaning r e s t r i c t e d to 'elephant's trunk'. Regarding the Kai-kou Zhi-she, Cantonese had also merged a l l the grades and rhymes. There are, however, two correspondents i n Cantonese as a r e s u l t of phonological conditioning: [ e j ] occurs afte r i n i t i a l s descending from LMC retroflexes and r e t r o f l e x s i b i -lants, and [ i : J elsewhere. On the basis of h i s t o r i c a l and cross-d i a l e c t a l data, i t appears that * - i had diphthongized to - e j under certain conditions i n Cantonese—a change which had not affected the Zhong-shan d i a l e c t . (That process i n Cantonese i s p a r a l l e l to the diphthongization of *-u >ow and *-y> 0q i n that dialect.) Thus, although a l l the words i n (67) have f i n a l / i / i n Zhong-shan, only zhi fa , y i ^ , shi ^ and shi 4 l L i n (67) have f i n a l [ i :J in Cantonese; the rest have f i n a l [ e j ] . Moreover, words with: i n i t i a l s descending from Ri * r - , Y i Ying *?-, and Yu IV (.< EMC * j - ) , have the s y l l a b l e [ j i : ] i n Cantonese; that i s , f i n a l [ i i ] i s accompanied by a [j] on-glide i n the absence of a conson-antal i n i t i a l i n the modern re f l e x of such words. Hence, Cantonese 21 5 X pronounces y_i fyfy , for example, as [ j i : ] (Zhong-shan [ i i ] ) . A number of words i n Kai-kou Zhi-she have f i n a l / a j / or / a : j / i n Zhong-shan (and Cantonese) which may r e f l e c t an older stage of the Chinese language. Pulleyblank (1977:120), for example, observes that the Zhi j L (EMC * - i a <; Old Chinese *-dj and *-al) rhyme had rhymed with the J i a (O.C. *-aj) rhyme during the Wei-Jin and early part of the Northern and Southern Dynasties period ( i . e . , 3rd-5th c. A.D.), and proposes the l i k e l i h o o d that the Zhi L rhyme s t i l l had a f i n a l glide at that time, and was * - i a j (_ > EMC * - i a > LMC * - i ) . A l i s t of Zhong-shan words i n this - 280. -rhyme with f i n a l / a j / or / a : j / i s given i n (68). 48-11 / s a j 5 5 / 'sieve' 49-11 % /sarj / 'Imperial signet' 49-12 41 / •13 / /sa:: / 'to move one's abode' 49-16 ik / l a j 1 3 / L . , / l a : j 1 3 / C . 'to l i c k 50-5 / n g a j 1 3 / 'an ant' 50-12 t / l a : j2 2 / 'lichee' Only a few words occur i n Grade II of He-kou Zhi-she, and even fewer appear i n the d i a l e c t survey. A l l the words belonging to the second grade have descended from words containing an EMC re t r o f l e x s i b i l a n t i n i t i a l . Zhong-shan has f i n a l / 0 j / as a regular correspondent to thi s set of h i s t o r i c a l f i n a l s (as does Cantonese). However, i n Zhong-shan there are, i n fa c t , only two words i n thi s category which have the / 0 j / f i n a l . They are: shuai -|L (63-7) 'to decrease' / s 0 j 5 5 / , and shuai (65-1) -'a 22 leader' /scij /. In Grade III and IV of the Zhi JL and Zhi rhymes, the Zhong-shan f i n a l s are: /jzij/ a fter descendents of EMC pa l a t a l s , and LMC dentals, dental s i b i l a n t s , retroflexes; and /waj/ afte r LMC velars and gutturals ( a l l from *-yj) . In the case of the Y i * r j - ) i n i t i a l , unlike the other velar i n i t i a l s , the l a b i a l medial i n words containing that i n i t i a l has been l o s t i n Zhong-shan. Exam-ples are presented below. (69) a. /jzSj/ a f t e r EMC pa l a t a l s , and LMC dentals, dental s i b i l a n t s and r e t r o f l e x e s: 60- 11 fj|^ / t s ' j r f j 5 1 / 'to follow' 61- 11 \ / 1 0 j 1 3 / 'to t i e ' - 281 -&~ 13 61-15 'f; / j ^ j / 'stamen' b. / a j / after /rig/: 4 51 61- 7 yj5j /ngaj / 'dangerous' 62- 9 \h / n g a j 2 2 / 'false* c. /waj/ elsewhere: 55 61-4 }&j /k'waj / 'regulations' 2 2 62-11 /waj / 'for, because o f 6 4-3 %fu /kwaj 1 3/ 'a track' Cantonese has the same pattern of correspondence as Zhong-shan i n the Zhi j L and Zhi f\$ rhymes. The Wei rhyme d i f f e r s from the other two He-kou rhymes of the Zhi-she not only i n i t s r e s t r i c t i o n to Grade I I I , but also i n i t s i n c l u s i o n of words containing labiodentals (with LMC f i n a l * - j i < * y j ) . In terms of words occurring with the i n i t i a l s also found i n the Zhi £. and Zhi AH rhymes, e s s e n t i a l l y the same f i n a l s and pattern emerge i n the Wei rhyme i n Zhong-shan. The only exception concerns words with i n i t i a l / f / descending from the Xiao (*x-) i n i t i a l . In fact, with, respect to the Zhi-she as a whole, i t i s only i n the Wei rhyme that one finds / f / corresponding to the Xiao i n i t i a l . The f i n a l i n t h i s case i s / a j / . F i n a l / a j / i s also the Zhong-shan correspondent i n Qu-sheng words with i n i t i a l / f / descending from LMC labiodentals. The remaining words i n th i s class of h i s t o r i c a l i n i t i a l s have f i n a l / i / i n Zhong-shan. Exam-ples of / a j / and / i / f i n a l s for the Wei rhyme are given i n (.70) . C70) a. Words with i n i t i a l s '•• descending from LMC labiodentals: / a j / a f t e r / f / i n Qu-sheng words: 66-12 >% / f a j 2 2 / 'pimples' 2 2 66-14 f / f a j / "to waste' - 282 -/!/ elsewhere: 65- 10 ^ / f i 5 5 / 'to f l y ' — 13 66- 5 |^ / f i / 'vagabonds' 66-8 / m i 1 3 / 'a t a i l ' 66-16 ^ Mi22/ 'not yet' b. / a j / a f t e r / f / descending from Xiao (*x-) i n i t i a l : 65-17 / f a j 5 5 / 'to move' 65-19 / f a 3 5 5 / 'honourable' Cantonese has f i n a l [ej] corresponding to Zhong-shan / i / above; Zhong-shan / a j / i n t h i s rhyme, however, i s likewise [B J] i n Can-tonese . 3.2.2.6. Xiao -she As mentioned e a r l i e r , the present study w i l l treat the Xiao-she as a rhyme group which contains only Kai-kou rhymes. On th i s basis, the Hao rhyme only occurs i n Grade I Kai-kou, and has the r e f l e x of f i n a l /ow/ (<.*-uaw a f t e r LMC b i l a b i a l s , *-aw elsewhere) i n Zhong-shan. .Examples are presented below. > 51 (.71) 67-14 %j /mow / 'hair of animals' 69-16 ^ /tsow 1 3/ 'early' 71-9 £ /kow 2 2/ 'to t e l l ' Cantonese also has f i n a l low] for the Hao rhyme. One exception i n Zhong-shan i s bao (67-12) 'to praise' 55 /paw /, with f i n a l /aw/ instead of /ow/. Three other words have f i n a l /a:w/ i n Zhong-shan (and Cantonese): kao (.70-2) 'to examine' /ha:w13/, kao ^ (.70-3) 'to roast' /ha:w13/, and kao | (71-12) 'to depend on' /k'a:w22/L. (/how 2 2/C). The l i t e r a r y - 283 -j reading for kao |£ i s probably borrowed from Mandarin. The Yao jfa rhyme i s i n Grade II only. The Zhong-shan correspondent for t h i s rhyme i s /a:w/ (< * - a i w ) , as i n Cantonese, with Zhong-shan examples presented i n (.72) . (72) 72-9 j$> /ma:w55/C. 'a cat' (./miw5:>/L•) 73- 6 ty /ts'arw 1 3/ 'to f r y ' 74- 16 #L /ha:w22/ 'a school' Grades III and IV (the l a t t e r having been combined under the former i n the d i a l e c t survey) of the Xiao ^ rhyme, and Grade IV of the Xiao ^ rhyme have merged i n the Zhong-shan f i n a l , /iw/ (_< *-iaw), corresponding to these rhymes. Examples are: (.73) 75-8 /miw 5 5/L. 'a cat' (/ma:w55/C.) 76-11 /k'iw 5 1/ 'a bridge' 80-11 9$j /hiw 1 3/ 'dawn' I t might be noted that mao %*9 i n Grade II (number 72-9) and Grade III (number 75-8) have been re-interpreted i n Zhong-shan i n terms of a c o l l o q u i a l versus l i t e r a r y d i s t i n c t i o n . As i n Zhong-shan, Cantonese likewise has [iiw] as the f i n a l corresponding to the rhymes i n Grades III and IV of the Xiao-she. 3.2.2.7. L i u j f f u -she The Liu-she has only Kai-kou rhymes. Grade I occurs only i n the Hou \% rhyme. Grades II and IV of the You rhyme are combined under t h e i r Grade III counterpart i n the same rhyme i n the d i a l e c t survey. The Y6u ^ rhyme has Grade IV only. The general correspondent of the Liu-she i n Zhong-shan i s /aw/, with. - 284 -the merger of the four grades. The basic exception i s words the i n i t i a l of which descends from the Ri i n i t i a l (which, only occurs i n Grade I I I ) , or words i n Grades III and IV with the LMC guttural i n i t i a l s . In such cases, the Zhong-shan r e f l e x i s the s y l l a b l e /jaw/: 174) 81-16 « I / t 1 aw / 'to s t e a l ' 82-21 / 13 / /maw / 'male animal' 83-21 / 13 / /aw / "to f i g h t with s t i c k / f i s t 85-5 /faw 5 1/L. 'to f l o a t ' (/p'u5l/C.) 87-1 / • 51 , /Daw / 'soft' 89-2 « ,. 13 , /Daw / 'to have' There are some exceptions, such as /u/ („£" *-aw afte r LMC b i l a b i a l s , and *-uw after LMC dent i l a b i a l s ) i n Zhong-shan, as shown i n the c o l l o q u i a l pronunciation of fou i n (74) . Other occurrences of /u/ are l i s t e d i n (.75) below. (75) 82-22 4-. 13 . /mu / 'mother' 84-1 /mu / • f i f t h ^ of the Heavenly 87-24 A, 2 2 / /hu / 'a wife' (/p' i 13 / • 1 u / m i / 55 , /san p' 13 'u / 'daughter--in-law') 87-25 I /u 2 2 / /hu / 'to bear' 89-7 /hu 2 2/ 'wealth' 89-8 b\ /u 2 2 / /hu / 'to aid' In Grades III and IV, there are also exceptions with f i n a l J3" 13 /iw/ C < *-iw)(e.g., zhou ^ (88-10) 'a broom' /tsiw / ) , and f i n a l /a:w/ (e.g., zhou f$ (88-6) 'elbow' /tsa:w-13/) . In the 55 case of the word qiu jt| (87-6a) 'surname' /hia:w /, the f i n a l - 285 -i s /ia:w/, which only occurs i n a couple of c o l l o q u i a l words. Cantonese correspondences to the Liu-she do not d i f f e r much from those i n Zhong-shan. There i s an occasional [AXW] f i n a l i n Cantonese instead of the regular IB W] f i n a l , and the d i s t r i b u - i. t i o n of f i n a l s [ow] and lux] i s dependent upon the i n i t i a l (e.g., A . 35 [ow] a f t e r b i l a b i a l s , such as mu -Hf [mow ] i n (.75) , but [ u i ] »>7 4 4 a f t e r labiodentals, such as fu ^ [fux ] i n (.75)). 3.2.2.8. Xian -she The Xian-she has a number of rhymes, most of which occur i n Kai-kou. The He-kou words are found i n the Fan J\* rhyme only. Furthermore, i n the d i a l e c t survey, only words with the h i s t o r i c a l labiodental i n i t i a l s are included i n the Fan rhyme. Note also that i n t h i s rhyme group, words occur i n a l l four h i s t o r i c a l tones — P i n g , Shang, Qu and Ru—with the l a s t containing a stop corres-ponding to the point of a r t i c u l a t i o n of the nasal ending i n the other three tones. To simplify the description of correspondences, the Ru-sheng stop ending w i l l be assumed i n the present and sub-sequent discussion of rhyme groups, a l l of which contain the four h i s t o r i c a l tones. Thus, a f i n a l such as "/om/" would also repre-sent the corresponding f i n a l , "/op/", and so forth. Discussing f i r s t the Kai-kou series of the Xian-she, the regular Zhong-shan correspondents to Grade I, Xian-she are: /om/ af t e r h i s t o r i c a l velars and gutturals, and /a:m/ elsewhere—namely, after dentals and dental s i b i l a n t s . Both sets of f i n a l s are descended from LMC *-am. Examples are shown onvthe.following page. - 286 -(76) a. /om/ after LMC velars and gutturals: 92- 11 /k'om55/L. 'to sustain' C/k'am55/C.) 93- 7 jfe /om22/L. 'to cover with hand' (/am13/C.) 94-16 ^ /hop 2/ 'a small box with l i d ' 96-4 ^ /kom 1 3/ 'to dare' b. /a:m/ elsewhere: 92-5 ,f) /na:m51/ 'south' 94- 1Q / t s a : p 2 / 'mixed' 95- 16 Jjjj /ta:m 1 3/ 'the g a l l ' As shown i n kan i j | and an ^ above, there are a couple of cases of c o l l o q u i a l versus l i t e r a r y forms i n which the c o l l o q u i a l has f i n a l /am/. I t may be the case that the c o l l o q u i a l forms were, borrowed into the d i a l e c t from Cantonese (more on t h i s point short-ly) • Th e word han (.96-6, 9 8-8) 'to c a l l ' occurs i n both 22 22 Grades I and II, with /horn / a l i t e r a r y form and /ha:m / the c o l l o q u i a l counterpart. The d i s t i n c t i o n between Grade I with the /om/ f i n a l and Grade II with the /a:m/ f i n a l , as we s h a l l soon observe, has been re-interpreted as a d i s t i n c t i o n of st y l e i n 44 Zhong-shan. The words i n the two grades are both [ h A i m ] i n Cantonese, although i n Grade I, Xian-she, Cantonese normally has [em] corresponding to Zhong-shan /om/ after h i s t o r i c a l velars and gutturals, and [ A i m ] elsewhere, as i n Zhong-shan. There are very few ir r e g u l a r f i n a l s i n Zhong-shan corres-ponding to Grade I, Xian-she. One i s s_a -H/ (97-3) ' t h i r t y ' / s a 5 1 / , perhaps borrowed from Mandarin. Another i s l a ifcfc (94-9) 55 L r 'to p u l l ' / l a : j /. Pulleyblank observes that l a jjL. i s also exceptional i n other d i a l e c t s , and suggests that the pronunciation - 287 -given i n Zhong-shan (and other dialects), may i n fact not be the word o r i g i n a l l y represented by the graph. The regular Zhong-shan (and Cantonese) correspondent of Kai-kou, Grade II Xian-she i s /a:m/ (.< * - a i m ) , as mentioned above and exemplified below. (.77) 98-5 ffy /tsa:m 1 3/ 'to behead' 100- 7 /la:m 2 2/ 'a warship' 101- 3 t £ /ka:p 2/ 'a small box' Besides nan vj£j^ , discussed e a r l i e r , two other words i n Grade II have f i n a l /om/ i n Zhong-shan: xian (98-3) 'together 51 i J. /horn /, and xian (.99-3) 'meat, f r u i t , etc. for pastry s t u f f i /horn 2 2/. There are three cases l i s t e d i n (78) where the l a b i a l end-ing had become dental i n both Zhong-shan and Cantonese. (.78). 98-10 J$f. / t s a : n 2 2 / 'to earn' 10.1-6 / a : t 2 / 'to press down' 101-7 / a : t 2 / L . 'to press, to crush' (/a:t5/C.) I t i s possible that the dental nasal a r t i c u l a t i o n i s due to Man-darin influence. The writer does not know what may be the source of the dental stop a r t i c u l a t i o n i n the two Ru-sheng words i n (.78) 2 There i s , moreover, a c o l l o q u i a l word /a:p / meaning 'to pressure or force someone' which can probably be associated with ya . Grades III and IV Kai-kou of the Xian-she have f i n a l /im/ (. < *-iam) i n Zhong-shan; for example: (.79) 102-6 /£$ /k'im 5 1/ ' p l i e r s ' - 288 -106- 6 J£ /ngip 2/ 'property' 107- 14 1$ / t s i m 2 2 / 'to usurp' Again, there are three instances where the l a b i a l ending had become dental i n a r t i c u l a t i o n i n the present Zhong-shan data. (.80) 102-13 / p i n 1 3 / 'to censure* 107- 10 / t i n 2 2 / *a shop' 108- 2 / t i t 2 / 'to f a l l down' The change of /m/ to /n/ i n the word bian i s due to the general process of l a b i a l d i s s i m i l a t i o n , which affected almost a l l the Chinese d i a l e c t s . None of those represented i n the Zi-hui (p.177), for instance, have f i n a l [m] for the word bian . Nasal /m/ for 2 2 dian jfc , i t may be noted, i s recorded by Chao (./tim /) . A b i -l a b i a l nasal ending i s also recorded for Cantonese. The dental stop for die i n Zhong-shan (as well as Cantonese) i s duetto the reading from the Shan-she (number ,13.2-10) . , C r o s s - d i a l e c t a l l y , Mei-xian and Xia-men show f i n a l I t ] for die : i t i s [ t i e t 4 ] i n 5 Mei-xian, and I t i e t ] i n Xia-men-(Zi-hui, p.32) He-kou Xian-she i s r e s t r i c t e d to the Fan rhyme, which only occurs i n Grade I I I . Moreover, a l l the words i n the d i a l e c t survey contain the LMC labiodentals. Due to l a b i a l d i s s i m i l a t i o n , the Zhong-shan correspondent to t h i s rhyme i s /a:n/ (.4. *-axm). There are no exceptions found i n the data. Examples are given below. (.81) 109-1 / f a : n 5 1 / ' a l l ' 109- 5 $J / f a : n 2 2 / 'to transgress' 109-9 ^ / f a : t 2 / 'to be i n want' - 289. --no 3.2.2.9. S h e n - s h e The Shen-she consists of only the Qin rhyme. Grades II and IV of that rhyme are combined under Grade IIT i n the d i a l e c t survey. The regular Zhong-shan correspondent of the Qin rhyme i n Grade II i s /am/ (.< *-am). In the case of Grades III and IV, words with the h i s t o r i c a l R i, Ying and Yu IV i n i t i a l s have the s y l l a b l e /jam/ i n Zhong-shan; otherwise, the f i n a l i s /am/. (.The f i n a l s i n Grades III and IV a l l descend from LMC *-im.) Examples are presented i n (_8 2) . (.82) 110-13 ^ /sam 5 5/ 'luxuriant vegetation' 112-9 fa /jam 1 3/ 'to drink' 114-11 ^ /kap 5/ 'anxious' Whereas two of the words i n the data from the Qin rhyme with f i n a l /a:m/ instead of the expected /am/ f i n a l can be traced 2 to readings i n the Xian-she, the pronunciation of /tsa:p / i n the l i t e r a r y word x i ^ (113-18) 'garment' cannot be s i m i l a r l y d i s -missed. Nevertheless, the same long nuclear vowel i s found i n 3 Cantonese for the word.(viz., [ t S A i p ] ) . The word i s l i k e l y to have undergone a sound change i n Zhong-shan s i m i l a r to i t s Canton-ese counterpart—whatever may have been the source of that change. (Alternatively, of course, i t could simply be a pronunciation borrowed from Cantonese.) Labial d i s s i m i l a t i o n i s again involved i n the change of /m/ to /n/ i n the two l a b i a l - i n i t i a l words: bin |j (.111-15) 'to 1 o 13 p e t i t i o n ' /pan /, and pin & (.111-16) 'personality' /p'an / t< *pira and *p'im respectively). - 290 -3.2.2.10. Shan -she The Shan-she has both Kai-kou and He-kou categories, as well as occurring i n a l l four grades. Turning f i r s t to the Kai-kou series, there are two regular correspondents to the Grade I rhymes i n Zhong-shan (and Cantonese): /on/ i n words with . :LMC velars and gutturals, and /a:n/ elsewhere. Both sets of f i n a l s are de-rived from LMC *-an. Examples are: (.83) a. /on/ aft e r LMC velars and gutturals: 115-25 % /on55/ 'quiet' 117- 9 ^ /ngon 2 2/ 'shore' 118- 4 fy\ /hot 2/ 'to hack' b. /a:n/ elsewhere: 55 115-14 fj/fo /sa:n / 'coral' 115- 8 ^ /na:n 5 1/ ' d i f f i c u l t ' 116- 20 $ / t a : n 2 2 / 'an egg' There are a number of exceptions i n which the f i n a l accom-panying the h i s t o r i c a l velar or guttural i n i t i a l i s /a:n/, as i n kan f*\ (.115-21) 'to carve' /ha:n 5 5/, kan /ijfu (116-12) 'straight-forward' /ha:n"'"~V, and han (116-13) 'rare, strange' /ha:n 1 3/. Grade I I , Kai-kou of the Shan-she has only one regular correspondent i n Zhong-shan (and Cantonese) : /a:n/ (.< *-a:n ) , as exemplified i n (84) below. (84) 118-10 ^ / s a : n 5 5 / 'a mountain' 120-1 /V /pa:t 2/ 'eight' 122-5 ^ /nga:n 2 2/ 'a clear sky' Note the homophony of shan ffi i n (83) and shan J * i n (.84) i n - 291 -Zhong-shan, where both f i n a l s are / a : n / i n s p i t e o f a d i f f e r e n c e i n grade: the former i s from Grade I and the l a t t e r Grade I I . Th is i s c o n t r a s t e d w i t h the d i s t r i b u t i o n of f i n a l s found i n an & Oj i n (83) and yan tj£ i n (84) , where an has f i n a l /on/ and yan jjp. f i n a l / a : n / . On the b a s i s of P u l l e y b l a n k ' s r e c o n s t r u c t i o n s , i t i s obvious t h a t a merger of Grade I (_*-an) w i t h Grade I I (*a:n) had taken p l a c e except a f t e r h i s t o r i c a l v e l a r s and g u t t u r a l s , where a d i s t i n c t i o n of the two grades has been preserved i n Zhong-shan, as i t has i n Cantonese. There are on ly two except ions to f i n a l / a : n / i n Grade I I , K a i - k o u Shan-she : x i a (122-8) ' b l i n d ' , and x i a ^ ( .122-9) 2 3 ' t o g o v e r n ' , both of which are /hat / i n Zhong-shan ( [het ] ^ n Cantonese) . With regard to Grades I I I and IV (the l a t t e r subsumed under Grade I I I i n the s u r v e y ) , K a i - k o u rhymes of the Shan-she , the Zhong-shan (and Cantonese) correspondent i s / i n / (.< * - ian ) , as shown i n (85) . (85) 122-15 jfy / m i n 5 1 / ' f l o s s s i l k ' 2 2 131-17 / h i n / ' t o m a n i f e s t ' 133-1 / k i t 2 / ' t o c o n t r a c t ' Bes ides the r e g u l a r / i n / f i n a l , the re are a number of words with, /yn/ f i n a l as the correspondent i n Zhong-shan. A l i s t of these words are presented below. (.86) 122-18 / l y n 5 1 / * ' t o connect ' . 127- 13 -iT / n g y n 5 1 / 'words' 128- 1 ftj^. / k y n 2 2 / ' a d o o r - b o l t ' 29.2 -128-2 22 'to e s t a b l i s h ' 128-3 /ky n 2 2 / -' strong' 128-6 /hyn 2 2/ 'to o f f e r up' 130-2 /hyn 5 1/ 'string of a musical instrument' 130.-3 ** /hyn 5 1/ 'string of a musical instrument' F i n a l /yn/ i n l i a n ^4J ^ , xian (or xuan) ttfh , and xian (or xuan) may have been He-kou o r i g i n a l l y , although they are categorized as Kai-kou i n Middle Chinese, as suggested by xie-sheng evidence (Pulleyblank, p . c ) . The phonetic occurs i n the He-kou word, guan (138-7) 'to shut' /kwa:n55/, for example, while the He-kou word, xuan ~% (147-2) 'dark' /hyn /, occurs as the phonetic i n xian and xian . Observe, also, that the alternative reading xuan i n Mandarin for the two words could r e f l e c t o r i g i n a l He-kou f i n a l s . Regarding the remaining words i n (86), Pulleyblank brought the writer's attention to the fact that a l l these words are from the Yuan fij rhyme (LMC *-ian < EMC *-ian). F i r s t of a l l , these words serve to preserve d i s t i n c t i o n s between the Yuan /Li and Xian \k (LMC *-ian< EMC *-lan) which.have largely been l o s t elsewhere. Zhong-shan f i n a l /yn/ i n these words suggests that they must repre-sent pre-LMC, at a time before the fronting of medial * i > *i a f t e r velars and laryngeals i n the Yuan ?tj rhyme. The d i s t i n c t i o n between the two rhymes i s maintained i n Fu-zhou: f i n a l [yoy] i s the ref l e x of the Yuan rhyme, while [ieij] that of the Xian rhyme. ->• 52 v t ^ 213 Thus, yan i s I g y o g ] / j i a n liL- and j i a n ^ I k y p g ] , and xian fjfrk. I x y o n 2 1 3 ] (Zi-hui, pp. 185 ,188 ,190) . An example of Fu-zhou correspondence to the Xian rhyme i s l i a n (122-17) 'to connect': 52 51 [liei] ] (Zhong-shan / l i n /) . - 293 -The Fu-zhou data for the Yuan rhyme would suggest that the nuclear vowel i n EMC * i a n became a rounded back vowel (or i n Pulleyblank (1962), i t was already analyzed as back rounded *.-d-);. Possibly, medial then became *-y- by assimilating the round-ing of the nuclear vowel. In Zhong-shan, the nuclear vowel, was . subsequently l o s t , r e s u l t i n g i n the compensatory lengthening of the medial, which became the new nuclear vowel i n the d i a l e c t . There are also some words with the i r r e g u l a r f i n a l /im/ corresponding to Grade III and IV Kai-kou i n the Shan-she. The writer can o f f e r no explanation for t h i s anomaly. However, i n th i s case, Hashimoto's Cantonese data likewise shows a couple of [i:m] f i n a l s here. Those i n Zhong-shan are l i s t e d i n (.87) . (.87) 123-14 /sim / 'a cicada' 123-15 / • 51 , /sim / 'Zen, Buddhist* 128-7 /im / 'an embankment' 132-11 & / n i p 2 / 'to fabricate' 132-15 & / t s ' i p 2 / 'to s t e a l ' A c r o s s - d i a l e c t a l comparison shows that other d i a l e c t s besides Zhong-shan and Cantonese have a l a b i a l f i n a l for at le a s t some of the words i n (87). Chan i s [siem ] i n Fu-an (Min; Norman, :; :. p.c.) . Chan ^ and nie ]jf£. are two words from C87) which appears i n the Zi-hui (.pp. 33,172) . Regarding the word chan , the dia-21 lec t s with a l a b i a l ending for that word are: Cantonese (.[sim ] ) , 11 55 Mei-xian (.[hiam J ) , and Chao-zhou U s i s m ] ) . The same d i a l e c t s show a l a b i a l ending i n the word, nie : Cantonese ( [ n i p 3 . ] ) , 4 1 Mei-xian (i!>iap J ) , and Chao-zhou (.[nisp ] ) . Thus, Cantonese and Zhong-shan (both, representing the Yue d i a l e c t group), Mei-xian - 294 -(Hakka), and Fu-an and Chao-zhou (_Min)—-all of which are southern Chinese dialects—show a l a b i a l f i n a l for words which, should have had'contained a dental one i n Late Middle Chinese. One would expect that these words i n the southern d i a l e c t s probably had the same o r i g i n . A highly i r r e g u l a r correspondent to Grade III/IV, Kai-kou Shan-she i n Zhong-shan i s the word jxie (.126-12) 'dirty , ragged' 22 / t s i /, which has} already been mentioned i n r e l a t i o n to i t s i r r e g u l a r / t s / i n i t i a l corresponding to the Xin (*s-) i n i t i a l . Another exception i n Zhong-shan i s nian (130-11) 'to 13 twist with the finger' /nan /. Hashimoto (p.529) records the 35 word as [ni:n ] for Cantonese, but elsewhere (.p. 246). states that f i n a l l i i n j i s the l i t e r a r y form; the c o l l o q u i a l reading of nian i s [ n e n 3 5 ] , p a r a l l e l to the form obtained i n the Zhong-shan 13 data. The s y l l a b l e /nin /, i n the r i s i n g tone, does not occur i n Zhong-shan. Hashimoto records nian as the only word with 35 35 s y l l a b l e [ni:n • ] . Note, moreover, that the s y l l a b l e [nen J, with the r i s i n g tone, i s also rare; nian constitutes the only word f i l l i n g that syllable-tone s l o t i n Hashimoto's data for Cantonese. In Zhong-shan, i t i s also the only word with s y l l a b l e 13 /nan /. C o l l o q u i a l l y , the word i s often used to mean 'to tease, to exasperate'. Turning now to the He-kou series of the Shan rhyme group, the regular Zhong-shan (and Cantonese) correspondents i n Grade I are: /yn/ afte r words the i n i t i a l s of which descend from LMC dentals and dental s i b i l a n t s , and /un/ elsewhere, as exemplified i n (88). Both Zhong-shan f i n a l s come from LMC *-uan. - 295 -(88) a. /yn/ after LMC dentals and dental sibilants;: 134-16 /tyn / 'short' 136-1 22 /lyn^ V 'disorderly' 137-6 it / t s y t 2 / 'a pinch of' b. /un/ elsewhere: 133-10 % 51 /p'un / 'a dish' 136-11 fa 22 /ngun / 'to play' 137-11 / u t 2 / ' l i v i n g ' Among the exceptions i n Zhong-shan, there are several with f i n a l /a:n/~/(w)a: t / , which are : l i s t e d below. They appear to have merged with, the Grade I I series of He-kou Shan-she, where f i n a l /wa:n/ occurs af t e r LMC velars and gutturals, and /a:n/ elsewhere. (.89) 133-12 Si / 55, /pa:n / 'a scar' 133-14 41 / 22 , /ma: n / 'steamed bread' 135-15 ML / 22 , /ma:n / 'water overflowing' 135-16 / 22 , /ma:n / 'a curtain' 136-19 /pa:t 2/ 'a small b e l l ' 136-20. Ui /pa:t 2/ 'to walk' 137-8 2 /kwa:t / 'a clamour' There are also a couple of other exceptions. Luan tyf 13 (.134-20) 'an egg' /10n /, for example, has f i n a l /0n/ i n Zhong-1# 22 shan (and Cantonese). Duo 4fK (137-1)'to gather up' /tsy ./and duo (137-2) 'to weigh, a thing i n the hand' / t s y 2 2 / also have alternative pronunciation of / t s 0 j 2 2 / . The word wo ^ (137-12)-an 13 'to revolve' i s pronounced /k'un / i n Zhong-shan, apparently based on the reading of guan , which occurs i n combination with guan 1? (134-1) 'an o f f i c i a l ' to form Guan-guan ffi t /k'un 1 3 kun 5 5/, - 296 -the t i t l e of an o f f i c i a l under the Han dynasty. Wong (.1954) gives the Cantonese p r o n u n c i a t i o n of wo jpj as I W A U J and guan i^j- as Iku:n 3 5J . Hashimoto, on the o t h e r hand, records the pro-n u n c i a t i o n of Ifui-n 3 5] for wo 1 ^ , w i t h no a l t e r n a t i v e p r o n u n c i -a t i o n s g i v e n . As mentioned above, Grade IT, He-kou s e r i e s of the Shan-she has two r e g u l a r correspondents i n Zhong-shan (and Cantonese): /wa:n/ a f t e r h i s t o r i c a l v e l a r s and g u t t u r a l s , and /a:n/ elsewhere C. *^waxn). Zhong-shan examples are g i v e n i n (90). In the p r e s e n t data, t h e r e i s one e x c e p t i o n to the above c o n d i t i o n i n g ; the v e l a r n a s a l i s f o l l o w e d not by /wain/, but by /a:n/. In both Egerod. (.1956) and Chao (1948)'s data on Zhong-shan, however, the s y l l a b l e /ngwatn/ does occur. As noted i n a more d e t a i l e d d i s c u s s i o n of the t o p i c i n Chapter 1.2.2 on medials, t h i s s y l l a b l e i s found 51 s o l e l y i n the word wan fCifL (.137-14, 138-8) ' o b s t i n a t e ' /ngwa:n /. The l a b i a l m e d ial has a p p a r e n t l y been l o s t i n the speech, of the informants i n the present study. (90) 137-14, 5?| / n g a : n 5 1 / ' o b s t i n a t e ' 138-8- % 138-4 $t» /wa:t 2/ 'to scoop out' 138-5 fj) / s a : n 5 5 / 'a b o l t (for door) ' S 2 2 /kwa:n / 'accustomed t o ' For Grades. I l l and IV (the l a t t e r i n c l u d e d i n Grade I I I i n the survey) of the He-kou s e r i e s of the Shan-she, Zhong-shan has two correspondents: /a:n/ It *-a:n) before words w i t h LMC l a b i o d e n t a l s , and /yn/ (..< *-yan). elsewhere, as shown i n (.91). - 2 9 7 -(.91) a. /a:n/ a f t e r LMC labiodentals: 1 4 4 - 3 H /fa:n / 'foreign' 1 4 5-2 &JL, /ma:n13/ 'evening' tip 2 1 4 6 - 7 |,J / f a : t / 'to punish' b. /yn/ elsewhere: 1 4 0 - 1 6 ifr / s y n 5 1 / 'a boat' 1 4 5 - 1 6 /ngyn 2 2/ 'to be w i l l i n g ' 1 4 8 - 5 ^1 / y t 2 / 'a cave' Cantonese has e s s e n t i a l l y the same set of correspondences as Zhong-shan except that Cantonese has let] i n Ru-sheng f i n a l s , and [ A i n ] i n non-Ru-sheng f i n a l s a f t e r LMC labiodentals. Thus, fan ^ and wan 0$LJ i n ( 9 1 ) are Cantonese [ f A i n 5 3 ] and ImAin2'*], with f i n a l tip 3 l A i n ] , as i n Zhong-shan. Fa ^t\ i n Cantonese I f e t ] , however, 2 contrasts with Zhong-shan [ f A i t ] . There are a few cases of f i n a l /un/ i n Zhong-shan instead of the regular f i n a l /yn/, thus behaving l i k e Grade I f i n a l s . ( 9 2 ) 1 4 4 - 1 8 & /nun 5 1/ 'a wall' 1 4 5 - 5 fa /un 1 3/ 'to y i e l d ' 1 4 5 - 7 J$j /un 1 3/ 'pleasant' 3 . 2 . 2 . 1 1 . Zhen -she The Zhen-she has both Kai-kou and He-kou rhymes. In the Kai-kou series, Grade I only occurs i n the H e n ^ rhyme. In Zhong-shan (and Cantonse), the regular r e f l e x of the Kai-kou Grade I rhyme i s /an/ (.<. *-an) . Examples are: (.93) 1 4 8-6 %r / t ' a n 5 5 / 'to swallow' - 298 -149-4 /han 1 3/ 'angry' 149-5 /ngan 5 1/ 'a l i m i t ' In the case of the Y i n g - i n i t i a l word, en ,z? (148-10) 55 'grace', the Zhong-shan pronunciation i s /jan / although one 55 would have expected the s y l l a b l e /an / i n Grade I. The excep-t i o n a l presence of / j / i n the word, however, has already been dealt with, i n Chapter 3.1.2.8. The Grade II Kai-kou rhyme i s represented by the Zhen rhyme, which i s included under the Zhen & (III) rhyme i n the d i a l e c t survey. The only set of words i n th i s Grade II rhyme i s words with. EMC r e t r o f l e x s i b i l a n t s . In Zhong-shan (and Cantonese), the correspondent of this grade i s /an/ (.< LMC *-an) . Only four words i n th i s series are represented i n the d i a l e c t survey: (94) 150-15 | £ / t s 0 n 5 5 / 'the utmost' 152-14 / t s ' a n 2 2 / 'to a s s i s t ' 5 154-9 fh. /sak / 'a stringed, musical instrument' 154-10 jg* / s a t 5 / 'a louse' Of the above words, two are exceptions i n Zhong-shan: the l i t e r a r y word, zhen ffe , with f i n a l /0n/ ( f i n a l [011] also i n Cantonese) , and se fp, , with f i n a l /ak/ instead of /at/ i n Zhong-shan (in t h i s 5 l a t t e r case, Cantonese has the regular pronunciation of [set ] for the word) . F i n a l /k/ for s_e f^- i n Zhong-shan i s l i k e l y due to d i a l e c t a l influence, since LMC * - t i s regularly preserved i n Zhong-shan. Observe, for example, the following Min forms for se recorded i n the Zi-hui (p.15), a l l of which show Ik] or [2] (re-c a l l i n g that Zhong-shan /k/ i s sometimes phonetically a g l o t t a l 32 1 23 stop): Xia-men [ s i k ], Chao-zhou I s e k ] and Fu-zhou [ s a l ? _ ] . Kai-kou Grades III and IV of the Zhe-she likewise have f i n a l /an/ (_<. *-in) as the regular correspondent in' Zhong-shan. In the case of words with the LMC Ying or Yu TV i n i t i a l , the Zhong-shan s y l l a b l e i s /jan/, as i n y i n |9 (.151-3). 'because' 55 / 3 \ 13 /jan /, and yin (152-2). 'to lead' /jan /. Cantonese has a further conditioning i n th i s s e r i e s : f i n a l [011] afte r LMC dentals, [0nj or Ivnl after LMC dentals and retroflexes, and [vn] elsewhere, * 22 Thus, x i n ^ff (152-9) 'to believe i n ' , for example, i s /san / i n 44 22 Zhong-shan, but I s 0 n J i n Cantonese. (Chao records /san / and 22 22 /sjzin / as alternatives, with /s0n / quite obviously a Cantonese borrowing.) There are a number of exceptions to f i n a l /an/ i n Zhong-shan for Grades III and IV. These include such i r r e g u l a r f i n a l s as /yn/, i n Grade III, exemplified by (95a), and / i n / , i n Grade IV (95b). (95) a. /yn/ f i n a l s : j*** 22 153- 10 j|; /hyn / 'to o f f e r blood i n s a c r i f i c e ' 154- 18 ZJ / y t 2 / 'one' b. / i n f i n a l s : 153- 14 <jb / p i t 2 / 'must' (also: / p i t 5 / ) 154- 8 ^ / t i t 2 / 'orderly' 154-17 -f& / k ' i t / 'to investigate' Aside from preserving grade d i s t i n c t i o n s , the f i n a l s i n (9 5) may also be traces of pre-LMC d i s t i n c t i o n between rhymes that have been l o s t by Late Middle Chinese. While rhymes Zhen and Yan |^ (LMC *-in) go back to EMC * - i n , the Zhen ^ rhyme (LMC *-in) - 30.0 -i s EMC * - i n . A l l the words i n (95). belong to the Zhen rhyme, suggesting perhaps s u r v i v a l s of an e a r l i e r d i s t i n c t i o n that echoes what was found between the Yuan 7L> and Xian / / j i rhymes i n the Shan-she. (Similar d i s t i n c t i o n s are also exhibited i n Fu-zhou.). Two other exceptions have f i n a l /ing/: (.96) 153-19 0$ / n i k 2 / 'familiar' 154-4 %s / s i k 5 / 'to comprehend' F i n a l /0t/ occurs i n l i f t (153-20) 'chestnut' /10t 2/, and i s probably a borrowing from Cantonese where [ 0 t j i s a regular f i n a l a f t e r dentals. Another Zhong-shan exception i s q i ^ L k (156-5 4) 'to beg', which has the regular r e f l e x [het ] i n Cantonese, as 5 opposed to Zhong-shan /ha:t /. The Zhong-shan form may have arisen on the analogy of such alternations i n the two d i a l e c t s as r_a -^-ij 3 2 (146-7) 'to punish': Cantonese [fet ], Zhong-shan IfAit ], and wa (146-8) stockings': Cantonese [met 3], Zhong-shan [ m A i t 2 ] , etc. In both Zhong-shan and Cantonese, the set of correspondences for the He-kou series of the Zhen-she does not reveal a very neat, clear-cut pattern. Although the f i n a l s /0n/, /an/, /un/ and /yn/ occur i n both d i a l e c t s , they are not always i n the same words, nor i s the conditioning with respect to the h i s t o r i c a l i n i t i a l s always re a d i l y apparent. Focussing f i r s t on the Grade I rhymes, the correspondent of words with the h i s t o r i c a l l a b i a l i n i t i a l s i s /un/ or /an/ (< *-an) • The LMC Ming (*m-) i n i t i a l only occurs with, the /un/ f i n a l . Examples are given i n (97). - 301 -(97) a. /un/ f i n a l s : 156- 11 f*[ /mun51/ 'a door' 157- 19 /pun 1 3/ 'root' 159-12 /mut2/ 'to die' b. /an/ f i n a l s : > 55 156- 6 /pan / 'to run away' 157- 20 4£ /pan 2 2/ 'stupid' /p.'an / 'to spurt' After LMC dentals, Hashimoto (p.658) presents l y i n j and [en] as the regular Cantonese reflexes i n t h i s set of h i s t o r i c a l f i n a l s . Although hal f of the words i n thi s category have f i n a l Ion] , for various reasons (.such as readings i n other grades, ana-l o g i c a l readings, and l i t e r a r y readings), Hashimoto does not regard [en] as one of the regular correspondents. In Zhong-shan, i t would be more appropriate to propose that there are two regular correspondents af t e r LMC dentals: /an/ and /0n/ (both from *-wan), with /an/ primarily i n c o l l o q u i a l readings, and /0n/ i n l i t e r a r y ones. Words exhibiting this s t y l i s t i c alternation are given i n (9 8) (98) 156-13 / t a n 3 / C , /ttfn /L. 'aheap' 156-16 A t , / t ' a n 5 5 / C , / t c i n 2 2 / L . 'dumpling' 156-18 .l/ft / t a n 2 2 / C , /t0n 2 2/L. 'to stew' 158-2 ;<k / t a n 2 2 / C , /t0n 2 2/L. 'confused' With regard to the dental i n i t i a l s , i n Zhong-shan, f i n a l /yn/ as a r e f l e x of Grade I He-kou Zhen-she are found i n only three words: (99) 156-15 UK A ' y n 5 1 / 'a suckling pig' 156-17 Ufa / t ' y n 5 1 / 'buttocks' 159-1 ^ /nyn 2 2/ 'tender' Of the three words i n (.99), the two words with i n i t i a l / t ' / are l i t e r a r y words, and may have been influenced by Cantonese. Nen 2 2 / nY n /i however, i s used i n d a i l y speech i n Zhong-shan, and would therefore be more d i f f i c u l t to explain away i n terms of d i a l e c t a l borrowing of l i t e r a r y pronunciations. It i s with LMC dental s i b i l a n t s that f i n a l /yn/ (.< *-wan) i s the regular Zhong-shan correspondent for Grade I He-kou Zhen-she. (.The same correspondence i s found i n Cantonese.) Examples are: (.100) 157-4 $ / t s ' y n 5 5 / 'a v i l l a g e ' 158- 6 i\ / s y n 1 3 / 'to injure' 159- 15 tf\ / t s ' y t 2 / 'abrupt' The remaining LMC i n i t i a l s i n t h i s set are the velars and gutturals. The regular f i n a l which combines with t h i s group of i n i t i a l s i s /wan/ i n Zhong-shan. In the case of words with, the Xiao (*x-) i n i t i a l , the s y l l a b l e i s /fan/; the l a b i a l medial *-w-had combined with the i n i t i a l to form the labiodental / f / . Exam-ples are: (101) a. /an/ a f t e r / f / : i 55 157-12 Jjf /fan / 'to marry a wife' 160-4 ,% / f a t 5 / 'suddenly' b. /wan/ elsewhere: 157-17 ;JL. /wan 5 5/ 'warm' - 303 -158-8 £0 /k'wan 1 3/ 'to bind' 160-1 ^ /kwat 5/ 'a bone' Grade II He-kou Zhen-she, which had been merged with Grade III i n the survey, i s only found i n two Ru-sheng words with, the EMC r e t r o f l e x s i b i l a n t i n i t i a l s : shuai (163-1). 'to lead' / s 0 t 2 / , and shuai 0-63-2). 'a c r i c k e t ' / s 0 t 2 / . I t should be noted, however, that i t was only from the Guang-yun onwards that such, words were placed i n t h i s rhyme. Both shuai <lj-; and shuai |fc^ were o r i g i n a l l y from the corresponding Zhen ||L rhyme, which i s not represented i n the d i a l e c t survey. Grades III and IV (the l a t t e r included i n the Grade III rhyme i n the survey), have the following pattern i n Zhong-shan (.and Cantonese) : f i n a l /an/ (. < *-un) afte r LMC labiodentals, /f6n/ after LMC dentals and s i b i l a n t s , and /wan/ af t e r velars and gutturals. With tie exception of the l a b i o d e n t a l - i n i t i a l words, the f i n a l i n the remaining words i s derived from LMC *-yn. Examples are given below. (.102) a. /an/ a f t e r LMC labiodentals: 163- 8 / f a n 5 5 / 'to divide' 164- 19 f a j /man 2 2/ 'to ask' b. /j6n/ after LMC dentals and s i b i l a n t s : 160-20 ^ /t s ' j r f n 5 5 / 'spring' 2 162-10 /ljzJt / 'a law' jj> 5 162-15 j\ /s0t / 'eleventh of the Twelve Branches' c. /wan/ after LMC velars and gutturals: 164- 1 % /kwan 5 5/ 'a r u l e r 1 165- 15 /($ /k'wat 5/ 'crabbed' - 304 -Aside from the above correspondences, there are a number of exceptions with f i n a l /an/. Words with the h i s t o r i c a l Ri i n i -t i a l , for example, have the s y l l a b l e /jan/ (e.g., run f£[ (162-9). 22 'extra 1 /jan / ) . There are three L a i - i n i t i a l , Ping-sheng words from the Zhun rhyme i n the survey which have f i n a l /an/, as i /— 51 exemplified by lun (160-6) 'constant' /lan /. Their counter-parts i n the Ru-sheng have f i n a l /fit/, as i n Iii above. Among the h i s t o r i c a l s i b i l a n t - i n i t i a l words, there are also those with. the exceptional /an/ f i n a l ; for example, jun j i . ^ (160-11). 'to com-22 . 22 plete' /tsan /, and xun (.16 2-2) 'quick' /san /. Exceptions with f i n a l [en] are not found i n Cantonese. Lun , for example, 21 i s [ l 0 n ] i n Cantonese. 3.2.2.12. Dang ^ -she In the Kai-kou series of the Dang-she, the Tang rhyme has Grade I rhymes only, and the Yang ffy rhyme has Grades I I , III and IV rhymes (with Grades II and IV included i n Grade III i n the d i a l e c t survey). Grade II i s only found i n those words with EMC r e t r o f l e x s i b i l a n t i n i t i a l s . The regular Zhong-shan (_and Cantonese) correspondent of Grades I and II Kai-kou series of the Dang-she i s /ong/ (.< *-ag). Examples are: j - 51 (103) 166-13 m /t'ong / 'name of a dynasty' 167-1 |2] /kong 5 5/ 'a mound' 170-17 /ok 2/ ' e v i l ' 175-12 /K. / t s o n g 2 2 / 'form' There are two regular reflexes for Grades III and IV, Kai -' - 30.5 -kou Dang-she i n Zhong-shan, conditioned by the type of h i s t o r i c a l i n i t i a l . In words with the EMC p a l a t a l Ri i n i t i a l , the Yi (*g-)_ i n i t i a l , and the Ying i n i t i a l ( a l l of which are from Grade III rhymes), and i n words with, the Grade IV Yu i n i t i a l (_<. EMC * j - ) , Zhong-shan has the s y l l a b l e /jong/ (c * - i a g ) . Cantonese has the s y l l a b l e [ijoeig] i n the same environment. The regular Zhong-shan re f l e x after other h i s t o r i c a l i n i t i a l s i s f i n a l /iong/ (.< *-iag)., phonetically Iasg] . i. The corresponding f i n a l i n Cantonese i s [ o B i - g ] . Zhong-shan examples are presented i n (104). (104) a. S y l l a b l e /jong/ with i n i t i a l s R i, Ying, Y i and Yu IV; 173-1 £^ / j o n g 5 5 / 'the center' a 51 173-4 *r /jong / 'a sheep' 176-16 £ / j o k 2 / ' i f 176-20 ^ / j o k 2 / 'cruel' b. F i n a l /iong/ elsewhere: 170-18 Jjrfi / n i o n g 5 1 / 'a g i r l ' 175- 21 £ j / h i o n g 2 2 / 'facing towards' 176- 18 f$f / k i o k 2 / 'foot' Turning now to the He-kou series of the Dang-she, t h i s series i s only found i n Grades I and I I I : the Tang rhyme i n Grade I, and the Yang ^ rhyme i n Grade I I I . The Zhong-shan f i n a l s of the Tang /% rhyme are: /wong/ (< *-wag) as the entire s y l l a b l e i n words with the h i s t o r i c a l Xia or Ying i n i t i a l , and /ong/ as the f i n a l (also <*-wag).elsewhere. In some cases of the Xiao i n i t i a l (and i n one instance of the Xia i n i t i a l i n the data), the i n i t i a l had changed to / f / i n modern Zhong-shan, having absorbed the l a b i a l *-w- medial. The Zhong-shan f i n a l a f t e r / f / - 306 -i s /ong/. Examples are: (105). a. Syllable /wong/ with, i n i t i a l s Xia and Ying: 177- 4 jfc /wong 5 1/ 'yellow' 178- 9 4l /wok2/ 'a b r o i l e r ' b. F i n a l /ong/ elsewhere: 177-1 /kong 5 5/ ' l i g h t ' 177- 13 "ifu / f o n g 5 5 / 'to l i e ' 178- 3 /kok 2/ 'a surname In Zhong-shan, the h i s t o r i c a l v e l a r - i n i t i a l words have l o s t t h e i r medial segment e n t i r e l y , thus merging with t h e i r Kai-kou counter-parts. The Kai-He d i s t i n c t i o n has been preserved i n most cases d i s t i n c t i o n , the pattern of correspondence i n Cantonese i s i d e n t i -c a l to that i n Zhong-shan. h i s t o r i c a l i n i t i a l s that occur here are the LMC labiodentals, velars and gutturals. The Zhong-shan correspondents are: f i n a l /ong/ after labiodentals (. < *-dn) and velars (.< *-ycuj), and the s y l l a b l e /wong/ (A *-yaij) a f t e r the Ying and Yu III ( <. EMC *w-) i n i t i a l s . The Grade III rhyme has e s s e n t i a l l y merged with the Grade I rhyme i n the He-kou Dang-she i n Zhong-shan (and Cantonese). Examples are given i n (.106) . (106) a. Syl l a b l e /wong/ with, i n i t i a l s Ying and Yu III; Grade III) rhyme, the only 179-5 X /wong 5 1/ 179-12 /wong 1 3/ ' a king ' useless' - 307 -b. F i n a l /ong/ elsewhere: 55 179-1 Ji /hong / 'to correct' 179- 18 §J /mong22/ 'to hope' 180- 5 &if / f o k 2 / 'to bind' In the case of the word guang i ^ i , (180-1) 'to roam', Zhong-shan 51 has a l i t e r a r y reading of /k'ong / and a c o l l o q u i a l one of 22 33 /k'wa:ng /. (It i s IkwAiij J i n Cantonese.) Summarizing t h i s rhyme group, the d i s t i n c t i o n has been maintained i n the Kai-kou series of the Dang-she between Grades I and II on the one hand, and Grades III and IV on the other, while such a d i s t i n c t i o n between grades has been large l y l o s t i n the He-kou series i n Zhong-shan. At the same time, the Kai-He d i s t i n c -t i o n i n v e l a r - i n i t i a l words, preserved i n Cantonese and the majority of Chinese d i a l e c t s , has been l o s t i n Zhong-shan. From the fusion of the Xiao i n i t i a l with the medial *-w- to form i n i t i a l / f / i n modern Zhong-shan Ce.g., huang -jffu i n [105)), one can assume that the loss of medial *-w- i n v e l a r - i n i t i a l words i s a more re-cent process than that of d e n t i l a b i a l i z a t i o n i n the Zhong-shan d i a l e c t . 3.2.2.13. Jiang jJ- -she The Jiang-she has only the Jiang rhyme which, i n turn, has Grade II rhymes occurring only i n the Kai-kou series. The regular Zhong-shan correspondent for t h i s set of f i n a l s i s /ong/ (.<• *-orj), with examples presented below. (Cantonese likewise has f i n a l ID i.n'J .) (.107) 181-2 $T / t s o n g 5 5 / 'a stake' - 308 -183-18 ^ /sok 2/ 'new moon' 183-23 $j[ A ' o k 2 / 'to pick out' In both Zhong-shan and Cantonese, there are a few excep-tions involving words with LMC r e t r o f l e x or r e t r o f l e x s i b i l a n t i n i t i a l s . Instead of f i n a l /ong/, these words have f i n a l /iong/ i n Zhong-shan (and f i n a l loenj] i n Cantonese) : 'a window' 'a pair' •a table' 'to establish' "to cut or po l i s h stones' 'to peck' (/tiong 5 5/C.) Pulleyblank (1977:111) i n fact proposes to reconstruct the Jiang-she as Early Middle Chinese * - c B g . In that proposal, the Cantonese [oesq] reflexes from the Jiang-she are a su r v i v a l of the EMC f i n a l rather than an exception. The absence of a medial glide i n Grade II rhymes would suggest that Zhong-shan /iong/ i n t h i s case i s the r e s u l t of the merger of t h i s f i n a l i n the Jiang-she with Grades III and IV f i n a l s i n Kai-kou Dang-she, i n which the regular Zhong-shan f i n a l /iong/ i s derived from the LMC f i n a l *-ian. Thus, the "breaking" of EMC *-cerj i n Zhong-shan to [0^ 0 ] has been phonemicized as /iong/ based on the f i n a l i n the Dang-she. By Northern Song times, the Jiang-she had merged with the Dang-she (Pulleyblank, 1971:238). There are also a few /uk/ f i n a l s i n Zhong-shan. Hashimoto (p.660;fn.l2,p.675) suggests that such f i n a l s (in Cantonese) may be traces of an e a r l i e r d i s t i n c t i o n that had once existed within 181- 5 & & / t s ' i o n g 5 5 / 181- 6 M /siong / 183- 7 1 / t s ' i o k 2 / 183- 8 1 / t s ' i o k 2 / 183- 9 / t i o k 2 / 183- 10 2 /tiok /L. - 30.9 -the Jiang-she; that i s , words with the IukG f i n a l are mostly from the Archaic (Old) Chinese Hou rhyme, while those with other f i n a l s came mostly from the Xiao p) rhyme. 3.2.2.14. Zeng A -she The Zeng-she has both. Kai-kou and He-kou rhymes. In Grade I of the Kai-kou rhyme, the regular Zhong-shan (and Cantonese) correspondent i s /ang/ (_* *-og), as exemplified below for Zhong-shan: 1 r a t t a n 1 'willing* 'to engrave' & 5 The word hei (186-16) 'black' i s /ha:k / i n Zhong-shan. 5 5 Cantonese has Ihek J i n the l i t e r a r y reading, and IhAik J as the c o l l o q u i a l counterpart. I t i s l a t e r i n the Geng $ L - s h e that one encounters a regular alternation between leg] and lAigJ for d i f -ferences i n s t y l e , r e f l e c t i n g d i f f e r e n t layers of the language. Grade II Kai-kou, together with Grade IV, i s incorporated into the Grade III Kai-kou rhymes i n the survey. The regular Zhong-shan (and Cantonese) correspondent for Grade I I , which only contains the EMC r e t r o f l e x s i b i l a n t series, i s /ak/ (.< *-ok) (only Ru-sheng f i n a l s appear i n the survey). There are only three words i n the data that show f i n a l /ak/: (.110) 190-1 #| /ts'ak 5/L. 'the side' (/tsak 5/C.) 190-2 / t s ' a k 5 / 'to fathom' 190-3 / t s ' a k 5 / 'to p i t y ' (.109) 184-7 Ufa / f a n g 5 1 / 185- 2 -j| /hang 1 3/ 186- 14 f/t\ /k'ak 5/ - 310 -In Cantonese, ce }$.\ and ce '{#'| i n (.110) have [ek] and { A i k J f i n a l ; as alternative readings r e f l e c t i n g l i t e r a r y and c o l l o q u i a l read-ings . Two words with the Shan k i n i t i a l have f i n a l / i k / i n this set of rhymes i n Zhong-shan (and Cantonese): se h U90-4). 5 j . 5 'colour' / s i k /, and se ^ (.190-5) 'miserly' / s i k /. The two remaining Kai-kou grades, III and IV, have /ing/ (.< *-iq) as the regular Zhong-shan (and Cantonese) f i n a l . Example are: (111) 187-16 / s i n g 5 5 / 'to ascend' 188-2 M / i n g 5 5 / 'ought' 190-22 #|> /ngik 5/ 'or* One exception i n th i s set of rhymes i s ping (.187-2) 51 21 'according to' /p'ang / (.Cantonese Ip'-et] ] ) . A second highly 22 ir r e g u l a r correspondent here i s /jan / as the pronunciation of yun jf. (.189-3) 'pregnant' (.Cantonese Ijen ] ) . The change of *-n to *-n i n yun seems to have affected a number of the Chinese d i a l e c t s . In the Zi-hui (p.223), four of the d i a l e c t s represented there appear to have preserved a vel a r nasal ending: Tai-yuan (Mandarin) , Wen-zhou (.Wu) , Chao-zhou (Southern Min) , and Fu-zhou (Northeastern Min). However, i n a l l four of these dia-l e c t s , the dental nasal ending has been l o s t ; i n Chao-zhou and Fu-zhou, f o r example, *-n had merged with * - r j . The other Min di a -l e c t represented i n the ,Zi-hui .is Xia-menwhere a dental, nasal . ending i s recorded for yun Q (.[in33].) . Given the dental..ending, i Xia-men./ one can be quite confident that, a change to dental a r t i c u l a t i o n must have also affected Chao-zhou and Fu-zhou before *-n>*-- 311 -The He-kou series of the Zeng-she i s not only li m i t e d to Grades I and II I , but also to words with the h i s t o r i c a l v e l a r or guttural i n i t i a l . There are four words i n the d i a l e c t survey representing the Grade I series ('.< *-wan). , and one only represent-ing the Grade III series (_.<• *-yk) . These words are l i s t e d below. (112) a. Grade I f i n a l s : . 191-1 \U /wang 5 1/ 'to enlarge' 191-2 |g /kok 2/ 'a nation' 191-3 /wa:k / 'perhaps' 191-4 & /wa:k2/ 'to mislead' b. Grade III f i n a l s : 191-5 /wa:k2/ „ 'a f r o n t i e r ' Cantonese has b a s i c a l l y the same pronunciation for the above words except that guo i s [kwoik 4], with a l a b i a l medial, and yu i s [wik ]. B a l l (p.530) gives /wik/ for yu i n S h i - q i . I t i s 2 l i k e l y that the pronunciation of /wa:k / for the word i n the present data i s due to the analogical reading of huo ^ i n (112). 3.2.2.15. Geng j$L -she There are both Kai-kou and He-kou rhymes i n the Geng-she. Only a few obscure words occur i n Grade I Kai-kou, and none i n Grade I He-kou. As a r e s u l t , Grade I i s not represented i n the d i a l e c t survey except for the curious case of da ( . 192-17) 'to 13 s t r i k e ' / t a /, which has been recorded i n the d i a l e c t survey as a Grade II word. (LMC dentals only occur i n Grade I.) Da %1 has the fan-gie s p e l l i n g of de ^ (186-5) /t a k 5 / and leng 'A (192-18) 13 13 /lang /L. (/la:ng / C ) . The regular r e f l e x one would therefore - 312 -13 13 expect i n Zhong-shan would be /tang / or /ta:ng /. Presumably, 13 the pronunciation i s more l i k e l y to be /tang / since there i s no s y l l a b l e /ta:ng/ i n the d i a l e c t ; the s y l l a b l e /tang/, on the other hand, i s very common. The Zhong-shan pronunciation of da , 13 however, i s /ta /, with no nasal ending. C r o s s - d i a l e c t a l l y , the nasal has been l o s t i n the majority of the Chinese d i a l e c t s . In the Zi-hui (p. 3 ) , for example, only i n the cases of Su-zhou (_Wu) 41 and Xia-men (S. Min) i s . there any trace of a nasal ending: [tag J i n Su-zhou, and [ t a 5 1 J i n Xia-men. Turning now to Grade II Kai-kou rhymes, i n Zhong-shan as i n Cantonese, there i s a clear pattern of an alternation between an /ang/ and /a:ng/ f i n a l (.< *-a:jg) corresponding to a d i s t i n c t i o n between a l i t e r a r y and a c o l l o q u i a l layer. In most cases, only one or the other f i n a l i s present i n Zhong-shan. There are, never-theless, a number of words i n which the alternation between the two f i n a l s i s s t i l l currently used to dis t i n g u i s h a l i t e r a r y versus a c o l l o q u i a l reading of the word. A l i s t of such words can be found i n Table 4 (b) i n Chapter 1 .4 .4 on c o l l o q u i a l versus l i t e r a r y forms. A few examples w i l l be repeated here: (.113) 192 -4 £ /sa:ng 5 5/C, /sang 5 5/L. 'raw* 1 9 2 - 1 3 " J /ha:ng 5 5/C, /hang 5 5/L. 'to pervade' 1 9 2 - 1 8 / l a : n g 1 3 / C , / l a n g 1 3 / L . 'cold' The word cheng ( 192 -3 ) 'vessel for warming wine' 55 55 /tong /, with f i n a l /ong / i n Zhong-shan, may not be an exception. F i n a l /ong/ i s based on the pronunciation of the word i n the Dang-she, d a n g , meaning 'a small gong', which i s h i s t o r i c a l l y homo-jj, 55 phonous with, dang % 0 - 6 6 - 8 ) 'ought1 /tong /. In Zhong-shan, - 313 -the two words are likewise synchronically homophonous, both, being pronounced / t o n g 5 5 / . In the case of f i n a l / i ng/ i n beng ..J^ L- (196-5). 'to scatter' 22 /ping /, also, there i s a counterpart i n another rhyme on which, the Zhong-shan pronunciation i s based. In t h i s case, the Zhong-shan pronunciation of beng i s derived from the reading added to Grade IV of the Qing rhyme i n the Ji-yun. No h i s t o r i c a l l y homophonous word i n the Grade IV Qxng "j| rhyme has been recorded i n the survey. In Zhong-shan, the pronunciation of peng (195-1) 'a 55 ditch" i s /pan /. The source of /n/ instead of /ng/ i s not clear. No pronunciation i s provided for Cantonese by Hashimoto, but i t 53 i s Ipoiq ] according to the Zhong-hua Xin Zi-dian and Wong (1954). In Grades III and IV Kai-kou, the c o l l o q u i a l reading has f i n a l /ia:ng/ i n Zhong-shan (.< *-iajq). , except after h i s t o r i c a l gutturals, where the s y l l a b l e becomes /ja:ng/. The l i t e r a r y counterpart i s f i n a l /ing/, or s y l l a b l e / i n g / af t e r LMC gutturals. 13 (Recall, however, that i n the r i s i n g tone, s y l l a b l e /ing / i s 13 phonetically I j ig ].) Cantonese has f i n a l [exqj for the c o l l o -q u i a l reading and I iqJ for the l i t e r a r y one. Syllables I j e i g ] ( c o l l o q u i a l layer) and I j ig] (.literary layer) occur af t e r LMC gutturals regardless of tone. Zhong-shan examples are presented below. A more extensive l i s t of /ing/ versus /ia:ng/ f i n a l s i s recorded i n Table 4 (c). i n Chapter 1. (.114) 198-15 |) /ja:ng 1 3/C. , / i n g 1 3 / L . 'a shadow' 199-4 /kia:ng 2 2/C., / k i n g 2 2 / L . 'to respect' 20.3-16 ^ / t s i a : k 2 / C , / t s i k 2 / L . 'to burn' - 314 -As with the other Kai-kou rhymes i n the Geng-she, i t i s not the case that every word i n these rhymes 'has both a l i t e r a r y and a c o l l o q u i a l pronunciation. The majority, i n fact, have only one or the other pronunciation preserved i n Zhong-shan. (The same i s true of Cantonese.) Aside from the / i n g / ~ /ia:ng/ alternation, there are three exceptions with f i n a l /ang/ i n Zhong-shan: (115) 197-19 $ /mang 5 1/ 'an oath' 13 198- 11 rgL /mang / 'a vessel' 199- 16 & /ngak 2/ 'to disobey' Meng tifi has a Grade II reading which would account for f i n a l /ang/, but the Grade II reading i s i n Qu-sheng. The Zhong-shan pronunci-ation of meng jjj^ should then be /mang /, with tone /22/ as the regular r e f l e x of the Qu-sheng. Perhaps i t i s a case of d i a l e c t a l borrowing. In the Macao d i a l e c t recorded by B a l l , and among some speakers i n Ku-chong, Zhong-shan I i y ] consistently corresponds to leg] i n that variety of Zhong-shan speech. Observe that ying ^ 55 (195-14) 'the Chinese o r i o l e ' i s recorded as /ang / i n the pre-sent data, although B a l l (p.518) notes that i t i s pronounced /ing/ i n S h i - q i , while /ang/ i s used i n Macao (and the rest of the county) The graphic variant, ying IJ^ iJjg (195-15) i s recorded with both, form, 55 55 /ing / and /ang / i n the present data. I t i s possible that the standard Zhong-shan speech once had [eg] instead of [ i g ] , with the l a t t e r used by the educated as a r e s u l t of Cantonese influence, as suggested by B a l l , who notes, for instance, that i n S h i - q i , there i s an attempt, amongst the educated class at least, to approximate Canton pronunciation (p. 510.).. Thus, instead of f i n a l /ang/ being - 315 -borrowed into the standard Zhong-shan d i a l e c t , perhaps i t would be more correct to regard the few cases of f i n a l /ang/ instead of /ing/ as survivals of the o r i g i n a l f i n a l i n Shi-qi speech. Before leaving the Kai-kou series of the Geng-she, i t should be observed that i n both Zhong-shan and Cantonese, the l i t e r a r y layer of the Geng-she has merged with the Zeng ^ -she. Thus, i t i s i n the c o l l o q u i a l layer that the o r i g i n a l d i s t i n c t i o n between the two rhyme groups i s preserved. Turning now to the He-kou rhymes of the Geng-she, the regular Zhong-shan (and Cantonese) correspondent of the Grade II rhymes i n the c o l l o q u i a l layer i s /wa:ng/ (< *-wai.j Q ).., with /wang/ in the l i t e r a r y layer. Unlike the Kai-kou rhymes, there are no words with both c o l l o q u i a l and l i t e r a r y pronunciations. Further-more, the Grade II He-kou rhymes are r e s t r i c t e d to words with i n i t i a l s descending from velars and gutturals. In the d i a l e c t survey, t h i s grade i s represented by only ten words. Examples of the two Zhong-shan f i n a l s for t h i s set of rhymes are given i n (.116) (116) a. /wa:rig/final - Col l o q u i a l layer: 208- 4 /wa:ng 5 1/ 'crosswise' 209- 5 j f . /wa:k2/ 'to take i n hunting' 209-6 |»1 /wa:k2/ 'to divide' b. /wang/ f i n a l - L i t e r a r y layer: * 55 209-1 i % /kwang / 'the rumbling of carts' 209-3 % /wang 5 1/ 'spacious 1 There are a couple of exceptions. Guo C208-7) 'name 5 5 of an ancient feudal state' i s /kwik / i n Zhong-shan (also Ikwik ] i n Cantonese). Kuang (old pronunciation gong) (208-5) 'metal - 316 -22 44 ore 1 i s /k'ong / i n Zhong-shan (Cantonese I k 'org ] ) . Hashimoto (p.662) suggests the analogical reading of kuang fyff^ with, kuang 8$, (178-2) 'a desert', from Grade I He-kou Dang % -she. Kuang •fo^ and kuang are likewise homophonous i n Zhong-shan, both 22 pronounced /k'ong /. However, as Hashimoto notes, neither are common words. I t seems doubtful that the analogical reading hypothesis i s correct here. The homophonous reading of the two characters i n Mandarin may be the reason f o r the p a r a l l e l homophony i n the two Yue d i a l e c t s . I t may be the case that the old pronun-c i a t i o n gong for kuang i n Mandarin was replaced by kuang as a r e s u l t of analogical reading with kuang flj^ , which i n turn influenced Cantonese and Zhong-shan. The Grades III and IV rhymes of He-kou Geng-she have the following correspondents i n Zhong-shan: the s y l l a b l e /wing/ ( C *-yajg) aft e r the LMC Xia and Yu III (.< EMC *w-) , and the f i n a l / i n g / elsewhere. Examples are: (117) a. S y l l a b l e /wing/ with i n i t i a l s Xia and Yu I I I : 13 .209-9 jfc^ /wing / 'perpetual' 210- 1 /wing 2 2/ 'to dive' 211- 7 /wing 5 5/ 'the dashing of waves' b. F i n a l / i ng/ elsewhere: 209-7 /<_> / h i n g 5 5 / 'an elder brother' 55 210-5 yjfcfl /k'ing / ' to overthrow' 210-7 / i n g 5 1 / 'an encampment' There are, however,. a number of exceptions i n Zhohg-shan i n which s y l l a b l e /wing/ i s also found i n some Grade IV words with i n i t i a l Yu ( i . e . , < EMC * j - ) , for example: - 317 -(118) 211-3 |^ /wing 2 2/ 'a sharp point' r»- 2 211-4 ^ /wik / 'pestilence' 211-5 /wik 2/ 'a j a i l o r ' Conversely, there are a couple of words with the loss of /w/ i n Xia and Yu III i n i t i a l s : (.119) 209-8 / i n g 5 1 / 'glory' 211-6 / i n g 5 1 / 'a glow-worm' Although i t i s clear that the l i t e r a r y layer of the Geng-she i n Kai-kou rhyme has merged with the Zeng-she synchronically, the paucity of data from the He-kou rhymes of the Zeng-she makes i t impossible to make a similar claim i n Zhong-shan with respect to the He-kou series of the Geng- and Zeng-she. Such a claim has been made by Hashimoto (p.662-3) for Cantonese, since what evidence does exi s t i s more conclusive i n Cantonese than i t would be for the Zhong-shan data on hand. Nonetheless, h i s t o r i c a l evidence point to a clear merger of the Geng- and Zeng-she. Pulleyblank (.1970-71:237) observes, for example, that by Northern Song, these two rhyme groups had been combined under the same table i n the Si-sheng Deng-zi & % 3r and the Qj^-yun Zhi-zhang Tu tj) ig^lM , as well as having been combined by Shao Yong 1^, . 3.2.2.16. Tong A- -she The present analysis of the Tong-she proposes that the Dong JL and Dong ^ rhymes occur i n the Kai-kou series, while the Zhong rhyme i s only found i n the He-kou serie s . The Kai-kou rhymes occur i n a l l grades, whereas the He-kou counterpart occurs i n Grades III and IV only. - 318 -The Grades I and II Kai-kou rhymes have /ung/ C < *-og). as the regular Zhong-shan (and Cantonese) r e f l e x . (Grade II rhymes, are found only i n words with the EMC p a l a t a l i n i t i a l s and have been recorded i n the d i a l e c t survey under Grade III rhymes.). Zhong-shan examples are provided below. (120) 215-5 jjj /ung 2 2/ 'an earthen j a r ' 216-7 ^ /uk 5/ *a room' 219-1 ^ /sung 5 1/ ' l o f t y ' There are two exceptions, as mentioned previously, i n 55 words with. LMC gutturals: weng jfa (.213-20) 'an old man' /jung /, and wo (218-3) 'to water' /juk 5/. In both these two words, the s y l l a b l e begins with a p a l a t a l glide instead of a smooth onset. Grades III and IV Kai-kou rhymes of the Tong-she have the s y l l a b l e /jung/ (.< *-iog). after LMC gutturals and the Ri i n i t i a l , and the f i n a l /ung/ elsewhere. (.Words with LMC labiodentals have f i n a l /ung/ from *-ug < LMC *-ywrj. See Chart 10 on LMC f i n a l s . Zhong-shan reflexes of the labiodentals i n t h i s environment have i n i t i a l /h/ and /m/, contrasting with I f ] and Im] i n Cantonese.) Zhong-shan examples are presented i n (.121) . (.121) a. Syl l a b l e /jung/ with i n i t i a l s Ri, Ying and Yu IV:: 219-13 / j u n g 9 1 / 'to fuse' 221-12 rt / j u k 2 / 'flesh, meat 221-19 \ /juk / 'to nourish' b. F i n a l /ung/ elsewhere: 218- 4 /hung 5 5/ 'wind' 219- 16 ^ /mung22/ 'a dream' 221-1 f t /tsuk 5/ 'bamboo' - 319 -Looking at the He-kou rhymes, reconstructed as *-oy 4 LMC *-yog after LMC labiodentals, and from *-yog elsewhere, i t can be observed that there i s no longer a Kai-He d i s t i n c t i o n i n Zhong-shan with respect to the Tong-she. Sy l l a b l e /jung/ (.4 *-yog) occurs after the Ri i n i t i a l and the guttural i n i t i a l s , and f i n a l /ung/ after LMC labiodentals (< *-dg< LMC *-yog) and the remaining LMC i n i t i a l s C 4*-yog) without d i f f e r e n t i a t i o n of Kai- He-kou rhymes. Similar changes i n other Chinese d i a l e c t s had resulted i n the l a t e r c l a s s i f i c a t i o n of the Tong-she as e n t i r e l y He-kou despite an e a r l i e r Kai-He d i s t i n c t i o n . The merger of the two types of rhymes had taken place by Southern Song. 3.3. Tones The four t r a d i t i o n a l tones or sheng f-g^ , as discussed i n Chapter 1 are: Ping 'even 1, Shang 'ascending', Qu 'departing', and Ru 'entering'. They are further dichotomized into Yin Cupper'), and Yang f - ^ (.'lower') r e g i s t e r s . There i s a c o r r e l a t i o n between the r e g i s t e r s p l i t and the nature of the h i s t o r i c a l i n i t i a l s , although the physiology behind t h i s i s s t i l l not f u l l y understood. In modern Zhong-shan, only the Ping- and Qu-sheng show r e g i s t e r - s p l i t , y i e l d i n g six tones i n accordance to t r a d i t i o n a l enumeration: Yin-ping \%, -f /55/, Yang-ping f% /5.1/, Shang }L /13/, Q U £ /22/, Yin-ru f t /5/, and Yang-ru f % /2/. These six tones are presented i n Chart 13 Ca). For comparative purposes, Cantonese correspondences to the h i s t o r i c a l tones are given i n Chart 13 (b).. In Zhong-shan, one might argue that the Shang- and Qu-sheng - 320 -Chart 13 (a) . Zhong-shan Correspondences to the H i s t o r i c a l Tones. Ping -f" Shang Qu Ru X_ Yin f £ 55 13 5 Yang f < | 51 22 2 Chart 13 (b). Cantonese Correspondences to the H i s t o r i c a l Tones Ping ^ . Shang J l Qu £ Ru > V Yin f t 53/55 35 44 Shang X- 5 Zhong rjz 4 Yang f | 21/22 24 33 3 did not develop a s p l i t to Yin and Yang; or conversely, one could propose that the s p l i t had once occurred, but had since re-merged. There i s no synchronic evidence to support the l a t t e r hypothesis. H i s t o r i c a l l y , nevertheless, i t would simplify the description of the development of regi s t e r i n Zhong-shan i f one can assume an i n i t i a l s p l i t into Yin and Yang of a l l four tones, and a subsequent merger of Shang and Qu i n Zhong-shan. Among the Yue d i a l e c t s , Zhong-shan has the smallest number - 321 -of tones. Of the twenty-two Yue d i a l e c t s surveyed by Hashimoto (pp.52-55), Zhong-shan i s the only d i a l e c t with an absence of Yin-Yang categories i n both. Shang and Qu. It has been suggested that the paucity of tones i n Zhong-shan may be due to the influence of the neighbouring Hakka d i a l e c t . The Hakka spoken i n Zhong-shan county recorded by Egerod (19 59), for example, has only four phonemic tones: two l e v e l , one f a l l i n g , and one r i s i n g , as i n the Zhong-shan d i a l e c t . Only the d i s t r i b u t i o n of these four Hakka tones with respect to the h i s t o r i c a l tones d i f f e r s from those found i n the Zhong-shan d i a l e c t . Given the outstanding number of tones which usually charac-t e r i z e the Yue d i a l e c t and the genuine p o s s i b i l i t y that the Zhong-shan d i a l e c t has been influenced by the number of tones found i n Zhong-shan Hakka, spoken i n the qu immediately south of S h i - q i , i t i s l i k e l y that at one time Zhong-shan had undergone a r e g i s t e r s p l i t of a l l four h i s t o r i c a l tones, with a subsequent merger of Yin and Yang i n Shang and Qu tones. The proposal that i s advanced here i s that h i s t o r i c a l l y Zhong-shan had nine tones, s i m i l a r to Cantonese, i n that the Yin-Yang b i f u r c a t i o n yielded eight tones. A further s p l i t of the Yin-ru into Shang X~ 'upper' and Zhong ^ 'mid' resulted i n a t o t a l of nine tones p a r a l l e l to modern Cantonese, as shown on Chart 13 (b) based on Hashimoto. (The modern Cantonese tones can be used to suggest how the h i s t o r i c a l tones i n Zhong-shan would have sub-divided. ) The s p l i t of Yin-ru into Shang and Zhong i n proto-Zhong-shan and i n present-day Cantonese can be understood from the i n t e r -action of tone, the various classes of h i s t o r i c a l i n i t i a l s , and - 322 -the d i v i s i o n of rhyme groups into Nei- and Wai-zhuan. It was mentioned that there i s a c o r r e l a t i o n between r e g i s t e r s p l i t and the nature of the Late Middle Chinese i n i t i a l s . Whether a s y l l a -ble has the Yin or the Yang re g i s t e r i n modern Cantonese i s con-^ ditioned by the various h i s t o r i c a l classes of LMC i n i t i a l s (_cf. Chart 4 naming these classes).. Syllables that had a 'clear' (l..e,, .1 h - J . Qing 'clear' or Ci-ging }%, jTfa 'second-clear') i n i t i a l contain the Yin, or upper re g i s t e r . Other s y l l a b l e s - - ! . e . , those with the 'muddy' (or Zhuo }% ), or 'not-clear-not-muddy' (Bu-qing-bu-zhuo -7 ~jf v /ft T ) i n i t i a l s — h a v e the Yang, or lower r e g i s t e r . Ru-sheng words are further conditioned by the Nei 1^1 ' i n - ; ner' and Wai $\ 'outer' zhuan (cf. Chart 9 for rhyme groups f a l l -ing into Nei- or Wai-zhuan). With respect to the Nei-zhuan, Ru-sheng words with the 'clear' i n i t i a l s are i n the Shang Yin-ru, while those with the non-clear i n i t i a l s are i n the Yahg-ru. In the case of the Wai-zhuan, Ru-sheng words with the 'clear' and non-clear i n i t i a l s are i n the Zhong Yin-ru and Yang-ru respectively. In other words, the non-clear s y l l a b l e s have Yang-ru (Cantonese /3/) regardless of whether they belong to the Nei- or Wai-zhuan. Yin-ru, however, s p l i t s according to whether a word belongs to Nei or Wai: words i n the Nei-zhuan have Shang Yin-ru (Cantonese /5/), and those i n the Wai-zhuan have Zhong Yin-ru (Cantonese /4/). It would probably be more precise to use the term Xia "1* or 'lower' Yin-ru. In Cantonese, t h i s tone /4/, being midway between /5/ and /3/, probably accounts for the choice of the term Zhong 'mid'. Chart 14 summarizes the d i s t r i b u t i o n of the three Ru-sheng i n Cantonese with respect to the classes of LMC i n i t i a l s and the Nei-and Wai-zhuan. Zhong-shan correspondences are also shown for Chart 14. Ru-Sheng Correspondences i n Cantonese and Zhong-shan. Dialect NEI-ZHUAN WAI-ZHUAN Clear I n i t i a l s Non-Clear I n i t i a l s Clear I n i t i a l s Non-Clear I n i t i a l s CANTONESE Shang Yin-ru /5/ Yang-ru /3/ Zhong Yin-ru /4/ Yang-ru /3/ ZHONG-SHAN Yin-ru /5/ Yang-ru /2/ Yang-ru /2/ Tone (Style) PING SHANG - Coll o q u i a l - L i t e r a r y QU RU Nei-Zhuan Wai-Zhuan Yin-ping /55/ (53/55) Shang /13/ (35) Yin-ru /5/ (5) Qu /22/ (44) Yang-ru /?/ (4) i MOT-CLEAR-NOT-MUDDY Nei- & Wai-Zhuan Yang-ping /5.1/ (21/22) Shang /13/ ( . 2 4 ) Qu / 2 2 / (33) Shang /13/ ( . 2 4 ) Qu /22/ (33) Yang-ru /2/ (3) CO CO - 324 -comparative purposes, and w i l l be discussed i n the following paragraphs. In modern Zhong-shan, the absence of a Yin-Yang s p l i t i n the Qu-sheng i s echoed i n the loss of a regis t e r d i s t i n c t i o n i n the Ru-sheng s y l l a b l e s with a clear versus non-clear i n i t i a l i n the Wai-zhuan. In the Nei-zhuan, Zhong-shan tone /5/ i s found i n s y l l a b l e s with a clear i n i t i a l , and /2/ i n s y l l a b l e s with a muddy or not-clear-not-muddy ( i . e . , non-clear) i n i t i a l . Hence, the correspondence of Zhong-shan /22/ i n the Qu-sheng to Cantonese /4 4/ and /33/ i s p a r a l l e l l e d i n the correspondence of Zhong-shan /2/ i n the Yang-ru to Cantonese /4/ and /3/. The elimination of a mid-tone therefore affected both the Qu and Ru tones. The absence of a regis t e r s p l i t i n the Wai-zhuan i n Ru-sheng words contrasted with the presence of such a s p l i t i n the Nei-zhuan further supports the contention that, at one point, i t would have been l o g i c a l for a reg i s t e r s p l i t to have occurred i n the Wai-zhuan i n Zhong-shan Ru-sheng words. The loss of a re g i s t e r s p l i t i n the Qu-sheng effected a si m i l a r loss i n the Wai-zhuan Ru-sheng. The merger of Yin-Yang i n the Shang-sheng i n Zhong-shan does not a f f e c t any other tones. The loss of re g i s t e r s p l i t i n the Shang-sheng i s prevalent among Chinese d i a l e c t s , and may have been a very old merger. Having c l a r i f i e d the picture with regard to Zhong-shan correspondences to the h i s t o r i c a l Ru-sheng, we w i l l proceed to an elaboration of the correspondences of the other h i s t o r i c a l tones. Chart 15 (.on page 323). i s drawn to f a c i l i t a t e discussion. (Ru-sheng i s also included to complete the chart.) In modern Zhong-shan, Ping-sheng has the r e f l e x of /55/ - 325 -(Tone 1) i n s y l l a b l e s with, a clear i n i t i a l , and /51/ (Tone 2) else-where . The Zhong-shan correspondence of the Shang-sheng i s /13/, except i n the l i t e r a r y layer of words with a muddy i n i t i a l , i n which case the r e f l e x i s /22/ (e.g., zai fo. (.29-12). 'to be at' 13 22 /ts'oj /C. .— /tsoj /!>.).. Thus, the tone of o r i g i n a l l y Shang-sheng words with a muddy i n i t i a l i n the l i t e r a r y layer has merged with the Qu-sheng. The Qu-sheng simply has the correspondence of /22/ (Tone 3) i n Zhong-shan, irrespe c t i v e of zhuan and h i s t o r i c a l i n i t i a l . The Ru-sheng has already been discussed. There i s only one other major complication to the discus-sion thus far, and that pertains to the effects of the merger of rhyme groups on the c l a s s i f i c a t i o n of Nei- and Wai-zhuan. As can be observed from Chart 14, whether a word belongs to the Nei- or Wai-zhuan concerns those i n the Ru-sheng only. Hence, the follow-ing discussion w i l l be r e s t r i c t e d to those rhyme groups that con-t a i n Ru-sheng words. Two mergers involving rhyme groups have been mentioned i n the previous section on f i n a l s : (.1) the merger of the Jiang jZ- -she with the Dang %. -she, and (2) the merger of the l i t e r a r y layer of the Geng -she with the Zeng % -she. In the f i r s t case, both the Jiang and the Dang rhyme groups belong to the Wai-zhuan, and hence do not cause problems with respect to tonal correspondences. The same cannot be said of the Geng- and Zeng-she. The Geng rhyme group o r i g i n a l l y belonged to the Wai-zhuan, while the Zeng-she belongs to the Nei-zhuan. The merger of the Geng-she with the Zeng-she only affects the l i t e r a r y layer, and not the - 326 -c o l l o q u i a l one. The r e s u l t i s that Ru-sheng words i n the c o l l o -q u i a l layer of the Geng-she i n Zhong-shan behave, s i m i l a r l y to those Ru-sheng words i n other rhyme groups i n the Wai-zhuan. The Zhong-;shan correspondent i n the c o l l o q u i a l layer i s /2/ regardless of i n i t i a l . The word bai i§ (.194-1) 'hundred' /pa:k /, for example, i s a c l e a r - i n i t i a l word i n the Wai-zhuan. I t has the Yang-ru tone /2/. However, the muddy-initial word bai (3 (194-8) 2 'white' /pa:k / likewise has the Yang-ru tone /2/. The i n i t i a l i s i r r e l e v a n t i n words i n the Wai-zhuan. The Ru-sheng words i n the l i t e r a r y layer, on the other hand, behave i n congruence with Ru-sheng words i n the Nei-zhuan; hence, Yin-ru /5/ occurs i n words with clear i n i t i a l s , and Yang-ru /2/ i n those with non-clear i n i t i a l s . . J i ^ (203-5) 'to 5 amass' / t s i k /, for example, i s a c l e a r - i n i t i a l word i n the l i t e r a r y layer of the Geng-she, and thus contains the Yin-ru tone /5/. In contrast, x i (203-12) 'a mat' / t s i k 2 / , which has a muddy i n i t i a l , i s i n the Yang-ru tone /2/. Even afte r taking into consideration the change of zhuan re s u l t i n g from the merger of rhyme groups, there i s s t i l l a large number of apparent exceptions to the h i s t o r i c a l categories i n the d i a l e c t survey for Zhong-shan correspondences of LMC tones. One large group of such exceptions i n fact results from these words belonging to another h i s t o r i c a l phonological category not indicated i n the survey. There are often words which occur i n two or more tonal categories but only appear i n one of these categories i n the Fang-yan Diao-cha Zi-biao. The singular appearance of such words i n the d i a l e c t survey can therefore be misleading. A number of these words which appear to have acquired an exceptional tone, - 327 -but are actually regular with, respect to t h e i r pronunciation as derivatives of another tonal category, are given i n Table 5.; They are l i s t e d according to the tone under which they have been recorded i n the survey. The alternative tonal category, which, yi e l d s a regular tonal correspondence i n Zhong-shan i s given for each of these words. Where possible, a word which was h i s t o r i c a l -l y , and i n Zhong-shan i s s t i l l , homophonous with i t i n the a l t e r -nate tonal category i s also included. Moreover, i t should be em-phasized that, unless s p e c i f i e d otherwise, with the exception of tonal differences, the counterpart to the word i n the alternate tonal category should be i d e n t i c a l i n i t s h i s t o r i c a l phonological c l a s s i f i c a t i o n to the word l i s t e d . Table 5. Words with Alternate H i s t o r i c a l Tonal Categories. a. Words Lis t e d as Ping-sheng: Alternate H i s t o r i c a l l y Homophonous Number Word Transcription Tone Word i n the Survey 72-13 ik / 4 . 13 , /tsa:w / Shang 73-4 13 / /tsa:w / 86-21 / 13 ,a /saw / Shang*3 — 9,1-2 , . 22 ,c /miw / Qu 91-6 if /maw / 157-6 1% / t s ' y n 1 3 / Shang'3 — 166-4 ti /pong 1 3/ Shang 167-14 /pong 1 3/ 166-23 / t s o n g 2 2 / Qu d 168-13 / 4 . 22 , /tsong / 172-12 # / • 13/ /siong / Shang 174-7 X / • 13 / /siong / - 3 2 8 -b. Words. L i s t e d as. Shang-sheng: . Alternate H i s t o r i c a l l y Homophonous Number Word Transcription Tone Word i n the Survey 3 2 - 6 it A a : j 5 5 / Ping 3 1 - 1 3 * 5 5 / k c u j ^ / 3 8 - 8 i t / t s a j 5 5 / Ping — 4 1 - 1 4 «• /u - 2 2 . /hu] / Qu 4 2 - 1 9 / h u j 2 2 / 6 4 - 7 Hi /waj 5 1/ Ping 6 3 - 1 6 /waj 5 1/ 6 6 - 7 i / f i 5 5 / Ping 6 5 - 1 1 / f i 5 5 / 8 3 - 5 / i 5 5 ,e /law / Ping 8 2 - 1 i f / i 5 1 -/law / 8 3 - 1 . 6 /k'aw22/ Qu 8 5 - 1 /k'aw 2 2/ 8 3 - 1 8 fl / u 2 2 / T f /haw /L. Qu 8 5 - 2 \* 2 2 /haw / 1 1 9 - 6 m /ka:n 2 2/ Qu 1 1 9 - 1 6 y r . 2 2 , /ka:n / 1 3 0 - 1 4 w / • 5 1 . /ngm / Ping 1 2 9 - 2 4 . / - 5 1 -/ngm / 1 7 3 - 2 4 / 4 - • 2 2 -/tsiong / Qu g 1 7 5 - 5 ft / * . • 2 2 -/tsiong / 1 7 7 - 1 3 / f o n g 5 5 / Ping 1 7 7 - 2 t / f o n g 5 5 / 2 1 3 - 2 2 /mung51/ Ping 2 1 2 - 4 /mung51/ C. Words L i s t e d as Qu-sheng 1 3 - 1 4 A /k'wa5 5/ Ping 1 2 - 1 8 5 5 /k'wa / 1 3 - 1 8 i f /wa / Ping 1 3 - 2 * - 5 1 . /wa / 6 7 - 4 /waj 1 3/ Shang — 7 1 - 1 $ / i 5 5 . /low / Ping 6 8 - 2 i f /low 5 5/L. 7 3 - 1 0 / i 2 2 ,h. /ka:w / Qu 7 4 - 8 /ka:w / 7 3 - 2 0 i 51 / p ' a ; w ° V Ping 7 2 - 5 / i 1 3 , i /p'a:w / 7 4 - 5 » /s.a:w / Shang 3 — 7 4 - 6 ft / 1 3 A /sa:w / Shang-1 — - 329 -(c. Words Lis t e d as Qu-sheng--cont1d) Alternate H i s t o r i c a l l y Homophonous, Number Word Transcription Tone Word i n the Survey 128-7 it /im / Shang 142-7 / s y n 5 1 / Ping 140-8 / s y n 5 1 / 142-12 & /kyn 1 3/ Shang 142-1 A y n 1 3 / 147-7 fa /hyn 5 1/ Ping 147-2 /hyn 5 1/ 168-6 % / . 51,1 /p ong / Ping 166-3 f /p'ong 5 1/ d. Words Lis t e d as Ru-sheng; 22 162-13 «Jt£_ / t s a n / 176-2 j$% / l i o n g 2 2 / Qu Qu 162-1 /fJL 174-19 -ft. /tsan / /n • 2 2 / / l i o n g / Notes to Table 5 a 55 C o l l o q u i a l l y /sa:w / i n Zhong-shan. k Added to the Ji-yun-rhyme dictionary. c / J-' • It also occurs i n the You /u rhyme, Grade IV, h i s t o r i c a l l y homophonous with mou 'tjj^ (95-7) /maw51/ and mao -f (95-8) /ma:w51/. B a l l records /miu/ for what i s e s s e n t i a l l y both miu ( h i s t o r i c a l l y homophonous with miu (91-6)), and mou i$L ( h i s t o r i c a l l y homophonous with mou (95-7)). He records /m^u/ (i . e . , /ma:w/) for miao |»^ - (91-2). Observe that the three-way d i s t i n c t i o n i s present i n modern Mandarin, as can be seen i n the Pin-yin romanization. cl This word should be i n the Shang-sheng. The Ji-yun has added i t - 330 -to the Qu-sheng category, thus making i t homophonous with zang Since this, word i s derived from the Lax i n i t i a l , which, xs a 'not-. clear-not-muddy 1 i n i t i a l , the Zhong-shan r e f l e x i n the Ping-sheng should be Yang-ping /51/, as i n lou (82-1) , rather than /55/. f 13 Also pronounced /haw /C. ^ H i s t o r i c a l l y , t h i s word should have been recorded i n the Qu-sheng. h 2 2 Alternate pronunciation of/k'a:w / i n combination with j ian IJ2 for 'scissors' : - i l f ) /k'a:w22 t s i n 1 3 / . Shang-sheng counterpart has been added to the Ji-yun. k 22 Alternate pronunciation of /sa:w / has the regular /22/ r e f l e x for the Qu-sheng. 1 22 Alternate pronunciation i s /pong /, with tone /22/ as a regular correspondent of Qu-sheng. Obviously, aside from the words l i s t e d i n Table 5 with, alternate tones, there s t i l l remains a number of exceptions i n the Zhong-shan tonal correspondences to the h i s t o r i c a l tones. Some of these exceptions w i l l be discussed b r i e f l y . In Ping-sheng, there are some words i n Zhong-shan derived from the non-clear i n i t i a l s which have Yin-ping /55/ instead of Yang-ping /51/. The following exceptions are r e s t r i c t e d to words the i n i t i a l s of which i n modern Zhong-shan are / l / , /m/, /n/ and /ng/. Such words should always be i n the Yang r e g i s t e r . The occur-rence of these words, i n the Yin r e g i s t e r i s therefore e a s i l y recog-nized as anomalous. - 331 -3-12 / 55 / / i t l O / ; 1 a demon' 3-14 55 /mo / 'to f e e l with hand' 3-15 55 /mo / 'to feed an infant by hand' 3-20 «?>. / l o 5 5 / 'prattle' 8-9 \% / 55, /nga / 'a c h i l d CmiTk-teeth not shed)' 14-8 /mo55/C. 'to f e e l f or' (/mo13/L, /mo51/L 68-2 a /low 5 5/L. 'to f i s h up' (/la:w51/C.) 68-4 /low 5 5/ 'to chatter' 72-9 /ma:w55/C., 'a cat' •75-8 / • 55,_ /miw /L. 156-11 m , 55 , /mun / 'to lay the hand on' 166-6, 178-17 $ , 55 , /mong / /mong55/ 'sharp-edged grass' '. : ; . / 166-7 'vast' The Ru-sheng likewise has words with non-clear i n i t i a l s which have the Yin-ru tone /5/ instead of i t s Yang-ru counterpart (v i z . , /2/> . As i n (.122)., only words with / l / , /m/, /n/ and /ng/ i n i t i a l s i n modern Zhong-shan are recorded i n (123), based on the present data e l i c i t e d from the d i a l e c t survey. a 23). 113-12 I / l a p 5 / 'a bamboo rain-hat' 113-13 /nap 5/ 'a grain' 189-5 I / n i k 5 / 'to hide' 19 0-22 #1 /ngik 5/ 'or' 220-19 % / l u k 5 / 'to seal, to stamp' The Yin r e g i s t e r i n a number of the words l i s t e d i n (.122) and (.123) i s also found, i n Cantonese. Norman (.p.c). observes that l i j£, i n C123)_ above i s also i n the Yin r e g i s t e r i n Min and Hakka. - 332 -There are also a number of c o l l o q u i a l , characterless words i n Zhong-shan which likewise have a l a t e r a l or nasal i n i t i a l occurring i n the Yin re g i s t e r . A few examples are given below. 55 5 (.124) / h i nuk / 'unsteady' 55 /lam / 'to shade over' 55 /mang / 'to p u l l ' / n u 5 5/ 'that' 5 /ngip / 'to blink; to b i t e (of i n s e c t ) ' Observe i n Chart 14 that words belonging to the Wai-zhuan have the Yang-ru tone /2/ as the regular correspondent i n Zhong-shan. The occurrence of such words with Yin-ru tone /5/ would therefore be an i r r e g u l a r tonal development. A few words with such i r r e g u l a r tonal correspondence are provided i n (.125) . (.125) 101-7 J§- /a:t 5/C. 'to press' 153-14 ip / P i t 5 / 'must' 156-14 IJ /ha:t 5/ 'to beg' 183-1 $'1 /mok5/C. 'to peel' 183-15 /tsok 5/ 'to seize' 186-16 %. /ha:k 5/ 'black' It may be noted that Wai-zhuan words have a tense, or long nuclear vowel i n Zhong-shan, while corresponding Nei-zhuan words have a lax, or short nuclear vowel. Thus, the i r r e g u l a r tonal correspon-dences are e a s i l y detected; such- f i n a l s as l i x t ] ( / i t / ) , lyxt] C/yt/). , l o i - p j (./op/)., lA:-kJ C/a:k/)., etc., would not normally co-occur with tone /5/. F i n a l l y , while the words i n (.125) more or less exhaust the - 333 -repertoire of exceptions to words l i s t e d i n the d i a l e c t survey, there are additional l e x i c a l items from the d a i l y speech, of Zhong-shan that f i t the above pattern of i r r e g u l a r correspondence, as exemplified i n (.126) below. Onomatopoeic s y l l a b l e s are not included. 5 13 (.126) /ngip s i / 'stingy' / l i t 5 / 'a knot' 5 / t y t / 1 to protrude' 5 / t s y t / 'to give a k i s s ' 5 /tiok / 'to chop (.e.g., i n h a l f ) ' 5 / t i a : p / 'to t r y a taste o f 5 / l i a : k / 'smart' 3.4. Concluding Remarks Of the Yue d i a l e c t s , only Cantonese has been studied to any great extent, because of i t s status as the standard, and because of the amount of material available r e s u l t i n g from that status. With the exception of Cantonese, data on the Yue d i a l e c t s tend to be rather l i m i t e d . E f f o r t s toward the reconstruction of proto-Yue, for instance, would require that we have a better know-ledge of the various d i a l e c t s that comprise the Yue group. Thus, i n providing an analysis together with, a reasonably large corpus of f i e l d data, i t i s hoped that the present thesis w i l l be a modest contribution to future comparative studies on Yue, and i n the process, lead to further insight into the Chinese language as a whole. - 334 -Notes to Chapter 3 1. Pulleyblank brought the writer's attention to an alternative proposal by Zu-mo Zhou who thinks that Ye-xia %^ 1* was the northern standard. 2. While i t i s true that the character readings of most non-Min dialects can be traced back to something similar to what Pulleyblank c a l l s Late Middle Chinese, J. Norman (personal communication) thinks that this i s true mainly because of the way d i a l e c t data are collected. The core of the language— the everyday lexicon—would y i e l d much more "troublesome" data which cannot be derived so regularly from Late Middle Chinese. Norman therefore questions to what extent a l i s t of character readings, " l i t e r a r y " and " c o l l o q u i a l " , can be considered "the language". I t i s the writer's contention that character readings constitute at least one important dimension of the language. The reconstruction of a proto-language which encompasses data from everyday vocabulary should be the major objective of future l i n g u i s t i c endeavours. In the past, gathering of the real core of the language, the everyday lexicon, has been haphazard, yi e l d i n g a very limited set of data. As a result, i t i s very d i f f i c u l t to conduct comparative studies. The use of the Fang-yan Ci-hui Diao-cha Shou-ce ^ "t ^ "t$ i J "ftfr* (.'Handbook of Chinese dial e c t vocabulary'), compiled by the Chinese Linguistics Project at Princeton University i n 1972, based on the Han-yu Fang-yan Ci-hui, or portions of i t for data-gathering would be a positive step towards achieving that goal. 3. Observe that the present treatment of [u] and lui] as - 335 -allophones of the same phoneme /u/ allows for a simple des-c r i p t i o n of the phonological conditioning involved i n the ref l e x of / f / or /h/ with respect to ce r t a i n f i n a l s . In contrast, Chao's treatment, which assigns ;[u:J to /u/ and [u] to /o/, would necessitate two separate rules to formulate e s s e n t i a l l y the same phonological conditioning. 4. See Hashimoto (pp.158,645) for a d i s t i n c t i o n between tense and lax vowels i n Cantonese and the conditioning of the Ru-sheng i n Cantonese by the nature of the nuclear vowel from a synchronic perspective. 5. In Chan (forthcoming), the writer presents various arguments to support the contention that the Sino-Portuguese data r e f l e c t s not Standard Cantonese, but the Zhong-shan d i a l e c t , and i t i s the difference i n d i a l e c t base and the accompanying difference i n t h e i r phonological system which invalidates Boltz' tentative conclusion on the dating of the d e n t i l a b i -a l i z a t i o n process i n Cantonese. Nevertheless, i t i s probably the case that not only has d e n t i l a b i a l i z a t i o n s t a b i l i z e d by mid-eighteenth century i n Zhong-shan, but the same i s l i k e l y true of Cantonese and other Yue d i a l e c t s . 6. The present text of the Yun-jing comes from Southern Song (1127-1279), although Pulleyblank (.1970-71:206) believes that th i s e d i t i o n represents reasonably f a i t h f u l l y a work of the late Tang period ( i . e . , l a t t e r part of the eighth century A.D.). 7. In Pulleyblank (1977:65), for example, such a treatment i s supported by the argument that the r e t r o f l e x glide - r - plays a role a f t e r dental i n i t i a l s analogous to that of - j - a f t e r - 3 3 6 -velars. 8 . The reconstructions that appear i n the paragraph i s based on a more recent theory of Pulleyblank 1s than that used i n thi s thesis. The important point, however, i s that the basic arguments presented are not affected by the change i n the reconstructing of certai n forms. 9 . Hashimoto (p. 6 4 8 ) c o n f l i c t s with other Cantonese sources (e.g., Wong, 1 9 5 4 ; Zhong-hua Xin Zi-dian, 1 9 7 6 ) i n providing [ l 0 q 2 1 ] for luo ^ ( 3 - 1 7 ) 'a mule', while agreeing with the other sources i n recording [ I D : 2 1 ] for luo ( 3 - 1 8 ) 'a conch'. As i n the word for 'mule', Hashimoto i s not i n con-sensus with other Cantonese sources on the pronunciation of the word for 'donkey', l u C 1 8 - 2 5 ) i n the Yu jJL -she. She 2 1 has [low ] , whereas Wong ( . 1 9 5 4 ) and the Zhong-hua Xm Z i -dian, for example, record [ l 0 n 2 1 ] for the same word. Lu ,1^ . 5 1 i s pronounced / l u / i n Zhong-shan. - 337 -CHAPTER 4. LEXICON (AS ARRANGED IN THE FANG-YAN DIAO-CHA ZI-BIAO). In the lexicon that follows, a few words have been added to the o r i g i n a l corpus. Where there i s a word already i n the survey that i s h i s t o r i c a l l y homophonous with the word to be included, a l e t t e r i s added to the number already assigned to a survey word. For example, lao Iffi i s assigned the number "68-1" in the survey l i s t . A word that i s h i s t o r i c a l l y homophonous with i t Cand other words i n the set) i s lao iiffi , which i s assigned the number "68-la". In the case of words added to the survey for which, there are no h i s t o r i c a l homophones, the word receives the following type of assignment. Using a concrete example, on page 63 of the survey, a R i - i n i t i a l word, r u i , i s added to the l i s t . Following the sequence of h i s t o r i c a l i n i t i a l s , the Ri i n i t i a l occurs aft e r the Chan -f^ 2 i n i t i a l but before the Jian i n i t i a l . Thus, i n front of the Jian i n i t i a l , ;.a footnote i s added. The new word i s assigned the number "68-10.1" following i n sequence after the C h a n - i n i t i a l word shui -%& (68-10). A few basic abbreviations are used i n the survey: C = C o l l o q u i a l L = L i t e r a r y N = Noun V = Verb In order to maximize easy access to the d i a l e c t survey l i s t for comparative d i a l e c t a l purposes, the o r i g i n a l pagination of the Fang-yan Diao-cha Zi-biao i s maintained. Pagination aft e r the survey l i s t continues as _if the pages i n the survey l i s t were num-bered i n accordance with, the thesis as a whole. That would also - 338 -f a c i l i t a t e the use of the index which follows the d i a l e c t survey l i s t . The index, the present facsimile of which i s from Hashimoto (1979), i s a product of the Chinese L i n g u i s t i c s Project at Prince-ton. - 339 -m n - : £ 1. t o 5 5 & 2. t'o 5 5 3. f a 5 5 K 4. f o 5 1 >~itB& . 5 . t'o 5 1 K 6. t'o 5 1 m & fi& 7. no 5 x ( c h a o : / n o 1 3 / ) S 8. l o 5 1 $ 1 9 . l o 5 1 ^ 10. l o 5 1 1 8 11. l o 5 1 ^ 12. ts'o 5 5(L.), t s ' a : j 5 5 (C.) 13. so 5 5 JL a' 14. ko 5 5 TV 15. ko b b m 17. ngo 5 1 H i a 51 *f 18. ngo f£ 19. ngo 5 1 Sn[ 20. ho 3 D M 21. ho 5 1 fa 22. ho 5 1 23. ho 5 1 * F 24. k'o 5 5 (Chao: also A ' o 5 5 / ) ^ a ~ % 25. a) o 5 5, b) a 2 2 - 340 -± 51 1. t'o m T W T ? 2. no 51-3. tso I T 4. k'o 5. ngo 13 6. ho 22 |7fJ • 5IA 7. ta:j .22 t'o 5 1 fcf. 1-51 9. no 51 10. tso 2~T 22 m ft - 341 -jt, •A ngo 22 ko 22 5 r 4. ho 22 Hi M m k'0 5 1 m n WL 6. PO 55 , 55 po , p'o 3 t -71 P'o13 9. po 5 5 55 po P'o51 55 mo 13. mo 55 14. mo33 55 mo , 51 mu 8. 10. 11. 12. (cf. 4-20) )l5 16 mo 51 17. '*& 18. 10 10 wo 51 51 ( ft JL ) ~<< 20. lo 55 - 342 -4 ft m & 1. — fit .2. so 5 5 3. so 5 5 4. so 5 5 5. s a 5 5 & m -Hi-m 6. koZ^ (cf. 6-10) 1$ 7. wo ;£ 8. ko 5 5 f4 9. f o 5 5 10. wo lf( '14 > 11. l o 1 3 , ko 1 3 (/p'o55/C.) 22 Sft 12. ngo m E f O ~ i ^ 13. wo51 ^ 14. wo51 •fg 15. wo55 M~7iP 16. wo55 S ( i'l ) 17. wo55 ± 18. p*o i j (L.), p a j 5 5 (C.) (1) & 19. po 2 2 (cf. 5-20) M 20. p'o 1 3 (1) Chen says that-/p'o / is rarely used. He only gave /paj / on the first elicitation. - 3 4 3 -1. t o 1 3 2. t o 1 3 3. t'o • 13 4. t'o 5. to (Ball: /to/ for both Zhong-shan & Cantonese) 7. to 22 m ft JL as 8. lo 9. 13 10. t s o 2 2 (L.), ts'o 1 3 (C.) (cf. 6-9) 11. so 5Jt~5£ 12. so 13 13 13. ko 14. ko 13 13 15. ko' 13 I I $c 16. lo 13 m 17. fo •X ) 1 8 . fo 19. wo I T 13 22 % 2 0 -21. 22. ffi ( * po^ po 22 22 22 p'o ) 23 — a &- 24. pok a) mo (cf 5 1 , b) mo 169-8) 22 (cf. 3-13) - 344 -6 m M i - t o 1 3 (/tiok2 c.) - 22 i S k § ~ f £ > 2. t'o + ( f f ) 3. 2fc M ~ 4. — m a* s. ts-o^ s± 6. ts'o £~JfE - ® 7. — H 8. tso ^ 9. tso 2 2 (L.), ts'o 1 3 (C.) (cf. 5-10) % rH S¥ ^ 10. ko 2 2 It 11. f o 2 2 & 12. ngo 2 2 PI K . 1 3 . f o 2 2 ^ P ^ f j 14. wo51 (cf. 4-13) JL l ¥ IS 16. — «fc 17. ho^ 18. — - 3 4 5 -55 1. pa 55 2. Pa 55 3. Pa 2fc 4. 5. .p'a' P'a 6. p'a' 51 51 51 SML 8. ma 9. ma 51 51 p'a" 51 10. mok 11. ma ( * ) l 2 . na 3 T 13. ts'a 51 & ( R ) l 4 . a t s ' a 5 1 (but (a) is / f u 5 1 / — cf. 14-14) 0J : 0 J ~ ( ) 15. t s a 5 5 : ( f e ) l 6 . t s a 5 5 lb 17. ts'a 55 18. ts'a' ^ 19 20. — 55 ts'a 55 22. 23. sa sa 21. 55 i 55 ts'a 51 - 346 -1. ka 2. ka' 3. ka: 4. ka n 5 55 55 55 55 ka 55 3P 6. 7. nga c nga" 51 i 51 8. nga' 51 f . f / J N ^ 1*3?)' 9. nga 55 is dig 10. 11. 12. 13. 14. 15. ha 55 ha ha; ha1 ha' ha! 55 51 51 51 55 16. 17. 18. "55" 55 i 55 —: It ± 19. pa 13 Hi 20. ma' 13 ( -1 ) 21. ma 13 22. sa' 13 - 347 -ka ka 13 13 nga 13 T i g - 22 13 4. ha , ha^ (1) 5. 6. 7. ha' 22 ha 22 (cf. 10-13) => • B S C 8. pa 22 22 ^ ^ 9. pa' 10. pa 2 2 t % j l | 11- — •fe 12. pa i < 22 ta 13. p'a" **R(lpfi) 14. p'a 2 2 £ . ~tt!l l ffi ) 15. p'a 5 1 16. ma' 22 f i 22 17. t s ' a " |£ 7K§ 18. ts'a 2 2 (L.), t s ' a 5 5 (C.) (2) 2T 22 19. tsa Li] #~ <ft 20. tsa 21. tsa 22 ^ K f t f f i 23. ts'a 2 2 24. tsa' 22 > Id Jgg&~ 25. ha' 22 (1) /ha 1 3/ is used in e.g., /jat 5 ha 1 3/ - ~p 'once', A a 5 5 ha 1 3/ "S- 'new'. 22 (2) Chen gave /tsa / as the literary reading the second time. - 348 -10 Ji fg&^ 1. k a 2 2 ^ 2. k a 2 2 as 22 S5 3. ka h& 22 4. ka ^ 5. ka /as 22 In. 6. ka ' is ' • 7 - — iif — 8 . — • 22 §<F^~ 9. nga 22 jSjgw 10. nga IS # ~ _ g f c l l . ha:k 2 (cf. 194-24) T ~ j$ - ± ~ 12. h a 2 2 W 22 13. ha HE , , ' 22 14. ha J£ S. 5. a 2 2 iff rt ft Hi- 16. t s i a 5 5 JH 17. s i a 5 5 3 f l 18. t s ' i a 5 1 §4 19. t s ' i a 5 1 m 20. (Tliis character should not occur here.) - 349 -11 1. 2. 3. 4. 5. tsia 55 ts'ia . 51 sia ts 1 ia' 55 55 s i a 55 51 51 sia (but (a) is /y /— cf. 20-3) . 51 7. Da . 51 8. Da . 51 9. Da Da 55 (1) ± m m ft 10. t s i a 1 3 11. t s ' i a 1 3 .13 12. sia - i r 13. tsia M ( j | ) l 4 . ts'ia 13 13 >9l 15. 16. sia . 13 sia B 17. ja 13 WO ft 19. ja' 20. ja' 18. ja 13 13 13 (Ball: /ya/ and /ye/) 55 2 55 ;/ (1) /ja / is used in /pa:k ja / )Q %f 'father' (used by some) . 5 - 350 -12 l i ft /i± ^ . 22 •IB 1. tsia I i 2. t s ' i a 2 2 (/ts'022/ C.) * § - ~ 5 X 3. tsik 2 (cf. 203-9) 4. tsia:k 2 . 22 m 5. sia * P • 2 2 ZtjJ 6. sia . 22 ESf 7. tsia m . = n j . • 22 i±F, 8. tsia 22 ii'f 9. sia 10. s i a 2 2 . 22 11. sia 4b , ~ .22 is 12. sia •fetij: ETJ ^ Fo 1 3 • ~~ W. 14. :a - : I t IK Ol 15. — JL A T JA 16. kwa55 ^ 17. wo55 (Chao: /o55/) 18. k'wa55 C^) 19. wa55 - 351 -13 1. fa 11 2. 51 wa im 3. wa51 4. 55 51 ko (/p'a / C.) 5. 55 wa 6. 55 wa 3 i i , I JfE ± ® LU 7. 51 so CM) 8. 13 sa % 9. kwa 10. / r l •if (* m) 11. k'wa55 •12. k'wa 13. 13 nga 14. k'wa55 JCtfjfs]15. n g a 1 3 (cf. 13-13) it 16. fa 2 2 " ,\h ' & 17. wa M 18. wa51 22 - 352 -14 V3. ~0~. pu 5 5 (Chao) ] 1. p'u 5 5 n , 51 2. p'u 3. p'u A , 22 , 13 j g g g - — - 4- p u , p'u , „ . 51 5. mu 51 6. mu 51 mu 13 M ( m ) p'u 55 Is -3E 7. 8. 5 1 n mo (L. mo 55 (L.) , mo (C.) j - a i • TS. 9. t ' 10. tu 55 55 A C 8fc 11. 12. 13. 14. 15. t'U fu; t'u1 t ' U t'u 51 51 51 51 51 51 51 51 51 51 51 16. 17. 18. 19. 20. 21. 22. nu nu lu lu ! lu lu-lu 1 fit 23. tsu 55" ( k ) 2 4 . ts'u 55 24.1 25. su 26. su 27. su ts'u 55 55 55 51 (Chao). ] - 3 5 3 -** 1 . ku 5 5 1 5 m- 2. ku 5 5 « 3 . k'u 5 5 * 4 . ku 5 5 & 5 . ku 5 5 (Chao: A ' u 5 5 / ; Ball: Au/) (.Ball: for a l l present /ng/ syllables: /ung/ for Zhongshan, /ng/ for Cantonese) 15 5 1 6 . ng 7 . 51 ng 5 1 8 . ng 9 . 5 1 nq 1 0 . hu 5 5 ® 1 1 . hu 5 1 m 1 2 . hu ffl 1 3 . i 5 1 hu n 1 4 . hu 1 5 . hu 1 6 . u 51 hu 1 7 . hu 1 8 . hu S8 1 9 . u 5 1 hu (1) v. 51 51 k'u , u (2) a? 20. u » 21. u 5 5 * » 2 2 . u 5 5 43" M l ± PS 23. 24. pu i J p'u &. 25. , 13 p'u 26. , 13 p'u I f 27. 28. 22 pu 22 pu (1) Recorded by Ball as /wu./ for Cantonese, and /fu/ for Zhong-shan. This is the only occurrence of syllable /fu/, which normally should not be found in Zhong-shan. (2) A ' u 5 1 / 'paste (edible), pudding'; /u 5 1/ 'paste, glue; to paste", - 354 -16 ± 1. 2. 4. 5. 6. 7. 4. 1 3 t u 4- 1 3 t u 3. 4-- 1 3 t ' u t ' u t u 2 2 t ' u 1 3 t ' u ' 13 (cf. 16-7) n u 9. )lO. 11. 12. 13. l u " l u -l u l u -l u ' 13 13 13 13 13 13 m. 14. 15. t s u t s u 13 13 16. u I3 k u -ft 17. k u 18. — • ft 19. . 13 k u as: 20. 1 1 3 k u 13 i x 21. k u 22. 1 I3 k u • f t 23. 1 55 k u 24. u 1 3 h u (Chao: /k'u13/) s. 25. • n g • I S 26. n g 27. ^ n g 13 13 13 - 355 -17 E 5? 1. hu 1 3 if^ 7k. -w 2. hu 1 3 & 3. hu 1 3 Z 5 4. hu 2 2 m s. hu 2 2 «S 6. u m m 7. pu" ffi 8. pu 2 2 22 Wg.^, 10. pu 2 2 ^ 11. 'pu22 St 12. pu 2 2 O a ) 13. faw2 2 (but (a) is /pu 2 2/—cf. 17-11) & 14. mu22 3£ K 2 2 ^ 15. mu I 2 2 16. mu # 1 7 - , , 2 2 18. tai" 22 M. 19. t u " P±pg^ 20. t'u 2 2 H 21. f u 2 2 F f r 22 £ 22. tu M 23. t u 2 2 22 ^ 24. t u " % m ifcX „ 22 25. nu & 26. l u 2 2 I& 27. l u 2 2 S 28. l u 2 2 M~M 29. l u 2 2 - 356 -18 *® ( AO 2. ts'u 3. ts'u 4. ts'o 22 5. su 22 6. su 22 7. su 8. 22 su tsu' 22 22 22 22 @ 9. 10. . 22 ku ku ijil,—• 11. L. 22 ku M 12. v, 22 ku m 13. 14. ku k'u 2 2 (L.), hu 1 3 (C.) (l) ny # ( 15. , 22 hu is 16. 22 ng tit 17. 18. 22 ng 22 ng m n ~ 19. v, 2 2 hu s - 20. u 2 2 hu n 21. u 2 2 hu jj 22. hu 5 1 (cf. 15-16) 22. 22 u -llj 24.. lu' 25. lu 51 51 (1) / h i i 1 3 / occurs in the name of the village /hu 1 3 ts'ung 5 5/ - 357 -19 lint* m rt ft £ ~ 1. • tsy 5 5, t s u 1 3 (Ball: /ch'ii/) :fl=L * ~ 2. tsy 5 5 (Chao: also /tsy 5 5/) 7 E _ 55 # 3. sy & 4. ts'y 5 1 & & 5. t s y b 5 Bfc 6. ts'y 5 1 7. sv 5 1 (/ts'u13/C.) iff O J ^ 8. ts'o 5 5 8ft 9. ts'o 5 1 (N.), t s o 5 1 (V.) 10. so 5 5 M~ar i i . so 5 5 ^ 12. so 5 5 BE ' i n i# I f It 13. t s y b 5 • 14. sy 5 5 '5? 15. sy 5 5 S W 16. ngy 5 1 rl m J£ 17. k'y 3 0 i f L ^ ^ g ^ 18. ky 5 5 ^—- rfT 19. hy 5 5 2g 20. k'y 5 1 f&flfe 21. k'y 5 1 & 22. ngy 5 1 M 23. ngy 5 1 m ^ 24. hy 5 5 iSgfc^ 25. hy 5 5 - 358 -20 tik 2. y , u 5 (1) £ 3. y 5 1 «* 4. y 5 1 5. -± era m 6. ny i J (Ball: /nil/ and /ngu/ on two separate pages.) fgg?£ 8. --fi* 9. l y 1 3 iS 10. l y 1 3 m ik & % _ 22 # 11. tsy & 12. ts y 2 2 ft 13. sy 1 3 & it |f 14. t y X J (Chao: /ty X J/, /ts'yx"7) 15. sy 1 3 <x~JK 16. t y 1 3 B %] Hi PI • 17. tso 1 J 7 § 18. ts'o 1 3 UttTS 19. ts'o 1 3 #f 20. so 1 3 (1) /y / 'muddy', /u / 'dirty'. - 359 -21 Bg ft 1. tsy 1 3 2. t s ' y 1 3 # 3. - (1) ^ 4. sy If, 5. sy * . 1 3 6. sy K — 7. s y 1 3 a 8. n i 1 J (Ball: /ngti/) JL m m |§ 9. ky i j (Chao: A y 1 3 / and A y J J / ) ' S 10. ky 2 2 #1 11. k'y 1 3 12. k'y 1 3 m ,, 13 p a 13. ngy > m rF 14. hy 1 3 B , ^ 15. y 1 3 fa M M 16. l y 2 2 (Ball: /Iii/ or /loi/) * M 17. l y 2 2 (Ball: /Iii/ or /loi/) lice rt ft & 18. s y 1 3 ' (1) Chen gives /ng / for the word, as in /ng tsok lew / 'undertaker' (alternative term: A u n ts'oj low / %$kk\£). He gives the character for the meaning of 'to pierce or jab with a long object', and the pronunciation of /ts'y 1 3/ for i t . - 360 -22 [ i l 1. tsy 22 • H i 1 2. tsy 22 3. tso 22 4. so 5 5 (cf. 19-11) (Chao: /so"/ (1) 22 I* '0f 5. ts'y 22 22 sy I S ' 6. 7. sy ' & 9. sy 22 51 13 13 8. sy (Chao: also /sy /) M 10. ky 22 22 22 fg^^i 11. ky ( A a:j / and A 0 / (C 22 £ 3fc~ > ~ H 12. hy 22 M 13. ky" $9 14. ngy 22 15. ngy '(2) ) 9> 16. 22 y 17. 22 y 18. 22 y 19. 22 y (1) /so"/ society .22 2 2 / in the word /siong 2 2 so 2 2/ J l 'upper stream, high 22 (2) /ka:j / is used for large objects, and A 0 / for small ones - 361 -23 1. 2. hu ' 3. 3k 4. hu 5 5 5. hu 5 5 » 6 . ^ 7. hu ft 8. hu & 9. hu 10. hu 5 1 /TUT, 7m 11. ma51 12. 13. 51 mu 51 mu 14. 51 ( B a l l : / m o / — i . e . , / m o w / ) m * 15. ts'y /(!•> 1 16. 17. 55 sy 55 su 18. sy 5 5 £ P 19. 4- 5 5 tsy m 20. ts y 5 5 21. tsy d m 22. ts'y 5 1 m 23. ts'o 5 5 m 24. ts'o 5 5 - 3 6 2 -2 4 SB IP >fc 1 . tsy W 2. t s y 5 5 * 3 . t s y 5 5 H 4 . k'y 5 5 ( C h a o : a l s o A'y 5 5/) $ f c 5 . sy 5 5 & 6 . sy 5 1 B I S 7 . y . H m #J 8. k'y" M 9 . k'y 5 5 1 0 . k'y 5 5 1 1 . k'y 5 5 H 1 2 . k'y 5 5 II 1 3 . k'y 5 1 m. , A 5 1 di 1 4 . ngy M 1 5 . ngy $ 1 1 6 . ngy 5 1 E I f f 1 7 . hy" JB 1 ? 18, hy 5 5, y 5 5 T 1 9 . y55-JS 2 0 . y 5 1 ( 1 ) 2 1 . y 5 1 M 2 2 . y 2 2 M 2 3 . y 2 2 srb 22 \ $ . 2 4 . y ( 1 ) E . g . , ^ ^ . /t'a:m51 y / 'spittoon'. - 3 6 3 ± Of 5<: hu hu hu 1 3 P'u13 C , l 3 5. p'u 13 pu v, 13 hu 6. 7. 9. hu 10. 11. 12. 13. hu' hu' hu mu 22 13 22 22 13 1 ,13 (Chao: /hu13/) 14. mu^ .13 muj 13 15. 16. mu"1 17. law' 13 & 18. ts'y 1 3 +1* 19. t s y 2 2 (L.), ts'y 1 3 (C E T * 22 fH 20. tsy 21. ft 22. ts'y 1 3 M 23. su' 13 - 364 -26 0 1. tsy 13 sy 3. .22 r±3-f m — i i i 4. J0j IS 13-(Chao: also /j0j /, Ball: /yu/; ky 13 6. y 7. 13 y m 8. 13 y 13 9. y 10 22 ft- n. 12. U 2 2 hu hu 13. 14. V. 2 2 hu , 22 hu th 15. v, 2 2 hu m 16. v, 2 2 hu 17. 22 m mu 18. 22 mu 19. luj .13 22 20. ts'y 21. tsukZ (cf. 225-17) - 365 -27 ftfc W ft 1. tsy 2. tsy 22 22 tsy 22 su 22 72" 22 22 22 ft 5. 6. 7. tsy tsy' tsy tsy ft; 9. 10. 11. sy c sy~ 2 _sy_ 22 55 r 22 (cf. 24-5) (1) H, 12. k 22 13. 14. 15. 16. ky ky 22 22 n g Y „ ngy^ 22 r 22 17 18 19 22 u y 1 3 (cf. 26-9) y 2 2 (Ball: /ngu/) on 22 20. y oi 22 2 1 . y 2 7 - II-1 (1) £) i n i t i a l : A J $ /y b l/, /y11/ (Chao) 51 22 28 • MP ( f l ) 1. ta:j 55-J£ 2. t ' o j 5 5 H 5 ^ ~ ' ~ M - ~3C 3. t ' o j 5 1 t'oj' .51 ^ ' -ff~ 5. t'oj & ) 6. t ' o j 5 1 .51 -, i - 5 1 7. 103 I — 3 5 — tsoj 4. - 5 5 tsoi . • 5 5 ' 51 8g 8. 9. 10. 11. 12. ts'oy ts'oj 51 13. ts'oj .51 14. ts'oj 15. — .51 16. 17. so3c soj" .55 .55 18. 19. koj hoj 55 ^ ( t ) 20. ngo j m 21. — 51 I* 22. , .51 , .51 ho;j , ha:] :53 •a mm, 23. 24. 25. a:} i a:j-.55 (1) .55 (1) Both /a:j / and /oj / were given on the first reading, but /oj was subsequently rejected. / a : j 5 5 / occurs in the colloquial word /la:k 2'a:j 5 5/ O 'dirty'. 'Dust' is /puk2 ts'an 5 1/ O A - 367 -29 MJJ Kffif ( n l )2. — 3. 4. 5. 4- 2 2 to: t'o: 4 - . -'13 t'o: 7^ 6. .13 na: j — — r ± 3 ~ tsoj 4. - 1 3 tsoj 4. - ' 1 3 ts'oj t s ' o j 1 3 ts'03 . .22 tso: m m 7. 8. 9. 10. 11. 12. (L.), t s ' o j 1 3 (C.) 7 T 3 -13. koj •13 14. oj .13 is. o: 16. hoj 7IT 17. hoj .22 - : P £ ft M U m 18. 19. 20. •' .22 ta:: 4.1 t'a:: t ' a : j 2 2 (Chao: / t ' a : j 2 2 / , /toj 2 2/) ft 21. 22. 4. - 2 2 toj 4. t o : — — — — - 368 -30 .22 1. no] 1.1. l o j 2 2 (Qiao) ] ft fo 2. tsoj 2T 3. 4. 5. 6. 7. tsoj .22 .22 ts'03 fc .22 tso] .22 S O ] (cf. 30-3) (1) 8. 9. k'oj k'oj 2 2 -22 11 !fe—• 10. k'oj 22 * K $ 0 i i . 12. ngoi. k ' a f .22 13. oj R ft 14. 15. r22 , .22' ft 16. — * 9 . " ft 17. 18. 19. T TZrZ —• ta:] t'a:] , .22 t'a:j .22 § !£ 20. tao aa -—-If 1 21. noj 22. 23. la: j la:j 22 22 (1) 30-7a. % /soj 2 2/. - 3 6 9 -31 #s m ft 1 . 4. . -22 ts'o] fo w E :M rrrt. 2 . 3 . 4 . k'oj Z Z (L.) , k o j Z Z (C.) v. -22 k'o] i i -22 k1 cr) 5 . 22 22 ngoj (Chao: /nga:j /) m * 6 . , . 22 ho] 7 . 8 . o j l j .13 .Im—• • ^ ft 10 9 . pa:] . 5 1 ma:] •55" 1 1 . tsa:j" 1 2 . ts'a:j 35" . 5 1 E 1 3 . 1 4 . 1 5 . ka: j ka: j ka: j 55 55 1 6 . ka:] 1 7 . ka:] 1 8 . k'a:j 5 5 - 370 -32 m m i u -51 5g 1. ha:] Si- o u '51 u -51 fx 2. ha:] , ho] ' 3. a : j 5 5 ± + 4. — & + SKBB±W~5& 5. — IHA w % fg1 6. ka:j 5 5 (Chao: A'a:j 1 3/) m E 7. ha:j 1 3 ts n ?f- 8. pa:] ft I i i 9. — i l 10. p i 2 2 w - 371 -W fo-lk ft 1. ka: j n 2. ka: j ft 3. ka: j 4. ka: j .5. ka: j M 6. ka: j 7. ka: j 8. ka: j 22 22 22 2^2 22 22 ,22 22 9. ha:j .22 m 10. P'a:j 11. — .51 m 12. ts a: .55 D 13. ts a: .55 D 14. ts a: .51 J m~^- ( n ) is. 16. ka:j sa] 55-.55 17. ka:j .55 18. nga:j .51 • 5 1 19. nga: j 51 55 'fi ' ~ H 20. nga:j , nga:j (1) 33 21. ha:j 51 (1) /nga:j 5 5/ in /nga:j 5 5 t s i 1 3 / $Jk 3r 'very poor people living near Shi-qi'. - 3 7 2 -34 ± m 3H 1. pa:] ft 2. pa 2 2 .=» „ .13 .Pi 3. ma:] ^ ( ^ ) 4. na:j 1 J m w U-ih 5. s a 1 3 (cf. 8-22) E :M m > ~ § S 6. ka:j X J m 7. ka:j 1 3 8. ka:j 1 3 g 9. ha:j 1 3 & ^ 10. a : j 1 3 ( L . ) , a j 1 3 (C.) : H m n. , .22 pa:] 12 13. .22 ma:] iff 14. tsa:] 0J 81 15. .22 sa:] - 373 1. h a : j 22 2. a j 22" m - : * 3. 4. p a : ^ m a : j ' .22 ] .22 5. — 6. t s a : j 22 ftfc 7. p a j 122-4* .22 8. p a n . .22 9. p a ] 10. .22 w p a ] 11. .22 H p a ] §1 .12. .22 p a ] - 374 36 m m 1. I a j 2 2 22 2. la] J ® 3. I a j 2 2 i f m & zre 4. tsa] m. .22 r& 5. tsa] 6. tsaj & m Mm 22 W&~ , S ~ 7. tsaj BE t \ \\ m /# $? ffi 8. tsaj* 2 $H 9. t s a j 2 2 -W 10. s a j 2 2 ^ 11. sa] =>=* .22 ¥ 1 2 . s a j z z & 13. s a j 2 2 f¥ " S J 14. nga] JL r l XlJ 15. — (Chao: /nga:j22/) - 375 -37 JL Mr 5* 1. 2. .55 Pi pa] .55 3. maj 51 m 4. 5. 6. 7. 8. 9. 10. taj t 1 aj t'aj t'aj t'aj f a j t'aj 55 .51 155 51 51 51 51 i i . 12. na] .51 Iaj .51 51 ft w m m 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. ts'aj ts'aj ts'aj ts'aj • i 5 ,55 ,55 55 55 :si . , .51 ts' i saj' x. . -55 ts'aj .55 saj s i 5 5 kaj k'aj 755" 55 k'aj .55 26. ngaj 51 27. haj 28. haj 51 51 - 376 -38 ± MIX •T l . 2. pa:_ .22 .13 m 3. 4. 5. 6. ta: .13 .13 13 ta: f a j *. '22 4., -13 ta: / f a : (1) laj 13 8. tsaj 55^ 9. ts' a j •13 10. sa: .51 (2) ' a 11. k'aj 13 1! 12. 13. .22 pa: -.22 pa: + i l ~ 7 ] f t , M 15. £ 7 ] .51 ma: . 14. — 22 (Chao: /maj /) (1) / t ' a j 1 3 / in / a 2 2 t'aj 1 3/_$-j? 1 'younger brother'—used by older generation. (2) 38-9a. / t s ' a j 1 3 / (Chao) - .377 --MJU 1. t a j 2 2 (1) W 2. f a ] 3. f a j 2 2 flj • 4. f a j 2 2 5 5 , I I — 5. • t ' i — 22 *£ 6. t'aj » 4. -22 13 7. ta] • 22 JS 8. -taj m M | § ~ 9. l a j 2 2 - 22 H 10. t a j z z l i m ft ?p ^ 11. t s a j Z z •ggj 12. t s ' a j 2 2 ffl-~m , 13. t s a j 5 5 I B 14. s a j 2 2 *r?£~ 15. s a j 2 2 JL & m f f 16. k a j z z B 17. k a j 2 2 S ? ~ 1 8 - ha j 2 2 (cf. 39-23) „ 22 # 19. k a j z z 20. k'aj 2 2 fg 21. ngaj 2 2 E ^ 22. h a j 2 2 |g Si—• _ 23. haj 24. h a j 2 2 ' £ 25. a j 2 2 so. .22 26. a] (1) 39-la. Bj>L- / t s ' i / y t s ' L / . loathe present data, i t occurs in the dissyllabic word /hait 3 t s ' i / O "fT ' a sneeze', / t s ' i / is recorded by Chao, as well as occurring in the word /p'an22 t s ' i 2 2 / «|f *jMcf. 158-13) 'to sneeze'. - 378 -40 m 1. puj 3 3 2. p ' i 5 5 5 5 J* + ~ 3. p'i , faw33 (Chao: /p'uj /) 51 i§ 4. p'uj m 5. p'uj 5 1 BS(PS) 6. p'uj 5 1 $| 8. muj 5 1 ft 9. muj 5 1 10. muj 5 1 ^ 11. muj 5 1 ik. 12. muj 5 1 M U m — — — — j# 13. t u j 3 3 #£ 14. t ' u j 5 5 *I~jf 15. t ' u j 5 1 & If 16. l u j 5 1 IS ft fi 17. ts ' u j 5 5 ^ t t 18. t s ' u j 5 5 m m —_ —— 22 ——— — Ij| 19. kwa] 15 20. — 21. huj 5 5 22. huj 5 5 ffc 23. huj 5 5 ft&JS^ff 24. waj 5 1 - 379 -41 m m 1K m m i . 2. 3. 4. hir) m 5. .55 to 1 ± 6. p'uj 13 7. muj :13 8. t ' u j 1 3 9. nuj I T 10. luj .13 fo $p. 11. tsuj 22 '{i 12. k'waj22, f a : j 2 2 13. 14. 15. huj . .22 hu] .22 - 380 -42 & _£ m Jfl 1. pu] W 2. puj 2 2 22 @_ 3. p'uj" ft 4. p'uj 2 2 F f ~ l t 5. puj 2 2 (cf. 42-2) + ft_f~f£; 6. puj 2 2 If 7. put2 2^ 8. mu_ 3^ 9. mui 2 2 — m io. t u j 2 2 m i i . -13. t ' u j 2 2 m i4. t u i 2 2 m m F*3 15. noj 2 2 I S 1 ( # ) 16. 10j 2 2 (cf. 62-6) ^ 17. — m ft 22 18. suj E J$ 19. f a : j 2 2 m 20. huj 2 2 (Ball: /nei/) H 21. huj 2 2 (Ball: /h6i/) 22. k'uj 2 2 - 381 -43 m £ . m~ 1. t ' u j 2 2 , t u j 2 2 # 2. t u j 2 2 — T > IS w ft fo B _ .zz — . ^ 3. tsirj m m # ~ f t 4. u j " (cf. 43-8) #J 5. k'uj 2 2 & 6. k'uj 2 2 ft 7. ngoj 2 2 * # § § ~ 8. u j 2 2 #~^~9. U j 2 2 (L . ) , waj 1 3 (C.) (Chao: /waj 2 2/ C.) /It 10. k'uj 2 2 — f * « . 9|E 11. kwa:j^ : @ 12. waj 5 1 (Chao: /wa:j51/) %g 13. waj 5 1 r t 14. waj 5 1 - 3 8 2 -4 4 tS m w E m 2 2 fS 2 . kwa:j + 3 . fa : j 2 2 IU ' 4 . -m m J*. c '2 2 j|| 5 . wa:j E 6 . wa:j b b & to JH 7 . wa (cf. 1 3 - 5 ) ± E 1 3 8 . kwa:j - 383 -4 5 : H £ m i . 2. —— 22 : ' '-k'wa 22 kwa m E m. 3. wa:k 2 (V. ) , w a 2 2 (N.) SI 3¥ ft m 4. 5. f a : j 2 2 k'wa:] E 6. 22 wa 7T\ f i m ft 7. t s ' ( _ j 2 2 8. 9. M - 2 2 S(_] • 22 22 wa] (Chao: a l so /wai /) (cf. 46-6a) w 10. 22 2 t s s _ j " , tsyt: - 3 8 4 -4 6 SE J\\\ ' » 1 . t s 0 j 2 2 I S - 2 . s 0 j 2 2 2 2 3 . s0j B 4 . — J L M 5. — " m |§J 6 . waj 2 2 ( 1 ) ft 7. W2 m JS 8 . f a j 2 2 W 9 . f a j 2 2 ^ 1 0 . f a j 2 2 m • t o n - 2 2 M 1 1 . waj ^ -i i . 1 2 . kwaj 5 5 H 1 3 . kwaj 5 5 f^e: , . 55 3 E 1 4 . h u j ( 1 ) 4 6 - 6 a . j ^ - / w a j 2 2 / . - 385 -47 m 1. k'waj 2. k'vaj .51 51 m 3. kwaj 4. 5. IT .22 w a j .22 #1 6. pi • # 7. p i 5 5 • 8. p i 1 3 55 ft 9. 10. 11. * W 12. f ~ ^ 13. • 3H 15. ' + H f ! r S ( P'i , .51 P i , .51 P i , .51 p'i 14. mi 51 51 ni ) 16. —Si) 17. H 18. 19. ^ r g j ~ 20. l i l i 51 51 51 55 l i (L.), l i (C. Iaj 22 (cf. 39-9) - 386 -48 m i . t s ' i &Vk¥;$i~2. t s ' i #f 3 J$r 4 + $f(#f) 5 55 55 .55 si .55 si .55 si 16. 17. s i 5 5 18. s i 5 1 , t s ' i 5 1 (2) S ft 19. ngi^ 1 JL m 20. . .Ob ki m % m 21. 22. i,f51 k'l , . • 51 k'ia 23. k'l • IS 24. , , .51 k'l m -H 25. 26. .51 ngi .51 ngi (1) (2) 55 22 55 *4 A i / occurs i n the combiiiation / l a : j k i / #g ^ i - ' lychees' 51 51 55 0 $~ / t s ' i / occurs, i n the craifoination / t s ' i kang / rf^ £ 'spoon1. - 3 8 7 -49 ?8 1 . hx m • & 2 . i 5 1 ± & ft m & 3 . p i 1 3 • ^ 4 . p i 1 3 5 . p ' i 1 3 6 . p ' i 1 3 f l 7 . - -•3? 8 . n i 2 2 IS w >£> ?P m - 9 . t s i 1 3 itfc 1 0 . t s ' i 1 3 ffi . 1 3 M. 1 1 . sa:j 1 2 . s a : j 1 3 m W 1 3 . t s i 1 3 R ~ W 1 4 - t s i 1 3 &m~ 1 5 . t s ' i 1 3 S f t i a i S ^ ^ 1 6 - l a j 1 3 ( L . ) , l a : j 1 3 ( C . ) m 1 7 . t s ' i 1 3 B 2 2 M: 1 8 . s i J5: 1 9 . s i 2 2 a i f 2 0 . n g i 1 3 m m - 388 -50 13 • £ 1. k'i ^ f t 2. k ' i 1 3 a 3. k i 2 2 & 4. k i 2 2 £8- 5. ngaj 1 3 m 0 i i ft 6. i 1 3 fa(ft)7. i 1 3 & ft m •W 8. p i ^ • W~m9. p ' i 2 2 , ~m io. p i 2 2 • *8 11. p i 2 2 m m HWft 12. l a : j 2 2 (1) g£~|fl^ 13. l i 5 1 (cf. 47-17) IS m ft fo $| 13.1. tsik 0 (Ball: /tsik/, tsek/) (cf. 2UJ-5) $•? 14. t s ' i 2 2 , ts'ia:k 2 (cf. 203-7.1) 14.1. (Ball: /tsiek/) (2) ] m 15. t s ' i 2 2 ft] f t ^ 16. t s i 2 2 se /# 1? M 17. t s ' i 2 2 18. s i 2 2 H 22 55 (1) The combination is pronounced/la: j ki / (cf. 48-13). (2) Zi ;$L 'to soak, to dye' is recorded by Ball as /tsiek/ when i t mians "spots', suggesting a possible parallel with ci$»] (50-14 and j i \\ (50-13.1), both of which have a Ru-sheng counterpart •i n f h o (Zar\n— c h » = . ~ 389 -m *F i v 2 2 1.. ki ^ 0 .51 s_L 2. ngi * - .22 3. ngi 3 * . .13 53$ 4. ngi m "jft 5. h i " " ; 0fxix r=> 22 • 6. l itm= : la ¥ If rf __: m m. 7. P i b b 2 8. p ' i 5 5 • ! § ~ § 9. p ' i 5 1 10. p ' i 5 1 ^ 11. mi 5 1 m 12. mi 5 1 ^ 13. mi 5 1 14. n a j b i (Chao: /ni^V) I 15. n i 5 5 ^ 16. l i 5 1 IS ft ft X 17. t s i w lc 18. t s i 5 5 $ 19. t s i 5 5 ' I£~g 20. t s ' i 5 1 21. — E 22. s i 5 5 - 3 9 0 -5 2 ® l i f i . , . 5 1 2. 3 . 4 . sx 55 si 55 saj . 55 m 5. £ t ~ i t 6. tsi" . 5 5 P 7 . 8 . s i s i 55 L . 5 5 f l ~ e 9'. * i i -I P 1 2 . 1 3 . , . 5 5 k i 55 ki i i - 5 1 k ' l . . . 5 1 k ' l i , i - 5 1 k ' l p 1 4 . i 5 5 * . % 1 5 . . 1 7 . . 5 1 l . 5 1 . 5 5 . . . 1 , 1 (1) . 5 1 l (1) / i 5 5 / in e.g., / a 2 2 i 5 5 / J£- 'mother's younger sister' - 391 -5 3 ± PJ3 lh' 1. p ' i 1 3 (1) ; ^ - t 2 . p i 1 3 . — i g 3 . p i 1 3 4t 4. p'an13 (cf. 151-7) .13 5. mi i 13 6. ly I T ffi fo 3 & p 7. tsi . 13 8. s i 9. 4- . - 2 2 t s ' i t i l H 10. tsi 11. 4- ' 1 3 t s i m 12. 13. ts' 1 .13 si 14. ki t s ' i 5 1 (2) 3-3 53-la. /maw55/ in the combination / t i 2 2 maw55/ ' 4& j$ (1) 'ruffians, bullies' (Chao). It also occurs as the Grade III counterpart of pin $\Li (53-4) . Also pronounced as /p'i /. 55 51 (2) Chen gave / t s ' i / and / t s ' i /the second time. - 392 -5 4 _£ m m i . p i 2 2 & ' 2. p i 2 2 3. — • XEE 4. p i 2 2 r -22 • SE,«&~ 5. pi • •22 • Jit 6 . p'i. «_ • 22 i i 7 . pi . *. 8. p i 2 2 , pat 2 (1) • H(#fc) 9. p i 2 2 £f 10. mi 2 2 • U -11- rai22 m % 12. t i 2 2 % M 13. n i " (L.), naj -" (C.) * i 14. l i 2 2 15. l i 2 2 m ft . 22 16. t s ' i .22 g 17. tsi . 22 jjg 18. si . 22 HI 19. s i m •st 20. t s i " 21. t s i 2 2 (1) /pat 2/ 'elephant's trunk'. 4. 2 2 1. tsi 7 F IS. 2. si 3. 22 ;1I pf 4. 5. .22 si .22 si B 6. 7. .22 ngi .22 ngi 8. i , . - 2 2 k'i I I 9. . .22 hi •• n 10. , .22 hi - 11. .22 1 t 12- .22 l to m 13. l i 18. 19. 20. 21. 22. 23. 24. 25. 51 S I P S S 14. l i 5 1 -4-4- 15. t s i 5 5 16. t s i 5 5 17. 4. ' 5 5 t s i t s ' i 5 1 t s ' i 5 1 .55 s i ,55 s i S I ,55 . ,.51 t s ' i . ,,51 t s ' i . ,.51 ts ' i - 394 -ft 56 1. 2. . , .55 t s ' i . ,.51 t s ' i W fo t s i 55" 2 f t 4. 5. . -5b t s i tsi 55 6. t s ' i .55 7. s i .51 8. s i .51 9. s i 55 10. ngi 51 11. . .55 k i 12. ki 13 55 . .55 k i 14. 15. 16. h i v. -51 k ' i k'i 17. k'i k'i k'i 51 51 51 51 18. 19. •51 20. n g i m E E %2k m m r 21. 22. 23. 24. 25. 27. 28. u-55 h i K '55 h i h i V , ' 5 5 h i "53-26. .51 i .51 ,51 - 3 9 5 -57 ± ih fa 1 . n i 1 3 $ 2 . l i 1 3 a 3 . u 1 3 X - 4 . l i 1 3 S 5 . l i 1 3 II 6 . l i 1 3 m ft & 7 . t s i 1 3 | $ 8 . t s i 1 3 u 9- 13,113 * E S ~ 1 0 . t s i 2 2 E S ^ i l . t s i 2 2 ft He It 1 2 . t s ' i 1 3 1 3 . t s i 2 2 1 4 . t s i 2 2 B W U-ik ^ 1 5 . t s o j i J ( 1 ) ± 1 6 . s i 2 2 tt 1 7 . s i 2 2 ffi 1 8 . t s ' i 1 3 2 2 $1 1 9 . s i if. 2 0 . s i 1 3 £ 2 1 . s i 1 3 1£ 2 2 . s a j 1 3 ( 1 ) C h e n g a v e / t s o j / o n t w o s e p a r a t e o c c a s i o n s . - 396 -58 at 1. 2. t s i 4. -13 t s i t i t ESS] 3. 4. . .13 t s i t s ' i 1 3 W &. W 5. 6. 7. t s ' i 1 3 .13 s i 4. i -51 • t s ' i a 8. . 13 • n g i a 9. , • I J 1 • - -k i tit •~s-' » i i . £ r ~ 1 0 . k i 1 3 h i 12. i - 1 3 k i l ¥ •nsv 13. .13 n g i j s l 14. . . I J " - • h i I E n — 15. h i m E i u 16. 17. 18. .13 l .13 i .13 i •zs: • ± " ;L\i\ 22 l i * 19. IS if l£ 20. 4 . ' 2 2 t s i 21. 4 . " 2 2 t s i to 22. •22, .,,.22 si ts 1 s 23. s i 5 5 (cf. 55-22) 24. . .22 tsi p If 25. 26. 4 - ' 2 2 t s i 4 . > - 2 2 t s ' i - 397 -w 1. tsi 22 2. tsi 3. tsik 2 (cf. 189-17) MO-—'^T ' 4. ts'i 22 5. si .22 0 E 6. tsi 7. tsi f£ 8. tsi 3> 22 9. 10. 11. .22 si .22 si .22 si 12. ni 22 22 13. ki 22 1 13.1. k ' i J 22 14. ki 15. .22 l 16. i 22 E 17. ki m. is. k i 5 5 55 19. ki 55 20. ki fT 21. k ' i 5 1 #f~/5j 22. — 55 - 398 -60 EH ft n i . 2. hi hi D i x : TO 3. 4. .bb l .55 l ± r£ i ¥ 5. 6. h i 1 3 -J l & •is 7. 8. ft) 9. . .22 ki V , 2 2 hi , .22 hi m . f t 10. .22 nga^ :£ fit VC to 11. t s V j 5 1 as - 399 ^ BE 61 1. ts'0j 2. ts'0j .55 55 3. sc^ j 51 4. 5. 6. k'waj 55" k'waj .55 k'waj .55 7. .51 nga^ E 8. * -55 faj TO » ^ 10. ~ 9. waj 1 3 .51 waj ± IS n. 10:13 ' ft it 12. • vt fo fit 13. ts'pj ® O J 14. ts'yn 1 3 B HI 15. - 400 -62 TO 1. kwaj 2. kwaj .13 22 3. waj 4. waj 7TT 13 5. waj ±3-m m 22 6. 101 M BE 7. 8. S0j S0J 22 22 S I ngaj .22 7T TO 10. waj 11. waj .22 - 401 -i L ^ H : m 63 la H /if ft Si i . s 0 j 5 5 m 2. sy 5 5 • (1)' ft — (ytj TB. 3. ts0 j $1 4. t s ' 0 j 5 1 5. t s ' 0 j 5 1 m- 6. — *± w lU ^ 7. ssrfj55 8. sjtft5, ts'yt 2 BE 9. .t s t f j " H io. s0j 3T & 11. kwaj 5 5 12. k'waj • ^ 13. k'waj51 H. 14. k'waj51 m m 15. waj 5 1 • 16. waj 5 1 (Chao: /naj 5 1/) • 17. waj 5 1 (Chao: /naj 5 1/) •31 18. wai 5 1 (Chao: /waj51/, / i 5 1 / ) (1) Chen gave /sy / on two separate occasions. (2) $ i n i t i a l : 63-10.1. / j 0 j 5 5 / (Chao). - 402 -64 <7b fc ± lb J H fc-j H M M 1. luj 1 3 (Chao: / l u j 1 3 / ; Ball: /Iii/) s e i i 3 •7k 2. s0j 13 3. kwaj 4. — 13 .51 .22 5. k'waj , kwaj 6. k'waj51 (Chao: A'waj13/) i l waj .51 • OB 8. 10 j 9. 10 j 22 22 m to it 10. . tS0j 11 722" .22 12. s 0 j 2 2 (Chao: /suj 2 2/) 13. s 0 j 2 2 , s u j 2 2 (Chao: /suj 2 2/) 14. 15. 16. S0j .22 S0j S0j .22 22 T . ^ j 2 2 (rh^o. / w * j 2 2 / 1 (cf—4n-na) £0 18. ts0j 22 6 5 m \h 1 . s^J 2. S(zfc55 (Qiao: also /s0t5/) JL 3. 4. 1,1 --22 k'waj 22 kwaj m m 5. i« -22 kwaj i ^ -22 kwaj i ~ -22 kwaj m m 6. 7. m m f i 8. .22 waj ZE - r 9. f i , 5 ft 10. f i 5 5 11. f i 5 5 12. f i 5 1 » 13. m i 5 1 i s 1.4. .51 mi i f 15. . .55 kwaj n n 16. .51 nqaj m ft 17. 18. 19. - -55 faj - .55 f a j - .55 f a j 3/, 7 * 1. .55 waj .51 2. wan .51 3. waj .51 4. waj m i t ^ H : i t ± IT H 5. f i 6. — » 7. f i 5 5 8. mi 13 9. kwaj 13 10. 11. waj .13 waj .13 i k £ = : f t 12. faj 22~ 13. faj .22 .22 r~M 14. faj 13 15. f i 16. mi 17. mi 22 22 - 405 -kwaj .22 67 TO m ft J5Z Mi 4 * m S 3 5r3 ngaj .22 .13 3. waj .13 4. waj 5. waj .22 6. waj 7. .13 waj 8. .22 waj .22 9. waj 10. .22 waj 11. .22 uj m 53 53— 12. paw (Chao: also /paw /) ft ft 13. p' ow 51 14. mow 51 55 15. tow ( *ffl) 16. 17. t'ow55 * (fs)l8. 55 77 i m m m 19. 20. t'ow" t'ow' 51 2]^ . t'ow '* ( ' #k) 22 51 23. 24. 25. 26. t'ow' t'ow t'ow' t'ow' 51 .51 51 51 t'ow 55 51 t'ow 51 (cf. 67-19) t'ow (cf. 67-26) - 406 -68 m \k 3 i . 2. low 51 (1) , 55 . 51 low , la:w ^ 3. low' .51 4— -low-55 is. m 5. 6. tsow' tscw .55 55 ft • ~ ® •8. 7. ts'cw ts OW iS m 9. ts 'ow m 10. 55 sow 11. 12. 4- . 2 2 ts 'ow 55 sow 55 m m 13. 14. ¥ 15. 16. 17. 18. kow kow' kow' kow' kow 55 55 .55 .55 55 51 ngow how 20. -how how' how' 19. iH 21. 22. 23. 24. 51 51 how22 (cf. 71-18), how55 t3 3£ 25. (1) 68-la. *f / l o 5 1 A i n A ' a n 5 1 l o 5 1 / * ^ * ^ 'spider'. - 407 -69 ± 1. T5 pow 13 2. pow 13 3. pow 4. p'ow : 7 / , — 1 3 i f f 5. t'ow it * 4. I 3 fir 6. to #flfT~ • « I ~ 7. tow13 % 8. — ft 9. t'ow13 22 i f 10. tow fa 22 11. tow — IT 12. now m to ta 13. now' 13 ,, 13 14. now 15. low 1 3 16. tsow ± 3 -*m 17. tsow' 13 18. tsow 13 22 t?k 19. ts'ow" ¥ 20. ts'ow13 W 21. — & 22 E 22. tsow 7 t ? t ~ 23. ts'ow22, tsow22 (cf. 71-7) i * f ~ * 24. sow22 (cf. '71-8) #j| .25. sow - 408 -70 m 13 1. kow' 2. ha:w' 3. ha:w 13 13 ±3-22 .22 22 4. 5. 6. 7. hew how' how' how' "cw22 (Chao: /ow13/) ow 13 10. . pow -'Ml 12. 13. 14. 15. 16. pow 22 pow 2 pow mow mow 22 2 22 22 -L4JJ •f : rm 17. 18. 19. 20. 21. tow tow 13 tow 22 (1) t'ow 22 22 tow 2 tow 22 (1) /tow22/ in e.g., /tow22 tiw 2 2 nga:n13/ A&\ $ f 8'cross-eyed' - 409 -71 as 1. 2. 3. 4. ~ * 6. 7. low i low .55 51 ts'ow' tsow 5. 22" ts'ow 2 3W 22 ts 'ow' 55 22 2 22 tsow , ts 1 ow sow 22" -4=-IS ^1 9. kow' 55 10. kow" (cf. 68-14) 22 11. kow 12. k'a:w22 (L.), how22 (C.) 13. — 14. ngow' MA 15. 22 22" 22 how' ~ 16 17 •» 18. how mow (1) 22 22" 22 22 k 19. ow' 20. ow' . 21. ow 22 (1) Chen gave /mow / on two separate occasions. - 410 -72 m Q 1. 'pa:w" Sf| 2. pa:w55 ^ 3. p'a:w22 5(2 4. p'a:w 5. p'a:w13 * & ! K f l f e 6. p'a:w51 + « 7. -^ 51 ^ 8. ma:w Iffi 9. miw55 ( L . ) , ma:w55 ( C . ) (cf. 75-8) C$g) 10. na:w51 M & ^ 11. — (Qiao: /na:w3X/) 12. — (Qiao: /na:w13/) (cf. 73-3.1) at w fo-il! ^ 13. tsa:w i J fc|>fcJR.~S 14. ts'a:w 5 5 15. ts'a:w 5 5 M , 16. ts'a:w 5 1 17. sa:w55 J H ^ W 18. — £ *¥ £ 19. ka:w55 ^ 20. ka:w55 M > 21. ka:w $ C ~ » 22. ka:w22 (cf. 74-8) M 23. k'a:w55 [ I t ffi 23.1. ha:w55 (Qiao) ] ^ 5 1 ^ 24. nga:w ia< 1 C 51 M 25. nga:w "ft 5 at a ^ L L I ~ 26. a:w22 (Qiao: /a:w55/) (1) i n i t i a l : 72-12.1. * ^ /tsa:w55/ (Qiao). - 411 -(1) £ 73 ± pa:w 13 2. a) pa:w , b) pa:w £ P 3. 13 ma:w t , 13 4. tsa:w C&J 5. 13 tsa:w # 6. B6 7. 4. . 13 ts 'a:w 13 ts'arw 55 22 (Chao: /pa:w /) w U4 m *5 8. 9. 10. 11. 12. 13. ka:w' ka:w' ka:w ka:w' ka:w' 1 3 " 13 22 13 k'a:w22 (2) 13 k'a:w 13 •»^(^)14. nga:w 13 %2k ftfr~gfr , 15. 22 ,T , a:w (L., a:w13 (C.) 16. pa:w 17. pa:w' 22 (iK) 18. p ' a : w " 22 /S^aTKS ( ) 19. p'a:w 20. p'a:w51 t£ 21. ma:w22 22 (1) (2) in i t i a l : 73-3.1. \%> /na:w13/ (Chao). 22 22 * A'a:w / used in A'a:w tsin / ^ 'scissors' - 412 -74 to na:w 22 tsa:w 22 04 if m Mi 3. — 4.1. ts'a:w' 22 5. 6. 1. 13 (Chao) ] sa:w 13 22 13 sa:w , sa:w , sa:m $1 Mi - f i b * 9. ka:w' 8. ka:w 22 2 2 " 10. ka:w' 22 55 11. ha:w~ 22 22 12. ka:w , kow 13. ka:w' 22 " 2 2 " 22 22 14. ha:w 15. ha:w" 16. ha:w" 17. ka:w22 (cf. 74-9) 18. — - 4 1 3 -7 5 31 m~ i • P i w &s ~ . 5 5 • 2 . piw 9H ' , . 5 5 • 184 3. p'iw • n 4. p^w 5 1 • I f 5 . p'iw 5 1 P£} I "H 6 . miw t r S 7 . miw 51 m i w 5 5 ( L . ) , m a : w 5 5 ( C . ) ( c f . 7 2 - 9 ) 7b fra H 1 C L tsiw J ! j . —- 1 1 - tsiw 5 5 $1 1 2 . tsiw 5 5 flfc 13. ts'iw 5 5 + 1 4 . -1 5 . ts'iw 5 1 1 6 . ' ts'iw 5 1 to | I 1 7 . siw 5 5 W 1 8 . siw 5 5 W 1 9 . siw 5 5 m 2 0 §1 2 1 & 2 2 . . 5 5 siw • 5 5 2 1 . siw . 2 2 S I W ft] I 2 3 . tsiw 5 5 ^ 2 4 . ts'iw 5 5 H ^ f t 2 5 . ts'iw 5 1 Si 2 6 . ts'iw 5 1 BE If ^ 1. ts'iw 5 5 2. tsiw 5 5 3. sxw an~SS 4. sxw51 0 s& _ . DX -5. ngxw. 6. — m m B . . . 55 m 7. kxw A» o i • 55 m 8. kxw m 9. k'iw51 -fit 10. k'iw51 M 11. k'iw 5 1 12. k'iw 5 1 , m 13. — M 14. hiw 5 5 m & 15. iw b b *fr 1 C . 13 . 55 ' ^ 16. xw , xw 17. iw 5 5 • 18. iw 5 5 f.S 19. — (1) M 20. iw 5 1 % 21. iw 5 1 H 22. iw 5 1 M 23. iw 5 1 (Ball: /ngiw/; or /iw/ by scholars) i i 25. iw 5 1 (1) The colloquial word is /ma:w t'aw ing / owl' (literally, cat-headed eagle). - 415 -77 mm—: w ± m ML 3s- , . 13 3x 1. piw m T . 55 m. 2. piw ' ^ .3. piw 5 5 (cf. 75-2) o /§§ , , . 55 , . 22 4. p iw , p iw • Hi 5. — • la 6. miw 7. miw13 • & 8. miw13 m j & / C ~ M 9. l i w 5 1 (cf. 75-9) n w ft & n l i 10. tsiw1"3 f^p 11.. ts'iw 1 3 / J N 12. siw 1 3 ft |§ 13. tsiw 2 2 3)u 14. ts'iw 2 2 fig K - 2 2 ftP 15. siw m if ?§f l6~ , ~ £ 16. tsiw 1 3 17. siw 1 3 #<g 18. siw 2 2 H -ta 13 19. ngiw Hfl~ 20. ngiw - 416 -78 JL I. kaw m m m 2. — — 3 - 3 — — iw 3. — • 3^? • F3 & • C I ) ] . 4 . g ~ 5 TT3 : — — piw , . 22 / T . . 55 ,„ . . p'iw (L.) , piw (C.) 6 , . 22 p' iw m 7. 8. . 22 miw . 22 miw law' 51 22r m to f l ~ , W— 10. tsiw' 22 in 1 ^ § J ~ 13. 11. ts'iw 12. — WW-14. siw' 15. siw 22 55 to 16. tsiw' 22 -jr 22 BE SO RH JL 17. 18. 19. J2SL tsaw tsiw' . 22 saw . 22 S I W - 417 -r 0 79 1. ngiw 1 3 (cf. 77-20) ~ . . 22 2. kiw 2 2 ~ J r — ' 3. iw 4. iw . .51,T . . 22 .„ 5. iw (L.) , iw (C. (1) 6. t i w J J 7. ^.55 tiw m 8. 5 5 tiw 9. tiw $§ 10. . . 55 tiw 11. t ' i w 5 5 i% 12. t ' i w 5 1 13. 4.1 • 51 t IW ft 14. 4 . . - 22 t'iw (L. t ' i w 5 5 (C.) (cf. 81-3) fc J 15. • 16. I 18. : 19-l i f t 20. liw liw' ,51 51 il 17. liw T 51 li w liw' .51 liw' 22 i t (i) 21. siw' 22. S I W 55 55 / i w 2 2 / i n e.g., / t s i 1 3 i w 2 2 / £ ' k i t e ' - 418 -80 1. , . 55 hiw 2. hiw % 3. . 51 ngiw & 4. iw i . ± 6. niw (Ball: /niu/, sometimes /liu/) t1) ] ^ 6.1. t'iw 1 3 (Chao) -J„m 7. l i w 1 3 (2) m s. i i w 5 1 m $ & ± ~ 9. kiw' \%~WlQ. hiw' T3~ 55 Si 11. hiw' ~T3~ 43-12. miw (1) 80-6a. %) (Ball: /niu/, sometimes /liu/). (2) Ball: ^ and words which are pronounced / l i u / in Cantonese are transcribed as /luui/ for Zhong-shan. - 4 1 9 -81 $$§23: i f -LJ S l m St 1 . tiw^ ^ 2 . tiw 2 2 3 . t'iw 2 2 m 4 . -£ 5 . tiw 2 2 6 . tiw 2 2 I f ~ f i j 7 . tiw 2 2 8 . — M 9 . niw 2 2 £f 1 0 . l i w 2 2 fem~m?- i i . — IS w ft & 585 1 2 . siw 2 2 m W j - 1 3 . kiw 2 2 Si 1 4 . k'iw 2 2 SB •fir 1 5 . 1 6 . 1 7 . 1 8 . taw 53" t'aw' t'aw t'aw' 5 5 5 1 5 1 - 420 ^ 82 m ® 1. law M~m 2. l a w 1 3 Mmmmftm^ 3. — § , i 13 Is 4. law _ & m m m •fej 5. ngaw 6. ngaw jag , 5 5 -iH 7. kaw ffipn 8 • * M &s~Cit) 9. « m ^ 10. haw 5 1 11. haw 5 1 M 12. haw 5 1 13. haw 5 1 m fib i A 55 W\ 14. aw SS . 15. aw 5 5 M 16. aw 1 3 17. aw 5 5 ± ,-g=- . — — • glj 18. p'aw 1 3 -tf. 13 ~ : 19. maw 13 p& 20. maw 4 t 21. maw13 « 22. mu 1 3 ft ?3. — • • - 421 -83 4 i . 4- 13 taw & ft. 2. 3. taw13 (1) taw U% 4. — law m IS 5. c a 1 ) 6. , 1 3 law \tntt Ira 7. tsaw m Vt fo 8. 13 saw * P p pp. 9. kaw' 10. kaw' 11. 12. kaw' haw' .13 .13 22 13 13. k'aw' 22 14. ngaw' 13 #§ @2~ , E ~ 15. ngaw13 IB 16. 17. k'aw' 22 2 T haw 22 18. haw"" (L.), haw' 22 19. hawzz (Ball: /hau/, A'au/) 13 (C.) P|I~ltfc 20. aw' ^ 21. aw' .13 13 (1) Chao used |>J- to record /t'aw13/ 'to rest' - 422 -84 vim-: ^ W JX 1 . inu 2 2 (Ball: /no/) 2 . maw22 22 M 3 . maw •HU A 4 . taw22 S 5 . t'aw22 -2. 6. taw22 7. taw •**• 22 -S.~ * l 3. taw fo m It 9 . naw22 ?s 2 2 •M 10. law M 1 1 . law 2 2 m m to ft 5fe • , „ . 2 2 ^ 12. tsaw c a t 22 22 ^ 1 3 . tsaw , ts'aw 22 V$L 14. saw JL m s a ± ~ 15 . — $ j ( f n 16 . kaw22 •m 22 m 17 . kaw m 22 m 18 . kaw z /n 55 19. ngaw r * 22 20. kaw f 0 ~ f t 2 1 . k'aw22 7& 22 . k'aw i 0 23 . k'aw22 l r i ~ 3 S 24. naaw13 (cf. 83-15) - 423 -2 22 1. k'aw 1^ 2. haw22 if: 7.K 3. aw22 # 5. faw5 1 (L.), p'u 5 1 (C.) B $ 6. maw51 jpp: 7. maw 5F 8. ma:w51 •Mil Vffif (This character should not be here at all.) US 10. law 5 1 11. law 12. law 5 1 13. law M^fi 14. law 5 1 %t~m(M) 15. law 5 1 - 424 -1. — § M ~ J S 2. — — 3 ^ 3. ts'aw55 $(~=f 4. ts'aw55 5. ts'aw55 Ate c 5 5 1P 6. saw 3g 7. saw # 3. saw55 EJ 9. ts'aw51 JHJ^ TK 10. ts'aw51 R=5P $ 11. ts'aw55 |lj 12. ts'aw51 H • 13. ts'aw51 14. ts'aw51 55 j g j 15. ts' aw £|3 16. tsaw + *KS!_h~ 17. — 22 3 § ~ |a3t5 18. tsaw 2& 19. saw51 SI 20. saw13 21. saw13 (/sa:w55/ C.) I I 1 5 - 7 22. saw13 (/suk5/ C.) 13 3 w M 2 4 . t s a w b b jir 25. tsaw55 'M 26. tsaw55 m 27. tsaw55 55 saw 4ft 28. 29. 30. ts' aw' #l(f£) 9. t s ' " 5 1 aw 51 - 425 -87 B m 1. jaw 5 1 •» • 51 1* 2. law r l m AS 3. kaw55 d?£~ 4. k'aw55 (Qiao: also A'aw55/) 5. taw13 (cf. 91-5) Jx 6. jaw 5 5 (1) ^ 7. k'aw51 3^ 8. k'aw51 10. k'aw51 11. k'aw51 4 1 12. ngaw51 fa 13. jaw 5 5 & 14. jaw jg* 15. jaw % 16- jaw 5 1 17. jaw 5 1 & 18. jaw 5 1 tft 19. jaw 5 1 W- 20. jaw 5 1 3@ 21. jaw 22. jaw 5 1 ± 9c m Siltt, m ^ 23. faw $f 24. hu 2 2, p'u 1 3 (2) 22 ^ 25. hu z jf. 26. faw 2 2 (1) 87-6a. Aia:w 5 5/ 'surname'. (2) /p'u13/ used in /san 5 5 p'u 1 3/ ([swn55 p'u:13]) $ i | Jj^ 'daughter-in-law'. - 4 2 6 -r 1 3 j f f i 1 - n a w £ g . 2 . n a w g £ 3 . n a w 1 3 $ P 4 . l a w 1 3 4. 1 3 tsaw T T - 1 3 ~ Fu 6. t s a : w 5 7 . t s ' a w 1 3 22 8 . t s a w can n 22 1 0 . t s i w ( / s o w / C . ) 1 1 . t s ' a w 1 3 ^ 1 2 . 1 3 s a w !•> 1 3 1 3 . s a w rtr 1 3 £ p 1 4 . s a w 3 « 2 2 5? 1 5 . s a w f l 1 6 . k a w A 1 7 . k a w 1 3 M M 1 8 . k a w 1 3 ^ f f ~ 1 9 . k a w 2 2 I ^ S ^ — T 2 0 . t s ' a w ' B 2 1 . k a w 2 2 s 1 3 ^ 2 2 . k ' a w 2 2 1 = 2 3 . k a w 2 2 - 427 -89 , , , 13 , 13 1 . k'a:w , ha:w •*r 0 • 13 /ff 2 . jaw /X 3 . jaw 1 3 m . . 13 £3 4 . jaw * - • 13 5 . jaw | f fi. jaw 1 3 MU 8. hu 2 2 9. huk2 (cf. 220-13) ?g 10 . law 5 1 fg 1 1 . law 5 1 12 . l i w 2 2 <cf- 7 9" 2 0> Iff ft & ?p St 13 . tsaw22 36 n 22 T y 14. saw It 15 . saw22 16 . suk5 (cf. 220-21) * l§ £S£"»i7. saw22 fa 18. tsaw22 m fi 19 . tsaw22 E& 20. tsaw22 - 428 -90 Iii 22 1. tsaw 2. tsaw22 (/ts'arw51/ C.) 3. 4. 5. tsa:w 22 saw 22 saw 22 ft 6. 7. tsaw 22" ts 'aw 22 m 8. 22 saw & 9. 10. 22 saw 22 saw 11. 22 saw 12. i 2 2 kaw 25 13. v 2 2 kaw & 14. V 2 2 kaw 15. i , 2 2 kaw (i) 16. ts 1 aw' 22-SI: 17. jaw' 18. jaw" 19. jaw-20. jaw" 21. jaw' ~ 22. 23. 22 22 22 22 51 (1) 90-12a. Aaw22/. - 429 -^ 91 : m ft w inH 1. piw 5 5 B 2. miw22 3. tiw 5 5 g£ 4. jaw 5 5 ± IS fJf^jE , 5. taw13 vamp.: m W m M 22 pJ 6. maw to W 7. jaw - 4 3 0 -9 2 i l Ft 1 . 2 . 55 ta:m t'a:m m m 3 . 51 t a:m 4 . t'a:, 5 1 m 5 . 51 na:m 6 . 51 na:m M 7 . i 5 1 la:m 35" W to 8. 9 . 1 0 . tsa:m ts 'a:m' i ts 'a:m' 55 51 1 1 . 1 2 . k ' o n 5 5 (L.) , k'arn 5 5 (C.) cm 55 1 3 . 1 4 . 1 5 . 1 6 . 1 7 . hom' horn' i hom' 5 1 51 i 51 om 55 om 22 - 431 -9 3 ± m — 5 5 ! • — Ira m & 2 . 13 to ft r l m 3 . 4 . 5 . kom13 (Ball: /kyam/) k'a^ 1 3 k'a:m13 ^ 6 . kom13 (Chao: /hom13/) to » *y # 7 . cm (L.), am13 (C.) , ( |5f ) 8. t ' a :m 22 r l 8.1. Aon 2 2/ (Chao)] ' R . ~ & 9 . hom' 22 - 432 -94 s. m - 22 1. hom ® I 2~2 TZ , 22 ,—;—±3 - 7 : — 2. om (Chao: /cm /, /om /) A : 3?te 3. 4. 2 ta:p ' 2 ta:p ^ ~ ~ i > « f ~ 6. ftt 7. 5. ta f a : p 2 2 f'arp 2 a 8. 9. 2 na:p la: 3 (Chao: / l a : j5 5 / , /la 5 5/) » ' ft n 10. 2 tsa:p it, 2 - ?j- 11. hop (cf. 94-15) pi 12. 13. kop . 2 kop 'fi 14 15. hop —• 16. hop^ hot 2 (cf. 118-7), ha:p 2 2 (C.) - 433 -•Mil fc fo Ji If A m • m va^a 1. ta:m' 55" m 3. 4. t'a:m' i farm' 51 51 JXTL 5. la:m 6. la:m' 51 51 7. ts'a:m' 55 8. sa:m 51 ffl io. r£f~7k n . kom' i kom' kom' 55" 55 i 55 12. 22 13. — (/ngong/ C.) .55 14. hom" 15. hon 51 ± 16. 17. 18. ta:m' 13 55 — (/tsin V C.) ta:m22 (L.), t*a:m13 (C.) (cf. 96-9) 434 -96 IS • 1. , 13 la:m 2. la:m 3, la:m 4. i 1 3 kom < 5. 22 kom 22 22 6. horn (L.) , ha:m (C.) JUUJ 8. ta:m (cf. 95-1) 22 (Chao: /ta:m /) 9. 22 ta:m (L.), t'a:m13 (C.) (cf. 95-18) i 22 la:m . 22 la:m 10. 11. Ira i t 22 tsa:m ft 12. 13. 13.1 22 sa:m (Chao) 1 * P 1L r l 14. u 2 2 nam - 435 -W [li m m - : t& A 9 7 11. 12. ts'a:m' i ts 'a:m' 13. ts'a:m' 51 51 22 22 £ (1) Chen gave A'oj / for on two separate occasions, but 22 £ gave A'oj / for on the .first reading, and nothing on the second elicitation of the word. - 436 -98 E J l ~ & ( Al ) 1. ka:m m £ f M ^ 2. m hom51 (Ball: /ham/, /ham/) ha:m m 3. M 4. ± at m ff 5. 1 3 tsa:m w a* UJ E r l M 6. 7. ka:m ka:n m m m 8. 22 22 hom (L.), ha:m (C.) PS to 9. 22 tsa:m m Jt ( 8 # ) 10. 22 tsa:n 22 tsa:m - 4 3 7 -99 I — ff'/& 1 . tsa:m' 22 2 . ha:m' 3 22 Joan' 22 A to — 12 4 . — i l K~IH .5. tsa W 4 ^ 6 . ts'aip PH(/jf ) 7 . 2 tsa:p • & ) 8. t s a 2 2 li] 2 sa:p JL 5*5 1 0 . 2 ka:p 1 1 . i 2 ka:p 1 2 . — r l 'I& 1 3 . *s i4. , 5 hap Bit E $i 1 5 . u 2 ha:p lit* 16- ha:p £ 1 7 . k'ap5 - 438 -100 mm-: % m m 1. ts'a:m fo [ii 2. 55 sa:m ft 'H , 4. + , i 55 3. ka:m 13 k'a:m m m 5. nga:m^''" m 6. ha:m ± .Irii 22 7. la:m s a n . 8. ts'im 5 5 rm 1 10. ka:m -22-11. ka:m' 22 - 4 3 9 -101 MiH: A 3¥ fei i . 2 . ka:p ka:p^ 2 . 1 . ha:r 3 . ka:p^ 4 . ha:p^ 5 . a:p 6 . a:t 7 . a:t 2 (L.), a:t 5 (C), a:p2 ( 1 ) 8 . tsa:p2 (cf. 9 9 - 7 ) >~*a' 1 0 . Iim 5 1 1 1 . I i m 5 1 1 2 . Iim 5 1 1 3 . I i m 9 . nim , nim55; (a) t s i n 5 5 ft 1 4 . ~M 1 5 . 1 6 . ( l i t ) 1 7 . ^ 1 8 . ts im' 5 5 ts'im' i ts'im' ts'im' 5 5 5 5 5 5 1 9 . 2 0 . 2 1 . ts'im' ts'im' ts'im' 5 1 L 2 2 i 5 5 ( 1 ) /a:pV 'to force, to coerce'. - 44 0 -to m 102 & ( S ) I - t s i m 3 ' Cte3~8A 2. t s i m 5 5 1 *. • 5 5 3. tsim 4. tsim 5. s i m £1 % 6. k'im51 im 1 3 (Chao: /im33/, /im 1 3/ (1) T . 55 i m . 51 am . 51 i m . 51 n g i m . 51 s i m 7. 8. 9. 10. 11. n 12. ± 13. . 13 pin m 14. l i m 1 3 . 55 55 51 -A-(1) /im / occurs in e.g., /im law / /<€J 82 'to tarry long', and /im 1 3/ in e.g., /im 1 3 mut2/ }£ 'to drown'. - 441 -103 IS fit ft & #f 1. tsim 2 2 ft IE3 2. — (Chao: /ts'im13/) SB i \ \ \ # If 3. sim 1 3 5^ 4. sim 1 3 s ^ 5. ngun . "ft 6. ngim13 JL 'It 7. k i n 1 3 + ^ 8. Iim 1 3 "fit? 9. kirn 2 2 in 17^ 13 PK 10. him m — ^ 11. im 1 3 • % 12. — a * aa m $C 13. Iim 1 3 (cf. 102-14) (Chao: /Iim22/) M 14. Iim 2 2 ^ 14.1. ts'im 2 2 (Chao) ] - 442 -104 BE 1. tsim 2 2 2. tsim' 55 m m 3. ngim m • M 4. . 22 im 5. . 22 im k 6. 7. . 22 im . 51 im 22 (cf. 102-10) (/ip / C.: A 8. sip^ (Ball: /nip/ for Zhong-shan and Cantonese) 9. . 2 nip 10. .2 .2 nip , sip 11. up2 Ira 12. z • ~ tsip it ic 13. j . i • 2 ts'ip ft m 14. tsip ft 15. ts'ip ft - 4 4 3 -BE 4- ' 2 tsip i m sip sip' 105 • ® & 2. tsip' 5 . — 6 . 7 . IP . 2 , . 2 ip hip [Ball: /hip/) r l o • 5 1 8 . ngim 9. — 1 C L im 5 5 (/ip2/ C.«cf. 1 0 6 - 8 ) mm ± m 1 1 . ngim51 ( 1 ) ( 1 ) Chen gave the same pronunciation on two separate occasions. - 444 -106 w it, m 1 . 22 kxm 2. , . 22 him 3 . — A t i BI-4. 5. i • 2 kip hip 2 m 6. . 2 ngip m 7. hip 2 m 8. . 2 10. n. a a . - 2 2 9. tim t'im 4 . . • 51 t'im 12. & ~ g * i 3 . . 55 • num - 4 4 5 -107 m m 1. kim •jf — 3 5 2. k i m 5 5 HI 3. him m iflK 4. him ± Mil m ££ 5. t i m 1 3 ^ El SIX ^ 6. t ' i m 1 3 ^ 7. t ' i m 1 3 8. — n f£ 9- h i p 2 /£ 10. t i n 2 2 (Chao: /tim 2 2/) 12. t ' i m 1 3 & A. , .22 v & 13. mm Ira ft ft i s 22 14. tsam - 4 4 6 -108 & * f t 1. hip z mmm : ^ A . 2. t i t 2 ( C f . 132-10) i f • i f f - 3. t«i P 2 8£ 4. t'ip 2 # 5. t i p 2 « £ ( * ) 6 . t i p 2 ^ 7. t i p 2 %k 8. t i p 2 m • 2 9. tip 10. -i& i i . — m *#5~a5 12. kip 2; a) ka:p2 m & 13. hip 2 2 2 $4 14. hip , ha:p $S~ftJ 15. hip 2 - 447 -109 JI ft m JI 1. fa:n 2. fa:n 51 51 ± fa 3. * 22 fa:n f§ 4. - 22 fa:n 3E 5. - 22 fa:n ft 22 6. fa:n i * 51 7. fa:n A 2 ' — i 1 8. fa:t 2 9. fart a - 44 8 -110 M- l . lam 5 1 jpfr^/S , ^ , j £ 2. lam 5 1 i 51 Ena 3. lam ~& 51 * t 4. 1 am m an ft >£? 5. ts'am 3 'L' 6. sam 51 7. ts'am £p fit (M i7c 55 8. sam ?ft 9. ts'am51 H± w OJ H 10. tsainf^ (cf. 92-8) 11. ts'a:m55) (cf. 92-9) Y J 13 5i •^r 12. sam (Chao: /sam /) 55 13. sam 14. s a m 5 5 BS 1? 15. tsam55 ffl- 16. tsam55 ^ 17. sam55 s 5Er 18. jam"11 f £ & 19. jam 2 2 (cf. 113-2) (Chao: /jam51/) - 449 -.111. r l m ^ 1. kam55 <at 2. kam55 3 > k a m2 2 ( c f . H3-5) 4. k'am55 & 5. k'am55 ^ 6. k'am55 W 7. k'am51 & 8. k'am51 It 9. k'am51 •t 10. iam51, ngam51 (1) tfc 11. -m at ... . 55 B 12. jam R A . 55 |S 13. jam M 14. jam 5 1 ± w M 15. pan i J ( 3 ^ — | 13 nn 16. p an U 17. lam 1 3 M 18. lam 1 3 19. lam 1 3 Ira ft & ft ft 20. ts'am13 I* 21. — (1) /nganTV 'to grumble*. - 450 -112 ft (1) m et 1. tsam 2. tsam' 22 IT 3. 4. 5. 6. 7. 8. sam sam 13 a 13 sam 2 sam kam' 13 a • 22 I3~ ~m 9. jam ^ t ^ r 10. jam' 13 2 2 " [ m_ i f ft 11. jam' 11.1. lam' 22 (Chao) 12. tsam' 13. — 7T [ W tli 13.1. ts'am22 (Chao) ] ® . sam 22 (1) Q i n i t i a l : 112-7.1. tffe* /nam13/ (Chao) - 451 -BE i \ \ \ 22 1 1 3 VtWlM 1. tsam 0 f i f t ~ 2. jam22 3. jam 2 2 £B 4. jam22 5. kam22 &S(+lfc) 6. kam22 7. k'am13 §z m^ mm HM^iS^g. jam 5 5 (Chao: /jam22/) (1) f 9. __ ( 2 ) flfc~.H 10. jam13 (cf. 112-9/10) A I t m 3t 11. lap 2 2zi 12. lap 5 /•S 13. nap :— Ira m ft >o n I t — ^ D . ( i — 14. ts'ap5 j** 2 Hi 15. tsap •1&m~ 16. ts'ap5 (Chao: /ts'ap2/) la 17. tsap 2 31 18. tsa:p2 (1) /jam / is colloquial. 22 (2) Chen gave /am / the first time, and nothing on the second reading. - 452 -114 OJ BS 1. tsik m Till 2. 3. 4. 5. 6. 7. ' 4 3 * 9 . tsap" c tsap" sap 2 sap sap' sap B A 10. jap (1) Ik 11. kap" JL-. 12. ' f i t t £ ~ 14. 15. 16. 17. 17.1. k'ap" k'ap" 13. k'a|5 jap kap2 k' ap (Chao) 18. k'ap' (Chao: A ' a p 5 / , ngap5/) (2) %2l B 19. jap . J ^ f i ^ — 20. jap" (1) 114-10a. -it /ngap / 'twenty* A - 5 (2) 114-18a. ^ /hap/ (Chao). - 4 5 3 -dim-: m% 115 ft a a 5 f ' 1. 2 . 3 . 4 . 5 . 6 . 7 . ta:n tarn 5 5 " 55 t'am t'am t'am t'am 55 55 5 1 5 1 8 . 9 . t'am 3T 51 nam lam' 5 1 m 1 0 . lam m 1 1 . larn IS 'tn 1 2 . ts'am ft 3H 1 3 . 5 1 ts'am & m 1 4 . 55 sam 1 5 . kon 1 6 . i 55 kon 1 7 . . 55 kon 1 8 . i 55 kon . 1 9 . v 55 kam m 2 0 . u 55 hon 2 1 . ham 2 2 . hon22" (L . j , honJJ- (C.) 5 1 2 3 . 2 5 . hon" J l D X L 5 1 on 2 6. on . 5 5 55 - 454 -116 Mil m f4m%;~* i . — ifi 2. t'ain 1 3' M 3. ta:n 2 2 4. f a : n 1 3 m fc 'ti 5. la:n 1 3 tint? i f ft fo ft Wt&*®~ T 6. sa:n 1 3 ^ 7. sa:n 2 2 i l If 8. hon 1 3 H IS— 9. kon 1 3 & io. — 11. kon 1 3 "Ou. 12. . ha:n 1 3 11 ^ 13. ha:n13 (Chad: also /ha:n13/ #^ 14. hon 1 3 i i m. B 15. ta:n #c 16. t'a:n 2 2 Ifc 17. t'arn 2 2 <B 18. ta:n 2 2 S i ? 1 — 19. ta:n 2 2 ^(W) 20. ta:n 2 2 $ 21. — - 455 -117 1. nam 2 2 ftj 2. l a : n 2 2 Ira i t ft jm 3. tsam 'IS® 4. t s a : n 2 2 ^ 5. t s ' a m 2 2 i f c ^— 6. s a m 2 2 E SI S¥ It ' 7. k o n " ~ JL 8. hon 2 2 u u 22 7^ 9. ngon E 22 III 10. hon 11. hon 2 2 § ? ~ ^ S 1 2 . h o n 2 2 «v ' 22 13. hon 14. o n 2 2 * i C 2 2 15. on A 16. — (i) 17. t a : t 2 — — — 2 — 18. na:t M 19- l a : t 2 + ( >H) 20 — (1) 117-16a. / t * a : t 2 / (Chao). - 456 -118 m 1. ts'a:t' fo ft • ~ S ( # ) 2. sa:t' 3. sa:t kot kot 2 hot'' . 7. hot 7 8. hot fo ^ S T K S 9. sa:h 10. sa:n 51 55 5 5 -11. ka:n f ^ 3 £ ~ » +— 12. ka:n 55 13. ham 14. ha:n 51 51 - 457 -lliffi ± - : Oil 119 fo 1. 2. 3. 4. tsam r r 13 ts'am' tsa:n 2 2 (cf. 121-9) 13 tsam 13" 22 13 m 5. kam 6. 7. kam kam 8. kam 13 9. ngam' 13 m 10. ham 22 ill HI — : Oi: m i i . 12. 13. 14. 22 pam p'am _ 22 fam 22 pam E L ' I t 15. tsam 22 > 16. kam 2 T a* 22 17. ha:n f hin 22 - 458 120 A A 1. pa:t 2. pa:t , ( 3. mart2, nut (cf. 136-23) (Chao: /mat /) #7 4. 5. 6. 7. tsa:t 2 tsa:t tsa:t 2 ts'a:t^ 2 8. sa:t $ J 1 7 7 & J .9. — 10. tsa:t 55 n. m. 12. 21 13. 14. & 15. ^ 16. S i 18. pam c pam" e pam" 3 5 " l 55 l 55 p'am 5 5 (cf. 120-16) p'am p'am 55 55 ts'iong" 51 mam 51 - 4 5 9 -w fo-il} 121 HU 1. sa:n .& m m -*>£F 2. ka:n 5 5 • , , 5 5 3 S 3 . ka:n MR 4 . ngam ± i t n m & 5 . . pam 1 3 iK. 6. pam 1 3 fJLili-JYf-7 . 8. • 22 mam O J a 9 . 10. 22 tsam 22 sam (1) i n i t i a l : 121-6.1. /nam13/ (Chao). - 460 -122 w. i . v 2 2 kam m 2. i , 2 2 kam J P U m - 3. • 2 2 kam m i l 4. 22 ngam m gafciM . 5. 22 am ium- i mm A ffl M 6. J. sa:t fo- $ i}~77 (llj) 7. — il} m i f 8. hat41" E 9. hat 2 m W fc /MfS it? 10 pin 11. p'm ,. 55 p'm ,. 55 p'm . 22 pin . 51 nun • 5 1 16. min 12. 13. 14. 15. 17. "lin' 18. 19. lyn lin' 51 51 51 (cf. 125-6) - 461 -m m i . 2. 3. 4. 5. 6. 7. tsin 55 123 55 ts'in ... 55 ts'in ts in . 55 sin . 55 sin • 51 , . 51, _ , in (/sa:n / C.) 8. ts'in 51 9. 10. tsin 55 m i i . >2. 13. 14. I S , 55 t s i n . 22 sin sin sim 22 i 51 sim 51 (cf. 125-17) (cf. 125-18) 16. 17. in in 51 l 51 51 51 'ty 18. k'in" ^ 19. k'in" H ( ) 20. kin 1" (1) 2i. • — 55 : %3 sfc ff #f 22. — l€ 24. 23. 51 in 51 in (Ball: /ngin/) . 51 i n (1) Chen gave the same pronunciation on two separate readings. - 462 -124 i m ± S i m w 1. Pin 2 2 «» • 22 2. pxn 3. min 1 3 ggj 4. min 1 3 13 £ft'33*~ 5. min • ffi 6- rain13 • r l — 7 - ~ fc 51 9. t s i n 1 3 r t ft fo ft M 11. t s i n 1 3 12. t s ' i n 1 3 13. t s ' i n 1 3 tf$~4> 14. s i n 5 5 (cf. 123-6) (Chao: /sin 1 3/) & 15. s i n 1 3 to m M. 16. t s i n 1 3 m / i n m n - 2 2 l l 17. sin £» no -22 j p 18. sin m + 5] 19. — •JA . 13 * m 20. hin # 21. kin & - 463 -125 m 1. i n 1 3 ft m m w m. 2. p i n 2 2 H ~ f | 3. p-in 2 2 4. g ' i n 2 2 rt 5. pin • $UJ~6. p i n 2 2 • W 7. min 2 2 m 5i 8. t s i n 1 3 . (cf. 124-9) Ira i f ft & as 22 HU 9. tsin 2 2 5 5 2 2 ^ ^ _ j | . 7 k 10. tsin , ts i n , tsam (C.) jj£ 11. t s i n 2 2 ^ - -22 i t ~ f T 12. tsin « * T O • 22 ^ 13. sin dfc i - i -22 =§| 14. sin 15. tsin H 16. t s i n 5 5 (1) js .22 .>gi 17. sm 'HH -t rt * 22 m 18. sm ^ 19. s i n 2 2 ¥tt 20. s i n 2 2 iP?~^ 21. sim 5 1 (cf. 123-15) $m 22 • » 22. sin (1) / t s i n 5 5 / i n e.g., / t s i n 5 5 hon 5 1 t s i n 5 5 l a m g 1 3 / J | % | | >^ 1 shivering cold, to have the c h i l l s ' . - 464 -126 ngin 22 VE fc m ft fo JL to A sus-f l i l mm-2. pit 3. pit* 4. — 5. pi t 2 . 2 6. nut l i t ' l i t 2 9. l i t 10. sit 11. s it 12 4. - 2 2 t s i (1) 4fr 13. tsit 1[SJ:^~A 14, 15. t s ' i r 16. ts'it* 17. ts'it 18. ts'it 22 (1) Chen reconfirmed the pronunciation of / t s i / on the second reading of the word. - 465 -B8 #T~gf 1. t s i t 2 2. t s i t 2 ^ 3. tsaj 2 5 4. s i t 2 t£ 5. t s ' i t 2 • # f # ~ 7 6. t s i t 2 a ^ 7. ngit 2 E m • 8. k ' i t 2 #t 9. k i t 2 2 5 m 10. ngit (Chao: /ngit /) ^ f f i 11. — E $ I f ' U t ^ ^ 1 12. — W 13. ngyn 5 1 (Ball: /ngun/ or /ngin/; Chao: /ngin51/) g § E ¥ F 14. h i n 5 5 ffi. 15. h i n 5 5 (Chao: /hin 5 5/, /hyn55/) (1) I r B S 16. — 55 S5 (1) Chen gave /hun / the first time, and /hin / the second time. - 466 -1 2 8 ± % \% m _ 22 H j 1. k y n O J P ^ H : Tuft SI JL m — — ; ; — 2 2 fit 3. k y n 2 2 fit 4. -m U k 5. h i n 2 2 H £ 6. h y n * 7. i m 1 3 [ A 8. k ' i t 2 (L.) , h i t 2 (C) m 9. k ' i t 2 $§ . , : ; - 467 -m 129 ffc 1. h i t 2 » 2. k ' i t 2 m H 3. k'it"6 (Ball: also /k'it/) m m ; t2 4. pin Ai 5. p'in 5 5 $f~tt£ 6. p'in 5 1 m 7. min 5 1 Mil m H 8. t i n 3 5 H 9. t i n 5 5 ^ 10. f i n 5 5 ES 11. t ' i n 5 1 * 12. t ' i n 5 1 51 (R 13. f i n m fc ^ 14. n i n 5 1 $ 15. l i n 5 1 i l 16. l i n 5 1 •let1. Ira ft ft fo ft ^ 17. tsin I 1 18. t s ' i n 5 5 tH 19. t s ' i n 5 1 & 20. s i n 5 5 % m i f W 21. k i n " M 22. k i n 5 5 55 23. hin f f 24. ngin 5 1 - 468 - 130 m ft 1. h i n 5 1 5% 2. hyn 5 1 $£ 3. hyn 5 1 S(*H) 4. i n 5 5 i ^ S ' St 5. i n 5 5 ± m m m 6. pxn IM -7 .13 IS 7. pin 8. p i n 5 5 H 9. t i n 1 3 m~*t*- 10. t i n 1 3 u. n a n13 (c.) ^(M*) ' 12. -If £g 13. k i n 1 3 (Chao: also Ain 1 3/) #f 14. ngin 5 1 M&nmyk is. __ II 16. hi n 1 3 - 469 -131 m m , H 22 : — — 1. p'in jgj~tti2 2. p'in 2 2 ?i= -j r 55 ,.22 i w 3. p'in , p'in 4. p'in 2 2 m t • 2 2 f5 5. min m e l = r . • 22 m> 6. tin •fe. 7. tin •ta- 22 3? 8. tin {ffl 9. t i n 2 2 , f i n 5 1 j&~8(#0 io. t s i n 2 2 (1) m fc M 11. ' l i n 2 2 /*+. 22 M 12. l i n 13. — Ira tn ft fo ft M 22 '"" ^ 14. tsin JL m ft 15. kin"" (Ball: /ngin/) 22 M 16. in m i% 17. h i n 2 2 (Ball: /ngin/) (1) Chen gave /tsin / on two separate occasions. - 470 132 . 22 1. in 2. i t 2 . 22 3. in i A. % 1 ( Sri ) 4. — m 5. p i t 2 6. p ' i t 2 7. mit 2 JrOJ i t m • 8. t ' i t 2 m 9; t i t 2 10. t i t 2 11. . 2 nip 2 (Chao: also /nip /) Ira m 12. t s i t 2 m 13. t s ' i t 2 m 14. — m 15. , i • 2 ts'ip (1) ft 16. t s i t 2 17. s i t 2 18. s i t 2 ' 19. s i t 2 (1) The same pronunciation was given on two separate occasions. - 471 -m i f #B 1. kit U 2. k i t 2 5 ^ Mil 3. — * 4. — (l) m m^ •mm P t ~ & 7 5. i t K ® 6. p'un55 * * ( * 0 .7. pun 5 5 'M Q , 55 < w 8 . p' un # f ~ # . 9 . p'un13, p'ing 2 2 ^ 10. p'un x ^ 11. p'un ^ 12. pa:n 5 5 13. man51 II 22 B S 14. ma:n •hOl m ^ 15. tyn 5 5 a 16. t'yn 5 1 + M 17. t'yn 5 1 m fc W 18. l y n 5 1 (Ball: /nun/) ^ 19. l y n 5 1 (Ball: /nun/) m /ff ft fo ft i ^ M ,~m 20. tsyn 2 2 (L.) (cf. 136-2), tsyn 5 5 (C.) 21. - ' 22. syn 5 5 (1) 133-4a. ^ /ngit / (Chao). - 472 -134 1. kun fit 2. kun 3. kun 4. kun % 5. hun 7U 6. 7. hun 55^ hun 8. yn 5 1 51 ' SP~ 9. yn # L ~ » 10. yn 5 1 51 '3. 11. un 12. — i r 0 J ^ - , : fe^ ± ft 13. p'un22 (L.), p'un13 (C.) (1) 5 ^ ( * # ) 14. p'un # 15. mun13 22 16. tyn 17. , — ' ~IS 18 IP 19. 20. nyn 10n 13 tyn 2 2 (L.)., t'yn 1 3 (C.) (1) /p'un13/ cxrcurs in e.g., /jaw13 p'un13/ ^ \^ 'ccxpanionable' ra l . 2. tsyn 22 - 473 -(1) 135 f|5 3. 4. kun' kun T T 13 k'un 13 5L hun' 22 $ t £ & ( $ c ) 7. k'un 13 -13-un V-'-1 ft n ft 9. 10. 11. 12. 13. 14. 15. 16. pun 22 p'un p'un p'un' p'un' p'un' ma:n 22 22 22 22 22 22 ma:n 22 if ik' 17. 18. tyn' tyn' 22" 22 £ 19. tyn' g ( ^ ) 2 0 . tyn : Lfi 21. tyn : 22 22 22 22 22 (1) /tsa:n / was given by Chen on the f i r s t reading, and /tsyn / on the second. - 474 -136 1. lyn 22 2. tsyn •3 4- • 1 3 3. ts'yn 22-W ( % ) 4. syn syn 22 22 M. 6. kun m 7. kun' kun Z2^ II 9. kun ¥ 10. kun' 22 22 22 22 3^ ' i i ' 11. ngun 22 22 15. hun 12 13 14. un' hun 22 22 un -13- yn -±3-A 16. put 2 17. put 2 18. p'ut2 & 19. .2 pa:t 20. ,2 part m 21. mut2 22. .2 mut & 23. 4 - 2 4.2 mut , mart (cf. 120-3) (Chao: /mat2/) - 475 -137 > m^~& 1. 2. , 3. t'yt 2 4. tyt 2 tsyt , tspj 4. 4.2 4. -^22 tsyt , tspj Jft #6. ts'yt 2 .0 7. k'ut f g ~ ? * 8 . kwa:t' M \ M 9. hut 2 iS—«S » ~ D 10. k'ut2 2 HQ 11. U t 12. k'un' 13 0 J £ - : U4,! 13. kwa:n 55 j B S ~ & • ~ H 14. pq^:n 5 1 (Chan: /nqwa;ng£A 51 - 476 -138 f f B J il 1. wa: n' 22 O J ^ — : O i l A ft 2. 3. wa:t wart' SO wa:t ' UJ^ * ID P! ( JL 3£ sa:n 55 ( ) 6. ts'yn 5 1 (cf. 140-4 kwa:n 5 5 -' ~@ 8. nga:n51 (Chao: /ngwam13'1/) 51 5Jt 9. warn 51 51 10. warn " (1) n 51 11. warn (1) T& is used colloquially in Zhong-shan, as opposed to Cantonese ' 4; 33 which uses %. [tsUij ] for ' s t i l l , yet'. - 477 -139 B m , 55 =f 1. wa:n rf 2. warn 5 5 t ! j ^ - : f j f ! . ± B fo- , 3. t s a m lk i f B ta « „ 22 * 4. sam fo iii 5. t s ' a r t 2 JL -rgr — 22 '|K 6. kwam rl $ 7. t s ' y n 2 2 ( c f - 1 4 2 ~ 1 0 ) If m m m a 22 • ® 8. warn ^ 22 g. 9. wam A i l Via 478 - 140 tu fo ts'a:t T kwa:t (i) m ft fo ft 3. ts'yn 5 1 , 51 i s 4. ts 'yn 1: s. . 51 5. ts yn 6. ts'yn 5 1 7. 55 E3 syn 8. 51 syn to 9. 10. ts'yn 51 51 BE 11. )12. 13. 14. 15. 16. tsyn 5 5 tsyn ^ 51 spn 4. . 55 ts 'yn . 55 ts 'yn 51 syn (1) ^ i n i t i a l : 140-2.1. '"il" / l y n 5 5 / (Chao) - 479 -141 3. '# 4. - (fc ) 2. k'yn k'yn k'yh 5. k'yn 1. 51 51 51 51 hyn 55 6. yn 7. yn" 8. yn 9. yn 10. yn 51 .51 51 51 51 11 55 ± r t ft fo ft 12. syn 13 14. syn 13. tsyn 22 IT B fljfc i5. ts'yn 13 17. ngyn 13 - 480 -m m 142 m^fe 1. kyn 55 Ha ?g ~ 2. hyn am 35 3. -IS M fc if* 4. l y n 1 3 i&~*s. ( *m ) 5. — Hi. m ft >o ft $E~n£~-f& 6. syn 5 1 (cf. 140-8) ifl—J|* 7. syn 5 1 ft] « 8. tsyn 2 2 f f ~ t E 9. tsyn 2 2 SB 10. ts'yn 2 2 #11 11. ts'yn 5 5 JL / l If m # 12. kyn 13. kyn 1 3 (1) ±.a i A i 22 ' $H 14. kyn # 15. kyn 2 2 (1) Also used as a classifier for books in Zhong-shan. An alternate pronunciation is /kun13/. - 481 -1. yn 22 A m ft ft 2. lyt 3. tsyt 4. 5 . tsyt J2 syt 6. tsyt 7. tsyt 1$ HP 8. syt ft 9. yt__ - 482 -144 7U 3 § 1. fa:n 5 1 (cf. 144-6) IB 2. fa:n 5 5 # « ~ 3. fa:n 5 5 4. fa:n 5 5 <I c * 51 5. fa:n ™ 6. fa:n & "7 4T 51 56= 7. fa:n & 8. fa:n 5 1 m m 9. ngyn 5 1 ® 10. ngyn 5 1 ^ 11. nqvn51 E It 12. hyn 5 5 % 13. yn 5 5 ^ 14. yn 5 1 M K 51 w 15. yn H u 51 1231 16. yn 17. yn 5 1 18. hun 5 1 - 483 -± 145 1. farn r r 2. ma:n' 3. warn 13 13 wo 4. ngyn I J 5. un 22 6. yn 13 7. un 8. 13 yn 13 m 9. MJ^TM IO. 'IS. 11. fam' 22 farh i 12.. m a m ' • JH -P 13. m a m ' 22 22 22 14. hyn' 15. hyn' 22 22 16. ngyn' 22 - 4 8 4 -1 4 6 I S - 1 . hyn St-' — 2 . 'yn" A 3 . fa:t 2 4 . fa:t 2 5 . fa:t 2 6 . fa:t 2 7 . fa:t 2 8 . ma:t2 err , . 2 % 9 . k'yt - ( # ) 1 0 . — I*? 1 1 . k'yt 2 m 1 2 . k'yt 2 m 1 3 . kwat2 (cf. 1 6 5 - 1 4 ) 1 4 . — Hi n . 1 5 . ncryt eg cm t'J 1 6 . 1 7 . 1 8 . 1 9 . 2 0 . - 4 8 5 -147 m if 1. kyn 5 5 m j£ 2 . hyn ftg- 3. hyn m n£» — Wi 4 . yn 5 ± m if X 5 . k'yn 1 3 IS Pi H 6. yn 2 2 IS 7. hyn 5 1 (Chao: /hyn13/) - 486 -148 A m m m 1. k'yt 2 (Chao: Ayt z/, A'ytV (1) ) 2. k'yt 2 ^ 3. k'yt 2 E ' 2 iflL 4. hyt K 5. y t 2 § 6. t'an 5 5 I f £1 7. kan b b S 8. kan 5 5 E g[ 9. han 5 1 (/ha:j51/ C.) ,H 10. jan b b (1) Chao: A y t 2 / occurs in e.g., A y t 2 tyn22/;A Sf| 'to decide', and A'yt 2/ in A'yt 2 haw13/ C2 'a rupture'. - 487 -149 mm— * ± m m 1. han 1 3 H 2. han 1 3 | g 3 g f i ~ 3. ngan51 (cf. 155-5.1) m E ffi ( © ) '4. han 1 3 mm-JL m m ^ 5. ngan m 'It 6. han 2 2 m ^ 7. pan 5 5 ' $ 8. pan 5 5 ' 9. pan 5 5 10. pan ft 11. p'an 5 1 • Jg~*L2. p'an 5 1 13. man13 * S 14. man51 - 4 8 8 -1 5 0 1 . 2 . 3 . l a n l a n i 5 1 l a n 4 -i 5 1 l a n n 5 . 4. u 55 t s p n m 6 . 4. . 55 ts an ft 7 . 4- >M 51 t s ' p n 8 . 9 . 55 s a n s a n 5 5 1 0 . 55 s a n ft £0 1 1 . 55 t s a n no 4- • 51 1 2 . t s a n 1 3 . t s ' a n 5 1 1 4 . — 1 5 . t s c i n 5 5 1 6 . t s a n 5 5 /jf$ 1 7 . s a n 5 1 % 1 8 . s a n 5 5 $ 1 9 . s a n 5 5 # 2 0 5 5 s a n 2 1 . t s a n 2 2 ( c f . 1 5 3 - 4 ) „ 5 5 2 2 . s a n M 2 3 . san 5 1 san 5 1 „ ' 5 1 2 5 . san 2 4 . A 2 6 . jan 5 1 j= 2 7 . i a n 5 1 - 4 8 9 -151 rt3 1 . k a n 5 1 n g a n " 5 5 ~ j a n c j a n " 5 5 5 . j a n 5 1 6 . j a n " 5 5 ± Ira m ft JL ft 7 . 8 . p ' a n 1 3 1 3 m a n m a n 1 3 M ( + * § ) 1 0 . m a n 1 3 1 1 . m a n ' 1 3 f t ~ S i J ( H ) 1 2 . t s a n ' 2 2 1 3 . t s a n 2 2 I T 1 3 It 1 4 . t s ' a n 1 5 . t s ' a n t f f i 1 5 . 1 . t s ' a n 1 3 ( C h a o ) ] 2 2 1 3 W ' 1 6 . s a n ( L . ) , s a n ( C . ) 1 7 . • 1 3 . 1 3 , _ j a n ( L . ) , n g a n ( C . - 4 9 0 -1 5 2 JL m • 5? 1 . kan mm • 51 2 . j a n 1 3 3*03= : K ( $ ) » ( « ) . J5S m m • & 3 . pan 2 2 *g , 2 2 • f f 4 . pan it 5 . l a n 2 2 Ira ffi ft ;0 ft i l 6 . t s a n 2 2 e 7. tsan 8 . ts'an 5 5 (L.)(cf. 1 5 0 - 6 ) , ts'an 2 2 (C.) 2 2 2 2 2 2 -fH 9 . san (Chao: /san /, s0n /) W\ 1 0 . san 2 2 'it- 1 0 . 1 . t s a n 1 3 (Chao) ] ft i * 6*. " 2 2 " m 1 1 . tsan + A 2 2 M 1 2 . ts'an W 1 3 . t s a n 2 2 ffi #7 fo-lk M 1 4 . ts'an 2 2 - 491 -153 Bg # i? Si 1. tsan 2 2 (Chao: /tsan13/) *m 22 m 2. tsan $1 3. tsan 2 2 M 4. tsan 2 2 (1) 1» 5. san 2 2 (2) s M. 6. jan 2 2 . . wL 7. jan 2 2 (/nging22/ C.) IS 8. ngan22 JL If & 9. kan 1 3 E f i S ~ 10. hyn^Z (3) m m^ mm • EP 11-' jan 2 2 A n m ^ 12. pat 5 * ^ 13. pat 5 * & 14. p i t 5 , p i t 2 • E — ~;7TJ ' Ji 15. p'at5 i$f 16. pat 2 17. mat2 • H! IS. T7T34-2 m M BI 19. nik 2 ^ 20. lszit2 (1) Chen gave /tsan / the second time. 22 (2) Chen gave /sjzih / the first tine. (3) Given by Chen on two separate occasions. - 492 - 154 ft ft JSL ft i . ts'at 5 ts'at 5 m 2. m 3. tsat 2 m 4. sik 5 m 5. sat 6. tsat 7. tsat 8. t i t 2 iff W BE 23. a 9. 10. sek" c sat" (1) 11. tsat" 2 12. sat 13. sat" 14. sat" 15. jat 16. fn~Hl7. kat k'it** £ 18. yt • — 19. jat" 20. jat (1) Same pronuncation given by Chen and Yang. 493 -155 JR ft j? Mb 1. 2. 3. 4. 5. 5.1 . 55 kan kan k'an k'an k'an ngan J E fa 6. hyn b D, jan 5 0 (1) 10 c i 5 St m 7. 8. . 55 -jan . 55 jan ± r l % m i f 9. 10. kan^ ' 22 13 kan (L.), k'an (C.) 11. 12. jan 1 3 . 13 jan (1) /hun55/ was given the first time by Chen. - 494 - 156 m M ~ ( P J i . king" (cf. 202-13). '—' ! J5 2. kan 2 2 (L)., k'an 1 3 (C.) (cf. 155-10) A ^ T " ' ^ % m m IZ* 3. ngat 4 — 5 5 o 4. ha:t (Chao: also /ha:t /) — — — . — ^ — ^ m & 5. ngat : *Jt& n m ^ 6. pan"'"' •f 55 7. pan 1f!~zK 8. p'an 2 2 (cf. 158-13) & 9. p'un51 PI 10. mun51 ff? 11. mun55 12. tph b b (Chao: /tanjb/) & 13. tph 5 5 (L.), tan 1 3 (C.) i£ 14. tan 1 3 (Chao: /t'yn51/) ®[ 15. t'yn 5 1 1$ (^\^Ja 16. tcin 2 2, a) t'an 5 5 W 17. t'yn 5 1 & 18. tdn2.2 (L.), tan 2 2 (C.) - 495 -157 fc l . 2. l a n 2 2 (cf. 159-2) (Chao: /lan 5 1/, /lan 5 5/) (1) lan m # 3. tsyn m 4. ts'yn 5 5 • f t fo m - & 5. 6. 7. ts 'yn ts'yn 1 3 (2) syn 5 5 ft & 8. 9. v . 55 k wan i , . 55 k wan i * 10. 55 k wan W : ±1. fan 5 5" B 12. fan 5 5 55 14. wan" ft 13. fan 51 '*6 15. wan22 (L.), wan51 (C.) 16. wan22 (cf. 158-10) mm ?S -> 55 ™- 17. wan + M 18. wan ± 1 Q 13 19. pun w on 2 2 20. pan m 55 55 22 22 2 (1) /lan / occurs in the expression, /lan siong lan lok / ;rip s «*7 ' to chatter incessantly'. (2) The same pronunciation was given on two separate occasions. HU m - 496 -158 m -1. tan 1 3 #6 2. tf^n 2 2 (L.) , tan 2 2 (C.) J l > ? ~ , i f c ~ 3. t'0n 1 3 (Chao: /t0n13/) m ft fo ft * 4. -ft 5. ts'yn 1 3 c 13 6. syn ——• 1 1 n* r l l¥ + 7. kwan^ fB(.fl) 8- k'wan13 E 22 iJ*^@— , ~ jtfi 9. wan 10. wan22 J » *8 n 13 i s 11. wan IS 12. pan 3 5 (cf. 156-6) ® — • # , ~ i g 13. p'an2 F<M . 14. mnn22 -UJLJ m H 15. t#n" ^ 16. -($13 17. t ' u j 2 2 £6 18. t0n 2 2 tee 22 - 497 159 -yjzj. 4M 22 m M 1. nyn 2. la n 2 2 (Chao: also /lan 2 2/ here—cf. 157-1) m m w. 3. 4- 1 2 2 ts yn ft 22 4. san ft ' 5. 22 kwan m 6. v . 2 2 k'wan A 7. pat 5 . m 8. 4 - 2 put &~± 9. JI put I I 10. — 11. .2 put 2 12. mut 13. tat 2 (/tyt5/ C.) Ira 14. tsszSt5 i t 15. 2 ts'yt ft ;[\ 15. 1. ts'yt 2 (L.), ts'0t 2 (C.) (cf. 162-14.1) ] ft - 4 9 8 -1 6 0 1 . k w a t5 2. 2 k w a t ( / f a t 5 / C . ) 3 . n g a t . r n g a t m & 4 . f a t 5 5 . h a t 2 ( L . ) , w a t 2 ( C . ) : WW 5? im 6 . 7 . 9 . 1 0 . 1 1 . l a n l a n 5 1 5 1 5 1 t s p h 5 5 tsan' 2 2 ft 1 2 . A 5 1 s p n 1 3 . s p ' n 5 5 1 4 . A 5 1 s p n i l 1 5 . S 0 n 5 1 1 6 . t s ' r f n ^ — 1 7 . 5 1 ft M~ffF is . H 1 9 . ~ # 2 0 . ^ 2 1 . s p n s p ' n 5T 5 1 t s ' p n s p ' n 5 1 5 5 2 2 . ^ 5 1 s p n 2 3 . ^ 5 1 s p n 7 4 . A 5 1 s p n - 499 -IL ?i m 161 • i^J 1. k'wan55 • i f 2. k'wan55 • -£j 3. wan ± m ft fo UJ 4. san J 5. san 1 3 •If 1? * 6. ts^n 1 3 <t 7. tscn 1 3 ^ 8. ts'jzin13 9. t'0n 1 3 (Qiao: /tjzSn13/) (cf. 158-3) JL rl *¥ If 10. k'wan22 S5 13 11. k'wan rcg 13 12. wan m i-a 1 3 2s 13. wan • ft 14. wan13 . a- 1 C 13 .-^ 15,—wan - 500 -162 i t ft fo ft & 1. tsan^ m 2. san 3. tsan 2 2 4. s0n 5 1 , son 5 5 # 1? jig 5. scin 2 2 6. son 2 2 g& 7. s0n 1 3 B fi4 8. jan^ IH 9. jan 2 2 A If ^ 10. 10t2 ' 10t m /t ft fo ft 4s 12. tsgit* J$t*~*7 13. 2 j s a n 2 2 ' f § ? ~ s i4. tsan $f 14.1. ts'yt 2 (L.) (cf. 159-15.1), ts'jzfc2 (C.) ] & 15. stft5 . 'fat 16. s0t 5 & a Iffi 17. tsyt 2 . t. ' >f~ 18. s0t 2 1 . S0t~ 2. sgfc~ 3 . t s ' ^ t 5 i l 4 . s0t 2 5 . , 2 6 . s0t 2 k w a t " 8 . f a n 5? 9 . f a n 5 5 IB 1 0 . * 5 5 f a n 1 1 . 1 2 . f a n f a n 5 1 1 3 . — £ . 1 4 . 5 1 m a n US . 1 5 . 5 1 m a n 1 6 . 5 5 m a n 1 7 . m a n ' - 502 -164 JL m i i 55 1. kwan 2. kwan55 3. k'wan51 4. k'wan _ _—gg 5. fan 6. fan 5 5 _ , 55 7. fan 51 55 55 8. wan , fan (1) (Chao: /fan /) wK3st s 9. wan ft 10. wan51 ± n- 11. fan' 13 -m i t 12. * 13 fan 13. * 13 fan 14. 13 man M 15, man13 22 16. fan 17. fan 1 3 ~ » ! ~ 18. fan 2 2 PB3 T O 2 2 | O J 19. man ^ ~ ? n . -(1) /fan 5 5/ occurs in e.g., /fan 5 5 t s ' o j 2 2 / ^ |L 'a meat diet' - 5 0 3 -16: 6 */ — }3fe m 1. 22 kwan m fi! 2. k wan E 3/, sy K ( .* J 3. — (/fang*"/ C.) 4. 5. 22 wan 22 wan 6. 51 22 wan (Chao: /wang V) *m : A 7. fat 5 m 8. 9. fat 5 fat 5 # 10. fat 2 #J 11. 12. 2 mat mat2 r£ 13. wat5 (Chao: A' wat5/, /wat5/) n • {/Hi ~ sii 14. 15. kwat2 k'wat5 I? 16. wat5 Mil - 5 0 4 -1 6 6 m m 1 . pong 2 . p'ong51 3 . p'ong i A 1 3 K^ * 4. pong $M I 'It c _ 5 1 n I IL- 5 . . mong 5 5 ^ 6 . mong (see also: 1 7 8 - 1 7 ) t M ~ 5 5 5 5 - I E - --2-—mong (Chao: also /mong—/) s % • 8. tong ' 5 5 W 9 . t'ong" ^ 1 0 . t'ong 5 1 ^ 1 1 . t'ong 5 1 * t ~ * & ' 1 2 . t'ong 5 1 Hf 1 3 . t'ong 5 1 i t 1 4 . t'ong 5 1 - e v r / B 1 5 . t' ong 5 5 'M , m 1 6 . nong51 ^ I IP 1 7 . long 5 1 JB 1 8 . long 5 1 li% 1 9 . long 5 1 t & 2 0 . long 5 1 Jf 2 1 . long 5 1 ft 5 5 2 2 . tsong C « ^ f t » ' * ~ 2 3 . t s o n g 2 2 2 4 . tsong / f f | £ 2 5 . ts'ong 5 5 ^ 2 6 . ts'ong 5 5 TJggfS— 2 7 . ts'ong 5 1 ,' 1 S o n 5 5 >b | 36 2 8 . song OQ 5 5 2 9 . song - 5 0 5 -167 r w . i m JfiE ( 1. kong 2. kong 3. kong 4 . kong kong kong hong hong 5 . J 6 . 7 . 8 . 55 55 55 55 55 55 55 55 ngong 51 $TL 1 1. hong 10. hong 5 1 51 1 2. hong 5 1 1 3 . k'ong 22^ ± 14. pong I T 1 5 . mong 1 6 . mong 13 13 X T 13 M 17• tong f |f 1 PIL~ 1 8 . tong' 13 19. t'ong 2 0 . f o n g 1 3 B&~ 71. tona 2 2 X T 2 2 . — (Chao: /nongJ"~'/) i 13 2 3 . long - 506 -ffi m ft .c> ft 168 •# fiT£ 1. — .*. , 55 Djjl 2. song * f* 3. --If lfi~ftJ 4. k'ong13 ' - i - i w ^ . . . — . . . 0 5. p'ong f?f 6. pong22, p'ong51 iS m &~r? , &~ Bm tong22 9. f o n g 2 2 CM} io. t'ong 2 2 11. -m & #t 12. lona 2 2 n /t ft & ft m 22 13. tsong -~- 22 ffl, 14. tsong *M, 15. tsong 2 2 16. song 2 2 . - 5 0 7 -m 169 Ctl] 2 . fit 3 . * A 4 . k ' o n g k ' o n g 1 . k o n g 2 2 2 2 2 2 MM A ft IB r 6 . p o k 8 . 9 . 2 , 2 p o k . 2 p o k , p a : k ( 1 ) . i 2 . i 2 p ' o k , p ' a : k ( c f . a l s o 1 6 9 - 7 ) 1 0 . 1 1 . 1 2 . 1 3 . 1 4 . m o k m o k m o k 2 m o k 1 3 m o ( L . ) , m o 5 1 ( L . ) , m o 5 5 ( C . ) ( c f . 1 4 - 8 ) f t 1 5 . t ' o k ft ( $ 5 ) M f e 1 7 . t o k 1 8 . t o k 2 ~ 1 9 . t o k 2 I§ 2 0 . n o k 2 2 1 . l o k 2 2 2 . l o k 2 2 3 . l o k 2 2 4 . l o k 2 2 5 . l o k 2 2 6 . l o k 2 2 7 . l o k 2 t ' o k ' m fc ( 1 ) 1 6 9 - 7 a . /pck2/, /fok 2/. - 508 -x ~ 1. 2. ts 'ok ft 3. 2 tsok ft 4. tsok2 fo 5. ~ K • tsa ft tsok 170 2 sok2 kok2 S 8. kok2 9. kok2 10. kok2 2 11. ngok 12. ngok2 fM • 13. fit*~A 14 15. ' k'ok2 16. hok2 (2) 17. ok2 fc 5±-^ 18. niong — 22 & 19. jong J§_ 20. liong 5 1 21. liong 5 1 M.~&%i 22. liong 5 1 11 23. liong 5 1 24. liong 5 1 ^ 25. liong 51 (1) /tsiok 2/ occurs in the word /tsiok 2 j a t 2 ' 5 / fc) 'yesterday' (2) 170-16a. (Ball: /lok/) . - 509 ft W 55 jf^---3f". 1 . tsiong" 55 $ft 2. tsiong 5 5 Ki 3 . ts'iong" rtic 4 . ts 1 iong 5 1 JHI 5 . ts 1 iong 5 1 5 - ~ 6. siong 5 5 A * r - _ _ 55 ^@ 7 . siong JSi 8. siong 9. siong 5 5 i n . 5 5 g 1 0 . siong 31 1 1 . siong If 1 2 . ts'iong 5 1 iff 1 3 . ts 1 iong 5 1 5S 1 4 . tsiong H—fa. 1 5 - ts'iong 5 1 |J§ 1 6 . ts'iong 5 1 j|§ 1 7 . ts' iong 5 IB. tsong 5 5 5* 1 9 . tsong 2 0 . tsong 5 5 2 1 . ts'ong 5 5 2 2 . ts'ong 55 2 3 . siong • 55 2 4 . siong w m fo- fo lk *@ m - 510 -172 BB | if: l . tsiong 5 5 2. tsiong 3. tsiong a < 4 . . , , 55 i=a 4. ts'iong 55 5. ts'iong" C 4- . • 5 5 6. ts 1iong F§f 7. siong 5 5 & o • 55 w 8. siong If 9. siong 5 1 f>.,, f si. s 10. siong 51 siong . 1 3 12. siong B I WLiXL— 13. jong 2 2 (L.), nong51 (C.) I * S 14. — JH i§ 15. kiong' 55 5T e 16. kiong" ft 17. kiong 5 5 i f i ^ S 18. kiong 5 5 19. kiong 5 5 ic 20. kiong 5 5 m I ^ 21. kiong 5 5 22. k'iong 5 1 H§ "H 23. hionab5 ong" 24. hiong - 5 1 1 -1. 55 jong at 2. 55 jong 55 3. jong 173 4. 51 jong 5. 51 jong I ft 6. 51 jong 51 7. jong 8. 51 ?aS jong 9. 51 jong Mm ~ 10. . 22 tong # 11. ts 1iong (1) 51 £133 =: ± fli nf~f® 12. niong 1 3 13. liong' 13 J» 14. tsiong 1 3" 5? 15. tsiong =^ 16. tsiong ie 17. ts' iong >l> | fcS 18. siong 1 3 t 19. — ft Al 22 22 20. tsiong . 22 21. tsiong 22 tsiong ft | 23. tsiong 1 3 * SI ( ) 24. tsiong 2 2 22 22 22 3t 25. tsiong 26. tsiong 27. tsiong' (1) The same pronunciation was given on two separate occasions. - 5 1 2 -1 7 4 fcfc W fo BE a 1 . ts1ong 2 . song 1 3 ( (a) is /ts'iong 1 3/—cf. 1 7 3 - 1 7 ) ) 0 tsiong' 4 . ts'ong 1 3 1 3 m ( i s ) 5 . siong siong 1 3 hiong' 1 3 1 3 8 . jong 9 . jong 1 0 . jong' 22" 2 2 2 2 m @ ( % ) & ~ ' « ~ 1 1 . k'iong 1 3 1 2 . jong 1 3 1 3 . hiong' 1 4 . hiong' 1 3 1 3 1 5 . 1 6 . jong 1 jong 1 3 J 1 3 2 2 # f ^ 1 8 . liong Il£ 1 9 . liong II ( ff ) 2 0 . liong 21. liong 2 2 2 2 1 3 2 2 ( c f . 1 7 3 - 1 3 ) - 513 -w ft fo ft IS 1. tsiong 2 2 7J^A"~ 2. tsiong 2 2 E 3. tsiong 2 2 ^-"IS, 4. siong 2 2 • f c ifeR 5. tsiong^-H ( H ) 6. tsiong 2 2 $H 7. tsiong 2 2 I I - ,8. t'iong 2 2 (1) 9. tsiong 2 2 w fo-lk 10. tsong 2 2 |IJ 11. ts'iong 2 2 (Chao: /ts'ong13/) T!7\ 12. tsong 2 2 m m i f * 13. tsiong 2 2 0~ 14. tsiong "ii 15. ts'iong 2 2 16. ts'iong 5 5 (cf. 172-6) |£j 17. hiong 1 3 22 fpj 18. siong 2 2 , SB"** 19. siong s H 20. jong IE fp] 21. hiong 2 2 & -Sri 22. jong 2 2 ^ 23. jong 2 2 ^ 24. iona 2 2 (1) Chen not only gave the same pronuncation the second time, but he also tried to explain its usage and"definition. - 514 -A 176 M Ira m ft /O I L ft m i . 2. l i c k 2 22 liona m 3. 2 tsiok 4. 2 tsiok 5. tsiok 2 eg 6. 2 tsiok m 7. siok 2 ( ) ~~ ^ 8. tsiok ( i i ) H ~ . r f t ~ 9. tsiok'' BS 10. tsiok / f ^ l l ~ i i . ts'iok 12. 13. ts'iok ^^j. 14. ts'iok 15. ts'iok' "2 0 16. jok 17. iok' 18. kiok" 19. kiok 2 20. jok 2 m. 21. jok 22. jok' 23. 24. 25, jok jok' - 5 1 5 -1 7 7 B m m yt 1 . kong^ PH. TTL 2 . fong * 3 . fong 5 5 & A 5 1 M 4. wong lit It ~ 5 . wong51 M 6. wong _ 5 1 7. wong '41 8 . wong51 Q 5 1 >^ 9. wong , N 5 1 M. 1 0 . wong m m& £E —~zR l i . wong55 ± m m He 1 2 . kong13 m wit (. 5w ; f& ~ ]_3, fong 'It 1 4 . fong 1 3 1 5 . fong 1 3 - 516 -178 — * i . — 2. k1 ong 22 (1) A IP kok* kokz C IS ) ~ 5g 5. k1 ong' 22 6. k'ok 7. k'ok* H 8. 9. wok2 2£ BT" 55 55 .51 51 51 '51 J 55 51 2f -ctr 10. ' 11. 12. ? 13. 14. 15. 16. 'JE17. fong fong' fong fong fong l fong mong" c mong" (2) 18. mong 13 22 (1) Chen gave /koaq / and Aong /; he did not know which was the correct pronunciation.and stated so. (2) 4l i s used for fuel in the cooking stoves. ( ^ /kon^/ is also used for the same purpose.) - 517 -179 m m e 1. hong HI 2. hong55 8i8£~ 3. hong55 EE 4. k'ong51 ar c 51 dc 5. wong ± St 6. fong i J ffl 7. fong 1 3 #J*@(K 8. fong 1 3 W~W 9. fong 1 3 |§ 10. mong13 11. mong13 Dia 12. wong13 *± 13 1± 13. wong -as ft m SC 14. fong 2 2 tfr 15. fong 1 3 )"£ 16. mong51 (cf. 178-18) 1-7 22 17. mong i » 22 M 18. mong - 518 -180 3£ ( !f± ) m 1 . 2. k'ong51 (L.), k'wa:ng22 (C.) k'ong If. m 3. fong'" (Ball: A'ong/) & 2 2 , 4. wong OR? \>] -Q "n —. • HOT ^= A m ^ 5. fok 2 (Chao: also /fok2/; Ball: /pok/) (1) m m m ^ M i i 7. k'ok2 % 8. pong" f J 4 ( | | ) s 9. p'un22 (cf. 182-5 and 135-13) ill 10. p'ong" ~M 11. -51 (1) 180-6a. U (Ball: /fok/, A'ok/) 180-6b. (Ball: /fok/) - 519 -181 fc ?fi 1. lung 5 1 to fl? 2. tsong Pt 3. — & 4. tsong 2 2 (L.), ts'ong 2 2 (C.) & w fo-lk ?i5 rr , , • 55 & 5. ts iong ft 6. siong 5 5 E m tL 7. kong 5 5 £E 8. kong55 (Chao: A'ong 5 5/) 2 l ~ a 9. — ST io. kong 5 5 . 11. hong55 m ft ( +|$) 12. -P£~tS - &~ 13 hong51 St 14. kong ± m m CflO 15. pong13 Jab 13 , 13 16. pong , p'a:ng 17. pong13 fo fc I f ( + ^ ) i i 7 : ^ is . — - 520 -1 8 2 1 . kong 1 3 - f t 3 . 2 . kong 1 3 4 . k'ong22, hong 2 2 ( 1 ) 3¥ ( # ) 5 . p'un 2 2 (cf. 1 3 5 - 1 3 ) 6 . t s o n g 2 2 ( L . ) , t s ' o n g 2 2 - ( C . ) fo 4" u _ _ s i o n g 5 5 (cf. 1 8 1 - 6 ) P § T ~ 8 . kong 2 2 5 1 5 1 hung 3 1 (cf. 2 1 3 - 1 9 ) ( C h a o : /p'ung /) 1 0 . hong 2 2 2 2 ( 1 ) A'ong / occurs in e.g ., / t s ' o1 3 k'ong22 y 1 3/ £ 1 * 8 . '.(nane of emperor, and /hong22/ i n e.g., A a j 5 5 hong^/ |^ rf-J^ 2 2 , a n 2 2 young chicken'. A'ong / has the wider distribution. 183 A ft n m • M W $!j 1. p'ok' (L.), mok"' (C), mok" (C.) (Chao: /luk"/) (1) © 2. pok2 $t 3. p'ok2 # 4. p'ok2 ^ 5. puk2 (2) m 22 S 6. pa:w to its: ^ 7. ts'iok^ 8. ts'iok 2 M 9. tiok 2 (Chao: also /tiok 2/) 10. tiok 2, tiong 5 5 , 11. — S M S £ O l 2 . luk 5 (cf. 220-19) (Chao: /ts'ok2/) ^ 13. tsuk2 14. t i k 2 (2) OJ J E 15. tsokb (Chao: also /tsok5/) li% ( ) , 16. tsuk5 /fcil 17. — 9$ 18. sok2 J L %!m~ 19. kok'1 M 20. kok2 t £ ( 7 % J ~ ? - 21. ka:w13 (L.), (a), kok2 (C.) fil 22. k'ok2 fil* • ~w 23. k'ok2 5& 24. hok M 25. ngok2 •§§ 26. ngok ^ H-"w 27. ngok (1) /mok / used for e.g., peanuts; /mok / used for e.g., clothes. (2) Same pronunciation given on separate occasions. - 5 2 2 -1 8 4 ^ . 1 . hok2 31 2 . ak3 n ft. 3 . pangJJ S3 4 . p'ang51 MU m ^ 5 . tang 5 3 fe. 6 . tang 5 $ 7 . fang 3 # a ^ . 5 1 S I 8 . fang US 9 . -^ 1 0 . f ang 5 1 1 1 . f u n g 2 2 (cf. 2 1 6 - 9 ) B 1 2 . f a n g 5 1 •'-is i 5 1 1 3 . nang ' f f i - 1 4 . -ft ft' •L> ft 1 5 . tsang 5 5 * i t 1 6 . tsang 5 5 'If 1 7 . tsang 5 5 # ~ S 1 8 . ts'ang 5 1 , „ . • 5 1 ' 3 1 9 . ts'ang ^ i n J . 5 5 i s 2 0 . tsang g 2 1 . hang 5 1 ( L . ) , ha:ng55 ( C . ) ( 1 ) ( 1 ) /ha:ng55/ occurs in the name of a village in the first qu in the Zhong-shan county: /ha:ng55 mi 1 3/ ' f j i . £ - 5 2 3 -185 ± MU 1 . tang' 13" 2. hang 1 3 I - : &m m -MJLI m 3 . tang i t® : ~ 4 . — §15 5 . tang -TK f¥ ' ~ • 2 2 - ~ 6 . tsing #s fa mm ~ 7 . — ft , l \ It 8 . 2 2 tsang JL m 9 . kang53 m m' - 524 -186 A 4b 1. pakD m 2. mak2 3. mak2 4. tak" 5. tak 5 m 6. t'ik 5 (Chao: /t'ak5/) 7. 2 tak 8. lak 2, la:k 2 (1) ® 9. lak 2 Ira 10. ij — tsak"' m ft 11. ts'ak2 & I I 12. sak5 £ 13. k'ak5 _=.if 5 . 14. k'ak 55 15. k'ak5 M 16. ha:k3 (1) /la:k 2/ occurs in /mok2 ts'a:k2 la:k 2/ £}\ O flft 'naked down to the waist'. - 5 2 5 -ft ft BE If 5 H I 187 1— . afcJJSfc 1 . p i n g ' 55" 2 . p 1 a n g 5 1 3 . 4 . 5 . 6 . 8 . 9 . 1 0 . l i n g l i n g l i n g l —ins. t s i n g 5 1 5 1 5 1 5 1 53" 5 1 t s ' . i n g ( 1 ) t s ' i n g 5 1 t s ' a r n g 5 1 ( c f . 1 9 5 - 4 ) 1 1 . t s i n g 55" • — i i 1 3 . s i n g 7~k 1 4 . ~ 1 5 . 1 6 . •a i 7 . 1 8 . 1 9 . to 2 0 . 2 1 . s x n g s i n g 5 5 . 5 5 s i n g . 5 1 s i n g s i n g 5 1 ~ 5 T ~ i n g 1 2 . t s 1 i n g ' 5 1 5 1 5 5 55" \-+*.—. 2 2 . k i n g 5 5 2 3 . k i n g 2 4 . n g i n g 5 1 ( L . ) , k ' i n g 5 1 ( C . ) ( 1 ) 1 8 7 - 8 a . / t s ' a n g J X / ( C h a o ) 5 1 - 526 55 ' j— -ff l 2. ing' 3. xng 55-4. ing 51 188 ± I T 93 5. ts'ing' B ITTJ (2) fc SE S i ~ g j l 6. tsing 2 2 (Chao: /tang 2 2 / ( 1 ) } 7. tsing ( M J 8. tsing' 22 22 9. tsing 22 — ~ 1 0 . ts' ing' f l l 11. sing 2 2 ^ ~ J f t 12. sing 2 2 22 'fig 3S T 13. nging 5 1 (L.), k'ing 5 1 (C.) (cf. 187-24) (1) Chao: /tang 2 2/ in /nga:n13 tang 2 2 tang 2 2/ $t 'to stare fixedly, to glare'. (2) jfc. i n i t i a l : 188-5.1. ^ /pang22/. - 527 -189 m •m M s - 1. 22 hing ' ' 2. ing z z & . 3. . 22 jan : m% A if m 4. pik m m m n 5. nik^ m 6. l i k 2 \\nti m 7. tsik 5 IP II 8. 9. tsik 5 tsik 5 J . io. n . hi. sik" c sik" c sik" ! i ^ ® ^ i 3 . t& 14. $fc 15. tt 16. i t ( E )l7. tsik sik 5 tsik tsik" - 5 2 8 -1 9 0 fo III J0¥ m 1 . t s ' a k m 2 . t s ' a k m 3 . t s ' a k & 4 . s i k 5 (HJ 5 . s i k 5 7 . t s i k m 8 . t s i k ft 9 . s i k 2 l i 1 0 . s i k 2 1 1 . s i k 5 1 2 . s i k 5 n 1 3 . s i k 5 1 4 . s i k 5 m 1 5 . t s i k 1 6 . t s i k n 1 7 . k i k 5 1 8 . k ' i 2 2 ( c f . 5 9 - 1 3 . 1 ) 1 9 . k i k 10 2 0 . i k b 2 1 . i k 5 2 2 . n g i k 5 M 2 3 . i k 2 X 7 4 . — - 5 2 9 -1 9 1 Is la Z£ : sm T^ a m m. 1 . wang51 » 1 . 1 . wang55 (Chao) (a word added to the Ji-yun) ] •<£r . B 1=3 A r l m [1] 2 . kok^ E I? 3 . wa:k v 2 4 . wa:k B P :M A m 2 5 . wa:k (Ball: /wik/ for Shi-qi) 85153-• iTx. PS 0 ^ Hi? ~M 6 . 7 . 8 . 9 . p'ang ( 1 ) 5 1 p'a:ng p'a:ng51 5 1 ma:ng ( ia ) 10. mong! ( 1 ) 1 9 1 - 6 a . ji) /p'a:ng55/ (Chao). - 530 -1 9 2 fc Wi If 1 . t s ' a m g 5 5 ( L . ) , t s ' a : n g 2 2 ( C . ) iH 2 . t s ' i n g 5 1 ( c f . 1 8 7 - 8 ) j £ W fo UJ ^tomma 3 . tong 5 5 £ 4 . sang 5 5 ( L . ) , sa:ng 5 5 ( C . ) # 5 . sang ( L . ) , sa:ng ( C . ) ^ 6 . sang 2J§ 7 . sang l ¥ H^-j* • i ~ 8 . kang D D ( L . ) , ka :ng 5 5 ( C . ) 9 . — 1 0 . kang 1 1 . kang 5 5 ^ 1 2 . ha:ng 5 5 ft ^ 1 3 . hang"" ( L . ) , ha:ng"" ( C . ) ( 1 ) ? T ~ j | 1 4 . hang 5 1 ( L . ) # ha:ng 5 1 ( C . ) & f 1 5 . hang 5 1 Ufa-:mm ± m m w 13 13 1 6 . meng ( L . ) , ma:ng ( C . ) Mil n 1 7 . t a 1 3 < § • 1 8 . lang 1 3 ( L . ) , l a : n q 1 3 ( C . ) ( 1 ) / h a : n g / o c c u r s i n t h e n a m e o f a v i l l a g e n e a r K u - c h o n g : , 5 5 . 1 3 , - £ g • / h a : n g m i / 3 l \ j > - 5 3 1 -w fo LU /it 1 . s a n g 2 . s a n g . 1 3 1 3 W~ 3 . k a n g i J ( L . ) , k ' a n g X J ( C . ) „ 13 4 . k a n g "? ' m. 5 . k a n g . 1 1 3 193 6 . h a n g 2 2 fc 7 . m a n g ' 2 2 - f t 9 . t s ' i n g 5 1 ( L . ) , t s * a : n g 5 1 ( C . ) 2 2 I 1 0 . t s a : n g 1 1 . k a n g ' TT 1 2 . n g a : n g 2 2 ? T S ~ 1 3 . h a n g ' 2 2 - 532 -194 :HiP g A pg I . pa:k 2. .2 pa:k 3. .2 park 4. pik 5 (Chao: /pik 5// /pa:k5/) m 5. p'a:k AS . , 2 6. p'a:k ia 7. . V 2 p'a:k & a 8. .2 pa:k ft 9. .2 pa:k w 10. Pg~£ ' 11. pa:k (1) to W 12. ts'a:k2 13. — il w 14. , 2 tsak " 15. tsak 2 ± 0 • 17. 2 tsa:k m 5±c — ' i 1 > 5£s. 18. — fo iii JL *& 19. ka:k2 2 20. ha:k ^ ( Ht) 21. k'at5 22. , 2 nqa:k . — 23. 2 2 ha:k (Chao: /hak /) 24. ha:k2 IS (1) Chen gave the same pronunciation on two separate occasions. He also contrasted the Zhong-shan pronunciation with the Cantonese one, which he cites as [m*k] in low tone. (Ball: /mak/) 194-lla. (Ball: /mak/). - 5 3 3 -195 mm-:mm ,m m m 1. pan b b (1) m 2. p'a:ng51 BM 3. mang5''" ft •it 4. ts'a:ng 5 1 w OJ ^ 5. tsang 5 5 (L.), tsa:ng J J (C.) 6. tsang (&f0 7. tsang 5 5 (L.), ts'a:ng 2 2 JI m 8. kang 5 5 (L.), ka:ng55 (C.) i i 9. hang55 • 41 10. — 11. -* ® 12. king 2 2 (1) m 22 S 13. king & # 14. ang 3 3 (Ball: /ing/ in Shi-qi, /ang/ rest of s i * ^ , C 55 . 55 the county) ,£§~fg i ^ - a j - 15. ang , ing 16. ing 5 5 (1) Same pronunciation was given on two separate occasions. - 5 3 4 -1 9 6 ± :{% £ * ( ^ ) l . p o n g 1 3 ( c f . 1 8 1 - 1 7 ) £ m & 3 . k a n g i J m 2 2 ^ 4 . h a n g p-^ f — " ' m 3£ i£~§£ 5 . p i n g 2 2 A m f s S / U ^ ^ p ] 6 . p ' i a r k ^ , m a : k 2 ( C h a o : / m i k b / ) , 7 . — T*X 2 ^ 8 . m a : k M 9 . m a k 2 fc t l 1 0 . t s a k ^ - 5 3 5 -197 fo-il! Pi 1 . tsa:k M 2. ts'a:k 2 ffl 3 . t s ' a : k 2 tin A 5 5 W-^f^ 4 . sa:n JL a* ^ 5 . kak b, ka:k 2 ( 1 ) M 6 . ka:k 2 SI 2 7 . hat ^ . ^ ^ ^ 8 . h a t 2 ( L . ) , wat 2 ( C . ) Wk 9 . — M 1 0 . ak 5 l i ' 1 1 . — / Q 1 2 . ak 5 ¥ ft PJ3 1 3 . p i n g 3 3 ¥ 1 4 . p ' i n g 5 1 ( L . ) , p ' i a r n g 5 1 ( C . ) if 1 5 . p ' i n g 5 1 |Zp 1 6 . p ' i n g 5 1 OH i-i - 5 1 Hf 1 7 . ming $M 1 8 . ming 5 1 5 i i« 5 1 -dS. 1 9 . mang ( 1 ) Aak / i s pronounced [kep ] i n the word Aak ming / i f «j> 'revolution' , with A / assimilating the point of a r t i c u -l a t i o n of the following /m/. - 5 3 6 -1 9 8 1 . 2 . 3 . k i n g k i n g 5 5 5 5 5 5 k i n g " " ( L . ) , k i a : n g 5 5 5 5 ( C . ) h i n g " 5 . k ' i n g 6 . 7 . k ' i n g n g i n g " " 5 3 " 5 1 5 1 5 1 8 . i n g m =-:HP5 ± 9 . p i n g 1 0 . p i n g _ 1 3 1 3 3£ 1 1 . m a n g 1 3 1 2 . 1 3 . k i n g k i n g ' 1 3 1 3 1 4 . k i n g 1 3 1 5 . i n g 1 3 ( [ j l n " ] ) ( L . ) , j a : n g " " ( C . ) 1 3 1 3 - 537 -199 aga=: SIPS sr m ft w ffi 1 . ping 2 2 (L.), piarng 2 2 (C.) 2 2 2 2 ffi 2 . ping (L.), pia:ng (C.) 22 22 •op 3 . ming (L.), mia:nq (C.) m m m • 22 22 4 . king (L.), kia:ng (C.) .aba. 13 J 3 » 5 . king f t 6. king 2 2 (L.), kia:ng 2 2 (C.) m 2 2 @£ 7. hing 22 " 8 . king & f & 9. ing i J ([jln 1 3]) mm - : HPS A PS m m ft w S ' 1 0 . pik 3 E m m ^ 1 1 . kik J |^ 1 2 . kwik5 (Ball: A'wik/ ( 1 ) ) 1 3 . ,k'ia:k 1 4 . k'ia:k2 S * ~ 1 5 . kia:k 2 J lJM— ' — I I 1 6 . nqak2 ( 1 ) Ball gives A'wik/ for Shi-qi, but A w i V or A'wik/ for Cantonese. - 538 -200 fa . 22 7 T 1. pmg I ft £ 2. ming51 (L.) , mia:ng51 (C.) 55 x ^ •——55-ffi 1 ffi 3. tsing"" (L.), tsia:ng" (C. B -55 BB 4. tsing §igg~ 5. tsing 55 $fc 6. tsing -/ra | rt 7. ts'ing 5 5 51 ra 8. ts'ing 9. ts'ing 5 1 (L.), ts'ia:ng 5 1 (C.) 10. £0 I It ' l Z tsing 5 3 M 12. tsing 5 5 | 13. ^ I H 15. ts'ing 5 1 - S 16. ts'ing 5 1 14. tsing & I 17. tsing 5 5 & 18. tsing 5 5 I m. 55 (L.), sia:ng , (C.) ~ | j$ 20. sing^ (L.), sia:ng 5 1 (C.) (L.), siarng (C.) 19. . 55 sing 20. sing 5 1 21. . 51 sing 22. . 51 sing 17 23. 22 sing (L.) (cf. 202-12), sing5 1 (C.) - 539 -m 201 . i . k'ing 5 5 (L.), hia:ng 5 5 (C.) "fey iiH i - 5 5 2. ing •am 55 ' i * 3. ing S 4. in g 5 1 M 5. i n g 5 1 (L.), ja:ng 5 1 (C.) . IHPfH I /la 1 J l • n & m m 6. pia:ng fo fc H 7. l i n g 1 3 (L.), lia:ng 1 3 (C), nia:ng 1 3 (1) U 8. l i n g 1 3 (L.), lia:ng 1 3 (C.) m ft fo # 9. tsing (L.), tsia:ng (C.) IS 10. ts'ing 1 3 (L.), ts'iarng 1 3 (C.) £g 22 HT 11. tsing 12. tsing 2 2 13. sing 1 3 fc s l ~ f i a ' ~ £ b 14. a) ts'ing 5 1, b) ts'ing 1 3 as /# if? 15. tsing (1) /lia:ng / 'to apply for'; /nia:ng / 'collar'. - 5 4 0 -202 m S| 1. kia:ng m ft ,w 2. • ping 2 2 H 3. p'ing 2 2 (1) !%• 4 . ling n ft ft # 5 . tsing 2 2 (L.), tsia:ng 2 2 (C.) '14 6 . sing 2 2 22 22 £4 7 . sing (L.), sia:ng (C.) ft] 22 f$ 8 . tsia:ng &S If IE 9. tsing"""' (L.) , tsia:ng~~ (C.) 22 |& 10. tsing S P - I - I • 22 il 11. sing 12. sing ^1— & 13. king 2 2 1 (1) 202-3a. -4-3 /P'ing / (Qiao). - 541 -203 .R3H : fait A . m 1. pik ffl 2. p'ik 5 3£ rH 3. p'ik 5 > m 4. — Ira i t 5. tsik s (Ball: /tsik/, /tsek/) 6. 7. tsik 2 tsia:k fa *-J 7.1 2 ts'ia:k (Ball: / t s ' i t / in Shi-qi (1) ] ft n 8. tsik 2 9. tsik 2 i t 10. sik 5 (L.), 5 2 sia:k (C.) H 11. sik i f 12. tsik 2 13. tsik 2 2 14. tsa:k 15 2 5 15. tsia:k 6 16. tsik 2 (L.), tsia:k 2 (C.) # 17. ts'jJ<5(L.), ts'ia:k 2 (C.) ft 18. ts'ik 5 ^ •» - v t 1 I * 19. ts'ia:k 2 20. s i a 2 2 (cf. 12-9) 5 7® 21. sik W 22. sik 5 l£ 23. sik 5 S" 24. sia:k . 2 25. sia:k (1) Ball records /tsak/ and /tsek/ for the rest of the Zhong-shan county. - 542 -204 &y i . i * 5 a ^ akl>l ft m 2. 3. i k 2 i k 2 4. 5. i k 2 i k 2 m 6. i k 2 • P3 $27 P3 7. 8. 9. 10. 11. 12. 13. 14. , . 55 p' ing , . 55 p'ing ,. 51 p'ing , • 51 P 'mg . • 51 p'ing . 51 ming . 51 ming . 51 ming -55-,. 22 ,13 , p'ing , p'un (L.) , p'ia:ng 5 1 (C.) (1) (2) -'-Oj (1) (2) (3) T IT E~T' 15. 16. 17. 18. 19. •m 21. 22. 23. 24. 25. 26. 27. 28. ting ting 5 5(L.) tiarng 5 5 (C.) tia:ng" 4.- 5 5 ting 55 20. t'ing 5 5 (L.), t'ing" (L.), t'ia:ng J J (C.) 22 55 t'iaing' 4.- 55 ting 4 . , • 51 t'ing 4 . , • 51 t ^ 5 1 t'ing 4 . . • 51 t'ing 4 . , • 51 t'ing t'ing 55 51 (L.), t'ia:ng^ (C.). (3) 51 /p^aing-'"'"/ only occurs in the combination /jawbi p'iamg^V ;rj? 'child brought to a second or subsequent marriage by a woman'. 204-14a. /ming13/ (Qiao). 51 22 51 /t'ia:ng / occurs in e.g., / t i t'ia:ng / of the house for drying grains, etc.—in the villages'. 'area in front - 543 -205 . & S 2. l i n g 5 1 (L.), l i a : n g 5 1 (C.) ^ 3. l i n g 5 1 (L.), l i a : n g 5 1 (C.) A T 51 ^ 4. l i n g 5. l i n g 5 1 n e. -' ^ 7. l i n a 5 1 IS m & & flf 8. t s ' i n g 5 5 (L.), t s ' i a : n g 5 5 (C.) fif~&£ 9. t s ' i n g 5 5 Mk 10. s i n g 5 5 ( B a l l : /sing/, /seng/ for S h i - q i (1) ) M 11. s i n g 5 5 (L.), s i a : n g 5 5 (C.) ( B a l l : /seng/ always) r l m im , , . 55 .12. king ^ 13. hing"""' M • 14. h i n g 5 1 M 15. h i n g 5 1 M 16. h i n g 5 1 ± 18. ping 22 (1) Except f o r S h i - q i which approximates Cantonese speech, B a l l records /ang/ f i n a l corresponding to both /ang/ and / i n g / f i n a l s i n Cantonese. - 544 -206 is l . . . 13 ting 2. 4.- 1 3 ting IT 3. 4- 5 5 ting 4. 5. 6. (L.), tia:ng 1 3 (C 13 t ia:ng t 1 ing t i n g 2 2 (cf. 206-12) rt ft 13 13 7. sing (L.) , ts'iamg (C.) j g m I IT~(± JT t i n g 5 5 (L.), tiarng 3 5 (C.) (cf. 204-16) 22 22 nT~fo 9. ting (L.), tia:ng (C.) • ~ £ 10. f i n g 2 2 (L.), t'ia:ng 5 5 (C.)(cf. 204-20) AE? 11. t i n g 2 2 J§ 12. t i n g 2 2 ; $ ~ ^ j ( S r ) 13. ning 5 1 (cf. 205-1) IA -22 14. ning * I C 5 ) !5. Una 2 2 r t ft 2B 16. sing 1 3 (L.), ts'ia:ng 1 3 (C.) (cf. 206-7) - 5 4 5 -IL im 7S: k i n g ' 2 2 2 0 7 ~m 2- kin<? 2 2 3 . k i n g 2 2 4 . k 1 i n g c i . • 2 2 5 . k ' i n g 5 5 ( c f . 2 0 5 - 1 2 ) A pik (Chao: /pia:k /) 6 . 7 . 8 . p'ia:k p'ik 5 mik 7l 2* m 1 0 . 1 1 . 1 2 . 1 3 . 1 4 . 1 5 . 1 6 . 1 7 . 1 8 . t i k " t i k 5 t i k 5 , t ' i a t k ^ t ' i k 5 2 t i a : k t i k 2 t i * 2 t i a : k 2 - 5 E 1 9 . n i k " 2 0 . 2 1 . ft 2 2 . t s i k " 2 3 . t s ' i k 5 2 4 . t s i k 2 2 5 . s i k 5 ( L . ) , s i a : k 2 ( C . ) 2 6 . s i k 5 - 546 -208 (1) m (be IM [f 4. wa: ng 8 ! ^ - : HPS ± £ U m ^ ( | g ) 5. k'ong2^ : HPS m 51 22 |Sil*~ 6. wa:ng (cf. 208-4), wa:ng (2) m^=.: mm A ! & J S I ~ . 7. kwikb (1) Chao: also /ja:k 2/. Ball records /yak/ for both Shi-qi and Cantonese, but /yiek/ for Macao. (2) /wa:ng22/ occurs in e.g., /wa:ng22 wa:ng22/ i% 'line after line' (to depict horizontal parallel lines, bars, etc.). % I m i . kik 5 j m 2. kik 5 m | P £ ( I H ) 3. ja:k 2 - 547 -209 m 8H I J i i u kwangbb a * 2. — ^ 3. wang51 . a -I f 55 4. kwang (cf. 209-1) A m cr ,2 as 5. wa:k tfl 6. wa:k2 m 5^ 7. hing b b ^ 8. i n g 5 1 ± 7J\ 9. wing 1 3 - 5 4 8 . -2 1 0 P-7JN 1 . 2 . . 2 2 w i n g . 2 2 w i n g • /ra a ¥ fa ft fa ft m 3 . — fo 4 . — JL i l 5 . k ' i n g 6 . k ' i n g 5 1 _ • 5 1 7 . m g • 5 1 8 . m g ± #S fa ft fo M i l m 9 . — 1 0 . — ' f t - 5 4 9 -211 E bf 1" aft ' -— ma 1 . k'ing 1 3 ^ — f i s f 2 . — m 2 2 j f c l 3 . wing : f t i t A e £ S i 4 . wik2 f x 5 . w i k 2 m^m: If £1 p f . . . m M 6 . i n g 5 1 z F t — • ( $ § 7 . w i n a 5 5 (2) _ _ ± ai m ' ~ M 8. — (i) i n i t i a l : 2 1 1 - 5 . 1 . J|) Awing/ or Aang/ for S h i - q i ( B a l l ) . ( 2 ) S a m e pronunciation given on two separate occasions. - 5 5 0 -212 A m m m H i . -m ft m 'w o , 51 . 51. . g£ 2. p'ung , hung (L.) 51 ^ 3. p'ung » , 51 ^ 4. mung Mil m m. T±T 55 ip. 5. tung ;S c 4. 55 M 6. t'ung * '-feUCS-?- 7. t'ung 5 5 ^ 8. t'ung 5 1 i l 9. t'ung Iff 10. t'ung 5 1 IR! 11. t'ung 5 1 51 -S. 12. t'ung SS t - ' n n g 5 1 • m f i 14. l u n g 5 1 H 15. l u n g 5 1 u& 51 T i l 16. lung - 5 5 1 -213 ice m fa ft fo # 1 . tsung 5 5 #f ( m) m~ - 2 . tsung 5 5 M 3 . ts'ung 5 5 & 4 . ts'ung 5 5 -1! 5 . ts'ung 5 5 \£\m~ 6 . t'ung55 & 7 . ts'ung 5 5 ' o 5 1 8. sung J L l¥ 9. kung33 1 0 . kung55 X 1 1. kung ZJJ 1 2 . kung55 $C~ m 1 3 . kung55 -^ — M 1 4 . k'ung55 1 5 . hung^ (/hong^/ C.) (cf. 2 1 5 - 3 . 1 ) E 1 6 . hung51 ^ 1 7 . hung51 $ 1 1 8 . hung51 £E 1 9 . hunq51 (Qiao: /p'una51/ C i ^ . 55 m 2 0. Dung ± m _ *J3 13 l f | ~ ' j j | 2 1 . mung m~& 5 1 5 * 4 8 1 2 2 . iming • • — - 5 5 2 -2 1 4 m M 1 • tung It 2 . tung 1 3 11 3 . t'ung13 7 4 . . — 5 . tung M ^tl 6 . lung 1 3 ( 1 ) - m m ft & $H 7 . tsung 1 3 ! m » ' ?L 8 . k'ung13 E (.V£}~m 9 - hung13 2 2 ^ 1 0 . hung i l ^ - : MM i t -H-U Kffif ^ 1 1 . tung 2 2 W. 1 2 . tung 2 2 00 •f 1 3 . t'ung • / § 1 4 . tung 2 2 & lt£&Hgi 1 5 . — = 2 2 Jf 1 6 . lung IS ft ft SB ^0t 1 7 . tsung Z Z 2 2 1 8 . sung ( 1 ) 2 1 4 - 6 a . j t | /lung / ( C h a o ) . - 5 5 3 -215 1 m M 2 2 1 . k u n g 2 2 2 . k ' u n g 3 . k u n g ( c f . 2 1 3 - 1 4 ) m m 2 2 3 . 1 . h u n g 2 2 4 . k u n g ( / h o n g " / C . ) ] m 5S 2 2 • a 5 . u n g A 0J3 6 . p u k ~ 7 . p ' o k + fhfj 9 p ' o k 2 1 0 . 1 1 . p ' o k p u k ' p u k , p o w p ' u k ( C . ) 2 2 •Tfc 1 2 . p u k , p o w ' 2 2 1 3 . 1 4 . m u k m u k ^ ITS fc am. 1 5 . 1 6 . 1 7 . 1 8 . ' ^ 1 9 . 2 0 . 2 1 . 2 2 . t ' u k " t u k 2 t u k 2 t u k 2 t s u k ^ t s u k 1 * l u k l u k ' - 5 5 4 -2 1 6 ft m 1 . , 2 tsuk & m 2 . ts'uk & m. 3 . kuk13 4 . kuk5 5 . huk5 E 6 . huk2 M 7 . uk5 8 . 4- 5 5 tung 2 2 t'ung 51 1 0 . n. ^ 1 2 . nung c nung" c nung' 5 1 J 5 1 7Js 1 3 . ts 5 5 1 4 . sung 5 5 - 5 5 5 -1. kung 5 5 217 iw m 3. t'ung m m ft fo 4 . sung A 5 . 6 . tuk" c tuk" tuk 8. tsung' 5 5 ft fo f P tsung 5 5 - 5 5 6 -2 1 8 JL i . k u k5 W m E 2 . ^ m 3 . j u k ^ -* m #$ 4. 5 . — 5 5 — — — — , _ h u n g h u n g ft * IE am -52. 6 . 7. h u n g , 5 5 h u n g m 8 . h u n g m #s 5 1 l u n g m 9 . fa ft SB fc 1 0 . l u n g 5 5 * 5 5 ri] 1 1 . s u n g _ 5 5 — C P ^ ^ 1 2 . t s u n g 5 5 * 1 3 . t s u n g — t 5 5 ^ 1 4 . t s u n g /41 1 5 . t s ' u n g 5 5 » . 5 1 II 1 6 . t s ' u n g 1 7 . t s u n g 5 5 ( c f . 2 2 2 - 1 9 ) - 5 5 7 -'219 w fo-lk u- 5 1 1 . s u n g ^ 2 . t s u n g b b 3 . t s ' u n g 5 5 0 2£ 4 . j u n g 5 1 5 . i u n a 5 1 E If ^7 6 . k u n g 5 5 13 7 . k u n g 5 5 s 8 . k u n g ^ 9 . k ' u n g SI 1 0 . k ' u n g 5 1 t 6 1 1 . h u n g 5 1 M 1 2 . h u n g 5 1 ra& 1 3 . j u n g 5 1 ^ M, 1 4 . h u n g X J ( 1 ) MK 1 5 . h u n g 2 2 -as- 2 2 1 6 . m u n g ft] It 1 7 . t s u n g 2 2 2 2 -fqfj 1 8 . t s u n g ( 1 ) Same pronunciation was given on two separate occasions. - 5 5 8 -BE it m 1. 2 . 2 2 0 t s u n g t ' u n g 2 2 . 13 3 . t s 1 a w ' 27" ( c f . 9 0 - 1 6 ) l c i -A m 4 . h u k b 5 . h u k 5 mm— 6 . n u k 5 Wt 7 . h u k 5 m 8 . h u k 5 9 . h u k 5 M 1 0 . h u k 2 1 1 . h u k 2 1 2 . h u k 2 1 3 . h u k 2 1 4 . m u k 1 5 . m u k 2 • & 1 6 . m u k 2 fo fc 1 7 . l u k 1 8 . l u k 2 #3( 1 9 . l u k 5 ML 2 0 . 2 1 . 2 2 . s u k " c s u k " c s u k " - 5 5 9 -221 ft l . t s u k 5 2 . t s u k 5 m i f 3 . 4 . 5 . t s ' u k 5 2 t s u k 2 t s u k u. OJ 6 . s u k 5 = — eg 7 . t s u k 8 . t s u k 5 m <&- 9 . s u k 5 2 1 0 . s u k 1 1 . s u k5 s m 1 2 . j u k Ira 1 3 . k ' u k ^ - j * 1 4 . k ' u k5 r l ' 1 5 . k ' u k 5 — g • m 5 ' • " f t 1 6 . t s ' u k J ( c f . 2 2 1 - 3 ) E 1 7 . t s ' u k 5 5-- • 1 8 . j u k ' Wat — W 1 9 . j u k 2 - 560 -222 m i t =£t 1. hung"" ^ 2. hung55 ^ :3. hung55 4. hung ^ 5. hung55 6. hung 7. hung51 '"l 51 m fc 8. nung"'x (L.), Jung" (C.) f| 9. lung 5 1 f i fa ft gj£ 10. tsung"" . . 11- tsung 5 5 ft~§ 12. tsung 2 2 ft^~ 13. ts'ung 5 1 I£ 14. sung l& 15. ts'unq 5 1 to j l ~ 16- ts'ung ^ r-r-H 17. tsung"'"' SU 18. tsung 19. tsung 5 5 55 IH 20. ts'ung 21. tsung 5 5 a 22. iuna 5 1 - 561 -223 r l m ^ 1. kung^ 55 . 2. kung m J$ ,• 3. hungib >^a~ 4. hung M 5. hung55 & 6. jung 5 5 i&E 7. jung 5 5 5 v 7 n • 51 igf. 8. jung TPS 51 9. jung £?r 10. jung 5 1 Im 11. jung 12. jung 5 1 ± # 13. hung13 (L.), p'ung13 (C), pung13 (C.) (1) lf§ 14. hung22 f t 15. lung 1 3 16. lung 1 3 #s rt ft & ft (i) %£~JB. 17. sung 1 3 •m- 18. sung 13 /pung / 'to hold up in both hands'. - 5 6 2 ft 1 . 2 . 3 . t s ' u n g t s ' u n g . 1 3 1 3 2 2 4 2 2 1 3 t s u n g ( L . ) , t s ' u n g ( C J n 4 . 5 . t s u n g t s u n g 1 3 " 1 3 B 7 . 6 . j u n g 1 3 IT j u n a ' ^ 8 . k u n g 1 3 9. 1 0 . k u n g 1 3 k ' u n g 1 3 1 1 . j u n g ' T 3 ~ ffl—il 1 2 . j u n g ^ 1 3 . j u n g ' 14. jung 1 3 1 3 1 3 i a -m 1 5 . h u n g • 1 6 . h u n g 1 3 5 1 ( c f . 2 2 2 - 7 ) m . fa ft fo - 5 6 3 -2 Z 5 £ f ; f t ~ 1 . tsung 2 2 . 2 2 HFB 2 . tsung 2 2 §1 3 . tsung §S 4 . tsung 2 2 eg I f j .£ 2 2 " 4_ll~ ^ 5 . tsung E • ± ~ 6 . kung33 (cf. 2 2 3 - 2 ) & 7 . kung22 sfef 5 5 5 5 a S ~ , f j E 8 . jung ( C h a o : '/ung /) 9 . jung 5 5 (cf. 2 2 3 - 6 ) ffi 1 0 . jung 2 2 A fc n. l u k 2 1 2 . I n k 2 1 3 . t s u k 5 fa is 1 4 . t s ' u k 5 fo M 1 5 . s u k 5 ft* 1 6 . t s u k 2 ( L . ) , s u k 2 ( C . ) W. 1 7 . t s u k 2 - 564 -226 m i . tsuk5 m 2. tsuk5 w$ 3. ts'uk5 Bra 4. suk2 5. ts'uk5 m 6. tsuk5 7. suk 8. juk 9. j u k ' 8 f e 1 0 . •if? > » n. 12. kuk2 13. juk 2 14. juk 2 i S ~ 0 15. juk" k'uk" / H i 16. 17. juk juk* 18. juk - 5 6 5 -INDEX This index covers all the characters representing the morphemes which are included in the lexicon of this dialect handbook. Characters are arranged alphabetically according to the p'in-yin spelling of their Mandarin pronunciation. Pagination follows that marked at the top right or left corner of the lexicon. This pagination will enable the reader to find out all the corresponding dialectal forms of a given character (hence morpheme) included in our handbooks, by referring to the same page number in the handbooks. Alternative forms (pronunciations) of a morpheme in Mandarin are indicated with an arrow. Alternative pronunciations of a character repre-senting different morphemes are given in parentheses indicated with an arrow; these will help the reader check the semantic and/or the morpho-logical derivation of morphemes, if any, in various dialects. The Mandarin pronunciation of characters adopted here is that listed in Ting Sheng-shu and Li Jung's Ku-chin tzu-yin tui-chao shou-ts'e(Peking: K'o-hsiieh Ch'u-pan-she, 1958). No normalization or correction was attempted on the part of the present authors. - 5 6 6 -Roman i za -t i on a 1 a 1 a i a i an an C h a r a c t e r m P5J - a a I X a I PH Page number 1 106 28 28 28 32 28 33 31 34 30 30 31 35 92 92 115 115 93 94 Romani z a -t i on an ang ang ao ao ao ba ba Charac ter T P ft e A ao ao Page numbe r 117 117 117 167 167 68 72 68 70 71 71 71 71 71 72 7 7 7 120 120 136 bo - 5 6 7 -IE m - * pa 8 ban 121 121 ba W\ 9 9 9 ban m m 119 119 119 • h i 9 120 9 ft 134 134 n 34 135 IE —• ba 135 ba i —* bo bang n 166 bai -> bo 34 180 195 16 -> bo bang w 167 - bo 168 bai ^ 32 m 181 I f 34 bang w 168 35 & 168 be i 181 ban m. 120 120 §3; —* bang 181 31 120 bao 31 67 120 72 I X 133 m 72 -JaSia 133 133 j f t i ( - > bo) !fcf3 —»• pao 183 - 5 6 8 -bao bao bao bei be 1 be i m ft) bo) bo) it pao bo 169 183 69 69 69 73 69 70 70 70 73 73 73 73 70 40 42 47 47 51 30 32 32 41 be i ben ben ben ben beng beng b l bi' bi 1* bi bi pi ben ben it fit 42 42 42 42 49, 50 50 54 156 156 157 157 158 184 196 196 189 54 49 53 53 53 - 5 6 9 -153 35 35 35 35 35 35 37 38 38 38 38 47 49 49 50 50 b i it —• b i m- --* mi m -> mi 54 54 cat) 54 b i b i an b i an b i a n 153 b i ao b i ao b i ao b i an ft 153 199 203 207 122 122 129 129 102 130 130 124 124 125 125 125 130 131 131 75 75 91 77 77 77 - 5 7 9 -b i e b i e b i n b i n b i n g £ ( b i n 126 126 132 126 149 149 149 195 152 152 ft 187 & 197 bTng 111 198 198 ^ -*• b i n g if 201 JPI P i n g b i ng m 199 199 n 200 202 m 205 bo 3 3 bo bo bo bo m W. -* b° & - be, 4b bo bo 136 136 159 183 136 159 169 169 169 183 183 194 194 194 194 194 4 4 186 196 - 5 7 1 -bo bu V bu bu ca c a i c a i ca i ca 1 m a m 196 159 215 17 215 •15 15 17 17 17 17 17 159 118 28 28 28 28 28 28 29 29 29 30 30 31 can can can cang c'ang cao cao cao ce cen cen m 92 115 92 95 115 93 117 166 166 166 68 71 71 68 68 69 69 59 190 190 190 197 197 110 110 - 5 7 2 -c e n g E 3 184 cha i i f 184 c h i n c e n g 185 chan c h a 7 cha c h a cha c h a i c h a i c h a i c h a i c h a s - zha cha cha cha m che ce 7 99 7 7 120 9 9 9 122 33 33 33 33 194 chan chan c h a n g c h a n g iim m c h a n g 35 100 97 97 102 123 123 123 118 103 j§| 119 St 11.9 H= —* chan II 125 $t 100 ^ 172 H 172 { | ( - > chang) 172 S 171 S§ 171 ^ 171 172 172 172 • 172 171 - 5 7 3 -chang chang chao chao chao chao che che che chen chen j iao chao M (^) K -> r l ch 1 174 175 175 175 175 72 72 76 72 75 75 73 73 11 11 194 126 127 110 110 150 150 150 chen chen cheng cheng cheng cheng 150 150 152 152 cheng m 187 & 192 192 193 OS 200 187 m 187 s 187 m 187 187 187 187 192 i n 195 200 n 200 200 w 200 i s 200 & -* cheng - 5 7 4 -cheng 2 0 1 cheng 1 8 8 1 8 8 ch i -18 SI 5 6 P & 56 ch i m 4 8 m 4 8 4 8 is 5 2 ch i ch i chon^ 0. -> ch i 5 6 & 4 9 5 0 m 5 7 TEXT W 5 8 K 2 0 3 m 5 0 1 8 9 189 B 2 0 3 ft 2 0 3 % 2 1 9 m 2 2 2 m 2 2 2 chong chong chong chou chou chou chou chu • chu ft fe-ns. -(jfj —* chong c^jj —* chong m m m chu 2 1 8 2 1 8 2 1 9 2 1 9 2 2 4 2 2 0 8 6 8 6 8 6 8 6 8 6 8 6 8 6 8 6 8 6 8 8 8 8 8 8 9 0 1 9 1 6 3 1 9 1 9 - 5 7 5 -chu chu chu chua i chua'i chuan chuan chuan chuan ft I S -5F ft chuai chuang 23 23 19 • 20 20 20 21 21 22 162 221 226 61 140 140 140 140 140 140 141 141 142 142 139 171 chuang chuang chuang chuang chui chu i chun chun chun chuo chuo f i 181 Pf 181 U — * zhuang M(B) 174 jig —> chuang HJ 175 (£§) ft I fit) cha c 1 61 61 61 63 63 160 160 160 160 160 160 161 183 176 176 143 c 1 48 48 - 5 7 6 -c 1 c i cong cong cou m n it m % ex fa JlO *5 \ S 1 ce ft ft 51 51 55 55 55 55 55 4 9 50 50 54 cong 213 213 213 213 213 222 213 222 84 14 cuan cua n c u i \ c u i cun cun cuo ^ - > 2 IS ft ft cuo zan dun cha 18 159 25 133 136 139 40 40 45 64 64 64 157 160 157 158 159 1 137 4 6 - 5 7 7 -cuo da da da da d a i da i dai 7£ d i e ft da i de ft IS 170 6 94 94 94 94 117 192 2 28 29 29 29 29 29 29 29 30 30 da n dan dan dang dang dang dao dao f t ft B m cheng la dang 71 115 115 96 116 95 96 116 116 116 116 116 166 167 167 168 167 168 67 67 69 69 69 - 5 7 8 -dao dao de d ¥ i deng deng deng d i d i IP to dao de d i 69 70 69 69 70 70 70 186 186 184 184 185 185 185 185 185 188 37 37 207 207 207 d i d i d i d i an d l an d i an d i ao ft M it to mi 207 207 207 38 38 38 39 39 39 54 207 106 129 129 107 130 107 107 131 131 131 131 131 107 79 79 - 5 7 9 -d i ao H 79 79 m 70 d 1 ao ,% 80 d 1 ao 8.1 81 81 m 81 d i e it 10 108 d i e 108 108 108 108 108 & d i e 132 T 204 f l 204 I f 204 4T 204 204 IT - ,v d 1 ng IT -11 ding 206 206 d i n g IT 206 £T 206 IT 206 J E 206 206 d i u ^ 85 ^ 91 dong ^ 212 4 - 216 d o n g g 214 til 214 dong ^ 214 t£ 214 t t l 214 M 214 dou ffS 14 % 81 dou i f 83 Pr 83 B £ 83 dou 84 J£ 84 84 H 84 |ft —> du dC .f[S 14 - 5 8 0 -du H 217 du Wl 215 IH 215 m 2 1 5 fl| 215 • '/Ji 215 S 217 du *£ 16 St 16 At 16 H 217 du tU 17 t t 16. ft 16 17 duan j/^ 133 duan £ s 134 duan Hi 134 i f f 135 IS 135 WL 135 135 duan $ 5 135 dul i(£ 40 dui % 42 fit 42 B£ 42 £ 43 dun 156 i t 156 $0$. 157 dun i£g 158 H 158 J f 153 ' Eg 158 158 §t£ 158 a l 158 tt 156 duo ^ 1 duo —* duo % 137 £1 169 duo %k 5 4$ 5 £ K S ) 5 duo f£ 2 - 5 8 1 -du du du m fit du 217 215 215 215 215 215 217 16 16 16 2 1 7 du iu 1 7 m 17 tt 1 6 at 16 17 m 17 m 17 duan in. m 1 3 3 duan m. 1 3 4 duan m 1 3 4 m 1 3 5 IS _ 1 3 5 1 3 5 m c m 1 3 5 duan du i d u i dun d u n duo duo duo ft ft duo duo IS IS 135 40 42 42 42 43 156 156 157 158 158 158 158 158 158 158 156 1 137 137 169 5 5 5 2 - 5 8 2 -f an 1)7 fang fx. fang m f ang 1/3 m fang a/3 fei US fei IE f e i fang f ang 109 109 109 109 145 145 178 178 178 178 178 178 179 179 179 179 179 179 65 65 65 65 66 66 f e i f e i fen fen fen fen feng 66 66 B 46 m 46 46 66 66 66 163 163 163 163 163 163 164 f en pen fit (7>) 164 164 164 164 164 218 218 218 218 - 5 8 3 -f i n g feng feng f eng f<» I O U fu m fu feng feng fu fu 222 222 222 222 222 218 222 222 219 219 223 224 224 165 87 23 23 23 23 23 23 23 fu fu ft ft fu fu fu io 23 23 23 85 165 165 165 180 220 220 220 220 220 220 25 25 25 25 25 25 25 5C fu - 5 8 4 -fu fu ga 2a i ft lb m m m fu fu pu j i e 25 .25 25 25 26 26 26 26 26 26 87 87 87 89 89 89 220 220 220 220 gai gai gen gan gan gang n -tt-W J l cm) Iff ft 28 29 30 30 31 31 95 95 95 j i an 115 115 115 115 93 96 96 116 116 116 116 117 167 167 167 167 5 8 5 -gang gang gang gao gao gao ge ft (ffi) n . -c m -m -kang SE -(•«* M can J i a n g / f t • P P 3Xr (18) 167 167 181 gang gang 194 182 169 169 68 68 68 68 68 70 73 71 71 71 1 1 4 118 170 170 ge ge ge gen gen gen geng ft M. w geng Hj| ge he ge ge gen J m g 94 94 94 118 170 194 197 197 no 3 170 114 148 148 149 185 192 192 192 194 193 - 5 8 6 -g t " n S gong gong gong gong gou m I J m g gen geng 193 193 196 kuang hong 213 213 213 213 213 217 219 219 219 223 223 214 224 224 215 225 225 82 gou gou gou gu m 82 m 82 83 83 as 83 84 84 84 m 84 84 m 84 15 15 15 15 5£ gu 16 16 16 16 16 16 16 16 160 5 8 7 -gua gua g u a guai guai guai guan h u G JW ft m 7t± gao W O kuo guo 216 216 18 18 18 18 12 140 13 13 45 45 43 44 44 134 134 134 134 137 guan guan guan guang gu¥ng guang guT gu 1 E5 m M yt $fc —* guang I f g u i 138 135 135 136 136 136 136 136 139 177 177 178 180 40 46 46 61 63 65 62 64 64 64 66 43 - 588 -Rui m 43 ha i m 32 m 43 h a i & 32 fflk 46 han m 95 47 & . 95 62 iff 95 M 65 Iff 115 - E E . 67 han 92 gun m 158 i i 92 gun 159 92 guo Wi 4 95 m 137 115 IB 178 f t 115 guo i l 191 f p - > han guo 5 han 96 5 98 H 5 116 guo 4 han 93 ha fa 8 « 94 ha m -> ha E3 116 m ~* x i a 117 & - ge 117 h a i — ke 117 h a i 28 117 ha i M 29 hang 181 h a i % 29 hang 167 31 m 167 - 589 -hang hang hao hao hao hao he he ft 2? ham he ft ft «.* TO 1 6 7 he 3S 94 9 7 6 8 & 1 6 0 6 8 ge 6 8 1 9 7 6 8 1 1 8 6 8 he ) 3 6 8 3 6 70 1 0 1 7 0 (1 -> he 7 0 IS 1 3 7 70 ^ -> hao 7 0 H 1 7 0 7 1 £• 170 71 1 9 4 7 1 i s 1 9 4 9 4 hei 1 8 6 1 1 8 hen 1 4 8 1 7 0 hen m 1 4 9 1 hen 'IS 1 4 9 1 heng 1 9 2 1 heng 1 8 4 4 ^ — heng 4 1 9 2 94 2 0 8 - 590 -heng 208 hong m 209 209 213 hong 191 209 213 213 213 HE 213 hong m 214 m - gong hong pa - hong m 215 hou 82 m 82 M 82 m. 82 hou 83 hou 83 m 83 83 i n 85 hu 15 160 m — hu hu hu hu ^ -*• he & m m m m c a n m m he m hu xu hu 15 15 15 15 15 15 15 15 15 215 17 17 17 hua ft 15 17 17 18 18 18 18 13 591 -hua hua hua hua hua i hua i huan huan huan huan it m ' C D . 33 . ( B ^ ) ft 13 13 13 138 138 13 13 13 45 45 209 43 43 43 44 134 134 138 138 135 135 136 136 136 138 huan huang huang huang huang hui hui ( 5m. ) huang ik M 0 hui 139 139 177 177 177 177' 177 177 177 177 177 177 177 177 40 40 41 61 65 65 65 41 41 41 41 - 592 -hui hui hun hun hun hun huo 5 § m ft n kui 5Tn hun 62 62 41 41 42 42 43 43 46 47 47 67 67 67 64 157 157 157 164 157 157 157 158 137 huo huo huo huo J i (*) m 178 137 5 5 5 6 137 178 178 178 191 191 209 37 37 48 52 52 56 56 56 59 59 59 59 593 -m ge ch i J i 203 203 207 208 113 113 114 114 114 154 154 189 190 190 190 203 203 207 J i J " J i ( - > j T ) 60 JI i t J.I gei J i q i 38 53 58 199 36 36 36 39 39 39 39 39 39 50 50 50 51 55 58 59 60 65 - 594 -J i j i j ia J ia J i a J i a 7JP # l ( iO as m j i a /3r, TJ51 J ia x i a J ia 38 189 189 113 j lan 1 (Ai ) 8 33 99 99 9 9 101 101 10 10 10 10 10 10 j lan j i an ft ft fa 98 100 101 101 107 107 115 118 118 121 121 123 123 127 129 129 98 98 103 103 119 119 119 119 122 5 9 5 -j ian j 1 an m JUL rau 3L Its j i an j i an 124 1 3 0 1 3 0 1 3 0 100 1 0 0 1 0 0 1 0 3 1 0 6 J • ang 1 2 2 1 2 2 124 1 2 4 1 2 5 1 2 5 1 2 5 1 2 5 1 2 8 1 2 8 1 2 8 1 3 1 1 3 1 1 0 7 171 J i ang as J i ang J iang I I m j i ao J l an J i ang 171 1 7 2 1 7 2 1 7 2 1 7 2 1 7 2 1 7 2 181 181 1 7 3 1 7 3 1 7 3 1 8 2 1 8 2 1 7 5 175 1 7 5 1 8 2 1 8 2 72 7 2 7 2 7 2 75 - 596 -j tao j 1 ao j i ao j i ao j i ao 5ft j i ao j i ao W -* jue % ( - » • jue ) ^ ( - > jue ) 8 ( f c ) C 3 75 75 76 76 80 176 73 73 73 73 j ue 78 80 80 183 183 .183 74 74 74 74 74 174 j i ao j i ao' j ie J ie [Uf-j i ao jue ) 7FP m J ie ge j ue 7 8 7 9 8 1 183 77 31 31 31 31 33 104 128 106 127 128 132 132 133 133 133 104 - 5 9 7 -§ ± •>4- „ . . ' ft 7r Ik m -+ j i e * * ( - J e ) 7 TN 146 154 11 34 12 12 12 33 33 33 33 33 33 33 33 J m ft 203 111 111 111 111 150 151 J m J m J mg ft •<m) 2£ j i n v j i n geng 155 155 187 112 151 152 153 155 112 113 113 155 156 156 B BH 187 192 195 198 198 198 198 200 200 - 598 -J m g J i n g J m g J i ong j i u Bit I S SIS IS Hi 200 200 205 198 198 201 202 198 199 199 199 199 201 201 202 202 207 207 207 161 211 86 87 87 87 90 J i u j i u J iu JU JU M M % a 91 J iu JU 88 88 88 90 90 90 90 19 19 ft 24 24 24 — • JU -*• ju 163 221 221 599 -J u J u j uan j uan j uan JU shen 226 21 26 22 24 25 22 22 21 21 21 27 27 27 199 226 19 147 141 142 142 142 142 j uan j ue j ue j ue j un j un j un g u i € ( f ) ft j ue j un j i o n g quan 142 10 74 143 146 146 146 146 148 148 165 165 176 180 183 183 161 161 164 164 162 161 - 600 -jun i$_ 162 IB 165 ka % 120 ka —> ke m ) 194 kl i 13 28 31 k a i 0 1 29 f a 29 « 30 32 kai '(fE kai mtM) (-^ke)30 kan j g 92 H 92 i t 115 flj 115 kan ^ 93 & 93 •f£ 116 kan @J 93 117 $$[ 92 kang Jf| 167 H 167 'tH 168 kang tL 181 kang ffjt —*• kang kang £r[ 169 */L 169 S3 169 kao 5^ 70 >fc£ ( ) 70 kao j j | 71 til 71 ke fiff 1 3f 1 •. m 4 * 4 S ( m ) 4 IS 5 i l 97 ke ^ 10 Pg ( 30, 194 IS -> Qiao ke PJ 2 711 118 ke oj —• ke' £lj 186 % 186 l r 194 6 0 1 -ken 149 H 149 m 185 king tr[ 192 $g 195 195 1 9 5 195 ffcl —> qmg kong Z£ 213 kong ?L • 214 S 224 kong Zg 215 ^ 215 kou $1 82 "91 ( ) 82 kou • 83 kou PP 83 ft 84 M 84 IP 84 ku t f i 15 H 160 5s. 216 ku ^ 16 ku • JM. 18 ku W C ^ ) 18 218 kua 12 kua m ( 3 * ) 1 3 i% 13 kua jg5 13 kua> Bl 44 kuai Ifa 42 ^(ift) 44 ft 45 ^ 45 # -> hui kuan J|[ 134 kuan 135 kuang | E 179 f g 179 kuang ^ 179 kuang gg 178 178 <ft 180 $ f 208 kuT | § 40 I ! 61 H 61 kui jfjj; 40 0 46 - 602 -k u i ku i k u i kun ku n kun kuo cm m la l a i I ) 63 63 63 64 41 42 65 65 65 157 157 157 158 159 137 137 178 178 94 97 97 97 117 117 28 I a i la i 1 an 30 30 32 92 95 95 115 115 115 96 96 96 116 96 96 m 117 1 an 93 l a n g as 166 JIB 166 166 • i t 166 m • 166 l a n g ® 167 1 ang 168 1 ao m 67 603 -1 ao 1 ao ao I e i l e i l¥i m 67 lei 67 lei ® ( D f ) 67 69 m 71 71 m -> 1 ao 1 eng m - 1 ao leng uo ) 169 uo ) 169 1 i U K - * 1 uo ) 169 *& - 1 uo ^ —* le yue 1 uo fa -+ 1 uo —*• 1 uo . f t ( - I e i ) iHffl 5!" lei le i 169 186 186 186 40 41 64 '13 im. leng 42 42 62 64 64 186 184 192 37 37 47 47 47 47 51 55 55 38 53 57 57 57 57 57 36 - 604 1 : an m •46: 1 ing i an 36 36 39 39 50 50 54 54 58 113 113 113 1 3 189 207 207 101 101 101 101 122 122 122 129 129 102 i an fit 103 i an m 103 m 103 m 131 m. 131 m . 131 m 142 m 142 iang & 170 170 M. 170 m 170 170 170 iang M 173 iang ft 174 174 m ( r l ) 174 - M L . 174 1 i ao 1 i ao I i ao 79 79 79 79 79 75 605 -i a o 7 1 i a o l u 78 77 80 80 81 81 i e 104 n 126 126 126 143 i n 110 110 m 110 110 n ISO m 150 150 150 i n n 111 111 m 111 V i n 110 % 112 & 152 i n g i n g , v 1 i n g i n g l u 1 i n g ft m ( m 1 i n g 187 187 187 187 204 204 204 204 204 204 201 201 202 206 85 85 85 85 85 85 88 89 89 89 - 606 -i i u 1 ong l o n g l o n g l o u I o u I ou m m m m m 220 181 212 212 212 218 218 222 214 223 223 (—> nong ) 214 —> 1 ou 82 82 82 83 83 84 84 14 14 14 14 14 l u 1 uan 1 uan 1 uan 1 un / N ( —> I i 11 ) m, am (" 7m l u n ) 18 16 16 16 16 16 17 17 17 17 215 215 220 220 220 225 225 133 133 134 136 157 157 160 160 160 6 0 7 -I un I uo I uo 1 uo lu 3FFB m - • 1uo m 0 IS 159 37 3 3 3 3 5 5 6 169 169 169 169 169 169 18 20 20 20 20 i ue ma ma ma mai mai ma i 25 26 53 137 21 21 162 ( -> s hua i ) 162 225 176 176 7 ft 2g m6 IK mo ) mo ) man 7 7 7 8 8 9 31 34 34 35 196 196 - 6 0 8 -( —* ma i ) ma n m a n g ma n g ma o mao ma o OS) t t t i c 31 120 133 133 134 121 135 135 man 180 166 166 166 wang ) 178 191 JtC ^ S ) - * m e n g 1 7 8 3$ 167 3EH Iffl 167 72 75 67 72 72 85 73 70 mo l me n m 70 & 73 m 84 84 40 40 ** 40 40 40 i i m 35 m i me ng ff5 me n 51 51 51 159 41 53 42 42 54 54 156 156 158 me n g - 609 -flit' n g me n g me n g i t 3 ^ me n g mi an 191 195 197 212 192 213 213 193 219 37 47 47 47 47 37 38 49 49 54 54 153 153 207 122 me an TI l a o m i ¥ o m l a o m i n IE # y ao mi n m 122 129 124 124 124 124 124 124 125 131 75 75 77 77 78 78 91 126 132 149 151 151 - 6 1 0 -m i n g m i n g mi u mo mo mo nu n m & mi ao ft(£l) ft ( « ) mo 151 151 198 196 197 197 200 204 204 204 199 91 169 3 3 3 3 3 14 169 120 136 136 mo mou mu mu ft mo mo mao mi ao mo n 136 159 169 169 186 186 194 196 196 85 82 82 82 82 82 17 17 17 17 169 215 - 611 -m u ft n a n a n a n n a n n a n g n a n g n a n g n a o m -(§15) 77 n a n a n g J i (in) 215 220 220 220 7 2 2 91 91 117 29 34 30 30 92 92 115 124 117 166 167 181 neng n i n I 1T5 [if iH rao ft no ng e r n a o 72 72 69 59 69 74 41 42 159 184 37 37 51 51 57 58 54 189 199 207 - 6 1 2 -n 1 an n i an n i ang n i ang n i ao n I ao n i e n ie n I e n i n g 3^-,% C —* d '^°) n l 153 123 101 106 106 129 124 124 108 130 130 108 125 170 174 80 81 132 108 104 104 104 133 187 204 n i n g n l u n o n g n o n g nou n u n u n u o n u o m m m i u ft n a n 187 206 206 87 88 88 88 216 216 216 222 214 214 84 14 14 16 17 134 1 6 169 20 - 6 1 3 -nun *$k(.~* n e n ) 159 n'ue j | | 176 S ' 176 0 u —• ou PR 82 82 82 | » 82 m 83 \y 0 u 83 « 83 nfD" Hjro 83 84 0 u 85 It 85 pa 7 E E 7 7 7 9 pa ta 9 9 pa i 194 ft 194 p a i 31 33 pai 33 pai ty\ —> pai pai ^ 34 pan ban ) 120 # ( - > fan ) . 144 • -*• ban IP 120 'A i m ) 120 $1 133 pan $g 133 S$ - * pang pan |5} 119 # 121 #f 133 135 Wi 135 pang 166 pang ffi. 167 3? 166 2$ 166 t^C-^ fang ) 179 i l l 180 $f - * bang pang fi| (-> pan ) 180 0 ( # ) 182 pao (-> pao ) 72 - 614 -pao pao pao pao pe i pe 1 pe 1 pen pen pen peng Peng 72 —• pao 72 72 67 j;g -»• pao m 72 73 •mm) 73 40 J £ - * P i M 40 40 40 n 40 30 30 42 42 54 156 156 158 191 184 peng peng p i Pi v Pi P i ft ITU be i ba b i Pi 191 191 195 212 212 223 37 40 47 51 207 207 47 47 47 51 51 154 50 54 203 203 - 615 -p i an p i an pi an pi ao pi ao p i ao pi ao p i e pie p i n p in v p in p I an p i an ( —* b i an) b i an b i an ft pu m m pan 122 122 122 129 131 125 C3 RP 75 77 75 75 77 78 78 132 132 204 204 149 149 111 p l n p ing Po po po PO fb ft m m n w-m feng pu bo bo f an ft pu pai pu 53 151 202 187 197 197 197 204 204 204 3 3 136 1 5 194 194 - 616 -po pou NX pou pu pu pu pu q i pou ffi C m) ffi PS bao bao 194 82 14 215 215 14 14 14 215 15 15 15 183 17 q i 215 215 37 37 37 37 q i j i 5& J 1 zh i IK zhi s 1 56 54 54 207 132 37 37 47 52 52 52 56 56 56 56 56 48 48 48 59 - 617 -q i q i q i a q i a q i an H . ffl q ' H ( & it; q i 13 38 50 58 . 58 60 156 39 39 55 55 60 60 114 114 156 156 207 99 99 99 101 101 101 q l an q l an q l an q i a n q i an q i a n g q i a n g -m. IB. q i an 107 123 123 129 98 129 141 101 102 123 123 123 123 124 124 m 100 106 ' 108 171 i t 173 % 172 IM 171 JTJC 171 172 - 6 1 8 -qi ang q 1 ao q i ao q> ¥ o qi ao qi e qie qie q i e qiang a i ang ^ ( —•que) Ire] *5 m K que que 1* 105 133 75 133 75 qln 110 72 111 176 111 152 75 i q l n i 111 75 111 76 111 76 m 150 76 W) - 155 76 if 155 73 m 155 77 q l n m 111 V qin 112 113 78 qing 200 78 I f 201 81 %f 205 183 m 205 132 m 210 3 qing f f 200 11 3t 200 12 S U 200 104 M 198 - 619 -qing qing qing qiong q iong q l u q i u q i u qu m 5. m •Pi sua 198 qing 201 211 211 q i n 199 207 207 qiong) 219 210 219 219 86 86 87 86 86 chou 87 87 87 87 87 X 1 U qu 19 qu qu qu quan quan (-* qu ) m qi 23 162 162 24 24 24 165 133 221 226 19 19 24 25 25 226 26 22 212 141 140 140 140 141 141 - 620 -quan sy quan quan que que que que qun ran sy ran r ang r ang It quan 14 q ue as sy r ang 141 141 147 145 145 146 148 6 106 176 176 176 183 183 183 164 164 123 123 103 103 172 172 r a n g r a n g r a n g rao sy rao ren r en r e n r e n g r e n g reng A 59 /rVv ft 39 Ob #5 shen 174 174 174 174 175 76 76 77 77 79 127 110 110 150 150 151 113 113 113 153 153 153 187 187 187 - 621 -rong rong rou ru ru B m 7L i t -t-t-iu frfff A A rong yue 154 209 219 219 219 222 223 223 223 223 224 224 87 87 221 19 24 21 26 26 114 141 ruan rui ruo sa sai sai san IJT0 M m 03 sa s 1 •sang 145 61 46 62 46 162 162 176 176 118 8 34 118 97 28 28 186 30 95 116 116 117 166 - 622 -sang sang sang sao sao se sen seng sha m sha ces-cm fi m i t sao NX sao shan cha 166 168 168 168 168 68 68 68 69 69 114 114 154 186 190 190 110 184 7 7 120 13 sha sha i shan shan IS C i ) fe(^sc) 0 chan (-> dan ) 9 99 33 190 34 97 100 100 118 121 123 123 123 115 103 103 121 124 124 125 125 125 125 125 - 623 -shan s h a n g s h a n g s h a n g shao shao shao shao she she ± ± m m m at s h a o zhao % ( # 0 104 172 172 174 174 174 ' 175 175 72 72 74 76 76 176 176 77 74 78 77 78 11 11 11 11 she she s he she i shen shen shen iff C- zhe P3 ft s h i s h u i can sen s h e n 127 127 11 12 12 12 11 105 105 127 203 110 150 150 150 150 150 150 112 112 112 - 624 -shen shen s h e n g s h e n g s h e n g s h e n g sh i 112 112 113 151 153 shi 154 f f -»• sh i fc 187 m 187 192 192 m 192 % 192 m 200 m 187 193 c h e n g 188 m 188 H 202 & 202 48 52 m 52 §fs 52 56 m 114 sh i sh i sh l 154 ft -s h i M - she 56 m 56 + 114 t t 114 ft 114 X 154 Ik 190 m 190 i i 190 203 m 203 49 53 m 53 1£ 57 j & 57 St 57 *& 58 i t 36 36 36 ft 36 - 625 -s h i 49 ft 49 ffi 49 TJN 55 /IH 55 ^ 55 ± 57 ft 57 ffi 57 rfc 58 'f# 58 I f 59 ^ 59 j & 59 # - 59 i l 154 190 5^ 190 tit 190 190 —» shi 9 M 203 203 U 203 shou 107^ 86 shou ^ —• shu 88 shou ^ 88 IT 88 Tp 88 shou ^ 88 31 90 m 90 ^ 90 g 90 H 90 s h " ffit 19 l i t 19 !^ 19 • 19 -§? 19 $ro 24 & 24 • M 221 shu ft. 163 — shu 221 221 M 226 shu 4i 21 Hi 21 H 21 - 6 2 6 -shu ^ 22 H 22 ft 25 10 226 ® 226 shu H • 21 ©fE —* shu JU: 22 8g 22 ft, 27 ^ 27 11 27 $ f 27 M 26 fa 90 $ f 163 £6 163 ^ 226 shuai ^ 63 $ 63 shuai f^lp 65 $ 65 c$ 163 163 shuin R ( £| ) 138 - & ( m ) 138 shuin 141 shuan M 139 shuang 'mm *§ 171 171 i t 181 shuang 174 shuang 182 t shui 63 shuY 64 shu i fit 46 46 62 shun J i - > dun shun III 162 ft 162 m 162 shuo m 143 shuo ft -> shu ^ -> shuai shao ) 176 shao ) 176 m 183 s i m 48 m 48 a s c » f ) 37, 51 627 -song song song I E E sh i oo 55 55 53 54 54 57 57 57 58 58 58 58 58 57 160 216 218 222 216, 222 223 223 214 217 song sou sou v S O U S U su prt su shu shu 225 225 225 86 86 86 86 83 84 14 14 14 225 18 18 18 18 89 216 220 220 220 225 6 2 8 -suan suan su 1 sui sun sun suo suo 5? n i ao 133 135 136 63 60 63 61 42 45 64 64 64 157 158 161 161 1 4 4 4 4 221 5 5 SUO ta ta ta tai tai tai tai t an x i e 7-K E 3 9& tan 20 1 97 97 97 117 94 94 94 97 28 28 28 28 28 29 30 30 29 29 92 95 115 115 92 - 6 2 9 -t an H 92 %% 95 95 ^ ( - > d a n ) 95 m 115 115 P P ( - d a n ) 115 • y tan 95 fi 116 116 tan 93 M-J 116 116 tang m 166 m 168 tang 166 166 166 166 166 166 tang 167 ^ c m) 167 tang 168 | | ( - > ting) 168 tao w 67 tao -/§ 67 ffi ( fB ) 67 tao ^ 67 i i 67 •M ( ^ ) 67 IM 67 H 67 tao §J 69 tao |F 70 te K£ 186 # 186 t e' ng fit 184 I f 184 H 184 184 i ? ' 184 % 216 t"" ^ 37 SI) 207 flj 207 £i 37 1 1 $| 37 ^ 37 37 - 6 3 0 -tV ^ 38 11 # 39 •/S 39 $W ' 39 M. 39 ' t ian '||S 106 t i a n ^ 129 f f t 106 ffi 129 • iflfL 129 ^ 129 7$ 107 t i a n §& 130 M> i ° 7 l i a n f-jS 107 t i a o #fc 79 t i a o 79 l i 79 gfe 79 t i"ao ^ —* t i a o t i a o m ( & ) 31 f l 81 t i e 108 lfj£ 108 ti"e —• t i e ^ 132 tie —> tie t i n g I S ,204 II 204 •7J 204 t ing ^ 204 # 204 204 M 204 204 3l 204 tYng $ § 205 $ g 206 § E ( i f ) ' 206 ting |g; 216 tong i l l 212 212 tong [W] 212 III 212 fiij 212 • 1aJ 212 J6 2 1 2 St 212 tong ^ | 214 frf 214 tong 0 214 nJ —> t ong - 631 -tong to u t ou t o u t o u tu tu tong tu tu t u a n t u a n t u l i t u i NX t u i t u i 'ftl fc ± p± t uo 81 81 81 84 83 215 14 14 14 14 14 1 6 16 17 1 7 133 133 134 40 186 40 41 42 43 t u n tu n tu o t uo t uo t u o w a wa w a Bi -* dun fa f t ft (15) ta wo m (£) K m. c ^ ) 148 156 156 156 156 1 1 137 169 169 1 1 1 5 5 6 13 13 44 138 13 13 146 44 - 632 -w a n w a n w a n w a n w a n g wa n g w a n g y a o wa 1 ft 7 C 0 f t SE w a n g 43 134 134 139 139 134 134 134 138 135 145 145 145 145 136 136 145 145 177 178 178 179 179 w a n g w a n g we i we i E m 3 3 : w a n g we l 179 179 179 179 179 180 41 61 61 63 63 66 65 65 40 61 64 66 66 66 63 65 we i - 6 3 3 -we i we 1 we n we n ( - > h u i ) y i 8fc m 3 f f l X - 62 66 66 67 66 28 46 46 62 62 62 65 66 66 67 67 67 67 67 157 157 163 163 we n we n we n we n g we n g wo wo w u X we n a ® ) fc ft 163 163 158 164 164 164 164 213 215 4 4 4 12 2 6 6 184 218. 137 15 15 15 23 23 - 634 -wu wu 216 23 0% 15 15 35 a 15 15 23 16 ffi 16 16 17 i t : 25 n 25 m 25 m 25 m 18 ' i n 18 m 18 18 26 26 84 160 165 165 m •9 XI M C S ) XI a 37 37 37 37 37 49 56 56 56 56 60 60 114 189 203 207 203 113 113 189 189 203 203 635 -xi XI x i a m 207 38 m 49 fc 49 58 58 M 39 39 M 39 % 39 51 m 199 x i a •9 X 1 122 8 m 8 fa 8 10 ft. 99 99 fa ' q i a 101 101 108 122 x i an ft T sha T C-^ he ) ft 151 108 9 9 10 10 10 10 194 101 101 105 123 123 129 127 98 93 100 107 118 118 123 130 130 130 - 636 -x i an 1 7 * x i an S Ci 103 124 124 130 99 ) 99 119 119 123 128 131 xiang i t . j i an m 147 xi ang 171 m 171 171 m 171 m 171 m 171 m 172 n 172 x l ang ft 171 m 171 ^ < [—• j iang) 182 x i"ang m 173 x i ang xiao xiao xiao 173 175 174 174 173 173 173 xiang ) 175 175 182 182 •75 75 75 75 75 76 79 79 76 76 80 176 yao yao 77 80 - 6 3 7 -xiao 74 74 74 78 81 xi e 10 129 i t 129 m 132 x l e % 10 m 10 m 32 m 33 m 47 # 106 108 108 . N X xi e 11 jfQ. (-> xue ) 148 x i e 12 12 m 12 33 35 34 M 34 x i e x I n x m g xing xi ng xing m 126 n 126 m 132 111 110 150 150 §? 150 ]% 155 m 205 xun xm M m ft 152 153 188 205 205 205 205 205 205 191 sheng) 201 206 189 193 - 638 -x i n g x i ong x 1 o n g x i u x 1 u m ft su ) ? g ( - * su ) chou x u 196 202 202 193 209 223 223 223 219 219 86 86 86 187 89 89 89 89 89 89 89 99 220 23 23 x u x u X l l x u a n x u a n 7E. /MR. m shu ft c h u ) ut 2: x u a n x u a n 23 19 19 19 19 21 20 20 20 21 26 39 162 221 221 225 140 127 144 140 147 147 141 142 639 -xuan xue xue xue xun A . JllL 35 x i ao x I e y I n 142 146 147 6 126 184 143 148 164 164 164 110 160 160 160 160 160 111 152 159 162 162 165 xun ya x i n ya y a y a ya l yan Y m m ya y a i f m ya ya 8 8 8 101 101 101 33 33 9 9 10 10 10 10 120 10? 102 105 - 640 -y an ang yan yan yan 3aE y 1 n 5 m q l an yan 123 127 130 130 100 105 102 102 102 102 121 123 123 127 129 141 St 103 103 105 119 125 142 m lot yan yang yang ft yang f£p 104 106 104 104 104 122 122 126 128 132 132 132 131 173 173 173 173 173 173 173 173 173 173 173 174 - 6 4 1 -y a n g y a n g y ao v ao y ao "2, yue va o F-3 y ao 174 174 175 175 175 78 76 76 76 76 80 80 72 72 76 76 76 76 76 80 73 78 80 74 y ao ye ye ye ye 79 79 79 176 176 176 FM — • n u e — • yue m y ao ye m. m 11 w> 11 M 11 — * x i e & 11 m 11 & 11 m she 12 106 105 105 105 129 204 - 642 -ye y i J3i y i •y l 204 52 56 60 60 114 154 43 48 49 52 52 52 56 56 56 56 63 56 50 50 50 58 58 58 y i y i 154 we i 35 36 M 36 m. 39 W 39 = 15 5H 39 It 51 m 51 m 51 j§ 51 m 55 %k a© 55 59 J§L 59 y i 60 114 & 154 190 ft 190 w 190 s . 190 190 204 204 - 643 -y i y i n y i n y i n 5%. B a i m m c 204 204 211 211 111 111 151 151 151 155 155 111 111 149 151 151 112 152 155 155 161 113 113 113 y i n y ing yi ng v ying ying yong ft fl m ik y in J mg 153 188 188 195 195 195 198 201 201 188 198 201 201 210 210 211 198 210 211 189 193 199 223 223 - 644 -v i n$ yong yong you you yc5u m m m m 7t yong 224 225 225 223 209 210 210 224 224 224 225 87' 87 87 91 87 87 87 87 87 87 89 89 89 89 you yu yu X vou 89 90 90 90 90 90 90 90 91 24 20 24 24 24 24 24 24 24 24 24 19 19 20 20 20 - 645 -yu yu yu yu m =4. gffj m m yu ao 20 21 21 26 26 26 26 26 27 27 27 27 27 27 27 22 22 22 22 22 99 191 221 yu yuan yuan yuan yuan 7 U M 9U 3S yuan 221 226 226 226 226 226 165 144 147 141 141 141 144 144 144 144 144 144 144 144 145 143 146 145 145 - 646 -yun yun yun 0 m m •ir it shuo yun 146 146 176 79 143 143 146 146 146 146 176 176 176 180 183 183 183 yun 161 164 164 161 161 161 165 yun aza i za I za i zan zan zang zang *8 m j un zha 165 165 165 189 120 94 28 28 29 29 29 30 30 92 110 135 117 117 96 96 166 166 168 - 6 4 7 -zang 'm. (-> cang) 168 m 168 zao 68 m 68 zao m 170 zao 69 69 69 <s 69 zao i 69 m. 69 m 71 « 71 ze 186 z e I zeng z e n g 0 186 # 194 J i . 194 z h a i ) 194 » 197 ce • t o 'ft zeng 186 184 184 184 185 zha zha z h a zha zha i : ( * & ) — c h a tL(3fe)~* ^ a 'jfp —> zha c ) IIIJ c m) $L - » ya —• zha ft TltiJ 193 193 120 120 120 99 99 , 101 122 101 99 7 9 9 9 9 9 197 31 196 - 6 4 8 -zha i zha i z h a i zhan zhan zhan zhang ze 7TZ J 1 194 32 194 34 35 (¥B) 102 ^ f i C ~ * n i a n ) 102 102 102 123 #§ 123 98 119 124 m 96 AS 96 99 104 119 121 125 H -*• chan 171 172 f~ zhang zhang zhang zhao zhao NX zhao zhao 3t ( i l ) zhu ffi (-> zhu) ffi /ft IE 172 172 173 173 174 173 173 173 173 175 175 175 175 175 76 76 76 21 73 77 74 74 74 77 77 - 649 -zhao zhe zhe zhe zhe zhen m iff iff ft Iff zhe zhe zha i zhe zha i 78 78 78 11 126 105 104 114 127 zhen 11 105 12 127 110 110 150 150 150 150 160 zhen zh en zhen zheng fll JE IE T zhen zhen zhen d i n g 200 200 200 112 151 151 113 152 152 153 153 153 153 112 187 187 195 195 195 z h e n g 200 200 188 - 650 -zheng zheng zheng z h i z h i zh i IE m #1 3i ft-zh i 201 188 188 202 202 202 48 48 48 48 48 48 52 56 56 114 190 203 189 114 154 154 189 189 zh i z h i m 190 fit 190 m 190 m 49 • 49 t o E3 53 53 ± 58 a t 58 i t 58 m 36 m 36 m 36 50 53 54 55 57 57 H 59 fa 59 59 59 59 59 154 - 651 -z h i | £ 154 W. 154 M 190 U 203 ^ 203 Ifc ' 54 zhong 218 ,t& 218 |K 218 #1 219 H 222 M 222 J£ ( Ut ) 222 zhong 224 f t 224 @ 224 zhong tp 219 219 | f c 220 A % 225 j | 224 zhou J§j 86 7% 86 ')M 86 #11 86 ^ 221 zhou 221 zhou M .88 H 88 zhou H 89 & 90 1 ^ 90 % 90 M 88 & 89 H 90 —• chou zhu ^ • 19 § # 19 jvfc 23 ^ 23 ^ 23 24 5*: 24 24 zhu 26 ^ 221 221 |$1 — * zhou 'M 226 zhu ^tt 21 652 -zhu 25 26 shu) 226 zhu 20 m 21 m 21 . m 22 25 a 27' l i 27 ft 27 27 27 us 27 n 221 221 PI 226 21 zhua 12 72 zhua zhao ) 73 z h u a i z h u a i zhua i ft 127 zhuan * • 140 ) 140 zhan zhuan zhuang zhuang zhu i z h u i zhun zhun zhuo m at' fit i t 140 141 98 139 141 142 • c h u a n) 142 171 171 171 181 175 175 182 63 63 45 46 64 63 160 160 161 161 143 183 183 - 6 5 3 -zhuo zhuo NX z i 183 zhuo z h u ) 21 176 176 176 183 183 183 183 183 183 183 51 51 51 55 55 55 56 49 53 57 57 57 Z 1 i 54 58 ft 58 z o n g 213 • # U t ) 213 216 217 217 222 222 z o n g m 214 z o n g $?F —» z5ng m 214 225 zou 86 NX z o u 83 zou m 90 84 zu 14 z u 159 216 225 NX zu 16 .IE IS 16 20 zuan i f 133 - 654 -ziian 135 zuan zu 1 \ zu 1 zun NX zun zun zuo zuo NX zuo zuo zuan zhuan du i ft " ft -ft -ft ft ft j un zuo \ zuo cuo 61 41 43 64 157 160 158 170 2 2 5 6 6 18 170 170 170 - 655 -BIBLIOGRAPHY BARALE, Catherine, Mae-chu Chang, Marjorie Chan, John D'Andrea, David Strecker and Moira Yip. 1977. "Aspects of the Kaiping vernacular: a f i e l d study report". ins. (1977 L i n g u i s t i c I n s t i t u t e , U. of Hawaii at Manoa) BAWDEN, CR. 1954. "An eighteenth-century Chinese source for the Portuguese d i a l e c t of Macao". S i l v e r Jubilee Volume of the Zinbun-Kagaku-Kenkyusyo (Kyoto U.) : 12-33. BEI-JING DA-XUE. 1962. Han-yu Fang-yin Zi-hui y% if j) i f % ;B . (Peking: Wen-zi Gai-ge Chu-ban-sheTT . 1964. Han-yu Fang-yan Ci-hui j | t ^ "t "t^ >$> . (Peking: Wen-zi Gai-ge Chu-ban-she). BOLTZ, William. 1977. "Notes on an eighteenth-century Sino-Portuguese^glossary". Romance Philology 30.3:442-453. . 1978. "Notes on dating the Cantonese d e n t i l a b i a l i -zation of Middle Chinese gutturals". Journal of the American Oriental Society 98.1:99-100. BU Y i ^ ^ . 1977. Ao-men Zhang-Gu ;|j ?*\ *f t1!^ • (Hong Kong: Guang Jiao Jing Chu-ban-she). CHAN, Marjorie K.M. (forthcoming). "A response to Boltz' notes on Cantonese d e n t i l a b i a l i z a t i o n " . CHAO, Yuen Ren. 1940. "Distinctions within Ancient Chinese" Harvard Journal of A s i a t i c Studies 5:203-233. . 1947. Cantonese Primer. (Cambridge, Mass.: Harvard University Press). . 1948. "Zhong-shan fang-yan" h 'T) \ . B u l l e t i n of the I n s t i t u t e of History and Philology. Academia S i n i c a XX:49-73. . 1968. A Grammar of Spoken Chinese. (Berkeley: U. of C a l i f o r n i a Press). CHEN, Leo and Jerry Norman. 1965. An Introduction to the Foochow Dialect. (San Francisco: San Francisco State College). CHENG, Teresa. 1968. "The phonological system of Cantonese". Project on L i n g u i s t i c Analysis. Reports Second Series. No. 5 (March 1968):C1-C77. CHEUNG, Hung-nin Samuel. 1972. Cantonese as Spoken i n Hong Kong. (Hong Kong: Chinese U. of Hong Kong). - 656 -CHINESE .LINGUISTICS PROJECT. 1970. Fang-yan Diao-cha Zi-biao J) s> ~f&\ % /*t-• • (Princeton: Princeton U.) 19 72. Handbook of Chinese Dialect Vocabulary. (Princeton: Princeton U.) CHOU Fa-kao (chief ed.). 1973. A Pronouncing Dictionary of Chinese Characters i n Archaic &^ Ancient Chinese, Mandarin St Cantonese. (Hong Kong: Chinese U. of Hong Kong). Deng-Yun Wu Zhong. -ffi £> . 1977 Reprint. (Taipei: Yi-wen Yin-sEu =guan). (Five rhyme tables included: Yun-jing DOUGLAS, Car s t a i r s . 1899. Chinese-English Dictionary of the Vernacular or Spoken Language of Amoy" (London: Presbyte-ri a n Church of England). EGEROD, S0ren. 1956. The Lungtu Dialect: A Descriptive and H i s t o r i c a l Study of a South Chinese Idiom. (Copenhagen: E. Munksgaard). . 1959. "A sampling of Zhong-shan Hakka". In: Studia Serica Bernhard Karlgren Dedicata. editted by S<rfren Egerod and Else Glahn. (Copenhagen). Guang-Yun % ^ . (Present copy used i s a corrected e d i t i o n of Chang family's e d i t i o n of the Guang-yun: Jiao-zheng Song-ben Guang-yun Jk. 2fL %\ ("includes an appended index), 1976. (Taipei: Yi-wen Yin-shu-guan).) HASHIMOTO, Mantaro J. 19 73. The Hakka Dia l e c t: A L i n g u i s t i c Study of i t s Phonology, Syntax and Lexicon. (Cambridge, England: Cambridge U. Press). HASHIMOTO, Oi-kan Yue. 1972a. Phonology of Cantonese. (Cam-bridge, England: Cambridge U. Press). 1972b. "Two features of Proto-Yue I n i t i a l s " . UnTcorn 9:20-40. HSU, Immanuel C.Y. 1970. The Rise of Modern China. (.London: Oxford U. Press). Hua-qiao Zhi: Ao-men % ^ ^ : ^ . 1914. (Taipei Hua-qiao Zhi Bian-zuan Wei-yuan Hui). ivio HUANG, Po-fei. „ Cantonese Dictionary; Cantonese-English, English-Cantonese" (New Haven: Yale U. Press). KAO, Diana L. 1971. Structure of the Syl l a b l e i n 'Cantonese. (The Hague: Mouton). - 657 -KARLGREN, Bernhard. 1954. "Compendium of Phonetics i n Ancient and Archaic Chinese". Museum of Far Eastern A n t i q u i t i e s , B u l l e t i n No. 26:211-367. KRATOCHVIL, Paul. 196 8. The Chinese Language Today: Features of an Emerging Standard. (London: Hutchinson & Co., Ltd.). KWOK, Helen. 1971. A L i n g u i s t i c Study of the Cantonese Verb. (Hong Kong: U. of Hong Kong). LI Fang Kuei. 1966. "The zero i n i t i a l and the zero s y l l a b l e " . Language 42:300-302. . 1973. "Languages and d i a l e c t s of China". Journal of Chinese L i n g u i s t i c s 1.1:1-13. . 1974. "Studies on Archaic Chinese phonology". Translated from Chinese by Gi l b e r t L. Mattos. Working Papers i n L i n g u i s t i c s , U. of Hawaii:171-282. (Original paper t i t l e d "Shang-gu Yin Yan-jiu" 5 % %j Tsing Hua Journal of Chinese Studies,N.S. IX, Nos. 1 and 2 (1971):1-61). LI Rong £ ^ . 1957. +"Fang-yan l i de wen-bai yi-du" 69 frfr-Iffi. Zhongguo Yuwen (April 1957) : 22-23 . LIANG Seng-bao >$l $ If (Qing). Si-sheng Yun-pu fijfc fa^t . 2 Vols. 1955 ed i t i o n . (.Peking: Gu-ji Chu-ban-she) . LIANG Shih-chiu (ed.-in-chief). 1972. A New P r a c t i c a l Chinese-English Dictionary. (Taipei: Far East Book Co., Ltd.). LIU Jun-ren % \^ . 19 67. Zhong-guo Di-ming Da Ci-dian «f IfD *<l> k • (Taipei: Wen-hai Chu-ban-she). LO Chang-pei ^ . 1956. Xia-men Yin X i M $ Second e d i t i o n . (Peking: Ke-xue Chu-ban-she). MATHEWS, R.H. (Comp.).„ Mathews' Chinese-English Dictionary. Revised American E d i t i o n . (Cambridge, Mass.: Harvard U. Press) . MATISOFF, James A. 1973. "Tonogenesis i n Southeast Asia". In: Hyman, Larry (ed.). Consonant Types and Tone. Southern C a l i f o r n i a Occasional Papers i n L i n g u i s t i c s 1:73-95. McCOY, William John, J r . 1966. Szeyap Data for a F i r s t Approxi-mation of Proto-Cantonese {with] Appendix. Ph.D. di s s e r t a -t i o n , Cornell U. . 1970. "Upper regi s t e r nasals and l a t e r a l s i n Canton-ese". Paper presented at the Third Sino-Tibetan Recon-struction Conference. NORMAN, Jerry. 1969. The Kienyang Dialect of Fukien. Ph..D. - 658 -d i s s e r t a t i o n , U. of C a l i f o r n i a at Berkeley. . 1970. "A characterization of the Min d i a l e c t s " . Unicorn 6:19-34. The P r i n c i p l e s of the International Phonetic Association. 1949. (London: International Phonetic Association, University College). PULLEYBLANK, Edwin G. 1962. "The consonantal system of Old Chinese". Asia Major 9:58-144, 206-265. . 1970-71. "Late Middle Chinese". Asia Major 15(2): 197-239, and 16 (1-2):121-168. . . 1973. "Some further evidence regarding Old Chinese -s and i t s time of disappearance". B u l l e t i n of the School of Oriental Studies XXXVI.2:368-373. . 1977. "Early Middle Chinese", ms. SCHANE, Sanford A. 1973. Generative Phonology. (Englewood C l i f f s , N.J.: Prentice-Hall). SMALLEY, William A. 1963. Manual of A r t i c u l a t o r y Phonetics. Revised e d i t i o n . (Tarrytown, N.Y.: P r a c t i c a l Anthropology). THOMPSON, Robert W. 1959. "Two synchronic cross-sections i n the Portuguese d i a l e c t of Macao". Orbis 8.1:29-53. VAN ETTEN, Albert Henry. 1931. A Study of the Phonology of Sixth Century (A.D.) Chinese, and a Comparison with that of the Modern Cantonese Dialect. M.A~. thesis, U. of C a l i f o r n i a , Berkeley. WHITAKER, K.P.K.A. 1956. "A study of the modified tones i n spoken Cantonese" i. Asia Major 5:9-36, 184-207. WONG, James Poy. 19 70. "A study of the T'ai Shan (Hoi San) d i a l e c t " , ms. (San Francisco State College). WONG,Sik Ling. 1954. A Chinese Syllabary Pronounced According to the D i a l e c t of Canton. (Hong Kong: Chung Hwa Book Co.) ( F i r s t e d i t i o n : 1940). Xiang-shan Xian Zhi ^ *^ . 1923. (1967 r e p r i n t of 1923 ed i t i o n used. Taipei: Cheng-wen Chu-ban-she). XIE Shou-chang et. a l . (ed.) . 1933. Zhong-guo Gu-jin Di-ming Da Ci-dian «f I2J £ & j^f . (Shanghai: Shang-wu Yin-shu-jguan) . Xin-hua Zi-dian $}l % %r • 1973. (Peking: Shang-wu Yin-sh.u-guanTI ' - 659 -YANG, Paul. 1971. "On the reconstruction of Old Chinese based on modern d i a l e c t data". Paper presented at the Fourth. Meeting of the Conference on Sino-Tibetan L i n g u i s t i c s , at Indiana U., Bloomington, Indiana. . 1972. "Prefix Ka- i n modern Chinese d i a l e c t s and Proto-Chinese". Paper presented at the F i f t h Annual Conference on Sino-Tibetan Language and L i n g u i s t i c Studies, Ann Arbor, Michigan. YIN Guang-ren and Ru-lin Zhang i\ fu M.%&~ikJtk. . C a . 1750. Ao-men J i - l u e jyi -t£j t?£- . (Present copy i s a 1968 reproduction of a 1800 edi t i o n of the text. T a i p e i : Cheng-wen Chu-ban-she) (Zhong-guo Fang-zhi Cong-shu [Hua-nan Di-fangJ 109). YUAN Jia-hua et a l . l^-^ 1 . 1960. Han-yu Fang-yan Gai-yao • (Peking: Wen-zi Gai-ge Chu-ban-she). YUE, Anne 0 . 1979. "The Teng-xian d i a l e c t of Chinese". Compu-ta t i o n a l Analysis of Asian and Afr i c a n Languages 12 (CAAL Monograph Series No.3l~^ ZHAO Chuan-cheng ^ • 1955. Qing-dai D i - l i Yan-ge Biao jj| \\ tt^if. }& £ jk^ • (Peking: Zhong-hua Shu-ju) . ZHONG-GUO KE-XUE YUAN YU-YAN YAN-JIU SUO (ed.) % • 1964. Fang-yan Diao-cha Zi-biao ~f t $ jL %- " (Revised e d i t i o n ) . (Peking: Ke-xue Chu-ban-she). Zhong-hua Xin Zi-dian ^ jp >%k . 1976. (Hong Kong: Zhong-hua Shu^u) I "Zhong-shan qiao-xiang feng-guang" ^ ffi !@L ^ ('Scenic spots of Chung-shan'—English t i t l e provided). 1975 (?). (Hong Kong: Scenery Publishers). Addenda: CHENG, Teresa. 1973. "The phonology of Taishan". Journal of Chinese L i n g u i s t i c s 1.2:256-322. SHU Xin-cheng et al.41" i\ If . 1947. Ci-hai jf$ fy. (Shang-hai: Zhong-hua Shu-ju).
Featured Collection
UBC Theses and Dissertations
Zhong-shan phonology : a synchronic and diachronic analysis of a Yue (Cantonese) dialect Chan, Marjorie K. M. 1980
Item Metadata
Download
- Media
- 831-UBC_1980_A8 C45.pdf [ 29.85MB ]
- Metadata
- JSON: 831-1.0100261.json
- JSON-LD: 831-1.0100261-ld.json
- RDF/XML (Pretty): 831-1.0100261-rdf.xml
- RDF/JSON: 831-1.0100261-rdf.json
- Turtle: 831-1.0100261-turtle.txt
- N-Triples: 831-1.0100261-rdf-ntriples.txt
- Original Record: 831-1.0100261-source.json
- Full Text
- 831-1.0100261-fulltext.txt
- Citation
- 831-1.0100261.ris
Full Text
Cite
Citation Scheme:
Usage Statistics
Share
Embed
Customize your widget with the following options, then copy and paste the code below into the HTML
of your page to embed this item in your website.
<div id="ubcOpenCollectionsWidgetDisplay">
<script id="ubcOpenCollectionsWidget"
src="{[{embed.src}]}"
data-item="{[{embed.item}]}"
data-collection="{[{embed.collection}]}"
data-metadata="{[{embed.showMetadata}]}"
data-width="{[{embed.width}]}"
async >
</script>
</div>
Our image viewer uses the IIIF 2.0 standard.
To load this item in other compatible viewers, use this url: https://iiif.library.ubc.ca/presentation/dsp.831.1-0100261/manifest